SPRING 2020

VOLUME LVI

THE QUARTERLY JOURNAL OF THE GEMOLOGICAL INSTITUTE OF AMERICA

Causes of Color in Corundum

Chinese Jade Carving
Characteristics of Blue to Yellow Beryl
History of the Chivor Mine, Part I
 Spring 2020
VOLUME 56, No. 1
EDITORIAL
1 Investigations into Corundum and Beryl Color, Jade Artistry, the
Chivor Emerald Mine’s History, and Swiss Rhodonite…
Duncan Pay

FEATURE ARTICLES
2 A Quantitative Description of the Causes of Color in Corundum
Emily V. Dubinsky, Jennifer Stone-Sundberg, and John L. Emmett
Presents a quantitative means of determining the causes of color in a gem material, using corundum’s
six major chromophores as an illustration.
p. 3
30 The Evolution of Chinese Jade Carving Craftsmanship
Mingying Wang and Guanghai Shi
Explores the art form of jade carving in China across five periods from the late Neolithic Age to the
present day.
54 Color Characteristics of Blue to Yellow Beryl from Multiple Origins
Yang Hu and Ren Lu
Discusses the features of alkali elements and their influence on the color of blue to yellow beryl.
p. 35 66 History of the Chivor Emerald Mine, Part I (1880–1925):
From Rediscovery to Early Production
Karl Schmetzer, Gérard Martayan, and Jose Guillermo Ortiz
Chronicles developments at the Chivor emerald mine in Colombia between 1880 and 1925, based on
archival records.
110 Rhodonite-Pyroxmangite from Tanatz Alp, Switzerland
Franca Caucia, Luigi Marinoni, Maria Pia Riccardi, Omar Bartoli, and Maurizio Scacchetti
Investigates the chemical, physical, and gemological properties of attractive rhodonite-rich rocks from
p. 92 this Swiss deposit.

REGULAR FEATURES
29 The Dr. Edward J. Gübelin Most Valuable Article Award
124 2020 G&G Challenge
p. 143
126 Lab Notes
Diamond with cavities showing radiation evidence • The “Matryoshka” diamond from Siberia
• Graphitic cavities on diamond • Corundum inclusions in diamond • Solid carved dark gray
diamond ring • Clarity-enhanced glass imitating emerald • Unusually large grandidierite
• “Electronic device” in an atypical bead cultured pearl • Fluorescence spectroscopy for colored pearl
treatment screening • Fossilized shell blister and blister pearl • Saltwater bead cultured pearl with
laminated nucleus • Bismuth glass-filled Burmese star ruby
140 G&G Micro-World
Actinolite in spinel • Bavenite in quartz • Diamond with mobile green diamond inclusion
• Reversible twinning in neodymium pentaphosphate • Pallasitic peridot with iridescent needles
• Sapphire with phenomenon resembling play-of-color • Staurolite in ruby • Triplite in beryl
• Quarterly Crystal: Unknown inclusion in triphane spodumene
p. 174 148 Diamonds from the Deep
An examination of the dissolution features that occur on diamond during storage in the mantle
and transport to the earth’s surface.
156 Gem News International
Tucson 2020 • Burmese star peridot • Top-quality demantoid • Russian emerald • Nephrite from
multiple sources • Nigerian gems and jewelry • Exceptional natural freshwater pearl • Sapphires
from Montana and Australia • Blue sapphire from Rakwana, Sri Lanka • Trapiche gems
• Carvings from Bali • Carvings, fantasy cuts, and master recutting • Fine-quality jadeite jewelry
• Nordic gems and jewelry • Ethical Gem Fair • Bicolor synthetic sapphire • Gianmaria
Buccellati Foundation Award • GIA Museum receives two awards • Weldon receives Bonanno
Award • DNA barcoding of freshwater pearls • Sunstone from Ethiopia • Dyed chalcedony imita-
tion of chrysocolla-in-chalcedony • Jadeite and serpentine doublet • Documentary series:
“Beautiful Gem Stories” • Dutrowite: New mineral species of tourmaline • G&G Facebook group
Editorial Staff
Editor-in-Chief Editors, Lab Notes Assistant Editor
Duncan Pay Thomas M. Moses Erin Hogarth gia.edu/gems-gemology
Managing Editor Shane F. McClure
Contributing Editors
Subscriptions
Stuart D. Overlin Editors, Micro-World James E. Shigley Copies of the current issue may be purchased for
[email protected] Nathan Renfro Raquel Alonso-Perez $29.95 plus shipping. Subscriptions are $79.99 for one
Associate Editor Elise A. Skalwold Donna Beaton year (4 issues) in the U.S. and $99.99 elsewhere. Cana-
John I. Koivula dian subscribers should add GST. Discounts are avail-
Brooke Goedert Editor-in-Chief Emeritus able for renewals, group subscriptions, GIA alumni,
Technical Editors Editors, Gem News Alice S. Keller and current GIA students. To purchase print subscrip-
Tao Z. Hsu Emmanuel Fritsch tions, visit store.gia.edu or contact Customer Service.
Gagan Choudhary Customer Service For institutional rates, contact Customer Service.
[email protected]
Christopher M. Breeding Martha Erickson
Jennifer Stone-Sundberg (760) 603-4502 Database Coverage
[email protected] [email protected] Gems & Gemology’s impact factor is 1.844, accord-
ing to the 2017 Thomson Reuters Journal Citation
Reports (issued July 2018). G&G is abstracted in
Production Staff Thomson Reuters products (Current Contents: Phys-
ical, Chemical & Earth Sciences and Science Cita-
Creative Director Photo/Video Producer Video Production tion Index—Expanded, including the Web of
Faizah Bhatti Kevin Schumacher Larry Lavitt Knowledge) and other databases. For a complete list
Pedro Padua of sources abstracting G&G, go to gia.edu/gems-
Production and Illustrator
Nancy Powers gemology, and click on “Publication Information.”
Multimedia Specialist Russel Samson
Juan Zanahuria Albert Salvato Manuscript Submissions
Multimedia Associate Betsy Winans
Photographer Gems & Gemology, a peer-reviewed journal, welcomes
Christopher Bonine the submission of articles on all aspects of the field.
Robert Weldon Please see the Author Guidelines at gia.edu/gems-
gemology or contact the Managing Editor. Letters
Editorial Review Board on articles published in G&G are also welcome.
Please note that Field Reports, Lab Notes, Gem News
Ahmadjan Abduriyim Jaroslav Hyršl George R. Rossman International, Micro-World, and Charts are not peer-
Tokyo, Japan Prague, Czech Republic Pasadena, California reviewed sections but do undergo technical and edito-
rial review.
Timothy Adams Dorrit Jacob Andy Shen
San Diego, California Sydney, Australia Wuhan, China
Copyright and Reprint Permission
Edward W. Boehm A.J.A. (Bram) Janse Guanghai Shi Abstracting is permitted with credit to the source.
Chattanooga, Tennessee Perth, Australia Beijing, China Libraries are permitted to photocopy beyond the lim-
its of U.S. copyright law for private use of patrons.
James E. Butler Mary L. Johnson James E. Shigley Instructors are permitted to reproduce isolated arti-
Washington, DC San Diego, California Carlsbad, California cles and photographs/images owned by G&G for
Alan T. Collins Stefanos Karampelas Elisabeth Strack noncommercial classroom use without fee. Use of
London, UK Manama, Bahrain Hamburg, Germany photographs/images under copyright by external par-
ties is prohibited without the express permission of
Sally Eaton-Magaña Lore Kiefert Nicholas Sturman the photographer or owner of the image, as listed in
Carlsbad, California Lucerne, Switzerland Bangkok, Thailand the credits. For other copying, reprint, or republica-
tion permission, please contact the Managing Editor.
John L. Emmett Ren Lu Fanus Viljoen
Brush Prairie, Washington Wuhan, China Johannesburg, South Africa Gems & Gemology is published quarterly by the
Gemological Institute of America, a nonprofit educa-
Emmanuel Fritsch Thomas M. Moses Wuyi Wang tional organization for the gem and jewelry industry.
Nantes, France New York, New York New York, New York
Postmaster: Return undeliverable copies of Gems &
Eloïse Gaillou Aaron Palke Christopher M. Welbourn Gemology to GIA, The Robert Mouawad Campus,
Paris, France Carlsbad, California Reading, UK 5345 Armada Drive, Carlsbad, CA 92008.
Gaston Giuliani Ilene Reinitz Chunhui Zhou Our Canadian goods and service registration number
Nancy, France Chicago, Illinois New York, New York is 126142892RT.
Lee A. Groat Nathan Renfro J.C. (Hanco) Zwaan Any opinions expressed in signed articles are under-
Vancouver, Canada Carlsbad, California Leiden, The Netherlands stood to be opinions of the authors and not of the
publisher.
Richard W. Hughes Benjamin Rondeau
Bangkok, Thailand Nantes, France

About the Cover

Distinguished gem carver Michael Dyber created “Flora” from a 3,070 ct Ukrainian heliodor found in the
Volodarsk-Volynskii mine. The carving weighs 1,754.70 ct and measures 165 mm tall, 50 mm wide, and 37 mm
deep. It was photographed during the 2020 Tucson gem shows, which are featured in the Gem News International
section of this issue. Photo by Robert Weldon/GIA; courtesy of Michael M. Dyber.

Printing is by L+L Printers, Carlsbad, CA.

GIA World Headquarters The Robert Mouawad Campus 5345 Armada Drive Carlsbad, CA 92008 USA
© 2020 Gemological Institute of America All rights reserved. ISSN 0016-626X
 Investigations into Corundum and Beryl
Color, Jade Artistry, the Chivor Emerald
Mine’s History, and Swiss Rhodonite…
Welcome to the Spring Gems & Gemology! Publication was significantly delayed by the COVID-
19 pandemic, but we’re very proud to present this issue, which offers a blend of science, art, and
history. In our lead article, authors Emily Dubinsky, Jennifer Stone-Sundberg, and John Emmett
illuminate the chemistry behind gem corundum’s rich array of colors to reveal the influence of the
concentrations of the six identified
chromophores found in natural “This article provides insight for the practic-
corundum on the depth of color. The ing gemologist on the influence of six specific
authors use quantitative visible absorp-
tion spectroscopy and chemical analysis
chromophores and their concentrations on
by secondary-ion mass spectrometry depth of color in corundum.”
(SIMS) to study specially chosen
synthetic and natural corundum samples to calculate each chromophore’s absorption cross section. This information is
presented as spectra and color-circle arrays, which we hope practicing gemologists will find informative.
In our second paper, Mingying Wang and Guanghai Shi of the China University of Geosciences (Beijing) present a
study of Chinese jade carving artistry ranging from time-honored, traditional techniques to state-of-the-art computer
numerical control and 3D replicate engraving. Using data from almost 2,500 contemporary pieces, the authors demon-
strate the art’s rich heritage and its continuing innovation.
Next, two more authors from the China University of Geosciences (Wuhan), Yang Hu and Ren Lu, examine the
chromophores in blue and yellow beryl using quantitative spectroscopy and trace-element analysis to explore their
color-causing characteristics. The authors show that different concentrations of iron ions with various valences and
occupancies are most responsible for the gem’s blue to green to yellow colors.
Our following paper offers a change of topic. Karl Schmetzer, Gérard Martayan, and Jose Guillermo Ortiz provide the
first of two installments detailing the checkered history of Colombia’s Chivor emerald mine: its abandonment, redis-
covery, and intermittent operations by a fascinating cast of characters, including German gem merchant Fritz Klein.  
In our final feature article, a team of Italian researchers led by Dr. Franca Caucia offers a study of the chemical, physical,
and gemological properties of attractive rhodonite-rich rocks from Switzerland’s Tanatz Alp.
Our regular sections have much to offer, too. In Lab Notes, we describe diamonds with graphitic cavities and corundum
inclusions, a clarity-enhanced glass imitating emerald, a large grandidierite, and curious pearls. Micro-World also
provides intriguing curiosities, including a diamond with a mobile green diamond inclusion and a pallasitic peridot
with iridescent needles. In Diamonds from the Deep, regular contributors Karen Smit and Steve Shirey demonstrate
that diamonds are not necessarily forever with an exploration of the dissolution many diamond crystals suffer during
transport in kimberlite magma. Our Gem News International section surveys gem materials and events from the 2020
Tucson gem shows, including star peridot, demantoid garnet, emerald, sapphire, trapiche gems, Tucson’s first ethical
gem fair, and the results of the 2019 Buccellati jewelry design competition. Please also see the winners of our Dr.
Edward J. Gübelin Most Valuable Article Award and this year’s G&G Challenge quiz.  

Finally, do make sure you visit our new G&G Facebook group at www.facebook.com/groups/giagemsgemology. Since

our February 2020 launch, over 7,000 of you have joined, so if you haven't seen it yet, please check it out. Thank you all
for your support and interest! 

Duncan Pay | Editor-in-Chief | [email protected]

EDITORIAL GEMS & GEMOLOGY SPRING 2020 1

FEATURE ARTICLES

A QUANTITATIVE DESCRIPTION OF THE

CAUSES OF COLOR IN CORUNDUM
Emily V. Dubinsky, Jennifer Stone-Sundberg, and John L. Emmett

The color of a gemstone is inextricably linked to its chemical composition, yet the quantitative relationship be-
tween color and chemistry is poorly understood in most cases. Here we use corundum to present a comprehen-
sive quantitative description of the causes of color in a gem material and illustrate its predictive power. Natural
corundum has six major chromophores that cause color: Cr3+, h•-Cr3+, Fe3+, h•-Fe3+, Fe2+-Ti4+, and V3+. We use
synthetic samples doped with a single chromophore to study their light absorption behavior in isolation. Natural
samples are used as well to study single chromophores, and we can subtract out the absorption of additional
chromophores that might be present. Combining quantitative visible absorption spectroscopy with chemical
analysis by SIMS, we are able to calculate the absorption cross section of each chromophore. The absorption
cross section information is used to determine the depth of color that would occur in corundum of a given size
(optical path length) containing a specific chromophore of a given concentration.

G
emstones are valued for their beauty, rarity, ferred to as trace elements, or other point defects in
and durability, and what typically captures our the crystal lattice that have been incorporated during
attention is their magnificent array of colors. growth or later equilibration in nature. The causes of
Corundum exhibits an extremely wide range of colors color in corundum are many and have been primarily
in nature (figure 1). From pigeon’s blood red ruby to addressed in a non-quantitative way for many years
cornflower blue and lemon yellow sapphire, nearly (see, for example, Fritsch and Rossman, 1987, 1988;
every color is represented. The only corundum color Häger, 2001; Emmett et al., 2003). Trace elements
not represented in nature is a saturated intense emer- themselves can be the direct cause of color. Cr3+, for
ald green. However, less intense olive green to teal example, creates pink and red coloration in corun-
green stones are often found in basalt-hosted corun- dum. Trace elements can also interact with each
dum deposits. other, creating a new chromophore. The Fe2+-Ti4+ pair
Corundum’s broad range of colors is related to its is such an example, strongly absorbing in the yellow
detailed chemistry. Some minerals possess inherent and red regions of the spectrum and thus creating
color because the chromophore is one of the basic magnificent blue sapphires.
chemical components of its makeup. Such stones are When beryllium-diffused corundum entered the
termed idiochromatic, meaning self-colored. For ex- marketplace, we were surprised by the wide range of
ample, turquoise, whose chemical formula is colors that were produced, seemingly by a single ele-
CuAl6(PO4)4(OH)8 • 4H2O, is colored by copper, a pri- ment (Emmett et al., 2003). Measurements of the
mary component of its structure. beryllium levels showed that the concentrations were
Other minerals such as corundum are, when very generally from a few to a few tens of parts per million
pure, completely colorless. In fact, pure corundum, atomic (ppma), yet the colors produced were often in-
with the chemical formula Al2O3, is absolutely trans- tense. For comparison, red coloration in corundum re-
parent from the deep ultraviolet region into the in- quires several hundred to a few thousand ppma of
frared. Such minerals are termed allochromatic. Their Cr3+, a concentration at least two orders of magnitude
colors in nature are caused by minor impurities, re- greater than Be2+, to produce strong color.
Our studies of the beryllium-diffused stones (Em-
mett et al., 2003) demonstrated that the Be2+ ion itself
See end of article for About the Authors and Acknowledgments.
GEMS & GEMOLOGY, Vol. 56, No. 1, pp. 2–28,
was not the cause of color. However, replacing a
http://dx.doi.org/10.5741/GEMS.56.1.2 trivalent aluminum ion with a divalent beryllium
© 2020 Gemological Institute of America ion required the creation of a trapped hole (h•) for

2 CAUSES OF COLOR IN CORUNDUM GEMS & GEMOLOGY SPRING 2020

Figure 1. Gem corundum is rarely colored only by a single chromophore; the vast majority are colored by a combi-
nation of two or three chromophores. This composite photo illustrates the colors of single chromophores as they
would appear in a faceted stone. We searched through a few thousand photos, courtesy of Lotus Gemology, to find
these few that closely matched the single-chromophore color circles in this article. These gems represent (clock-
wise from top left) V3+ (first two stones), Cr3+, h•-Cr3+, h•-Fe3+, Fe2+-Ti4+, and Fe3+. Photos by Lotus Gemology.

CAUSES OF COLOR IN CORUNDUM GEMS & GEMOLOGY SPRING 2020 3

charge compensation. A trapped hole is an oxygen To determine the visual color of a corundum sam-
ion with a valence of –1 rather than –2. It is this O–1 ple from the chemical analyses, we need to know
ion that is the strong absorber of light (Kvapil et al., four additional factors: the absorption spectrum of
1973). In natural stones, this trapped hole associating each of the chromophores, the absorption
with either iron or chromium creates intense golden “strengths” of the chromophores, the thickness and
yellow or intense orange colors, respectively. crystallographic orientation of the sample, and the
In natural corundum, beryllium only rarely, if color temperature of the illumination.
ever, exists in solution in the corundum lattice and All of these factors are well known in gemology
thus does not produce these colors. Instead, natural with the exception of the absorption “strength.” The
corundum usually contains some amount of magne- term for the absorption “strength” in physics is the
sium. Magnesium, like beryllium, is divalent (Mg2+, absorption cross section (see box A on pp. 6–9; we
Be2+) when it replaces Al3+ in corundum and thus re- strongly advise reading box A before proceeding). The
quires charge compensation. If the stone is acceptor- absorption cross section quantitatively determines
dominated1 with [Mg + Ni] > [Si + Ti + H] (square the effectiveness of a single ion or ion pair in a partic-
brackets denote concentrations in ions/cm3), and ular host material (such as corundum) in absorbing
formed in conditions of relatively high oxygen fugac- light of a given wavelength. The larger the absorption
ity, the charge compensation is by a trapped hole (h•). cross section, the stronger the chromophore. The unit
Since it is the trapped hole in association with iron of the absorption cross section is centimeters squared
or chromium and not the beryllium or magnesium (cm2), which represents an area. One can conceptually
that produces color, the colors produced naturally by visualize this as the size of the absorbing area that a
magnesium or by diffusion of beryllium are very sim- single chromophore particle presents to the light
ilar (Emmett et al., 2003, 2017b; Kröger, 1984). beam. However, note that it is not the actual physical
Having identified the h•-Fe3+ and the h•-Cr3+ as two size of an ion or an ion pair.
additional chromophores in natural corundum, we
have six major chromophores that are responsible for
the multitude of colors in natural corundum: Fe3+,
Cr3+, V3+, Fe2+-Ti4+, h•-Fe3+, and h•-Cr3+. Individually, In Brief
their colors in corundum are as follows: • Natural corundum has six major color-causing chro-
mophores: Cr3+, h•-Cr3+, Fe3+, h•-Fe3+, Fe2+-Ti4+, and V3+.
Pink, Red: Cr3+
Orange: h•-Cr3+ • The depth of color produced by each chromophore
depends on three factors: the size of the gem, the con-
Yellow: Fe3+, h•-Fe3+
centration of the chromophore in the gem, and the ab-
Blue: Fe2+-Ti4+, V3+ sorption strength of the chromophore.
Green, Purple: V3+
• The absorption strength of each chromophore is ex-
These chromophores can occur singly in a corun- pressed by its absorption cross section. The cross sec-
dum sample (again, see figure 1), though it is com- tions of the various chromophores in corundum vary
mon for natural corundum to contain more than one by a factor of approximately 500, meaning they range
color-causing agent, as all natural corundum gener- from weak to strong colorants.
ally contains measureable levels of trace elements • Using the cross section data provided in this paper, the
Mg, Si, Ti, V, Cr, Fe, and Ga that have been incorpo- color produced in a sample of any specified size and
rated into the lattice. When multiple chromophores chromophore concentration can be calculated.
are present, the apparent color of the corundum sam-
ple results from the sum of the light absorption by
each of the chromophores present (Emmett et al.,
Knowing the absorption cross sections is ab-
2017a).
solutely critical to determining the origin of color of a
given sample. Before detailed chemical analyses were
1
available, it was often assumed that iron was primarily
We use the terms “acceptor-dominated” and “donor-dominated” and
also refer to chemical reactions among trace elements here without responsible for yellow color in sapphire. However, we
detailed explanations. These matters are discussed in detail in Emmett now know that there are yellow sapphires with 3000
et al. (2003), with corrections and extensions in Emmett et al. ppma Fe and yellow sapphires with 200 ppma Fe with
(2017a,b). Rather than repeat these discussions, we refer the reader to
these references and the extensive references therein. Furthermore, similar depth or intensity of color, so clearly these are
we will not repeatedly restate these references throughout this text. not colored by the same chromophore.

4 CAUSES OF COLOR IN CORUNDUM GEMS & GEMOLOGY SPRING 2020

 Figure 2. Shown are some
of the many crystallo-
graphically oriented sam-
ples used in this study.
Clockwise from top left,
they represent the follow-
ing chromophores: h•-Cr3+,
h•-Fe3+, Fe3+, Cr3+, V3+, and
Fe2+-Ti4+ (the two blue sam-
ples at lower left). The
sample in the far lower left
corner is a natural blue
sapphire from Yogo Gulch,
Montana, and the sample
in the upper right corner is
a natural yellow sapphire
from the Subera deposit in
Queensland, Australia; the
rest are Czochralski grown.
Sample numbers clockwise
from upper left are 991,
992, 1121, 1110, 1102,
1065, and 1014; detailed
information on the sam-
ples can be found in Ap-
pendix 1.

The objective of this paper is to present the results and single-doped synthetic crystals grown by the
of our efforts over more than a decade to determine Czochralski method (figure 2). Unlike many other
both the E⊥c (O-ray) and E||c (E-ray) (see box B) ab- minerals, corundum contains only a single small
sorption cross sections for all six chromophores in cation site. Additionally, the energy required to force
corundum and to describe how they were determined. a cation into an interstitial site is very high (Mat-
We also illustrate the range of colors these chro- sunaga et al., 2004). These facts, together with the
mophores produce. Finally, the digital files at 1.5 nm fact that the crystal must be rigorously electrically
resolution and 1 nm wavelength intervals from 200 neutral, very strongly constrain what trace elements
to 1100 nm for each of these chromophores are avail- and what valence states of these trace elements can
able to the reader to download for their own use (see exist in corundum. For example, iron in corundum
h t t p s : / / w w w. g i a . e d u / d o c / s p 2 0 - c o r u n d u m - by itself will be Fe3+ because the crystal must be elec-
chromophores-absorption-cross-section-data.xlsx). trically neutral. That is, the valence of Fe must equal
Determining the absorption cross sections relies the valence of Al3+. Unless there is a tetravalent
upon accurate determination of the chromophore donor such as Si4+ or Ti4+ or an H+ interstitial (El-Aiat
concentrations in the samples. These determinations and Kröger, 1982; Norby, 1989; Beran and Rossman,
were made possible by secondary ion mass spectrom- 2006; Li and Robertson, 2014) to charge compensate
etry (SIMS) analysis at the California Institute of it, Fe2+ will not exist. If there is such a donor, the Fe2+-
Technology (Caltech). SIMS was calibrated with sin- Si4+ and Fe2+-Ti4+ pairs will form. Thus the concentra-
gle-element ion implants in sapphire standards, elim- tion of Fe2+ is limited by the concentrations of H+,
inating any matrix effects. The calibration of SIMS, Si4+, and Ti4+. This is why we thoroughly analyze
a major effort in itself, is described in detail in Stone- even the singly doped synthetic crystals.
Sundberg et al. (2017). If a very high-purity synthetic crystal doped only
The chromophore concentrations in this paper with Fe3+ is processed at high temperature in a highly
have largely been determined by SIMS analyses from reducing atmosphere, oxygen vacancies with a
synthetic corundum crystals grown doped only with charge of +2 with regard to the lattice may form and
a single chromophore. This is true of Cr3+, V3+, h•-Fe3+, thus charge compensate two Fe2+ ions. It is for this
and h•-Cr3+. The Fe3+ chromophore and Fe2+-Ti4+ chro- reason that we process at high temperature in a pure
mophore data were determined using both natural oxygen atmosphere many of the synthetic crystals

CAUSES OF COLOR IN CORUNDUM GEMS & GEMOLOGY SPRING 2020 5

 BOX A: THE ABSORPTION CROSS SECTION
Highly pure allochromatic gem materials are colorless.
Color in these gems results from the incorporation of a
variety of trace elements and other point defects into the
lattice. Unlike idiochromatic gems, whose chro-
mophores are part of the fundamental composition of the I0 I
mineral, the concentration of trace elements incorpo-
rated in allochromatic gems can vary widely. Corundum,
an allochromatic gem, displays a very wide range of col-
oration: Nearly every color with the exception of emer- t
ald green is available. In what follows, we identify and
quantify the six chromophores that produce this wide
range of coloration in natural corundum by incorporating
either singly or in combination in the corundum lattice. Figure A-1. This simple diagram represents a transparent
Three of these chromophores are single transition material of thickness t. Light with intensity I0 strikes its
metals (Fe3+, Cr3+, and V3+), one is a pair of metals (Fe2+- surface at left. Some of the light is absorbed as it travels
Ti4+), and two are comprised of a lattice defect and a tran- through the material, so the beam has intensity I as it
sition metal (h•-Cr3+ and h•-Fe3+). As we demonstrate in exits at right. The light transmission T of the sample is
this article, the effectiveness of each of these chro- represented by the ratio of the exiting beam intensity to
mophores in producing color varies greatly because of the entering beam intensity, I/I0.
their very different electronic structures. For example,
nearly 2000 ppma of Fe3+ in a 1 cm thick sample is
needed to produce a saturated yellow, while only 1.5 All real samples have a portion of the probe beam re-
ppma of the h•-Fe3+ will produce a similar depth of color.
flected from their surfaces. Thus a spectrophotometer
In gemology we are familiar with the absorption spec-
does not directly measure the true absorbance, A, but
tra of the majority of these chromophores as single enti-
rather A*.
ties in corundum, but we do not have a simple
To determine the true absorbance A, the measured
methodology for characterizing their “strength.” The
absorbance A* must first be corrected for the loss from
strength determines the concentration of a chromophore
multiple reflections between the two polished surfaces
(in ppma) required to produce a given level of color satu-
of the sample. This is done by summing (T.S. Hemphill,
ration. Knowing the strength of the chromophores and
pers. comm., 2011) all the reflections between the two
the chemical composition determined by SIMS or LA-
surfaces and using a three-term Sellmeier equation for
ICP-MS analysis, we can correctly identify the sources of
corundum to determine the sample’s refractive index as
color in a sample. In physics the strength of an absorber
a function of wavelength (Tatian, 1984). The index input
of light or X-rays or neutrons is often referred to as an ab-
data to the Sellmeier equation is a combination of all rel-
sorption cross section. We can use this simple concept to
establish relative strengths of our chromophores. This atively recent data from the suppliers of high-purity syn-
paper presents our determination of the absorption cross thetic corundum material for the fabrication of
sections for each of the six chromophores. high-purity optical elements. The Sellmeier equation fit
To define the terms we will use, consider a transpar- to these data was performed for us by John Trenholme
ent material of thickness t, as shown in figure A-1. using TableCurve 2D software. Correcting for this loss
If an incident beam of light with intensity I0 strikes we have the true absorbance, A. This correction is im-
a sample that absorbs light, the intensity of the light portant because it is greatest where the absorption is
transmitted through it is I, assuming no reflection from least, and it is just this region of the absorption spectrum
the surfaces. Now the transmission, T, of the sample is: that transmits the most intense light and thus con-
tributes most to the color. This reflection correction has
T = I/I0 been made for all spectra presented in this paper and for
Or, if expressed as a percentage: all spectra used in deriving these results.
The true absorbance, A, for E⊥c of a synthetic sap-
T (%) = 100 I/Io phire 6.26 mm thick containing 116 ppma V3+ is shown
The decadic absorbance, A, of the sample is then: in figure A-2. Note that the vertical axis is unitless be-
cause absorbance is simply a ratio.
10–A = I/I0 The true absorbance A depends in turn on the thick-
ness of the sample, t, the concentration of the chro-
or
mophore, c, and its absorption cross section, σ. With the
A = –Log10(I/I0). true absorbance, A, we can write:

6 CAUSES OF COLOR IN CORUNDUM GEMS & GEMOLOGY SPRING 2020

 V3+ E⊥c V3+ E⊥c
1 3.0

ABSORPTION COEFFICIENT (cm–1)

0.9
2.5
0.8

0.7
ABSORBANCE

2.0
0.6

0.5 1.5

0.4
1.0
0.3

0.2
0.5
0.1

0 0
200 300 400 500 600 700 800 900 1000 1100 200 300 400 500 600 700 800 900 1000 1100

WAVELENGTH (nm) WAVELENGTH (nm)

Figure A-2. UV/visible/near-infrared (UV-Vis-NIR) ab- Figure A-3. UV-Vis-NIR infrared absorption coefficient
sorption spectrum of synthetic sapphire 6.26 mm thick spectrum of the V3+-doped synthetic sapphire, produced
containing 116 ppma V3+, as measured by a spectropho- by dividing the spectrum in figure A-2 by sample thick-
tometer and corrected for internal reflections. Note that ness (0.626 cm) and multiplying by 2.303. Note that the
the spectrophotometer measures light absorption in ab- shape of the curve is the same but the y-axis units are
sorbance units, which are dimensionless. cm–1.

I/I0 = T = 10–A = e–αt = e–σct vide the absorption coefficient of each chromophore by
Where I/I0 and T are here associated only with the ac- its concentration in absorbers/cm3 (note that 1 ppma in
tual absorption, and corundum is a concentration of 1.178 × 1017 ions/cm3):

T = transmittance σ = α/c
A = decadic true absorbance (dimensionless) The result, σ, is termed the absorption cross section.
t = thickness of the sample in cm Figure A-4 shows the absorption cross section for the
α = naperian true absorption coefficient in cm–1 same vanadium-doped sapphire sample.
c = chromophore concentration in ions/cm3 With absorption coefficient units of cm–1 and concen-
σ = true absorption cross section in cm2 tration units of ions/cm3, the units of the cross section
e = base of the natural logarithm = 2.71828… are cm2/ion, or just cm2, which is an “area” per absorber.
The reflection-corrected absorption coefficient is simply One can conceptualize this as the size of the absorbing
area that a single chromophore particle presents to the
α = 2.303A/t light beam. In other words, imagine that the light beam
(In gemology publications, the uncorrected absorbance is an arrow and the absorption cross section is a target; a
A* or an absorption coefficient derived from A* is usually larger absorption cross section means the target’s area is
presented.) larger. However, note that it is not the physical size of
The absorption coefficient of the V3+-doped sapphire an ion or an ion pair. The vertical axis in figure A-4 is la-
sample above is shown in figure A-3. Because A, a unit- beled “absorption cross section × 1019 cm2.” This means
less ratio, is divided by the dimension t in cm, the units that the actual values plotted are from 0 to 2 times 10–19
of absorption coefficient are cm–1. The absorption coeffi- cm2. It is important to observe that the spectra in figures
cient is independent of the sample thickness, making it A-2, A-3, and A-4 are exactly the same. Only the vertical
useful for comparing samples. axis values and units are different.
The absorption coefficient is, of course, dependent on Each chromophore has its own absorption cross sec-
the concentration of the chromophore. If we want to tion that is unique to that particular absorber. A sample
compare two different chromophores but the samples with multiple chromophores of different concentrations
have different chromophore concentrations, we can di- is best characterized by an absorption coefficient, as we

CAUSES OF COLOR IN CORUNDUM GEMS & GEMOLOGY SPRING 2020 7

 The concept of absorption cross section is very useful
V3+ E⊥c in understanding the types of optical transitions involved
ABSORPTION CROSS SECTION × 1019 (cm2)

2.0 with the individual chromophores. It is also useful in

1.8
comparing similar chromophores in different environ-
ments. Mattson and Rossman (1988) used this concept
1.6
in comparing the strength of the Fe2+-Ti4+ absorption
1.4
bands in different minerals. Their absorption cross sec-
1.2 tion was in units of (moles)–1 cm–1, which is simply a dif-
1.0 ferent system of units but the same concept.
0.8 This paper presents the color calculations for a range
of concentrations of each of the six chromophores. The
0.6
color coordinates were calculated using Thermo Scien-
0.4
tific’s GRAMS/AI spectroscopy software. To calculate
0.2 the color coordinates with Grams/AI, the spectral data
0 is input as the single-pass transmission spectrum, T. T
200 300 400 500 600 700 800 900 1000 1100 is related to the single-chromophore absorption cross
WAVELENGTH (nm) section and concentration as follows:
T = 10–A = 10–(ctσ/2.303)
Figure A-4. UV-Vis-NIR cross section spectrum of the V3+-
doped synthetic sapphire, produced by dividing the spec- where the product, ct, is termed the areal density in
trum in figure A-3 by the V3+ concentration of the sample. units of ions/cm2. Under the color circles in this paper,
Note that the shape of the curve is the same but the y- we present the areal density in ppma-cm. This factor is
axis units are cm2.
ppma-cm = ct/(1.178 × 1017 ions/cm3)
where 1.178 × 1017 is the number of ions in 1 cubic cen-
do with all natural samples. This absorption coefficient timeter of Al2O3 determined by
is the summation of each chromophore’s cross section
1 ppma = 5Ar/M = 1.178 × 1017 ions/cm3
multiplied by its concentration. For example, a sample
containing three chromophores of three different con- where A is Avogadro’s number (6.022 × 1023 mole-
centrations has an absorption coefficient of cules/mole), r is the density of corundum (taken as 3.99
grams/cm3), and M is the molecular weight of corun-
α = σ1c1 + σ2c2 + σ3c3
dum (101.961 grams/mole).
The magnitude of the absorption cross section of Finally, it should be stated that presenting the defini-
the chromophore, and the portion of the visible spec- tions and units of A, α, and σ is very important, as there
trum that its absorption spectrum covers, taken to- are multiple definitions in use in the scientific literature.
gether provide a rough indication of chromophore
strength. The portion of the visible spectrum is esti-
mated by the full width half maximum (FWHM) of the
absorption bands in the visible region, divided by 300 TABLE A-1. Comparison of various chromophores’ E⊥c
nm (the 400–700 nm approximate visible range). Table absorption values.
A-1 compares these values for the E⊥c absorption of the
various chromophores in order of increasing absorption Peak absorption
Approximate fraction of
cross section in the
cross section. Chromophore the 400–700 nm range at
400–700 nm range,
Examining the peak absorption cross sections first, chromophore FWHM
cm2 E⊥c
note that the values span a factor of about 500. This is a
huge difference. Next, examining the portion of the visi- Fe3+ 2.3 × 10–20 ± 8.0% 0.09
ble spectrum spanned by each chromophore, note that
V3+ 1.0 × 10–19 ± 8.9% 0.37
they are within a factor of two and a half with the excep-
tion of Fe3+, which is about one-fifth of the average of the Cr3+ 1.6 × 10–19 ± 7.6% 0.40
other five. Taking into consideration the very low absorp- 2+ 4+ –18
Fe -Ti 1.9 × 10 ± 25% 0.66
tion cross section and the very narrow absorption band,
the conclusion is that Fe3+ is a very weak chromophore h•-Fe3+ 1.3 × 10–17 ± 12.0% 0.28
in corundum. Figure 16 presents a visual method of com-
paring the strength of the six chromophores. h•-Cr3+ 1.3 × 10–17 ± 12.5% 0.44

8 CAUSES OF COLOR IN CORUNDUM GEMS & GEMOLOGY SPRING 2020

we have grown. This eliminates any oxygen vacan- surfaces of the wafer by summing all the reflections
cies, assuring a valence of +3. Additional possibilities (see box A). Wafers from four different boules were
for alternate valences of other ions, or interferences fabricated and measured, and the one with the lowest
from different pair absorption spectra, have been ad- absorption at approximately 226 nm was chosen as
dressed using the variety of synthetic crystals indi- the most pure. 226 nm is between the two highest-
cated in Appendix 1 at https://www.gia.edu/doc/ energy absorption bands of Cr3+ yet below the first
SP20-corundum-chromophores-appendix1. Cr3+ charge transfer band. However, at this wave-
Such considerations, plus the processing of syn- length the charge transfer bands of Fe, Ti, and other
thetic crystals in a high-temperature oxygen atmos- metals are very strong (Tippins, 1970), and thus min-
phere and thorough analyses for other trace imal absorption at this wavelength is a good relative
elements, assure that the chromophore concentra- indicator of crystal purity.
tions equal the SIMS-determined concentrations. Quantitative data on this sample with the mini-
In the following sections, we will first describe mum 226 nm absorption (and all of the other samples
the Cr3+ chromophore and how its cross section was characterized in this study) were obtained using a
determined. Much of this same detail will apply for Cameca IMS 7f-GEO SIMS instrument. SIMS is a
the other five chromophores and thus will not be re- highly precise analytical technique that can detect
peated. We present the origin of the samples (again, all elemental masses to the sub-ppma level. To
see figure 2), their crystallographic orientation, the achieve accuracy, however, relative sensitivity fac-
absorption cross sections determined for both E⊥c tors (RSF) need to be developed to correctly convert
and E||c, color circle arrays for different areal densi- the secondary ion signal for each trace element in the
ties of the chromophore, viewing directions, and il- matrix of interest into concentration. This is best
lumination sources. Detailed information on these done by creating ion implant standards. For each
samples can be found in Appendix 1. trace element we wish to quantify, a chosen isotope
of this trace element is implanted in a high-purity
THE Cr3+ CHROMOPHORE wafer of the intended matrix. Our developed RSF val-
Cr3+ is the trace element primarily responsible for col- ues were based upon multiple measurements vali-
oring ruby red and sapphire pink (McClure, 1962) and dating the “dose” of each of these ion implants, and
in doing so creates some of the most beautiful gems upon multiple SIMS depth profiles of these same im-
in the world. This coloration results from Cr3+ ions re- plants (Stone-Sundberg et al., 2017). The calculated
placing some of the Al3+ ions in the corundum lattice. RSF values account for all of the identified and quan-
High-purity synthetic sapphire doped with Cr3+ tifiable sources of uncertainty; for the final SIMS-
has been available since the invention of the laser in generated concentration values for all of the
1960. The very first laser employed synthetic ruby as subsequently measured samples, we report a total
the active element (Maiman et al., 1961). Interest in uncertainty that is equal to the square root of the
ruby lasers for many applications drove the develop- sum of the squares of the error associated with each
ment of large, exceedingly high-optical-quality and source. In total, there are four separately quantified
high-purity synthetic ruby crystals. Our samples sources of error: the ion implant standard dose meas-
were from excess material salvaged from several dif- urement, determination of the peak depths for the
ferent laser-quality single-crystal ingots (boules) ion implants, differences between the multiple depth
grown by St. Gobain Crystals and Detectors in profiles for each ion implant, and the set of measure-
Washougal, Washington. ments of the trace element of interest in the sample.
Wafer samples cut from these boules were crys- For the Cr3+ chromophore, the concentration and
tallographically oriented to an accuracy of approxi- total combined error is 134 ppma ± 7.6%. The deter-
mately 1.5° using an optical instrument (Thomas et mined detection limit for Cr3+ in corundum to the
al., 2014) with the c-axis in the plane of the wafer. 95% confidence level using our conditions was 0.002
The absorption spectra were measured with a reso- ppma (2 parts per billion atomic, ppba).
lution of 1.5 nm using a Hitachi Model 2910 spec- The absorption cross sections thus determined
trophotometer modified by Tim Thomas at GIA to over the 200–1100 nm range are shown in figure 3.
enable rotation of the polarization plane of the probe The data range includes the visible region of the elec-
beam, thus allowing the separate recording of both tromagnetic spectrum, which extends from 400 nm
E⊥c and E||c spectra. The spectra were corrected for (violet) to 700 nm (red). The vertical axis is labeled
the multiple reflections between the two polished “absorption cross section × 1019 cm2.” This means that

CAUSES OF COLOR IN CORUNDUM GEMS & GEMOLOGY SPRING 2020 9

 Figure 3. Cross section
4.0
spectra of the Cr3+
E⊥c
E||c
chromophore measured
3.5 from a synthetic ruby
ABSORPTION CROSS SECTION × 1019 (cm2)

grown by the Czochral-

3.0
ski method. The color
band shown along the
top of these spectra and
2.5 those in following fig-
ures is the visible spec-
2.0 trum that would be
observed through an
optical spectroscope,
1.5
shown for reference.
Also shown is a 1.30 ct
1.0 unheated Mozambique
pink sapphire, represen-
tative of the color that
0.5
would be produced by
the Cr 3+ chromophore
0 in isolation. Photo by
200 300 400 500 600 700 800 900 1000 1100 Lotus Gemology; cour-
WAVELENGTH (nm) tesy of Nomad’s Ltd.

the actual values plotted are from 0 to 4 × 10–19 cm2; tive of the color that would be produced by that chro-
we will use the same labeling convention for subse- mophore in isolation.
quent cross section spectra presented in this article. Note that the E⊥c peak cross section of 1.62 × 10–19
Figure 3 shows the absolute absorption cross section cm2 ± 7.6% we determined for the 560 nm U band
spectra for both the E⊥c (ordinary) and E||c (extraordi- represents a moderately strong absorber, and thus Cr3+
nary) rays for Cr3+ in the high-purity synthetic ruby. is a moderately strong chromophore. It is the magni-
These two spectra fully characterize, at 1.5 nm reso- tude of this cross section in the visible region of the
lution, the absorption characteristics of Cr3+ in corun- spectrum, and to a lesser extent the width and posi-
dum over the range from 200 to 1100 nm. Whether in tion of the bands, that determines the “strength” of
a high-purity synthetic crystal containing only Cr3+, a the chromophore (see box A). The position and width
natural ruby containing Cr3+ and iron, or a purple sap- of the absorption bands determine the color. Approx-
phire containing Fe2+-Ti4+ pairs and Cr3+, the absorption imately speaking, the larger the cross section and the
cross section of Cr3+ is the same. The broad bands near wider the bands in the visible region, the more color
560 nm are termed the U bands, while the two near produced.
400 nm are designated the Y bands. At a wavelength It is interesting to note that apparently the first
of 694 nm, there is a weak, narrow absorption feature. determination of Cr3+ cross section was published in
This is known as the R-line. In actuality, the R-line is 1961 in the seminal paper reporting the achievement
two weak lines, R1 and R2, which are separated by 1.4 of the first laser—the ruby laser (Maiman et al.,
nm. Since our instrumental resolution is 1.5 nm, these 1961). His value of the peak cross section for the E⊥c
two lines appear as a single weak feature. The weak U band was approximately 1.8 × 10–19 cm2. The small
lines near 660 nm most clearly observable in the E⊥c difference between his value and ours probably re-
spectrum are called the S-lines, while the weak lines flects the improved analytical instrumentation for
in both spectra near 470 nm are referred to as the B- determining Cr3+ concentration available to us today.
lines (Powell, 1998). Shown along with the spectra is Subsequently, the absorption and emission spectrum
an image of a faceted sapphire we selected as represen- of ruby has been extensively studied by many re-
tative of a stone that is very close to being colored by searchers; see, for example, Powell (1966, 1998) and
only the Cr3+ chromophore. In the following sections, Henderson and Bartram (2000).
we show a representative faceted stone alongside the While the absorption cross section provides the de-
spectra for each chromophore, selected as representa- finitive characterization of a chromophore, it is not

10 CAUSES OF COLOR IN CORUNDUM GEMS & GEMOLOGY SPRING 2020

easy to guess the color from simply examining the sources for the reader. Having determined the color
spectra. It is, however, straightforward to calculate the coordinates for a particular sample thickness and
color from these spectra. The color coordinates (Berns, chromophore concentration, a color circle is then cre-
2000) are usually calculated from the sample’s trans- ated by specifying the color coordinates to the drawing
mission spectrum and the characteristics of the light program. We have used Photoshop, EasyDraw, and
source. Calculation of the color coordinates for a given other graphics programs to produce the color circles.
color temperature was performed via Thermo Scien- Visualizing the color from the absorption spectrum of
tific’s GRAMS/AI spectroscopy software, a general- the sample has broad application in gemology. We first
purpose spectra manipulation code that includes color used this technique to determine the amount of dif-
coordinate calculations. There are many color coordi- fused beryllium that would cause a visible reduction
nate systems, and we chose the CIE 1976 L*a*b* sys- in the depth of color of a dark blue sapphire.
tem for its approximately uniform color space. We Figure 4 (left) shows an array of color circles cor-
used color temperatures of 6500 K (D65 illuminant) responding to a range of Cr3+ areal densities. Pre-
and 2856 K (A illuminant) in order to maximize the sented are the E⊥c color, the E||c color, and a color
color temperature difference and the color difference, representing equal proportions of E⊥c and E||c (we
thereby maximizing the visual effects of different light will call this E⊥c + E||c), as viewed with illuminant

Figure 4. Left: Areal density color circles of the Cr 3+ chromophore under illuminant D65. Right: Illuminant com-
parison color circles of the Cr3+ chromophore.

Cr3+ Chromophore

CIE-D65 Illuminant
Illuminant Comparison
(daylight; approx. 6500 K)
E⊥c

CIE-D65
E⊥c (~6500 K)

CIE-A
E||c (~2856 K)

E⊥c + E||c
E||c

100 200 375 750 1500 3000

CIE-D65
Cr3+ areal density in ppma-cm (~6500 K)

CIE-A
(~2856 K)

E⊥c + E||c

CIE-D65
(~6500 K)

CIE-A
(~2856 K)

100 200 375 750 1500 3000

Cr3+ areal density ppma-cm

CAUSES OF COLOR IN CORUNDUM GEMS & GEMOLOGY SPRING 2020 11

 BOX B: CRYSTAL OPTICS OF UNIAXIAL MINERALS
Glass is an optically isotropic material, meaning its
physical properties such as absorption coefficient spec- Electric Field Component
trum and refractive index are the same in any meas-
ured direction within the glass, regardless of the
polarization of the measurement beam. Consequently,
light traveling in any direction in an isotropic medium
behaves the same way. Many crystals are not optically
isotropic—they are anisotropic, which means their op- Magnetic Field
Component
tical properties vary with direction in the crystal.
Corundum crystals are anisotropic, and they are
termed optically uniaxial: The optical properties meas-
ured in two of the three spatial directions (x and y) are Pro
pa
the same, but properties in the z direction are different. ga
tio
We refer to the unique z direction of a uniaxial crystal n
as the c-axis, also called the optic axis. The other two
directions, x and y, are identical and are termed the a- Figure B-2. Light is a combination of electric and mag-
axes (see figure B-1). netic components that oscillate perpendicular to each
Light is a combination of electric and magnetic fields other and to the direction of propagation.
that oscillate perpendicular to the direction of the light
propagation as shown in figure B-2.
Linear polarized light is light whose electric field, E, a light beam is termed unpolarized and is necessarily
oscillates in a single direction perpendicular to the direc- perpendicular to the propagation direction.
tion of propagation, as shown in figure B-3. By conven- For optical measurements of uniaxial crystals, there
tion the direction of E is the polarization direction. are two unique geometries for a light beam passing
A light beam from a thermal source (such as the sun through a crystal that yield two different absorption
or a heated filament light bulb) comprises light rays spectra. This property is termed dichroism. It is this
whose electrical fields oscillate in all directions perpen- dichroism that accounts for the different colors we see
dicular to the direction of propagation (figure B-4). Such looking through crystals in different directions. The first
of these geometries is termed the ordinary ray; its polar-
ization direction, E, is perpendicular to the c-axis. This
Figure B-1. Corundum is a uniaxial mineral, meaning it is often designated by E⊥c. Thus a beam of light traveling
has one unique crystallographic axis out of three. This along the c-axis, polarized in any direction or unpolar-
drawing shows the three axes of a uniaxial crystal. The ized, is E⊥c. A linearly polarized light beam traveling at
two a-axes are the same while the c-axis is different.
any direction through the crystal is also an ordinary ray
Properties of uniaxial crystals are different when meas-
as long as the polarization direction E is perpendicular
ured along the c-axis rather than perpendicular to it.
to the c-axis (E⊥c). It is called ordinary because this holds
Optical absorption, and thus color, are different when
viewed with light propagating along the c-axis rather
true for any propagation direction through the crystal.
than along either a-axis. The second unique geometry is where the polariza-
tion direction E of the light beam traveling through the
c

Figure B-3. A light ray’s electric field direction, which is

also called the polarization direction, is perpendicular to
its propagation direction. In this diagram, the light ray is
propagating toward the right side of the page and its elec-
tric field is oscillating up and down in the plane of the
a1
page.
a2

Electric Field Direction, Propagation

E = Polarization Direction

12 CAUSES OF COLOR IN CORUNDUM GEMS & GEMOLOGY SPRING 2020

 Figure B-4. Unpolarized
light is composed of rays
E whose electrical fields os-
cillate in all directions per-
pendicular to the
propagation direction (left
and center). Polarized light
has an electric field E that
is constrained to oscillate
in a single direction (right).

crystal is parallel to the c-axis. Visualizing this geometry, spectrum (figure B-5, top row). This is because all polar-
it is clearly a much greater restriction in that the light ization directions of that light beam are E⊥c.
beam must be both linearly polarized and propagating When we view a corundum crystal along any a-axis,
perpendicular to the c-axis. Thus it is termed the ex- the situation is more complex. The light incident on the
traordinary ray and designated E||c. crystal is unpolarized. When entering the crystal, it is
For any other combinations of propagation and po- split into two colinear beams—one E⊥c and one E||c.
larization directions, what is measured is some combi- Each of these sees the corresponding absorption spec-
nation of these unique spectra or indices of refraction. trum, and thus when the beam emerges we see a color
When we view a corundum crystal down the c-axis, that is the average of the colors produced by the two ab-
the color we see is determined by the E⊥c absorption sorption spectra, E⊥c and E||c (figure B-5, bottom row).

Figure B-5. Shown at left are the ruby crystal from figure B-1 and two crystallographically oriented ruby wafers with
polished parallel surfaces corresponding to the shaded crystal faces. One wafer has the c-axis perpendicular to its sur-
faces (top), and one contains the c-axis in the plane of the wafer (bottom). A beam of unpolarized light contains rays
that are polarized in all directions perpendicular to the direction of propagation. Unpolarized incident light striking the
left side of the top wafer passes through unaltered except for absorption by the E⊥c absorption spectrum shown at the
right. Thus only E⊥c color is observed, and the dichroscope also shows only E⊥c color (both top and bottom dichro-
scope views regardless of dichroscope rotation). Unpolarized light striking the bottom wafer is split into two beams po-
larized at 90º to one another. One beam is polarized parallel to the c-axis (E||c) and one is polarized perpendicular to
the c-axis (E⊥c). The different absorptions of these two orthogonal beams are shown to the right. The observed color is
equal proportions of E⊥c and E||c. When the wafer is viewed with a dichroscope that is rotated to maximum color dif-
ference, the E⊥c and E||c colors are resolved individually (bottom dichroscope view).

Ruby Incident Ruby Transmitted Observer Observed Dichroscope Absorption

crystal light wafer light color view sprectrum
4.0
E⊥c
SECTION × 1019 (cm2)
ABSORPTION CROSS

3.5
3.0
2.5
c c-axis
E⊥c 2.0
hν hν
1.5
1.0
E⊥c 0.5
E⊥c 0
200 300 400 500 600 700 800 900 1000 1100
WAVELENGTH (nm)
a1
a2 4.0
SECTION × 1019 (cm2)
ABSORPTION CROSS

hν hν 3.5 E⊥c
3.0 E||c
E⊥c 2.5
is
ax

2.0
c-

E||c 1.5
E⊥c + E||c 1.0
0.5
0
200 300 400 500 600 700 800 900 1000 1100
WAVELENGTH (nm)

CAUSES OF COLOR IN CORUNDUM GEMS & GEMOLOGY SPRING 2020 13

 A calcite dichroscope is simply an instrument that We have chosen to use these two designations for sim-
splits a beam into two rays whose polarization direc- plicity, because they say exactly what they are and we
tions are at 90° to each other. If we observe through a don’t have to remember what they mean. The same des-
dichroscope a beam of light that has propagated along ignation issues arise with the two indices of refraction.
an a-axis of the crystal and rotate the dichroscope so In table B-1 we have listed the more commonly used des-
that the color difference between the two sides is at a ignations for both rays and indices of refraction.
maximum, we have found the orientation where its
two polarization directions match the two directions of
the crystal. We are observing both a pure E⊥c color and TABLE B-1. Alternative designations for rays and indices of
pure E||c color (figure B-5 bottom, dichroscope view). If refraction.
we view the crystal in this orientation without the
Rays
dichroscope, we see a single color that is equal propor-
tions of E⊥c and E||c color, which we call E⊥c + E||c (fig- E⊥c E||c
ure B-2 bottom, observed color). If instead we view the Ordinary ray Extraordinary ray
crystal with the dichroscope along the c-axis, we ob- O-ray E-ray
serve the same color on each side, and the color stays σ or sigma Π or pi
the same as we rotate the dichroscope. In this case we w ε
are observing pure E⊥c color (figure B-2 top, dichroscope Indices of Refraction
view). In this paper we present absorption cross section n⊥ n||
spectra for E⊥c and E||c, and color circles for E⊥c, E||c, no ne
and E⊥c + E||c. nσ or ns np or np
Crystal optics is a venerable field of study, and there nw nε
are many different designations in use for E⊥c and E||c.

D65 (daylight equivalent). The E⊥c colors are seen viewed as presenting a stone of fixed concentration
looking through a stone along the c-axis. The E⊥c + of increasing thickness, or of fixed thickness with in-
E||c colors are what is seen looking through a stone creasing Cr3+ concentration, or any combination of
perpendicular to the c-axis. With a properly aligned the two.
dichroscope viewing in this same direction, we see The color we perceive also changes with the type
the pure E⊥c and E||c colors separated (see box B for of light source. Figure 4 (right) shows the three sets
a basic introduction to uniaxial crystal optics). of color circles—E⊥c, E||c, and E⊥c + E||c—but also
L*a*b* color coordinates for these color circles and for each of two illuminants—D65 (daylight) and A
all subsequent color circles presented in this article (tungsten bulb). We have chosen these two light
are available in Appendix 2 at https://www.gia.edu/ sources to maximize the observable color difference.
doc/SP20-corundum-chromophores-appendix2. Thus we can see a substantial change in the per-
Along the horizontal axis of the color circle array ceived color with a change in light source. While this
is the areal density of the Cr3+ in units of ppma-cm. color change is significant, ruby is not described as a
Areal density is the product of the chromophore con- color-change gem. A color-change gem is usually
centration in ions per cm3, times the path length thought of as one where the hue changes signifi-
through the stone in cm. The concentration of chro- cantly (more than one adjacent hue position), such
as blue to red or green to red.
mophore ions is 1.178 × 1017 ions/cm3 times the con-
centration in ppma (see box A). For a 1 cm thick
sample, the numbers under the color circles are the THE V3+ CHROMOPHORE
Cr3+ concentration in ppma. For a 2 cm thick sample Vanadium (V3+) commonly occurs in trace element
showing the same color, the concentration is one half analyses of natural corundum, particularly ruby.
the numerical ppma-cm value. For a 1/2 cm thick Generally, the vanadium concentration is not high
stone, the concentration is twice the ppma-cm value. enough (<20 ppma) to have a significant impact on
Thus the numbers under the color circles can be the color. However, recently published analyses of

14 CAUSES OF COLOR IN CORUNDUM GEMS & GEMOLOGY SPRING 2020

corundum from the Mogok Stone Tract in Myanmar lence state, as previously discussed. Previous multi-
(Harlow and Bender, 2013; Zaw et al., 2015) have ple heat treatment experiments on this material had
shown concentrations as high as 2000 ppma and re- shown these conditions were effective.
ported sapphire colors of “slate” (grayish blue to gray- Figure 5 shows the absorption cross section for
blue) to purple as well as color-change in some cases. both the E⊥c and E||c for V3+ in sapphire. It is inter-
For concentrations at or considerably less than this esting to note that the V3+ spectra in the visible range
extreme, vanadium will indeed impact the color of show two absorption bands with similar positions
ruby and sapphire. V3+ spectra in corundum have and widths to those of Cr3+. However, their relative
been studied for decades (Pryce and Runciman, 1958; magnitudes are very different, leading to very differ-
McClure, 1962; MacFarlane, 1964). ent colors. Vanadium is a weak chromophore in
Several decades ago, the former Union Carbide corundum since its absorption band, centered at 580
Corporation crystal growth facility in Washougal, nm, is much weaker than that of Cr3+ at 560 nm. We
Washington, grew a large vanadium-doped sapphire determined that vanadium’s E⊥c peak cross section
crystal by the Czochralski technique for research for the 580 nm band is 1.0 × 10–19 cm2 ± 8.9%.
purposes. A slice of this unique crystal was provided Figure 6 (left) shows the three sets of color circles
for our research. SIMS analysis showed a vanadium for E⊥c, E||c, and E⊥c + E||c under D65 illumination
concentration of 116 ppma with a total combined un- for V3+. There is a significant hue difference between
certainty of ± 8.9%. The detection limit determined the E⊥c color, which is a grayish blue, and the E||c
for V3+ in corundum at the 95% confidence level was color, which is distinctly green at the lower concen-
0.016 ppma. Wafers were fabricated from this boule trations. Nearly the same amount of color differ-
with the c-axis in the plane of the wafer so that both ence also exists between the E⊥c color and the E⊥c
E⊥c and E||c absorption spectra could be recorded. + E||c color. The color of a faceted stone can there-
Since vanadium can exist in multiple valences, the fore be quite different depending on the orientation
wafers were heated at 1750°C in oxygen for 10 hours of the cut. While the hue differences between the E⊥c
to assure that all of the vanadium was in the 3+ va- color and E||c color for Cr3+ illustrated in the previous

6
E⊥c
E||c
Figure 5. Cross section
5 spectra of the V3+ chro-
ABSORPTION CROSS SECTION × 1019 (cm2)

mophore, measured
from a synthetic color-
change sapphire grown
4
by the Czochralski
method. Also shown is
a 1.17 ct color-change
3
sapphire in daylight
(left) and incandescent
light (right), representa-
tive of the color that
2 would be produced by
the V3+ chromophore in
isolation. Photo by
Wimon Manorotkul;
1
courtesy of Lotus
Gemology collection.

0
200 300 400 500 600 700 800 900 1000 1100

WAVELENGTH (nm)

CAUSES OF COLOR IN CORUNDUM GEMS & GEMOLOGY SPRING 2020 15

 V3+ Chromophore
CIE-D65 Illuminant
Illuminant Comparison
(daylight; approx. 6500 K)
E⊥c

E⊥c CIE-D65
(~6500 K)

CIE-A
E||c (~2856 K)

E⊥c + E||c
E||c

110 210 430 850 1700 3400 CIE-D65

(~6500 K)
3+
V areal density in ppma-cm

CIE-A
(~2856 K)

E⊥c + E||c

CIE-D65
(~6500 K)

CIE-A
(~2856 K)

110 210 430 850 1700 3400

3+
V areal density in ppma-cm

Figure 6. Left: Areal density color circles of the V3+ chromophore under illuminant D65. Note the significant color
difference between the E⊥c color, which is grayish blue, and the E||c color, which is quite green at lower concentra-
tions. As with Cr3+, the E⊥c and E||c color difference is significant. Right: Illuminant comparison color circles of
the V3+ chromophore for illuminants A and D65. For two of the orientations (E||c and E⊥c + E||c), the hue difference
under A and D65 and illuminants certainly appears substantial enough to qualify it as a color-change gem.

section are certainly noticeable, the hue difference hue. The A light source is richer in orange to red
for V3+ illustrated in this section is very substantial. wavelengths, so the stone transmits proportionally
Figure 6 (right) presents the E⊥c, E||c, and E⊥c + more of these wavelengths, giving the stone a pur-
E||c data for illuminants A and D65. While all three plish color. Given the hue differences, it would be in-
orientations show a noticeable hue difference for the teresting to cut some faceted stones from this
two illuminants, the E||c color and E⊥c + E||c color synthetic crystal, and to search for unusually colored
show a very strong hue difference, which would cer- natural high-V pieces from Myanmar.
tainly qualify as a color-change gem when properly
cut. The hue difference in these orientations is pro-
duced because the D65 light source is richer in blue Fe3+: MULTIPLE CHROMOPHORES
to green wavelengths than the A light source. Con- Iron is nearly ubiquitous in gem corundum, present
sequently, the stone transmits more green light over a very wide range of concentrations from less
under D65 illumination, giving the stone a greener than a few ppma to nearly 5000 ppma. While this

16 CAUSES OF COLOR IN CORUNDUM GEMS & GEMOLOGY SPRING 2020

makes it difficult to characterize, the fact that Fe3+ is that the fraction of pairs and larger clusters would
can exist as single ions replacing aluminum, Fe3+ ion certainly be dependent on the thermal history of the
pairs, and perhaps larger Fe3+ clusters pose yet greater crystal, which would not necessarily mimic nature.
problems. Because the Fe3+ spectrum is composed of For these reasons we too are constrained to using nat-
all of these species whose individual concentrations ural crystals for part of our characterization.
depend in a nonlinear way upon the total iron con- Figure 7 shows the absorption cross section spec-
centration in the sample, the spectrum’s dependence tra of a natural corundum sample with a very high
on concentration can be quite complex (Ferguson and iron concentration determined with SIMS of 4730
Fielding, 1971 and 1972; Krebs and Maisch, 1971). In ppma, with a total combined uncertainty of 8.0%,
addition to these problems, the Fe3+ chromophore is the highest concentration among our samples. The
the weakest chromophore in natural corundum. Its detection limit to the 95% confidence level for Fe
peak absorption cross section at 450 nm in the visible was 0.003 ppma. The highest concentration was cho-
region does not exceed 2.5 × 10–20 cm2 ± 8.0%, and it sen as the cross sections will most easily be meas-
is quite narrow. Thus a variety of samples with a ured. At such a high iron concentration, one can
wide range of total Fe3+ concentrations is required to easily observe and compare the weak broad features
characterize the spectra. of both the E⊥c and E||c absorptions. This sample is
Directly growing synthetic sapphire at its melt from the Subera deposit in Queensland, Australia.
temperature (~2050°C) doped with iron is limited by Yellow sapphires of similar iron concentrations are
the fact that iron and its oxides all have very high found in the Garba Tula deposit in Kenya, the Chan-
vapor pressures at this temperature and thus vaporize thaburi area of Thailand, and other alkali basalt-
from the melt. This imposes a practical limit of a few hosted deposits, but they are not common. Usually
hundred ppma of iron for crystals grown this way. Flux such sapphires also include a few ppma of Fe2+-Ti4+
growth at temperatures of 900–1300°C would seem to pairs, which shifts the yellow to a greenish yellow,
be an attractive alternative, and it has been tried (Fer- or they contain a few tenths of a ppma of h•-Fe3+ pairs,
guson and Fielding, 1972), yet the spectroscopy applied which adds some golden yellow coloration.
still employed only natural crystals. Another problem The spectra of Fe3+ containing corundum is quite

6
E⊥c
E||c Figure 7. Cross section
spectra of Fe3+ and Fe3+-
5
Fe3+ pairs measured in a
ABSORPTION CROSS SECTION × 1020 (cm2)

natural yellow sapphire

from the Subera deposit
4 in Queensland, Aus-
tralia, with a remark-
ably high Fe3+
concentration of 4730
3
ppma. Also shown is a
2.33 ct unheated Aus-
tralian yellow sapphire,
2
representative of the
color that would be
produced by this chro-
mophore in isolation.
1 Photo by Lotus Gemol-
ogy; courtesy of Terry
Coldham/Intogems PL.
0
200 300 400 500 600 700 800 900 1000 1100

WAVELENGTH (nm)

CAUSES OF COLOR IN CORUNDUM GEMS & GEMOLOGY SPRING 2020 17

complex. In figure 7, the narrow peaks at 377 and 450 with C1.5, where C is the total Fe3+ concentration.
nm are attributed to Fe3+-Fe3+ pairs, as is the weak but This implies that the absorption cross section of the
broad band at 540 nm. The narrow peak at 388 nm is 450 nm band increases approximately with the
attributed to single Fe3+ ions. The weak but broad square root of C. If there was not a concentration de-
bands at 700, 1050, and 1085 nm are only really quan- pendence of this cross section, there would be a sin-
tifiable at high Fe3+ concentrations. The lack of tem- gle curve for E⊥c in figure 8, just as we see for Cr3+
perature dependence in the absorption coefficient of and V3+. This dependence of the cross section on con-
these bands implies that they arise from single Fe3+ centration is currently under study by the authors
ions (Ferguson and Fielding, 1972). There is also a with over 100 natural samples analyzed by laser ab-
broad band at 330 nm that is identified as an Fe3+-Fe3+ lation–inductively coupled plasma–mass spectrom-
pair band, but it is not visible in this spectrum. It can etry (LA-ICP-MS).
be seen in two of the spectra presented in figure 8. Because the absorption cross section is not inde-
The relatively narrow spectral features at 377, 388, pendent of the iron concentration, our usual presen-
and 450 nm are observable in the spectra of most tation of color circles as a function of the areal
corundum samples with a Fe concentration of 150 density, ppma-cm, is not a valid characterization.
ppma or higher. This is because the absorbance, ab- Thus we present some E⊥c color circles at a few fixed
sorption coefficient, and absorption cross section of concentrations and as a function of the path length
different species are additive, and thus a weak narrow through the sample. Figure 9 (left) compares the E⊥c
feature can be easily seen on top of a much stronger color circles for D65 illumination. Figure 9 (right) also
wide band. shows the same color circles but under illuminant A.
The ratio of the peaks of the 388 nm band to the Fe3+ exists in natural gem corundum over a much
377 nm band changes dramatically with concentra- wider concentration range than any other chro-
tion. This is because the 388 peak is primarily a sin- mophore. At a few thousand ppma, Fe does indeed
gle Fe3+ ion, while the 377 nm peak is a pair of Fe3+ cause strongly colored yellow sapphire. However, it
ions. Our earlier studies of about 40 different natural is such a weak chromophore that only at these higher
samples with widely varying Fe3+ concentrations in- concentrations does it contribute much color to a
dicate that the peak absorption coefficient at 450 nm, gem. Presenting a more detailed discussion of the
which is a pair of ions, increases very approximately color contributed by iron will require a much better

10

9 4730 ppma
ABSORPTION CROSS SECTION × 1020 (cm2)

2960 ppma
8 1320 ppma Figure 8. Unlike the
580 ppma
other corundum chro-
7 mophores, the cross
section of the iron chro-
6
mophore is dependent
5
on the iron concentra-
tion. Shown here are
4 the E⊥c cross sections
for four different Fe3+
3 concentrations: 4730
ppma, 2960 ppma, 1320
2 ppma, and 580 ppma.
1

0
300 400 500 600 700

WAVELENGTH (nm)

18 CAUSES OF COLOR IN CORUNDUM GEMS & GEMOLOGY SPRING 2020

 Fe3+ Chromophore Comparison
E⊥c

CIE-D65 Illuminant CIE-A Illuminant

Fe content (daylight; approx. 6500 K) (incandescent; approx. 2856 K) Fe content

4730 ppma 4730 ppma

2960 ppma 2960 ppma

1180 ppma 1180 ppma

447 ppma 447 ppma

0.25 0.5 1 2 4 0.25 0.5 1 2 4

Path Length (cm) Path Length (cm)

Figure 9. The color produced by various iron concentrations and path lengths through the stone. The iron chromo-
phore is very weak, so high concentrations are needed to produce significant color. There is also a quite noticeable
color difference between illuminants D65 and A.

physical model of iron in the corundum crystal, as the The most commonly proposed explanation for
pair or larger cluster concentrations will also be de- the absorption spectra of blue sapphires is that of in-
pendent on the thermal histories of the samples. We tervalence charge transfer (IVCT) between the two
encourage our colleagues to study this matter. cations as follows:
hν
Fe2+ + Ti4+ Fe3+ + Ti3+
THE Fe2+-Ti4+ CHROMOPHORE
Fe2+-Ti4+ pairs are the primary chromophore of blue where hn is Planck’s constant, h, times the optical fre-
sapphire (Townsend, 1968; Mattson and Rossman, quency, n. This a common way to indicate that a pho-
1988; Moon and Phillips, 1994). The defect chemistry ton has been absorbed to enable the chemical reaction.
that creates this pair is interesting. Titanium in It would seem straightforward to determine the ab-
corundum is a donor, and iron in corundum can be sorption cross sections of this transition by just ana-
an acceptor. They electrostatically attract each other lyzing a fairly uniform piece of natural blue sapphire
and locate on adjacent Al3+ sites in the crystal, creat- and measuring the Fe and Ti concentrations to deter-
ing a donor-acceptor pair. Neither Fe2+ nor Ti4+ alone mine the concentration of Fe2+-Ti4+ pairs. However, as
has absorption features in the visible region. Ti4+ is a discussed elsewhere (Emmett et al., 2017b), it is diffi-
closed-shell electron configuration, and Fe2+ is ex- cult to accurately determine the amount of titanium
pected to absorb only in the near infrared, but that ab- available, [Ti]available, to pair with Fe2+ creating the blue
sorption has not yet been observed. Yet when these color. Since the Si donor state lies above the Ti donor
two ions are located on neighboring aluminum sites, state in the corundum band gap, it will preferentially
strong absorption bands arise across the visible and pair with acceptors such as Mg and Ni. Thus the Ti
near infrared region of the spectrum. For E⊥c the band available to pair with Fe can be expressed as:
peaks at 580 nm, while for E||c the peak is at 700 nm. [Ti]available = [Ti] + [Si – Mg – Ni] for [Si – Mg – Ni] < 0
While the theory of the energy levels of a single tran-
[Ti]available = [Ti] for [Si – Mg – Ni] ≥ 0
sition metal ion in a crystal is quite well developed,
the theory for ion pairs or clusters in a crystal is not. Note that the square brackets around ions denote

CAUSES OF COLOR IN CORUNDUM GEMS & GEMOLOGY SPRING 2020 19

atomic concentrations. The standard deviation (SD) We have observed microscopic iron crystals in some
of [Ti]available is the square root of the sum of the synthetic Fe-Ti doped crystals. Presently we have no
squares of SDs of Ti, Mg, Ni, and Si. Thus the stan- reliable way to determine what portion of the iron in
dard deviation of [Ti]available can be very large, often ex- the crystal is actually in solution as Fe3+. The syn-
ceeding the value itself, making accurate estimates thetic crystals did, however, give us very high-quality
of [Ti]available difficult. E⊥c and E||c spectra without the interference from
It would also seem straightforward to just grow a high concentrations of iron and other trace elements
synthetic crystal doped only with Fe and Ti, measure (Si, Cr, V, and Ni).
the concentrations, and calculate the Fe2+-Ti4+ pair To achieve reasonably good approximation of the
concentration. Such crystals were grown by Milan cross sections of the Fe2+-Ti4+ pair, we have chosen a
Kokta at the Union Carbide Corporation crystal combination of both synthetic crystals and natural
growth factory in the 1980s for synthetic gem mate- crystals. After examining a wide variety of natural
rial. However, two problems arose with the use of crystals, we have chosen Yogo sapphires for their
synthetic crystals. First, at the Czochralski growth rather extreme optical and chemical uniformity,
temperature of sapphire (~2050°C), the Fe2+-Ti4+ pairs which far exceeds all other natural crystals we have
are largely dissociated. They only partially re-form studied, and for the fact that there is no 880 nm ab-
as the stone cools over a day or two from growth sorption band in this material. Since the synthetic
temperature. Thus there are fewer pairs than the con- crystals have no interference to the shape of the spec-
centrations of [Fe] and [Ti] would imply, and we could trum from other chromophores, they have been used
not determine what fraction of the possible pairs to accurately determine the shape of the Fe2+-Ti4+ pair
were indeed present. absorption spectra.
In nature the growth temperature is much lower, We have carefully measured the trace element
so there is less dissociation, and the cooldown is pre- composition of 14 Yogo sapphires by SIMS at Cal-
sumably extremely slow, so the pairs should more tech. The SIMS technique gives the smallest SDs of
fully form. If so, the concentration of pairs should be our current measurement techniques, and only SIMS
determined by the ion concentrations. can quantify Si. These analyses provided accurate de-
To re-form the pairs, we heated the synthetic terminations of Fe, Ti, Mg, Si, Ni, Cr, and V. The de-
crystals in air for a few thousand hours. The temper- tection limits for Cr, V, and Fe were previously given,
ature of 750°C was chosen to be low enough that and the determined detection limits for Ti, Mg, Si,
many of the pairs would re-form, yet high enough and Ni were 0.039, 0.001, 0.552, and 0.030 ppma, re-
that the diffusion coefficients of the ions were still spectively. The higher detection limit for Si is par-
large enough that clustering would take place at
tially due to its larger RSF value, but mainly due to
some finite rate. The absorption coefficients were re-
the difficulty of removing all traces of Si from the
measured at intervals of 50 and 100 hours to track
chamber and mass spectrometer. Because [Si – Mg –
the increase. This process was carried on until no fur-
Ni] is always negative for Yogo sapphires due to their
ther increase with time was noted, and thus equilib-
high Mg concentrations (~100 ppma), the [Ti]available as
rium was reached. For our samples with the highest
noted above was calculated as
Ti concentration, equilibrium was reached in 1500
hours, with the absorption coefficient at 580 nm [Ti]available = [Ti] + [Si] – [Mg] – [Ni]
being 2–2.5 times higher than before heat treatment.
It is likely that some additional pairs would form if The absorption spectra of the 14 Yogo samples were
the temperature were lowered further and the time measured and the contributions from Cr3+ and V3+
greatly increased. Just what percentage we have subtracted from each using the absorption cross sec-
achieved of the low-temperature equilibrium pair tions of Cr3+ and V3+ previously determined. The
concentration is difficult to estimate. shape of the remaining Fe2+-Ti4+ absorption band
Another problem with the growth of synthetic closely matched the shape determined from the syn-
crystals doped with Fe2+-Ti4+ is determining what thetic crystals. The resulting absorption cross sec-
fraction of iron in the crystal is actually in solution tions are shown in figure 10. The band shapes are
rather than present as microscopic iron particles. from the synthetic crystals, and the band peak values
Under the oxygen partial pressure of Czochralski are determined from the averages of the 14 Yogo crys-
growth from an iridium crucible, all the iron present tals studied. The peak E⊥c absorption cross section
may not be trivalent and thus soluble in the melt. at 580 nm thus determined is 1.94 × 10–18 cm2 ± 25%.

20 CAUSES OF COLOR IN CORUNDUM GEMS & GEMOLOGY SPRING 2020

 3.0
E⊥c
E||c

2.5 Figure 10. Cross section

ABSORPTION CROSS SECTION × 1018 (cm2)

spectra of the Fe2+-Ti4+

chromophore measured
from a synthetic blue
2.0
sapphire grown by the
Czochralski method.
Also shown is a fine
1.5 4.10 ct unheated Sri
Lankan blue sapphire,
representative of the
color that would be pro-
1.0
duced by the Fe2+-Ti4+
chromophore in isola-
tion. Photo by Lotus
0.5
Gemology; courtesy of
Ruby N’ Sapphire.

0
200 300 400 500 600 700 800 900 1000 1100

WAVELENGTH (nm)

These spectra have been terminated at 330 nm in ing that the concentration of Fe2+-Ti4+ pairs along the
the UV. This was necessary because the first O → Fe3+ optical path can vary substantially, something our sur-
charge transfer band (Tippins, 1970) peaks at 259 nm. face-level measurements would not be able to capture.
We have determined the peak cross section for this Additional research on this topic is definitely needed.
transition as 1.66 × 10–17 cm2, which is extremely large. It is interesting to note that in the synthetic and
The sloping band edge on the long-wavelength side of Yogo crystals, the peak of the E⊥c cross section at 580
this band mixes with the short-wavelength side of the nm of 1.94 × 10–18 cm2 ± 25% is about 12 times larger
Fe2+-Ti4+ pair and rapidly exceeds the dynamic range than that of Cr3+ at 560 nm. The large cross section
of the spectrophotometer. Thus we terminated the and the width of the band in the visible region make
Fe2+-Ti4+ spectra before the charge transfer absorption the Fe2+-Ti4+ pair a strong chromophore. Less than
became significant. one-tenth the amount of it relative to Cr3+ is needed
In the course of our study of the Fe2+-Ti4+ band in to produce strong coloration.
many natural sapphires from many locations, we have Figure 11 presents the color circles calculated
found some whose Fe2+-Ti4+ cross sections appear much from the cross section determinations. As in the dis-
larger (by factors of two or three) than those deter- cussion of the Cr3+ chromophore, we have used the
mined here using synthetic and Yogo sapphires. Nearly areal density in ppma-cm to present the data. Figure
all of these non-Yogo samples contain very high iron 11 (left) shows the three sets of color circles for E⊥c,
content and substantial OH, as revealed by FTIR meas- E||c, and E⊥c + E||c under illuminant D65. As we saw
urements. We do not know if this is a correct determi- with the Cr3+ chromophore, the E⊥c and E||c colors
nation and thus represents an enhanced Fe2+-Ti4+ cross are quite distinct. This is because the E⊥c absorption
section due to the presence of iron pairs or clusters spectrum is more effective at blocking the blue-green
with the Ti4+, or whether it results from the fact that and green portions of the spectrum than the E||c
our chemical analyses only measure a very thin layer spectrum, thus transmitting a purer blue color.
on the stone’s surface rather than the bulk composi- Figure 11 (right) compares the color observed
tion. Many of these samples demonstrating anomalous under illuminants A and D65. While the color differ-
Fe2+-Ti4+ cross sections are strongly color zoned, mean- ences shown between illuminants D65 (daylight) and

CAUSES OF COLOR IN CORUNDUM GEMS & GEMOLOGY SPRING 2020 21

 Fe2+-Ti4+ Chromophore

CIE-D65 Illuminant
Illuminant Comparison
(daylight; approx. 6500 K)
E⊥c

CIE-D65
E⊥c (~6500 K)

CIE-A
E||c (~2856 K)

E⊥c + E||c E||c

1.5 3 6 9 12 15 CIE-D65
(~6500 K)
Fe2+-Ti4+ areal density in ppma-cm

CIE-A
(~2856 K)

E⊥c + E||c

CIE-D65
(~6500 K)

CIE-A
(~2856 K)

1.5 3 6 9 12 15
2+ 4+
Fe -Ti areal density in ppma-cm

Figure 11. Left: Areal density color circles under illuminant D65 of the Fe2+-Ti4+ chromophore. As with the Cr3+ chro-
mophore, the E⊥c and E||c colors are quite distinct. This is because the E⊥c absorption spectrum is more effective at
blocking the blue-green and green portions of the spectrum than the E||c spectrum, thus transmitting a purer blue
color. Right: Illuminant comparison color circles of the Fe2+-Ti4+ chromophore. While the color differences shown be-
tween illuminants D65 (daylight) and A (tungsten lamp) are not dramatic, the color circles for the D65 illuminant
exhibit a lighter, brighter blue resulting from a greater portion of blue light in the D65 spectrum.

A (tungsten lamp) are not dramatic, the color circles treated in air or oxygen, the excess acceptors will be
for the D65 illuminant exhibit a brighter blue result- charge compensated by trapped holes.
ing from the greater portion of blue light in the D65 h• = [acceptors] – [donors] = [Mg + Ni] – [Si + Ti + H]
spectrum.
If the sample also contains iron but not chromium,
the h•-Fe3+ pair can form, producing a strong yellow
CHROMOPHORES INVOLVING coloration. If instead the sample contains only
TRAPPED HOLES chromium, the h•-Cr3+ pair can form, producing a
When the concentration of magnesium and nickel strong orange coloration. If the sample contains both
exceeds the sum of the concentrations of silicon, ti- Fe3+ and Cr3+, the hole will preferentially pair with the
tanium, and hydrogen, the corundum sample is ac- Cr3+. The association of the hole with Fe3+ or Cr3+ is
ceptor dominated. If this occurs in a relatively unusual, as both are isoelectronic with Al3+ in the
oxidizing natural environment, or if the stone is heat corundum lattice. However, isoelectronic dopants

22 CAUSES OF COLOR IN CORUNDUM GEMS & GEMOLOGY SPRING 2020

 3.0
E⊥c
E||c

2.5 Figure 12. Cross section

ABSORPTION CROSS SECTION × 1017 (cm2)

spectra of the h•-Fe3+

chromophore measured
from a synthetic yellow
2.0
sapphire grown by the
Czochralski method.
Also shown is an 11.41
1.5 ct unheated Sri Lankan
yellow sapphire, repre-
sentative of the color
that would be pro-
1.0
duced by the h•-Fe3+
chromophore in isola-
tion. Photo by Lotus
0.5 Gemology; courtesy of
F.M. Gems.

0
200 300 400 500 600 700 800 900 1000 1100

WAVELENGTH (nm)

that trap electrons, holes, or excitons are well known rity was high enough that we did not need to account
in wide gap semiconductors (Pajot and Clerjaud, 2013). for Ti and Si that would scavenge available Mg. Fol-
Isoelectronic traps in wide-band-gap oxides have re- lowing growth, the crystals were annealed in oxygen
ceived some study as exciton traps that can greatly in- at 1750°C for 10 hours to assure that all magnesium
crease the fluorescence from UV excitation (Shtepliuk was charge compensated by holes. Previous experi-
et al., 2011; Zorenko et al., 2011). ments under a variety of conditions indicated that
To study the chromophores composed of trapped this annealing condition maximizes the hole concen-
holes paired with either Fe3+ or Cr3+, two crystals tration. The resulting crystals were a deeper yellow,
were grown by the Czochralski technique at St. Gob- almost golden color (iron and magnesium), and a
ain Crystals and Detectors; one was doped only with deeper orange color (chromium and magnesium).
iron and magnesium, while the other was doped only The cross sections presented for these two chro-
with chromium and magnesium. As grown, the crys- mophores are determined from the absorption spec-
tal containing iron and magnesium had a bright tra recorded on these double-doped samples. Since
orangy yellow color, and the crystal containing the h•-Fe3+ sample also contains the Fe3+ chro-
chromium and magnesium had a bright orange color. mophore by itself, that spectrum was subtracted be-
The concentrations with total combined uncertainty fore calculating the h•-Fe3+ cross section. Since the
for the h•-Fe3+ crystal were Fe = 180 ± 7.7% and Mg = Fe3+ chromophore is so weak, it only modifies the
2.13 ± 9.2%. If quantifiable concentrations of Si and color produced by the h•-Fe3+ chromophore at high
Ti were present in this crystal, the total available Mg Fe3+ concentrations. Similarly, since the h•-Cr3+ sam-
and its associated total combined error would have ple also contains the Cr3+ chromophore by itself, that
to be adjusted to account for these ions. However, spectrum was subtracted before calculating the h•-
both Si and Ti were below the quantification limits, Cr3+ cross section. However, Cr3+ is a moderately
eliminating the need to account for these trace ele- strong chromophore, and thus any gemstone con-
ments. For the h•-Cr3+ crystal, the Cr concentration taining the h•-Cr3+ chromophore will also exhibit a
with total combined uncertainty was 12.6 ± 7.7%, pink to red component from Cr3+ alone. It is these
and the Mg concentration with total combined un- two chromophores together that create the magnifi-
certainty was 0.813 ± 9.8% Similarly, the crystal pu- cent pinkish orange padparadscha sapphire.

CAUSES OF COLOR IN CORUNDUM GEMS & GEMOLOGY SPRING 2020 23

 h•-Fe3+ Chromophore

CIE-D65 Illuminant
Illuminant Comparison
(daylight; approx. 6500 K)
E⊥c

CIE-D65
E⊥c (~6500 K)

CIE-A
E||c (~2856 K)

E⊥c + E||c
E||c
0.4 0.9 1.7 3 7 14
CIE-D65
h•-Fe3+ areal density in ppma-cm (~6500 K)

CIE-A
(~2856 K)

E⊥c + E||c

CIE-D65
(~6500 K)

CIE-A
(~2856 K)

0.4 0.9 1.7 3 7 14

h•-Fe3+ areal density in ppma-cm

Figure 13. Left: Areal density color circles for the h•-Fe3+ chromophore. Right: Illuminant comparison of the h•-Fe3+
chromophore. Little color change is observed with the change in illuminant for all orientations.

The spectrum in figure 12 shows the absorption nated at 330 nm to avoid interference by the O →
cross sections for E⊥c and E||c of a sample cut from Fe3+ charge transfer band. This is not necessary for
the crystal containing iron and magnesium. Note the the h•-Cr3+ spectra, as the first O → Cr3+ band is well
extremely large cross section of the h•-Fe3+ chro- below 200 nm (Tippins, 1970).
mophore, 1.3 × 10–17 cm2 ± 12.0%, which is about 6.8 Figure 13 presents the color circles for the h•-Fe3+
times that measured for the Fe2+-Ti4+ pair or about 80 pairs. The E⊥c and E||c spectra are quite similar.
times larger than that of Cr3+. It is thus a very strong Thus the E⊥c, E||c, and E⊥c + E||c color circles under
chromophore in corundum. The h•-Fe3+ and the h•- D65 illumination shown in figure 13 (left) are also
Cr3+ are the strongest chromophores in the visible re- similar, unlike some of the other chromophores we
gion of the spectrum for corundum. As discussed in have studied. What is interesting to note here is the
the section on the Fe2+-Ti4+ pair chromophore, the extremely small amount of h•-Fe3+ pairs required to
spectrum of the h•-Fe3+ chromophore has been termi- create a strong color. Only a few ppma are necessary

24 CAUSES OF COLOR IN CORUNDUM GEMS & GEMOLOGY SPRING 2020

in a 1 cm thick sample to produce strong coloration, that strong absorption is present below 560 nm for
whereas nearly 3000 ppma of Fe3+ alone would be re- both E⊥c and E||c. Figure 15 (right) shows the color
quired to produce similar coloration (see figure 9). Ex- circles for both D65 and A illuminants. There is little
amining the color circles in figure 13 (right) for the chromatic difference resulting from the two illumi-
two different illuminants, D65 and A, we see that the nants. The color circles for illuminant A, which con-
color differences for the two illuminants are greater tains proportionally more yellow and red light, may
than the E⊥c and E||c color differences. be slightly brighter than their D65 counterparts.
The h•-Fe3+ chromophore is the primary cause of The cross sections and color circles presented for
color for yellow sapphires from Sri Lanka and other the h•-Cr3+ chromophore are for the chromophore it-
sources where the iron content is not extremely high. self and do not include a contribution from the Cr3+.
It also occasionally occurs in some high-iron sap- Since Cr3+ is a moderately strong chromophore, very
phires from Thailand, where stones with the combi- little is required to shift these hues toward a more
nation of the two chromophores are referred to as pure orange. Concentrations of only 20–40 ppma Cr3+
“Mekong Whisky” color sapphires. are sufficient to cause a noticeable color shift.
Figure 14 shows the E⊥c and E||c absorption cross
sections of the h•-Cr3+ pair system. Again, note the COMPARING THE CHROMOPHORES
extremely large cross sections (1.3 × 10–17 cm2 ± Looking closely at the absorption cross section spec-
12.5%), which are similar in magnitude to those of tra for the various chromophores, one can observe
the h•-Fe3+ pair. These two systems are by far the a wide range of values in the visible region of the
strongest chromophores in corundum. spectrum of a factor of about 500. However, the
Figure 15 illustrates the color circles for h•-Cr3+ variability of the “strength” of the chromophore is
pairs. At left are shown the color circles for E⊥c, E||c, even greater because the width (full width half max-
and E⊥c + E||c under D65 illumination. From the imum, FWHM) of the cross section features in the
color circles we can see that this chromophore ex- visible range varies from ~25 nm for Fe3+ to ~250 nm
hibits little dichroism, as the E⊥c and E||c circles are for the Fe2+-Ti4+ pair (see box A). Figure 16 shows the
very similar in appearance. This results from the fact approximate concentrations of the six chro-

3.0
E⊥c
E||c
ABSORPTION CROSS SECTION × 1017 (cm2)

2.5 Figure 14. Cross section

spectra of the h•-Cr3+
chromophore measured
from a synthetic orange
2.0
sapphire grown by the
Czochralski method.
Also shown is a 14.62
1.5
ct heated Sri Lankan
orange sapphire, repre-
sentative of the color
1.0 that would be produced
by the h•-Cr3+ chro-
mophore in isolation.
0.5 Photo by Lotus Gemol-
ogy; courtesy of Ceylon
Fine Gems.
0
200 300 400 500 600 700 800 900 1000 1100

WAVELENGTH (nm)

CAUSES OF COLOR IN CORUNDUM GEMS & GEMOLOGY SPRING 2020 25

 h•-Cr3+ Chromophore

CIE-D65 Illuminant
(daylight; approx. 6500 K) Illuminant Comparison

E⊥c

E⊥c CIE-D65
(~6500 K)

CIE-A
E||c (~2856 K)

E⊥c + E||c
E||c

0.4 0.8 1.6 3 6 12

CIE-D65
• 3+
h -Cr areal density in ppma-cm (~6500 K)

CIE-A
(~2856 K)

E⊥c + E||c

CIE-D65
(~6500 K)

CIE-A
(~2856 K)

0.4 0.8 1.6 3 6 12

• 3+
h -Cr areal density in ppma-cm

Figure 15. Left: Areal density color circles for the h•-Cr3+ chromophore. Right: Illuminant comparison of the
h•-Cr3+ chromophore. As with the h•-Fe3+ chromophore, little color change is observed with the change in illu-
minant for all orientations.

mophores required to create roughly similar levels engendered by the fact that the existence of, strength,
of color saturation for E⊥c in samples 1 cm thick. and absorption spectrum of the h•-Fe3+ chromophore
Concentrations for the chromophores were selected was unknown. Additionally, the analytical tech-
by visually comparing color circles and forming a niques available before SIMS and SIMS-calibrated
consensus opinion among the authors about which LA-ICP-MS with matrix-matched standards were in-
circles appeared to have similar saturation levels. capable of measurements down to the ppma level.
Comparing the chromophore concentrations in Another key to our current understanding of chro-
ppma under each circle emphasizes the rather ex- mophores in corundum was the capability of growing
treme range in “strength,” which is greater than a synthetic sapphire containing only one of the chro-
factor of 1000. mophores we wished to study. This allowed us to
The belief that yellow sapphires were colored confirm what we observed in natural samples.
only by iron with or without some undefined “color Determining the absorption cross sections is a
center” (Schmetzer et al., 1983; Nassau, 1991) was powerful technique in the study of color in allochro-

26 CAUSES OF COLOR IN CORUNDUM GEMS & GEMOLOGY SPRING 2020

 Fe3+ V3+ Cr3+ Fe2+-Ti4+ h•-Fe3+ h•-Cr3+

2960 400 100 4 0.9 0.8

Chromophore concentration in ppma

Figure 16. These color circles show the concentration of each chromophore necessary to produce similar saturation
levels in corundum, thus illustrating the drastic difference in absorption strength among each of the chromo-
phores. The circles represent the color produced by each of the chromophores in a 1-cm-thick sample viewed down
the c-axis (E⊥c orientation) under CIE-D65 Illuminant. The chromophore concentration varies from nearly 3000
ppma for a weak absorber like Fe3+ to less than one ppma for a strong absorber like h•-Cr3+.

matic minerals, as their magnitudes provide infor- insights. Our current trace element analytical capa-
mation on the types of absorbing species present. Ap- bility, at the ppma level or below, provides the criti-
plying it to other minerals might yield some new cal capability to support such research.

ABOUT THE AUTHORS Milan Kokta from the former Union Carbide crystal growth division
Ms. Dubinsky is president and head gemologist and jewelry de- and Zachary Coles’ team at Scientific Materials Corporation grew
signer at Emily Emmett, Inc. in New York City. Dr. Emmett is direc- many specially doped synthetic sapphire crystals for this work
tor of Crystal Chemistry in Brush Prairie, Washington, and a which were key to elucidating our understanding of the individual
consultant to GIA. Dr. Stone-Sundberg is a technical advisor on chromophores. We are very grateful to Yunbin Guan for perform-
GIA education operations and a technical editor of Gems & ing the SIMS measurements and for many discussions on the
Gemology located in Portland, Oregon. data. John Trenholme graciously performed a fit to the Sellmeier
equation using refractive index data for corundum, which allowed
ACKNOWLEDGMENTS us to correct our cross section spectra for multiple reflection
The authors would like express our appreciation to Tom Moses losses from the polished sample surfaces. We would like to thank
and Ken Scarratt for their long-term support that made this work John S. Harris for supplying the photograph of the visible spec-
possible. We would like to acknowledge our many discussions trum that appears at the top of all cross section spectra pre-
with George Rossman, which helped resolve key technical issues. sented in this article. We would also like to thank Dan Dell for his
We greatly appreciate the opportunity to present some of this assistance in producing the composite illustrations using the visi-
work in Dick Hughes’ book Ruby & Sapphire: A Gemologist’s ble spectrum and cross section spectra, and in preparing final
Guide as well as his support of this publication. He and Wimon drafts of figures B-1 and B-2.
Manorotkul also provided many of the photos of faceted gems.

REFERENCES
Beran A., Rossman G.R. (2006) OH in naturally occurring corun- Emmett J.L., Scarratt K., McClure S.F., Moses T., Douthit T.R.,
dum. European Journal of Mineralogy, Vol. 18, No. 4, pp. 441– Hughes R., Novak S., Shigley J.E., Wang W., Bordelon O., Kane
447, http://dx.doi.org/10.1127/0935-1221/2006/0018-0441 R. (2003) Beryllium diffusion of ruby and sapphire. G&G, Vol.
Berns R.S. (2000) Billmeyer and Saltzman’s Principles of Color 39, No. 2, pp. 84–135, http://dx.doi.org/10.5741/GEMS.39.2.84
Technology. John Wiley & Sons, Inc., New York, 272 pp. Emmett J.L., Dubinsky E.V., Hughes R.W., Scarratt K. (2017a)
El-Aiat M.M., Kröger F.A. (1982) Hydrogen donors in α-Al2O3. Jour- Color, spectra & luminescence. In R.W. Hughes, Ed., Ruby &
nal of Applied Physics, Vol. 53, No. 5, pp. 3658–3667, Sapphire: A Gemologist’s Guide. Bangkok, pp. 107–163.
http://dx.doi.org/10.1063/1.331150 Emmett J.L., Stone-Sundberg J., Guan Y., Sun Z. (2017b) The role

CAUSES OF COLOR IN CORUNDUM GEMS & GEMOLOGY SPRING 2020 27

 of silicon in the color of gem corundum. G&G, Vol. 53, No. 1, Moon A.R., Phillips M.R. (1994) Defect clustering and color in
pp. 42–47, http://dx.doi.org/10.5741/GEMS.53.1.42 Fe,Ti: α-Al2O3. Journal of the American Ceramic Society, Vol.
Ferguson J., Fielding P.E. (1971) The origins of the colours of yel- 77, No. 2, pp. 356–357, http://dx.doi.org/10.1111/j.1151-
low, green and blue sapphires. Chemical Physics Letters, Vol. 2916.1994.tb07003.x
10, No. 3, pp. 262–265, http://dx.doi.org/10.1016/0009- Nassau K. (1991) The seven types of yellow sapphire and the pro-
2614(71)80282-8 posed Ponahlo test. Zeitschrift der Deutschen Gemmologis-
——— (1972) The origins of the colours of natural yellow, blue, chen Gesellschaft, Vol. 40, No. 4, pp. 247–251.
and green sapphires. Australian Journal of Chemistry, Vol. 25, Norby T. (1989) Hydrogen defects in inorganic solids. In Ø. Johan-
No. 7, pp. 1371–1385, http://dx.doi.org/10.1071/CH9721371 nesen and A.G. Andersen, Eds., Defect Chemistry of Solids. El-
Fritsch E., Rossman G.R. (1987) An update on color in gems. Part sevier Science Publishers B.V., Amsterdam, pp. 101–142,
1: Introduction and colors caused by dispersed metal ions. http://dx.doi.org/10.1016/B978-0-444-88534-0.50011-4
G&G, Vol. 23, No. 3, pp. 126–139, http://dx.doi.org/10.5741/ Pajot B., Clerjaud B. (2013) Optical Absorption of Impurities and
GEMS.23.3.126 Defects in Semiconducting Crystals: Electronic Absorption of
——— (1988) An update on color in gems. Part 2: Colors involving Deep Centres and Vibrational Spectra. Springer-Verlag Berlin
multiple atoms and color centers. G&G, Vol. 24, No. 1, pp. 3– Heidelberg, 512 pp.
15, http://dx.doi.org/10.5741/GEMS.24.1.3 Powell R.C. (1966) The Interaction of Chromium Ions in Ruby
Häger T. (2001) High temperature treatment of natural corundum. Crystals. Physical Sciences Research Papers No. 299. Air Force
International Workshop on Material Characterization by Solid Cambridge Research Labs, Hanscom, Massachusetts.
State Spectroscopy: The Minerals of Vietnam, pp. 4–10. ——— (1998) Physics of Solid-State Laser Materials. Springer-Ver-
Harlow G.E., Bender W. (2013) A study of ruby (corundum) com- lag, New York, 423 pp.
positions from the Mogok Belt, Myanmar: Searching for chem- Pryce M.H.L., Runciman W.A. (1958) The absorption spectrum of
ical fingerprints. American Mineralogist, Vol. 98, No. 7, pp. vanadium corundum. Discussions of the Faraday Society, Vol.
1120–1132, http://dx.doi.org/10.2138/am.2013.4388 26, pp. 34–42, http://dx.doi.org/10.1039/df9582600034
Henderson B., Bartram R.H. (2000) Crystal-Field Engineering of Schmetzer K., Bosshart G., Hänni H.A. (1983) Naturally-coloured
Solid-State Laser Materials. Cambridge University Press, Cam- and treated yellow and orange-brown sapphires. Journal of
bridge, UK. Gemmology, Vol. 18, No. 7, pp. 607–622.
Krebs J.J., Maisch W.G. (1971) Exchange effects in the optical-ab- Shtepliuk I., Lashkarev G., Khyzhun O., Kowalski B., Reszka A.,
sorption spectrum of Fe3+ in Al2O3. Physical Review B, Vol. 4, Khomyak V., Lazorenko V., Timofeeva I. (2011) Enhancement
No. 3, pp. 757–769, http://dx.doi.org/10.1103/PhysRevB.4.757 of the ultraviolet luminescence intensity from Cd-doped ZnO
Kröger F.A. (1984) Electrical properties of α-Al2O3. In W.D. Kingery, films caused by exciton binding. Acta Physica Polonica A, Vol.
Ed., Advances in Ceramics: Structure and Properties of MgO 120, No. 5, pp. 914–917, http://dx.doi.org/10.12693/APhysPolA.
and Al2O3 Ceramics. American Ceramic Society, Columbus, 120.914
Ohio, pp. 1–15. Stone-Sundberg J., Thomas T., Sun Z., Guan Y., Cole Z., Equall R.,
Kvapil J., Perner B., Súlovský J., Kvapil J. (1973) Colour centre for- Emmett J.L. (2017) Accurate reporting of key trace elements in
mation in corundum doped with divalent ions. Kristall und ruby and sapphire using matrix-matched standards. G&G, Vol.
Technik, Vol. 8, No. 1-3, pp. 247–251, http://dx.doi.org/ 53, No. 4, pp. 438–451, http://dx.doi.org/10.5741/GEMS.
10.1002/crat.19730080125 53.4.438
Li H., Robertson J. (2014) Behaviour of hydrogen in wide band gap Tatian B. (1984) Fitting refractive-index data with the Sellmeier
oxides. Journal of Applied Physics, Vol. 115, No. 20, p. 203708, dispersion formula. Applied Optics, Vol. 23, No. 24, pp. 4477–
http://dx.doi.org/10.1063/1.4878415 4485, http://dx.doi.org/10.1364/AO.23.004477
MacFarlane R.M. (1964) Optical and magnetic properties of triva- Thomas T., Rossman G.R., Sandstrom M. (2014) Device and
lent vanadium complexes. Journal of Chemical Physics, Vol. method of optically orienting biaxial crystals for sample prepa-
40, No. 2, pp. 373–377, http://dx.doi.org/10.1063/1.4755936 ration. Review of Scientific Instruments, Vol. 85, No. 9, p.
Maiman T.H., Hoskins R.H., D’Haenens I.J., Asawa C.K., Evtuhov 093105, http://dx.doi.org/10.1063/1.4894555
V. (1961) Stimulated optical emission in fluorescent solids: Tippins H.H. (1970) Charge-transfer spectra of transition-metal
Spectroscopy and stimulated emission in ruby. Physical Re- ions in corundum. Physical Review B, Vol. 1, No. 1, pp. 126–
view, Vol. 123, No. 4, pp. 1151–1157, http://dx.doi.org/ 135, http://dx.doi.org/10.1103/PhysRevB.1.126
10.1103/PhysRev.123.1151 Townsend M.G. (1968) Visible charge transfer band in blue sap-
Matsunaga K., Nakamura A., Yamamoto T., Ikuhara Y. (2004) The- phire. Solid State Communications, Vol. 6, No. 2, pp. 81–83,
oretical study of defect structures in pure and titanium-doped http://dx.doi.org/10.1016/0038-1098(68)90005-7
alumina. Solid State Ionics, Vol. 172, No. 1-4, pp. 155–158, Zaw K., Sutherland L., Yui T-F., Meffre S., Thu K. (2015) Vana-
http://dx.doi.org/10.1016/j.ssi.2004.01.044 dium-rich ruby and sapphire within Mogok Gemfield, Myan-
Mattson S.M., Rossman G.R. (1988) Fe2+-Ti4+ charge transfer in sto- mar: Implications for gem color and genesis. Mineralium
ichiometric Fe2+,Ti4+-minerals. Physics and Chemistry of Min- Deposita, Vol. 50, No. 1, pp. 25–39, http://dx.doi.org/
erals, Vol. 16, No. 1, pp. 78–82, http://dx.doi.org/10.1007/ 10.1007/s00126-014-0545-0
BF00201333 Zorenko Y., Zorenko T., Voznyak T. (2011) Luminescence centers
McClure D.S. (1962) Optical spectra of transition-metal ions in in Y3Al5O12:La single crystals. Journal of Physics: Conference
corundum. Journal of Chemical Physics, Vol. 36, No. 10, pp. Series, Vol. 289, No. 1, p. 012028, http://dx.doi.org/
2757–2779, http://dx.doi.org/10.1063/1.1732364 10.1088/1742-6596/289/1/012028

28 CAUSES OF COLOR IN CORUNDUM GEMS & GEMOLOGY SPRING 2020

 First Place
GEOGRAPHIC ORIGIN DETERMINATION OF BLUE SAPPHIRE
WINTER 2019
Aaron Palke, Sudarat Saeseaw, Nathan Renfro, Ziyin Sun, and Shane McClure
Aaron Palke is senior manager of research at GIA in Carlsbad, California. He
holds a PhD in geology from Stanford University. Sudarat Saeseaw is senior
manager of colored stone identification at GIA in Bangkok. Ms. Saeseaw
Aaron Palke Sudarat Saeseaw Nathan Renfro
received her master’s degree in analytical chemistry from Mahidol University
in Thailand. Nathan Renfro is manager of colored stone identification at GIA
in Carlsbad. He earned a bachelor’s degree in geology from Appalachian
State University in North Carolina. Ziyin Sun is a research associate in the
gem identification department at GIA in Carlsbad. He holds a bachelor’s
degree in chemistry and a master’s degree in analytical chemistry from
Nanjing University in China. Shane McClure is global director of colored
stone services at GIA in Carlsbad and co-editor of G&G’s Lab Notes section.
He is well known for his many articles and lectures on gem identification. Ziyin Sun Shane McClure

Seconad Place (tie)

A DECADE OF RUBY FROM MOZAMBIQUE: A REVIEW
SUMMER 2019
Wim Vertriest and Sudarat Saeseaw
Wim Vertriest is supervisor of field gemology at GIA in Bangkok. Mr. Vertriest
has visited gemstone mining areas worldwide and oversees the field gemolo-
gy department and curates GIA’s colored stone research collection in
Wim Vertriest
Bangkok. He obtained a master’s degree in geology (geodynamics and
geofluids) from KU Leuven in Belgium. Sudarat Saeseaw was profiled in the
first-place entry.

Second Place (tie)

GEMSTONES IN THE ERA OF THE TAJ MAHAL AND THE MUGHALS
FALL 2019
Dona Mary Dirlam, Chris L. Rogers, and Robert Weldon
Dona Dirlam is librarian emerita at GIA’s Richard T. Liddicoat Gemological
Library and Information Center in Carlsbad, California, where she established
and expanded the collection known as the world’s leading resource center for
Dona Dirlam Chris Rogers Robert Weldon
gems, gemology, and jewelry. Ms. Dirlam holds a master’s in geology and geo-
physics from the University of Wisconsin, Madison. Chris Rogers is a research librarian at GIA’s Richard T. Liddicoat
Gemological Library and Information Center. She holds a bachelor’s degree in English from San Diego State University
and a master’s in library and information science from San Jose State University. Robert Weldon is director of GIA’s
Richard T. Liddicoat Gemological Library and Information Center. He is well known for his gem photography and has
contributed to scores of international publications and several gem-related books.

Many thanks to the members of G&G’s Editorial Review Board for voting this year.

MOST VALUABLE ARTICLE AWARD GEMS & GEMOLOGY SPRING 2020 29

FEATURE ARTICLES

THE EVOLUTION OF CHINESE

JADE CARVING CRAFTSMANSHIP
Mingying Wang and Guanghai Shi

Craftsmanship is a key element in Chinese jade carving art. In recent decades, the rapid development of tools
has led to numerous changes in carving technology. Scholars are increasingly focusing on the carving craft in
addition to ancient designs. Five periods have previously been defined according to the evolution of tools and
craftsmanship, and the representative innovations of each period are summarized in this article. Nearly 2,500
contemporary works are analyzed statistically, showing that piercing and Qiaose, a technique to take artistic
advantage of jade’s naturally uneven color, are the most commonly used methods. Current mainstream tech-
niques used in China’s jade carving industry include manual carving, computer numerical control engraving,
and 3D replicate engraving. With a rich heritage and ongoing innovation in jade craftsmanship, as well as in-
creased automation, the cultural value and creative designs are both expected to reach new heights.

J
ade carving is one of the oldest and most impor- tools. In 1978, China opened its economy to the out-
tant art forms in China, a craft steeped in history side. With the support of the government, influenced
and tradition that reflects Chinese philosophy and by jade merchants and carvers from Taiwan and
culture (Thomas and Lee, 1986). Carving, called Hong Kong and driven by an expanded domestic con-
Zhuo and Zhuo mo in Chinese, represents the ardu- sumer market, China’s modern jade carving industry
ous process of shaping and decorating an intractable has seen a period of vigorous development.
material with abrasives to create the desired object
(Hansford, 1950; Sax et al., 2004). During its more
than 5,000-year history (table 1), Chinese jade carving In Brief
has experienced five peaks—the late Neolithic Age
• With the prevalence of power-driven systems, China’s
(circa 3500–2070 BCE), the late Shang Dynasty
modern jade carving technology and industry have
(1600–1046 BCE), the Zhanguo Period (475–221 BCE), been vigorously developing for more than 40 years.
the Han Dynasty (206 BCE–220 CE), and the Qing
• For thousands of years, innovations in carving tools
Dynasty (1644–1911)—and was influenced by poli- have continuously promoted the development of Chi-
tics, economics, culture, and war (Kong, 2007). After nese jade carving craftsmanship, in a direction toward
Emperor Qianlong abdicated in 1795, the develop- increased diversity, complexity, and delicacy.
ment of jade carving craftsmanship was relatively • Modern Chinese jade carving technology includes
stagnant (Thomas and Lee, 1986). With the fall of the three mainstream methods: power-driven manual carv-
Qing Dynasty in 1911, the royal demand for jade ing, computer numerical control engraving, and 3D
carving faded. During that period, the folk workshops replicate engraving.
mainly focused on the carving of common jade wares • The value of jade cultural connotation and design cre-
such as jewelry and vessels. After 1949, state-owned ativity will improve with growing industrial automation.
jade carving factories were established one after an-
other, with an increasing number of employees and
the gradual popularization of electrically powered There is a saying in China that “workers must first
sharpen their implements when they want to do their
best.” Indeed, the emergence of each peak in jade carv-
ing across its history in China is inseparable from the
See end of article for About the Authors and Acknowledgments.
GEMS & GEMOLOGY, Vol. 56, No. 1, pp. 30–53,
innovation of tools (Xu, 2011). Currently, jade carving
http://dx.doi.org/10.5741/GEMS.56.1.30 tools are undergoing great changes, influenced by ad-
© 2020 Gemological Institute of America vances in manufacturing. However, little is known in

30 CHINESE JADE CARVING CRAFTSMANSHIP GEMS & GEMOLOGY SPRING 2020

 Figure 1. The white
nephrite Buddhist carv-
ing by modern master
Xi Yang is displayed in
the British Museum. It
presents an openwork
papercut-style profile of
Guanyin on the left
and a three-dimen-
sional carved lotus on
the right, jointly mani-
festing the Buddhist
culture with oriental
characteristics
(Michaelson et al.,
2016). Photo © The
British Museum
(2017.3034.1).

the West today about the modern craft compared with EVOLUTION OF JADE CARVING TOOLS
ancient Chinese jade carving (Michaelson et al., 2016). In the process of fashioning stone tools, human an-
Since the founding of the People’s Republic of China cestors gradually mastered the method of distin-
in 1949, papers on jade carving have been published guishing jade from other “non-jade” stones. The
in the West, but the majority have focused on tradi- emergence of the rotary machine (circa 3500 BCE),
tional carving methods and only a few on the modern in which jade was held against a grinding tool being
methods that emerged leading up to the 1990s (Hans- rotated by the other hand or another person, indi-
ford, 1950; Tucker, 1982; Markbreiter, 1985; Thomas cated that jade carving had completely separated
and Lee, 1986; Desautels, 1986). In 2017–2018, from general stone craft to become an independent
nephrite jade works by Xi Yang and four other con- category (Yang, 2004, 2006). The innovation of the
temporary masters were collected by the British Mu- rotary tool is the most important factor in the devel-
seum for permanent display (box A), indicating that opment of Chinese jade carving. As the key compo-
modern Chinese jade wares have begun to attract in- nent of the jade carving craft, the tool displays
terest in the West (figure 1). different characteristics in different development
The main research interests of jade wares are on stages. Yang (2006) divided Chinese rotary machines
their texture, shape, pattern, and craft (You, 2001). Pre- into five generations, and Xu (2011) divided the evo-
vious studies on Chinese jade carving have focused on lution of the jade carving craft into five periods. The
the exploration of shapes and patterns. In recent years, present authors combined the results from these
with the rapid development of jade carving tools and studies and offer the following five generations of
craft, research on craftsmanship has increasingly at- jade carving rotary tools in Chinese history.
tracted the attention of scholars (Kong, 2007; Xu, 2011).
The authors of the present article conducted field visits First Generation (circa 3500–circa 2070 BCE). The
to carving and trading sites such as Beijing, Shanghai, primitive rotary jade carving machine appeared in
Suzhou, Yangzhou, Urumqi, Sihui, and Zhenping, in- the Hongshan and Lingjiatan cultural sites of the Ne-
cluding requisite visits to carving workshops and tech- olithic Age (Yang, 2006). These jade carving tools
nical schools, to gain a more complete understanding were made primarily from natural materials such as
of the current state. This article analyzes the develop- stone, wood, and bone. Yang (1989a,b) proposed two
ment of jade carving tools over five periods and the rep- main features of the primitive rotary carving ma-
resentative carving craft during these periods. Finally, chine. The first is that it allowed the carver to as-
the processing methods and procedures of Chinese jade sume a sitting position. The second is that it was
carving in the modern context are introduced. manually rotated.

CHINESE JADE CARVING CRAFTSMANSHIP GEMS & GEMOLOGY SPRING 2020 31

TABLE 1. A brief chronology of Chinese history.

Period/Dynasty Common English Name Time Span

Neolithic Period Neolithic Period circa 10,000 years ago–2070 BCE
Xia Dynasty Xia Dynasty 2070–1600 BCE
Shang Dynasty Shang Dynasty 1600–1046 BCE
Xizhou Dynasty Western Zhou Dynasty 1046–771 BCE
Chunqiu Period Spring and Autumn Period 770–476 BCE
Zhanguo Period Warring States Period 475–221 BCE
Qin Dynasty Qin Dynasty 221–207 BCE
Xihan Dynasty Western Han Dynasty 206 BCE–8 CE
Donghan Dynasty Eastern Han Dynasty 25–220 CE
Sanguo Period Three Kingdoms Period 220–265 CE
Xijin Dynasty Western Jin Dynasty 265–316 CE
Dongjin Dynasty Eastern Jin Dynasty 317–420 CE
Nan-bei Dynasty Period of the Northern and Southern Dynasties 420–589 CE
Sui Dynasty Sui Dynasty 581–618 CE
Tang Dynasty Tang Dynasty 618–907 CE
Wu-dai-shi-guo Period Period of the Five Dynasties and Ten Kingdoms 907–960 CE
Beisong Dynasty and Liao Dynasty Northern Song Dynasty and Liao Dynasty 960–1127
Nansong Dynasty and Jin Dynasty Southern Song Dynasty and Jin Dynasty 1127–1279
Yuan Dynasty Yuan Dynasty 1271–1368
Ming Dynasty Ming Dynasty 1368–1644
Qing Dynasty Qing Dynasty 1644–1911
Republic of China Republic of China 1912–1949
People’s Republic of China People’s Republic of China 1949–present
From Yi (2001)

Second Generation (circa 2070 BCE–circa sixth cen- Fourth Generation (circa 581–1960). The table-type
tury BCE). The bronze rotary machine emerged and iron rotary machine was used from the Sui and Tang
thrived from the Xia and Shang Dynasties to the Dynasties through the Ming and Qing Dynasties and
early to mid-Chunqiu Period (Xu, 2014) as life in this into the 1950s (Yang, 2006; Xu, 2014). In the Sui and
period gradually centered on bronze technology. It is Tang Dynasties, interior furniture changed people’s
speculated that operators were kneeling and manu- habit of sitting on the ground, and they instead sat
ally drove the machine to faster speeds than achiev- on stools or chairs (Yang, 2006). As a result, the rotary
able before, based on the kneeling posture depicted machine in this period was of the table-type or tall-
in unearthed human stone carvings from this period legged table-type and driven by a foot treadle that fur-
and in the tomb of Fuhao (Yang, 2006). ther increased its working efficiency (figure 2).

Third Generation (circa sixth century BCE–circa 581 Fifth Generation (1960–present). The modern rotary
CE). The iron rotary machine appeared from the late machine has been used in the contemporary jade
Chunqiu Period until the Nan-bei Dynasty. Carvers carving process since the 1960s (Yang, 2006; Xu,
still knelt on the ground while operating the machine 2014). It was initially introduced from Europe, but
(Xu, 2014). As smelting technology improved, iron with the emergence of its manufacturing industry,
tools were increasingly used and the jade carving ma- China began to independently research, develop, and
chine evolved accordingly. It was still manually driven produce these machines (Read, 1981; Markbreiter,
but provided increased efficiency because iron tools 1985). The body of the jade rotary machine has
were sharper and more wear-resistant (Xu, 2014). changed from wood to iron, the grinding heads are

32 CHINESE JADE CARVING CRAFTSMANSHIP GEMS & GEMOLOGY SPRING 2020

 Figure 2. A restoration
of a fourth-generation
table-type iron rotary
machine. The artisan
sits at the table and
carves the jade using
the rotary tool and
abrasive sands, driven
by the foot treadle.
Photo by Mingying
Wang, from Tianjin
Museum.

now carborundum-coated, and foot power has been chine is small, portable, and has a handheld electric-
replaced by motor power (Deng, 2011; Xu, 2014). At powered rotary tool with a flexible drive (Michaelson
present, there are two main types of equipment for et al., 2016), which makes it convenient to change
manual carving: grinding machines and flexible-shaft different grinding heads (figure 3). The emergence of
machines. The grinding machine is a large apparatus, motor power and the improvement of tools have not
and the seated carver holds the jade while using tools only increased carving efficiency but also benefited
of various shapes. The flexible-shaft engraving ma- jade carvers. The new tools are suitable for detailed

CHINESE JADE CARVING CRAFTSMANSHIP GEMS & GEMOLOGY SPRING 2020 33

 BOX A: MODERN JADE CARVINGS SELECTED FOR THE BRITISH MUSEUM

Figure A-1. Hongwei Ma’s nephrite works “Wine Cup (Jue)” (left) and “Bronze Ram Zun” (right). Photos © The British
Museum (2017.3032.1 and 2018.3019.1).

I
n 2017–2018, eight nephrite works created by Figure A-2. Guang Yang’s “Incense Pot” in nephrite.
Chinese modern masters Hongwei Ma, Guang Photo © The British Museum (2017.3033.1).
Yang, Xi Yang, Ting Yu, and Yiwei Zhai were
selected by the British Museum for its perma-
nent collection. These works are on permanent
display to the public as jade carvings by living
Chinese artists of the twenty-first century.

34 CHINESE JADE CARVING CRAFTSMANSHIP GEMS & GEMOLOGY SPRING 2020

 Figure A-4. Ting Yu’s nephrite “Eggshell Teapot.” Photo ©
The British Museum (2017.3036.1).

Figure A-3. Xi Yang’s nephrite work “Autumn in

Suzhou.” Photo © The British Museum (2017.3035.1).
Figure 1 shows the artist’s other piece on exhibit at the
British Museum.

Figure A-5. Yiwei Zhai’s nephrite works “Beside is a bosom friend, thou. Everywhere is paradise, enow” (left and middle)
and “The Rhyme of Spring” (right). Photos © The British Museum (2017.3037.1 and 2018.3020.1).

CHINESE JADE CARVING CRAFTSMANSHIP GEMS & GEMOLOGY SPRING 2020 35

 Figure 3. Top left: The
driveshaft is fixed in a
horizontal direction on
the jade grinding ma-
chine; at the end of the
shaft is the turnbuckle
or bayonet where the
grinding head or the
long horizontal shaft
can be fixed. Various
grinding heads (top
right) can be exchanged
according to the carv-
ing needs. Photos cour-
tesy of Yongtao Chen.
Bottom left: The object
is moved by hand to
contact the stationary
rotating tool. Photo by
Mingying Wang. Bot-
tom right: The carver
works the jade piece
using the hand-held
burr of the flexible-
shaft machine. Both
the jade and the tool
are movable. Photo by
Eric Welch; courtesy of
Mazu Jade Co.

ornaments and small jade wares as well as larger ings. Given cost and efficiency concerns, three- and
wares. Furthermore, motor-driven equipment has four-shaft CNC machines are more common in the
driven the introduction and refinement of tools for Chinese market. CNC technology can be applied to
specific uses: for example, various types of cutters, a wide range of processing types and are mainly used
ultrasonic drilling equipment, mechanized grinding for engraving medium- and low-end jade and stone
beads, and vibratory polishers (Xu, 2014). materials of various sizes. Its emergence and popu-
Since the 1990s, automated computer numerical larity means that carvings can be produced in bulk,
control (CNC) processors have gradually been ap- further improving efficiency while maintaining the
plied to jade carving. Together with design software high intricacy of jade craftsmanship (figure 4).
and three-dimensional data processing, CNC engrav- In addition to rotary tools, abrasive sands and
ing and milling machines can be used not only to powders are an important element in jade carving.
manufacture original jade carvings but also to make Abrasive sands used in the primitive carving period
1:1 replicas of existing wares or historical relics. were ordinary sands that contained a large amount
Three types of automated CNC engraving and of quartz particles, which are harder than jade. With
milling machines are mainly used for jade carving: the development of jade carving craftsmanship, abra-
three-shaft, four-shaft, and five-shaft. The number of sive sands have gradually been refined (Hansford,
shafts represents the degree of processing freedom 1950; Sax et al., 2004; Xu, 2011). Today they include
and in turn the quality of the resulting curved surface natural abrasives (such as corundum, quartz, and di-
details. The three-shaft CNC machine is suitable for amond) and synthetic abrasives (such as carborun-
plane-relief engraving, while the four- and five-shaft dum and synthetic diamond). In general, artificial
CNC machines can execute three-dimensional carv- abrasives are superior to natural ones (except for nat-

36 CHINESE JADE CARVING CRAFTSMANSHIP GEMS & GEMOLOGY SPRING 2020

 Figure 4. With a CNC
engraving and milling
machine, a jade carv-
ing that takes 20 days
by manual carving can
be made in three to
four days. Photo by
Mingying Wang.

ural diamond) in quality, hardness, and performance. forward (Xu, 2011). Therefore, we will now analyze
Therefore, modern jade carving mainly utilizes arti- and summarize representative crafts used in different
ficial abrasives, which are often used for coating the tool development periods (table 2).
grinding heads (Zhao and Zhang, 2000). Table 2 shows that over 5,000 years of Chinese
Throughout the evolution of Chinese jade carv- jade culture, the development of carving tools has
ing, every tool innovation has experienced a long de- featured two trends: tool speed becoming increas-
velopment cycle. The transition from the prehistoric ingly faster, and grinding powder becoming increas-
stone tool to the bronze tool to the iron tool to the ingly harder. The evolution of tools improved
modern machine was accomplished over thousands processing conditions for carvers. As a result, Chi-
of years. However, due to regional variations and eco- nese jade carving moved toward greater diversity,
nomic constraints, new jade carving tools were not complexity, and delicacy. Consider the shapes, for ex-
suddenly applied in a given dynasty but were gradu- ample. During the Xia, Shang, and Xizhou Dynasties,
ally adopted over a longer period (Xu, 2011). For ex- jade wares were mainly two-dimensional tablet
ample, during the Xia and Shang Dynasties, bronze shapes. With advances in tools and craftsmanship,
tools did not completely replace the primitive stone the forms gradually became three-dimensional. In
tools, and the two were used together for a long time the Ming and Qing Dynasties, most jade wares dis-
(Lu et al., 2014). Nephrite, the mainstream material played rounded shapes.
of Chinese jade culture, has a high toughness, which Although each of the five stages in table 2 corre-
makes jade carving more difficult but promotes the sponds to a certain historical period, any craft or tech-
innovation of tools. In early jade carving, several nology progresses gradually through exploration and
weeks were needed just to saw through a boulder creation. The accumulation of experience takes a long
(Hansford, 1950), making the development of tools time. The refinement of a craft is first tested on a
and technology more important. small scale and then gradually promoted and popu-
larized. Therefore, a change in dynasty does not nec-
EVOLUTION OF JADE CARVING essarily represent the immediate progress of a craft.
CRAFTSMANSHIP
If the tool is the fundamental force driving the evo- Primitive Jade Carving (circa 3500–circa 2070 BCE).
lution of jade carving, then craft is also an important In the Neolithic Age, stone tools were fashioned by
factor. The two forces are inseparable. Every innova- cutting, sanding, and drilling. These three techniques
tion in tools allows the carving craft to take a leap were inherited by the primitive jade carvers and fur-

CHINESE JADE CARVING CRAFTSMANSHIP GEMS & GEMOLOGY SPRING 2020 37

TABLE 2. Periods of Chinese jade carving tools and corresponding craftsmanship.

Period Representative Jade Carving Tools Representative Jade Carving Craftsmanship

Primitive jade carving period
Jade collecting, cutting, grinding, carving, drilling,
Late Neolithic Age Manually driven piercing, inlay, eggshell carving, and polishing—the
(circa 3500 BCE–circa 2070 BCE) basic procedure and various techniques used by
Grinding tools made of nonmetals, such
later generations—initially emerge in this period
as seashells, stones, bones, and wood
Inlay, Qiaose (coloring design art), chain carving,
Xia, Shang Dynasties to Early to Mid- Bronze rotary machine period and slope carving (oblique grinding for jade
Chunqiu Period Manually driven patterns)
(circa 2070 BCE–circa sixth century BCE) Piercing becomes sophisticated and hollowing
Bronze grinding tools
reaches a high level

Techniques related to burial jade—Han Badao

(concise carving)
Iron rotary machine period
Late Chunqiu Period to Nan-bei Dynasty Metal-related techniques—inlay and stitching
Manually driven
(circa sixth century BCE–circa 581 CE） “Hair-like carving,” flexible connecting, and Bi
Iron grinding tools carving (a disc with a central round hole)
Piercing is widely applied

Inlay with gems and gold, eggshell carving,

Table-type iron rotary machine period antiquing, 3D picturesque carving, large jade carv-
Sui and Tang Dynasties to the 1950s
Driven by foot treadle ings, calligraphy engraving, and jadeite carving
(circa 581–1960)
Iron grinding tools Qiaose is commonly applied, and importance is
attached to polishing

Modern jade machine period

Motor-driven power The same crafts are used in the modern era, but with
automation and motor-driven tools
Modern Era Iron or steel grinding tools with
(1960–present) carborundum coating Qiaose, piercing, calligraphy engraving, chain carv-
ing, eggshell carving, miniature carving, and gem or
CNC machines gold inlay
Other automated tools include auto-cut- Plane-relief engraving, 3D engraving, and 3D
ting, auto-polishing, laser replicate engraving
engraving, and ultrasonic drilling

Modified from Xu (2011, 2012), You (2001), Yang (2006), Kong (2007), Deng (2011), and Lu et al. (2014)

ther developed (Yang, 2006). Based on evidence from play outstanding intricacy. These inlays were mainly
unearthed relics, it can be speculated that jade carv- combinations of bronze and turquoise, jade and
ing procedures in the Neolithic period mainly com- turquoise, or bronze and jade. Jade vessels emerged
prised collecting, cutting, grinding, carving, drilling, in this period, with a corresponding rise in skilled
and polishing either by hand or by rotary machine. hollowing (Xu, 2011).
These same basic procedures were inherited by later The jade carving craft in the Shang Dynasty,
generations. Such representative techniques as pierc- which included cutting, grinding, carving, drilling,
ing, inlay, and eggshell carving (to produce thin polishing, design, and shape creation, reached a high
“skins”) also made their first appearance (You, 2001; level (Technical Research Group of Beijing Jade Fac-
Kong, 2007; Xu, 2011). However, the quality of jade tory, 1976). During this dynasty, the Qiaose tech-
carving craftsmanship varied from place to place due nique to artistically take advantage of jade’s naturally
to different rates of cultural development (You, 2001). uneven or different color first appeared. In the
process of Qiaose craft, the shape and placement of
Bronze Rotary Machines (circa 2070–circa sixth cen- the colors on the rough jade should be taken into ac-
tury BCE). Although they emerged in the Neolithic count so as to utilize the colors skillfully. The jade
period, inlays in the Xia, Shang, and Xizhou Dynas- turtle in figure 5, unearthed at Xiaotun in Henan
ties outnumber those before the Xia Dynasty and dis- Province, is a successful example of Qiaose (Yang et

38 CHINESE JADE CARVING CRAFTSMANSHIP GEMS & GEMOLOGY SPRING 2020

 also appeared. The craft even exerted influence dur-
ing the Han and Wei Dynasties (circa 206 BCE–265
CE). Techniques integrating jade and metal, such as
inlay and stitching, also surpassed those in the Shang
and Xizhou Dynasties. Here, inlay refers to embed-
ding metals such as copper, silver, iron and gold into
the jade, while stitching means using metal threads
to connect jade pieces together. These two crafts
reached an advanced level during the Zhanguo Pe-
riod, as evidenced by the number and quality of art-
works applying them (Xu, 2011). For example, the
jade clothes sewn with golden or silver thread often
buried in the tombs of nobility in the Han Dynasty
are elaborate and complex in craftsmanship. They are
typical of stitching craft at this time and exemplify
burial jade culture in the Han Dynasty (figure 7).

Figure 6. This jade cicada is a typical example of the

Han Badao (concise carving) craft highlighted in the
Han Dynasty (206 BCE–220 CE). This technique un-
Figure 5. This meticulously designed jade turtle un- folds the full vitality of the jade cicada through just
earthed in 1975 in the Yin Ruins in Xiaotun is an ex- a few concise lines. As the cicada implied rebirth in
ample of Qiaose craft from the late Shang Dynasty ancient China, cicada carvings were often found as
(1600–1046 BCE). The grayish white part of the jade burial items in Han tombs, placed on the tongue of
is carved into the head, neck, and abdomen, and the the body. Photo by Mingying Wang, from Tianjin
dark part forms a vivid turtle shell, displaying the Museum.
artist’s skillful use of color. As a symbol of longevity
in China, the turtle was often used during ancient
burials. Photo from Yang et al. (2005, p. 158).

al., 2005). Chain carving, a very difficult craft carved

from a single piece of stone, also emerged in the
Shang Dynasty, laying the foundation for its devel-
opment in later generations (Xu, 2011).

Iron Rotary Machines (circa sixth century BCE–circa

581 CE). Burial jade thrived in the Chunqiu Period
and culminated in a set of special carving systems
formed in the Han Dynasty. Burial jade shows no sig-
nificant difference from daily jade in terms of the
carving procedure but it is more concise and less ex-
quisite (You, 2001). Take the Han Badao (concise
carving) craft in the Han Dynasty as an example.
Typical of burial jade, the lines from Han Badao are
bold, vigorous, and concise, seemingly carved by
swords (figure 6).
What epitomizes the craftsmanship of the
Zhanguo Period is the wide application of the open-
work technique (Lu and Ouyang, 2014). In this pe-
riod, piercing was mainly applied to flat jade wares,
but a small number of three-dimensional openwork

CHINESE JADE CARVING CRAFTSMANSHIP GEMS & GEMOLOGY SPRING 2020 39

 Figure 7. Sheng Liu’s
jade burial suit (Xihan
Dynasty, 206 BCE–8
CE) was unearthed in
Mancheng, China, in
1968. Approximately
1.88 m long, the suit is
composed of 2,498 jade
tablets sewn together
with approximately
1,100 grams of gold
wire. It is the earliest
and best-preserved jade
burial suit in the his-
tory of Chinese archae-
ological excavation
(Yang et al., 2005). Pho-
tos by Lin Xu (top) and
Tian Wang (bottom),
from Hebei Museum.

Flexible connecting, which has the same engraving Table-Type Iron Rotary Machines (circa 581–1960).
principle as chain carving, was found in tombs of the Gold and jade were already being combined prior to
Shang Dynasty but thrived and matured in the the Qin Dynasty (221–206 BCE), and the inlay of jade
Zhanguo Period, when many quality artworks ap- with gems and gold gained popularity during the Sui
peared (Xu, 2011). Among them, the sixteen-seg- and Tang Dynasties. In the late Ming Dynasty, this
mented jade ornament excavated from the tomb of craft was widely applied and became a striking fea-
Marquis Yi of State Zeng is the best-known application ture. In the Qing Dynasty, influenced by Islamic
of the flexible connecting craft in this period (figure 8). jades that originated in central, western, and south-

Figure 8. This dragon-

phoenix pendant with
16 grayish white com-
ponents (48.0 × 8.3 × 0.5
cm), which was carved
from a single piece of
jade, is from the early
Zhanguo Period (475–
221 BCE). The movable
and foldable compo-
nents are carved into
dragon, phoenix, or Bi-
type disc openwork de-
signs. The ornament
embodies not only the
flexible connecting
technique but also the
mature piercing tech-
nique (Gu, 2005). Top
photo from Gu (2005, p.
93). Bottom photos by
Yuetong Li, from Hubei
Provincial Museum.

40 CHINESE JADE CARVING CRAFTSMANSHIP GEMS & GEMOLOGY SPRING 2020

Figure 9. This white nephrite bowl with inlaid gems and gold tracery (Qing Dynasty, 1644–1911) is engraved with
a lotus pattern on the outside and six sets of pomegranate patterns inside. It shows both the eggshell carving and
inlay techniques and is a typical Islamic jade from China’s western regions (Gu, 2010b). Photos by Rui Zhang
(left) and Maiying Dong (right), from Shanxi Museum.

ern Asia in the late fourteenth century (figure 9), imals, and landscapes were often accompanied by
inlay often depicts various plant patterns, which gave the distinctive openwork carving of the time. The
off a sense of exotic charm (Deng, 2007; Xu, 2011). piercing technique surged, and flat carvings were
During the Song, Liao, Jin, and Yuan Dynasties gradually replaced by three-dimensional picturesque
(circa 960–1368), influenced by the mature arts of jade wares (figure 10). In the Ming Dynasty, the in-
painting and sculpture, jade carvings became three- tellectual “literati paintings” of southern Chinese
dimensional. Themes of human figures, flowers, an- culture, which are distinct from folk and royal paint-

Figure 10. This gem-

inlaid gold base in the
shape of a lotus is
topped with a carved
jade dragon and cloud
(Ming Dynasty, 1368–
1644). The white
nephrite adopts the
multi-layered three-di-
mensional piercing
technique with exqui-
site details. Photo by
Yuetong Li, Hubei
Provincial Museum.

CHINESE JADE CARVING CRAFTSMANSHIP GEMS & GEMOLOGY SPRING 2020 41

 Figure 11. The nephrite
Shan Zi sculpture “Em-
peror Yu Leading People
in Subduing Floods”
(224 × 96 cm, weighing
at least 5,000 kg) was
carved in Yangzhou
from 1781 to 1787 CE.
In the 53rd year of the
Qianlong reign (1788
CE), the court jade
craftsmen engraved the
imperial poem on its
back (Zhu, 2007).
Photo from Zhu (2007,
p. 71).

42 CHINESE JADE CARVING CRAFTSMANSHIP GEMS & GEMOLOGY SPRING 2020

ings, highlighted romantic charms and reflected the
literati’s personal feelings. This led to the incorpo-
ration of landscapes and texts in the decorative pat-
terns of jade wares (Xu, 2011). In the Qing Dynasty,
the use of royal court painters in the design of jade
carvings in combination with fine craftsmanship
created a peak of artistic integration (Yang, 1982; Xu,
2011).
During this period, the improvement and combi-
nation of tools and craft promoted the emergence of
large jade carvings (Kong, 2007). For example, the
massive bowl “Dragons Amid Waves” from the Yuan
Dynasty was the first large jade carving that we
know of. In the Qing Dynasty, jade carvers created
several large works weighing thousands of kilograms,
such as the complex three-dimensional landscape
(called Shan Zi in Chinese) “Emperor Yu Leading
People in Subduing Floods” (figure 11) and “Nine
Elders Meeting During the Huichang Period” (Xu,
2013). These large carvings are mostly themed with
landscapes, human figures, and animals, often ac-
companied by skilled calligraphy engraving on the
back or sides. To create these pieces, several jade
carvers had to work at the same time in an organized
fashion. Under the processing conditions of the time, Figure 12. This modern work (9.3 × 4.2 × 2.8 cm),
such a large work would take several years to com- which retains the beautiful natural color of the jade
plete. Thus, the production of large jade carvings crust, depicts ancient tree branches and a monk sit-
ting in meditation underneath, showing a Zen sensi-
marks the peak of the craft in ancient China (Xu,
bility. The ingenious application of traditional
2011). Chinese painting techniques to the grain carving of
Many other techniques also emerged during this the branches demonstrates a natural artistic style (Ao
period, such as the full development of Qiaose and et al., 2014). Photo courtesy of Gems and Jewelry
eggshell carving, and the emergence of jadeite carv- Trade Association of China (GAC).
ing in the Qing Dynasty (You, 2001; Kong, 2007; Xu,
2011).
changed jade carving, and pieces that once took years
Modern Machines (1960–present). After the Qing Dy- can now be finished in months or even weeks
nasty, the traditional foot treadle was used until the (Tucker, 1982).
early days of the People’s Republic (Xu, 2014). In the Modern jade carving inherits the characteristics
late 1950s, with the policy reform of the national of craftsmanship from past dynasties (Xu, 2014) and
planned economy, government-owned jade carving is further developed on the basis of various ancient
factories sprang up, led by the manufacturing centers techniques. With the emergence of power-driven
of Beijing, Shanghai, Yangzhou, and Suzhou (Ma, tools, jade carving has become more elaborate, with
2014). Jade carvings were produced under industrial smoother lines. The common techniques of the mod-
management using assembly lines (Ji, 1984). It was ern craft include Qiaose, eggshell carving, piercing,
at this time that motor-driven carving machinery chain carving, calligraphy engraving, gem embed-
began to be employed. ding, gold and silver inlay, miniature carving, and
With the popularization of power-driven systems others. Artisans will often make a final decision on
in the 1960s and the continuous refinement of the which techniques to use based on the color, shape,
tools (Xu, 2014), the modern carving industry ush- and toughness of the rough and the theme of their
ered in a new era following the reform and opening creation. Among these techniques, Qiaose (figure 12)
up of trade in 1978. Modern tools and methods have and piercing (figure 13) are widely used in jade carv-

CHINESE JADE CARVING CRAFTSMANSHIP GEMS & GEMOLOGY SPRING 2020 43

Figure 13. Because of its toughness and fine texture, Figure 14. The gold inlay used on this dark gray
nephrite is often subjected to piercing. The modern nephrite vessel is elaborate and exquisite. Accompa-
nephrite work “Holy Lingshan, Journey to the West” nied by traditional auspicious animal patterns, it
(38 × 26 × 8 cm) displays the scene of Xuanzang and makes the work vivid and graceful, paying homage to
his three disciples passing through the Lingshan holy an antique style. Photo courtesy of Jingui Ma Jade
land (Ao et al., 2009). Photo courtesy of the Gems and Carving Studio.
Jewelry Trade Association of China (GAC).

ing and showcase the carvers’ creative talents. Many traditional vessels and thus formed his distinctive
of these techniques are highly specialized, and as a style (figure 14).
result jade carvers often devote their whole life to one To illustrate modern jade carving, the authors sta-
technique. Jingui Ma, for example, devoted himself tistically analyzed the 2,464 finalist works from 2012
to the craft of inlay with gems and gold in the way of to 2017 for the Tiangong Awards (Wang and Shi,

44 CHINESE JADE CARVING CRAFTSMANSHIP GEMS & GEMOLOGY SPRING 2020

2020), China’s most prestigious competition. As
shown in figure 15, Qiaose and piercing are the two
most common, particularly the former, followed by
calligraphy and chain carving. Other crafts are more
specialized and mastered by only a few. Chain carv-
ing (figure 16) and eggshell carving (figure 17) are still
applied mainly to vessels; calligraphy engraving
rarely appears alone and is used to provide inscrip-
tions for the jade carvings (figure 18). By and large,
two or three techniques are frequently used in com-
bination. For instance, gem embedding is usually in-
tegrated with gold and silver inlay and eggshell
carving to heighten the exotic appeal (figure 19).
Techniques such as 3D and relief carving are funda-

Figure 15. Among modern jade carving techniques in

China, Qiaose is by far the most common. In this
chart, “Others” refers to works that contain only
basic techniques, such as three-dimensional engrav-
ing or relief carving. Yearly data from Ao et al. (2012–
2017).

Techniques Used in Tiangong Prize–Winning

Artworks, 2012–2017

0.6%

0.9%
18.0%

1.0%

1.4%
3.0% 46.1%

4.4%

24.6%
Figure 16. A modern jade bottle with chains (16 × 5
cm) showing the chain carving and piercing tech-
niques. The shape of the work is neat, delicate, and
elegant, manifesting the carver’s high level of crafts-
manship and ingenuity. Photo courtesy of GAC.
Qiaose Miniature carving

Piercing Gemstone embedding

mental and thus used in almost all pieces, which is
Calligraphy carving Gold and silver inlaying
why they are not mentioned separately in this article
Chain carving Others (Ao et al., 2012–2017).
Although inheriting the ancient craftsmanship,
Eggshell carving
modern jade carving attaches more importance to
personal style and has made breakthroughs in ex-

CHINESE JADE CARVING CRAFTSMANSHIP GEMS & GEMOLOGY SPRING 2020 45

 Figure 17. This
eggshell-thin nephrite
vase with dark green
color by Ting Yu, meas-
uring approximately
34.6 × 11.0 cm, is uni-
form and exquisite. Its
wall is even, thin
(around 0.1 cm), and
transparent, and its
body is decorated with
fine interlocking lotus
branches that lend a
dignified appearance
(Ao et al., 2015). Photos
courtesy of GAC.

pressing it with modern features. For instance, the and carved nephrite with dark green color into wrin-
traditional eggshell carving craft is mostly used in kled paper (Wang and Shi, 2020), which vividly
vessels. But master carver Xi Jiang broke tradition shows the thinness and transparency of the work and

Figure 18. Wei Hu’s

modern nephrite work
“Profiling the Red
Cliff” (7.5 × 8.0 × 1.0
cm) adopts calligraphy
and miniature carving
to present Shi Su’s
prose and resorts to
line-drawing carving to
display the landscape
of the Red Cliff, clearly
demonstrating the
artist’s mastery (Ao et
al., 2016). Photo cour-
tesy of GAC.

46 CHINESE JADE CARVING CRAFTSMANSHIP GEMS & GEMOLOGY SPRING 2020

Figure 19. This jade pot is light, thin, and embedded Figure 20. “To Mountain” (19.6 × 11.0 × 5.2 cm) takes
with gold wire, ruby, and sapphire in an exquisite Is- full advantage of the high toughness of nephrite and
lamic style. It combines white nephrite with embed- uses eggshell carving to show the thin (0.85–2.53 cm)
ded gems, gold inlay, and eggshell carving techniques, and transparent feel of creased paper. Its theme, uti-
fully reflecting the superb craftsmanship. Photo cour- lizing modern art methods, is “diligence as the path
tesy of Jingui Ma Jade Carving Studio. to the mountain of knowledge” (Ao et al., 2013).
Photo courtesy of Xi Jiang Jade Carving Studio.

breaks new ground for the eggshell carving technique Power-Driven Manual Carving. This method inher-
(figure 20). Master Desheng Wu, who excels at figure ited from ancient carving still enjoys a dominant role
carving, incorporates the concept of Western sculp- in modern China (figure 22). The procedure has re-
ture. He pays great attention to the innovation and mained essentially the same (Deng, 2011; Michael-
exploration of artistic style and boldly uses exagger- son et al., 2016). Material selection and examination
ated ways to endow traditional jade carving crafts- are targeted to maximize beauty. Design seeks to in-
manship with modern beauty (figure 21). tegrate the shape and patterns according to the fea-
tures of the jade material (such as color and texture).
JADE CARVING PROCESS ILLUSTRATED BY Carving, usually the most time-consuming step in
MODERN WORKS the process, is where design ideas turn into artwork
As mentioned above, ancient jade carving includes through the use of tools. Polishing refers to the fine
five steps: material selection, examination, design, grinding of a jade’s surfaces to make them smooth
carving with abrasives, and polishing (Xu, 2014). De- and bright, with an aesthetic appeal. It is interesting
pending on the tools and techniques, modern jade to note that artists sometimes use different finishes
carving can be divided into three mainstream meth- (degrees of polish or lack of polish) to create stunning
ods, each with a different procedure. designs.

CHINESE JADE CARVING CRAFTSMANSHIP GEMS & GEMOLOGY SPRING 2020 47

Figure 21. “Snake Girl,” a representative work from Master Desheng Wu’s “Naked Woman” series, combines the
smooth and warm luster of white nephrite jade with a modern stylized design applying openwork and Qiaose
crafts. Photo courtesy of Desheng Wu Jade Carving Studio.

48 CHINESE JADE CARVING CRAFTSMANSHIP GEMS & GEMOLOGY SPRING 2020

Figure 22. These four photos of a bodhisattva illustrate the modern manual carving process, which includes four
steps (left to right): material selection and examination, design, carving, and polishing. Photos by Xiaodong Wang.

Computer Numerical Control (CNC) Engraving. Finally, the program is imported into the CNC ma-
This method is composed of two parts: model design chine to control its operation (figure 23).
and processing by machine. The model is designed The CNC machine can achieve plane carving (fig-
using computer software. The designed model is first ure 24) and 3D carving, but complex three-dimen-
processed and programmed to determine the ma- sional shapes and patterns can only be achieved
chining coordinate system and model size. The cor- through a combination of CNC machine and manual
responding machining tools are then selected, and carving. Moreover, the CNC machine is rather lim-
the program is written according to the machining ited with respect to the special techniques listed
requirements. In China, a CNC engraving and above: Only simple piercing can be achieved. The
milling machine is frequently used in jade carving. eggshell carving process has very specific require-

Figure 23. A three-shaft

CNC machine engrav-
ing tabular nephrite.
The machining process
is monitored by the
computer program. Be-
cause the cutting speed
is extremely high, the
machine needs fluid to
keep cool. Photo by
Mingying Wang.

CHINESE JADE CARVING CRAFTSMANSHIP GEMS & GEMOLOGY SPRING 2020 49

Figure 24. Exquisite details of this dark nephrite ink slab (used to grind ink for Chinese calligraphy) were
processed by a three-shaft CNC machine. Photos by Mingying Wang.

ments for jade materials, and it is difficult to use with to jade carving history. Against the background of
materials with poor toughness that are prone to split- power and tool transformation, the reform and open-
ting. Qiaose, chain carving, and inlay with gold or
gems rely on manual carving and cannot be achieved
with the CNC machine.
Figure 25. This piece of Manas gray-green nephrite is a
1:1 replica of the imperial jade seal with calligraphy
3D Scanning and Auto Carving. The first step of this engraving in Qianlong’s reign (1736–1795). It was
method is to scan the object through the 3D scanning processed by 3D scanning and auto carving. Photo by
device, which results in a 3D model profile. The 3D Mingying Wang.
model profile is then reconstructed using reverse de-
sign software to generate a new model. The coordi-
nates of the 3D model are determined with the CNC
design software, and the tool parameters are com-
piled into a final machining program. Under the guid-
ance of the final program, the CNC machine tool
conducts the 3D carving. This method is mostly used
in the replication of ancient jade relics (figure 25).

THE NATURE OF CHINESE JADE CRAFT’S

EVOLUTION
You (2001) proposed that the fundamental method of
studying ancient jade is to analyze it in relation to its
historical and cultural background. This method ap-
plies to the study of not only ancient jade but also
the modern carving craft and culture. After all, mod-
ern jade carving is integral to the history of jade.
What modern power-driven implements and auto-
mated CNC engraving and milling machines are to
jade carving today is no less than what iron tools are

50 CHINESE JADE CARVING CRAFTSMANSHIP GEMS & GEMOLOGY SPRING 2020

ing up of China, rapid economic development, and formation. Influenced by the evolution of carving
mass consumption, contemporary jade art demon- tools, Chinese jade carving craftsmanship has under-
strates more diverse themes, more open art forms, gone long-term transformation and still continues to
more meticulous craftsmanship (figure 26), and a develop. It has distinctive characteristics—the diver-
wider audience (Wang and Shi, 2020). sity of techniques, the gradual nature of innovation,
The progress of any craft or technology is a and the adoption of foreign culture and craft (You,
process of gradual exploration and long-term accu- 2001; Kong, 2007). Thus the evolution of carving
mulation of experience (Xu, 2014). Analysis of Chi- craftsmanship not only helps to clarify the develop-
nese jade carving proves a basic law—the craft learns ment of China’s jade culture but also provides mean-
from the past and sets a new course for the future ingful enlightenment for the growth of the modern
with heritage and innovation. In general, innovation jade carving industry.
outweighs heritage. Innovation refers to multifaceted
improvement and transformation based on heritage
(You, 2001). Modern jade carving inherits and im- THE FUTURE
proves upon ancient techniques, but more impor- A look at the evolution of Chinese jade carving tells
tantly creates innovations in power-driven and us that every step forward in tools and production
machine engraving tools, contributing to an increase technology is immediately reflected in craftsman-
in efficiency and further development of the art. ship. In the contemporary era, the development of ar-
Thus, modern jade art is the best illustration of her- tificial intelligence and industrial automation will
itage and innovation working in harmony. continue to boost mechanization and large-scale pro-
In Chinese jade carving, the tool plays a funda- duction in the Chinese jade carving industry. These
mental role and is the root of all craftsmanship trans- trends will further contribute to the sustained im-

Figure 26. Left: A nephrite pendant (6.0 × 4.2 cm) with a poem and drawing from the Qing Dynasty. On the front
is a senior, a child, and scenery. The back demonstrates calligraphy engraving, with a traditional form of expres-
sion. Photo from Gu (2010a, p. 1). Right: A modern nephrite pendant “Free Cloud and Crane” (10.0 × 5.2 cm) also
features a carved landscape on the front and calligraphy on the back. Although similar to the traditional form of
expression, it uses a more freehand style that retains the material’s natural color, shaping it into the half red sun
with a touch-of-autumn feel reminiscent of Chinese ink painting (Ao et al., 2015). Photo courtesy of GAC. A
comparison of the details reveals that the modern craft is more delicate, with smoother carving lines, thanks to
motor-driven tools.

CHINESE JADE CARVING CRAFTSMANSHIP GEMS & GEMOLOGY SPRING 2020 51

provement of the creative design and cultural value manual carving, especially of high-end jade wares,
of jade carvings. which represents the inheritance of 5,000-year-plus
In this era of advanced science and technology, au- tradition. The advancement of modern jade carving
tomated CNC engraving has become a trend, espe- can only be achieved through enhancing creativity,
cially for middle- and low-end carvings. In the future, deeply understanding jade culture, and balancing
however, the dominant method will continue to be manual and automated machine methods.

ABOUT THE AUTHORS carvings. They are also grateful for the assistance of Xi Jiang, Xi
Ms. Mingying Wang is a doctoral candidate at the School of Yang, Hongwei Ma, Yiwei Zhai, Desheng Wu, Ting Yu, Guang Yang,
Gemmology, China University of Geosciences in Beijing. Prof. Xiaodong Wang, Yongtao Chen, Shangfeng Xie, Wentao Du, Yue-
Guanghai Shi ([email protected], corresponding author) is a tong Li, Tian Wang, Sisi Zhang, Huan Wang, Maiying Dong, and
professor at the School of Gemmology, and director of the Jour- Rui Zhang, who provided photos and valuable industry information.
nal Center, China University of Geosciences in Beijing. Special appreciations are given to Lin Xu, Tao Hsu, Xiuyan Xu,
Xiaochong Zhang, Xuemei Zhang, and Ying Jiang for their con-
ACKNOWLEDGMENTS structive discussions. The authors would not have been able to fin-
The authors thank Qing Zheng, Xing Tong, and GAC (Gems & Jew- ish this article without the help of these people. The review
elry Trade Association of China) for images of the recently awarded comments by anonymous referees are gratefully appreciated.

REFERENCES
Ao Y., Sun F.M., Zhao Z.L., et al. (2010) Classic Collections of 2009 Co., New York, 118 pp.
China Jade Carving & Stone Carving Works Awarded with Gu F. (2005) The Complete Collection of Jades Unearthed in
Tian Gong Prize. Geological Publishing House, Beijing, 295 pp. China (10—Hubei and Hunan Volume). Science Press, Beijing,
——— (2013) Classic Collections of 2012 China Jade Carving & 240 pp.
Stone Carving Works Awarded with Tian Gong Prize. Geolog- ——— (2010a) Chinese Jades in Traditional Collections (6—Qing
ical Publishing House, Beijing, 378 pp. Dynasty). Science Press, Beijing, 256 pp.
——— (2014) Classic Collections of 2013 China Jade Carving & ——— (2010b) Chinese Jades in Traditional Collections (7—Qing
Stone Carving Works Awarded with Tian Gong Prize. Geolog- Dynasty). Science Press, Beijing, 256 pp.
ical Publishing House, Beijing, 366 pp. Hansford S.H. (1950) Chinese Jade Carving. Lund Humphries Co.,
——— (2015) Classic Collections of 2014 China Jade Carving & London, 145 pp.
Stone Carving Works Awarded with Tian Gong Prize. Geolog- Ji L. (1984) Crafts and Arts in Contemporary China. China Social
ical Publishing House, Beijing, 402 pp. Sciences Press, Beijing, 650 pp.
——— (2016) Classic Collections of 2015 China Jade Carving & Kong F.A. (2007) The Study of Technology of Chinese Ancient Jade
Stone Carving Works Awarded with Tian Gong Prize. Geolog- Producing. PhD Thesis, Shanxi University, Taiyuan, 332 pp.
ical Publishing House, Beijing, 354 pp. Lu J.F., Fang X.M., Zhou X.J. (2014) General History of Chinese
——— (2017) Classic Collections of 2016 China Jade Carving & Jade (the Northern Volume of the Neolithic Age). Haitian
Stone Carving Works Awarded with Tian Gong Prize. Geolog- Press, Shenzhen, 303 pp.
ical Publishing House, Beijing, 340 pp. Lu J.F., Yu Y.J., Fang G. (2014) General History of Chinese Jade
——— (2018) Classic Collections of 2017 China Jade Carving & (the Xia & Shang Dynasties). Haitian Press, Shenzhen, 336 pp.
Stone Carving Works Awarded with Tian Gong Prize. Geolog- Lu J.F., Ou-yang M.Y. (2014) General History of Chinese Jade (the
ical Publishing House, Beijing, 400 pp. Zhanguo Period). Haitian Press, Shenzhen, 358 pp.
Bao Y.Z. (2008) Research on the Situation and Inheritance of the Lu J.F., Li Y.D. (2014) General History of Chinese Jade (the Qin &
Chinese Miniature Sculpture Art. Master’s Thesis, The Central Han Dynasties). Haitian Press, Shenzhen, 392 pp.
University of Nationalities, Beijing, 44 pp. Ma G.Q. (2014) Xinjiang Hetian Placer Nephrite (White Jade)
Barnes G.L. (2018) Understanding Chinese jade in a world context. Grading Standards and Illustrations. Xinjiang People’s Press,
Journal of the British Academy, Vol. 6, pp. 1–63, http://dx.doi.org/ Urumqi, 299 pp.
10.5871/jba/006.001 Markbreiter S. (1985) Jade carving in two cities. Arts of Asia, Vol.
Deng S.P. (2007) Exquisite Beauty: Islamic Jades. National Palace 15, No. 1, pp. 63–73.
Museum, Taipei, 287 pp. Michaelson C., Sax M., Wu H. (2016) The renaissance of jade carv-
——— (2011) Art in Quest of Heaven and Truth: Chinese Jades ing in China today. Arts of Asia, Vol. 46, No. 3, pp. 61–71.
through the Ages. National Palace Museum, Taipei, 172 pp. Niu Q.B. (2015) Jade carving craftsmanship during the Neolithic
Desautels P.E. (1986) The Jade Kingdom. Van Nostrand Reinhold Age and the Shang Dynasty. Journal of Nanyang Normal Uni-

52 CHINESE JADE CARVING CRAFTSMANSHIP GEMS & GEMOLOGY SPRING 2020

 versity (Social Sciences), Vol. 14, No. 11, pp. 46–51. The Forbidden City Publishing House, Beijing, pp. 324–338.
Read P. (1981) Travels in China. Canadian Jeweller, June, pp. 134– ——— (2014) Technology of jade-making is helpful to appraisal of
135, 137. ancient jade. Collectors, No. 8, pp. 28–34.
Sax M., Meeks N.D., Michaelson C., Middleton A.P. (2004) The Xu X.D. (2013) Classics of the Forbidden City (Jade in the Collec-
identification of carving techniques on Chinese jade. Journal tion of the Palace Museum). The Forbidden City Publishing
of Archaeological Science, Vol. 31, No. 10, pp. 1413–1428, House, Beijing, 300 pp.
http://dx.doi.org/10.1016/j.jas.2004.03.007 Yang B.D. (1982) Court jades of Qing Dynasty. Palace Museum
Technical Research Group of Beijing Jade Factory (1976) Some Journal, No. 1, pp. 49–61, 101–104.
ideas on jade carving craftsmanship in Shang Dynasty. Archae- ——— (1989a) Aspects of Chinese ancient jadeware. Palace Mu-
ology, No. 4, pp. 229–233, 286–287, 290. seum Journal, No. 1, pp. 32–48, 99–100.
Thomas S.A., Lee H.W. (1986) Gemstone carving in China: Winds ——— (1989b) Aspects of Chinese ancient jadeware. Palace Mu-
of change. G&G, Vol. 22, No. 1, pp. 24–34, http://dx.doi.org/ seum Journal, No. 2, pp. 31–40, 97–98.
10.5741/GEMS.22.1.24 ——— (2004) Discussion on characteristics of the prehistoric jade
Tucker E. (1982) Jade forms from ancient China. G&G, Vol. 18, carving and their differences. Archaeology, No. 10, pp. 62–68.
No. 1, pp. 20–31, http://dx.doi.org/10.5741/GEMS.18.1.20 ——— (2006) A review of jade carving tools. Yang Boda’s Views
Wang M.Y., Shi G.H. (2020) Characteristics of modern Chinese on Jade—Ba Zhi Selected Works. The Forbidden City Publish-
jade carvings: The nephrite jade works that won the Tiangong ing House, Beijing, 575 pp.
Awards from 2003 to 2014. Arts of Asia, Vol. 50, No. 3, pp. 88– Yang B.D., Qi J., Zhao G.D., et al. (2005) Chinese Jade Complete
102. Works. Hebei Fine Arts Publishing House, Shijiazhuang, Vol.
Wen G., Jing Z.C. (1992) Chinese Neolithic jade: A preliminary 1, 316 pp.
geoarchaeological study. Geoarchaeology, Vol. 7, No. 3, pp. Yi F. (2001) A Brief Chinese Chronology. The Heritage Press, Bei-
251–275, http://dx.doi.org/10.1002/gea.3340070304 jing, 267 pp.
Xu L. (2011) Chinese Ancient Jade-Making Process. The Forbidden You R.D. (2001) An Introduction to Chinese Ancient Jade. The
City Publishing House, Beijing, 269 pp. Forbidden City Publishing House, Beijing, 367 pp.
——— (2012) An overview of jade carving technology in ancient Zhao Y.K., Zhang J.M. (2000) Chinese Jade Carving Technology.
China—A preliminary comparison with the jade culture and Beijing Industrial Art Publishing House, Beijing, 481 pp.
jade carving technology in ancient Central America. Essence Zhu J.J. (2007) Classics of the Forbidden City (The Forbidden City
of Nature – Civilization of Ancient Jade in China and Mexico. Treasure). The Forbidden City Publishing House, Beijing, 247 pp.

For online access to all issues of GEMS & GEMOLOGY from 1934 to the present, visit:

gia.edu/gems-gemology

CHINESE JADE CARVING CRAFTSMANSHIP GEMS & GEMOLOGY SPRING 2020 53

FEATURE ARTICLES

COLOR CHARACTERISTICS OF BLUE TO

YELLOW BERYL FROM MULTIPLE ORIGINS
Yang Hu and Ren Lu

Aquamarine and heliodor are colored by Fe ions, an important coloring agent for beryl. Blue to yellow gem
beryl was studied by quantitative spectroscopy and trace-element analytical techniques to explore color char-
acteristics and chromophores. Blue color was caused by a 600 nm absorption, while yellow color was attributed
to an absorption edge in the violet-blue region. Color ranged from blue to green to yellow due to different pro-
portions of Fe ions with various valences and occupancies. Mn content was positively related to Fe, but abundant
Mn ions showed no impact on color (unlike Mn in morganite and red beryl). The arrangement of alkali ions and
water in channel and the charge compensation mechanism of beryl are discussed. Alkali ions (mainly Na and
Cs) and water were localized in the peanut-shaped channels, and all alkali elements (Li, Na, K, Rb, and Cs) were
relevant. Though alkali ions and water interacted with transition metal Fe and Mn ions, their influence on blue
to yellow color was indirect and rather weak.

G
em beryl is a significant gem species, includ- MATERIALS AND METHODS
ing color varieties such as emerald, aquama- Beryls from different origins were gathered and 14 of
rine, heliodor, goshenite, morganite, and red them with various color and alkali content were se-
beryl. Blue to yellow beryl has been found in numer- lected for this study (see table 1). They were classified
ous locations, including Brazil, South Africa, Russia, in the following color varieties: goshenite (colorless to
Ukraine, Canada, Myanmar, the United States, near-colorless), aquamarine (greenish blue to blue),
Afghanistan, and China (Belakovskiy et al., 2005). green beryl (green to yellowish green), and heliodor
Blue color in aquamarine and yellow color in he- (greenish yellow to yellow). With the exception of two
liodor are attributed to abundant Fe ions (Wood and faceted stones and one rough stone, the samples were
Nassau, 1968). Fe ions are also present in all other
color varieties of beryl, though Fe content is rela-
tively low in morganite. Although discussions on the
role of Fe ions in blue to yellow beryl are not new, In Brief
they have mainly focused on crystal physics and • Color in beryl ranges from blue to green to yellow due
chemistry. This article explores the color character- to different proportions of Fe-related absorption.
istics and chromophore ions of blue to yellow beryl • The color of green beryl can also come from Cr3+
using quantitative chemical and spectral analysis. and/or V3+ ions.
The crystal structure of beryl is unique for having
• Alkali elements and water in beryl were found to play
a peanut-shaped “channel” along the c-axis, and al- complex roles, but their influence on blue to yellow
kali ions in this channel interact with transition color is indirect and likely weak.
metal ions. Therefore, we will discuss the features of
alkali elements and their roles in beryl color, based
on analysis of the channel mechanism. This research
was part of a series of ongoing studies on the color fabricated as optical wafers perpendicular (PK-7 and
characteristics of beryl. PK-8) or parallel (PK-5, PK-10, PK-9, RUS-8, MOZ-2,
BM-1, MOZ-1, AF-3, and AF-2) to the c-axis (figure 1).
All samples were investigated by standard gemo-
logical testing, Raman spectroscopy, Fourier-trans-
See end of article for About the Authors and Acknowledgments.
GEMS & GEMOLOGY, Vol. 56, No. 1, pp. 54–65,
form near-infrared (FT-NIR) and ultraviolet/visible/
http://dx.doi.org/10.5741/GEMS.56.1.54 near-infrared (UV-Vis-NIR) spectroscopy, and laser ab-
© 2020 Gemological Institute of America lation–inductively coupled plasma–mass spectrome-

54 COLOR CHARACTERISTICS OF BLUE TO YELLOW BERYL GEMS & GEMOLOGY SPRING 2020
 PK-7
PK-5 PK-10 PK-8

BRA-3
PK-9
BM-1
MOZ-2
RUS-8

AF-3 AF-2 UK-11 UK-10

MOZ-1

Figure 1. The 14 studied beryl samples (0.63–3.50 ct) from various geographic origins. Color ranged from blue to
yellow, as well as near-colorless and colorless. Photo by Yang Hu.

try (LA-ICP-MS) chemical analyses. Inclusions were rate. Each analysis incorporated a background acqui-
captured using a Leica M205A microscopic system sition time of approximately 20–30 seconds followed
with oblique fiber-optic illumination. Raman spectra by 50 seconds of ablation. A multi-standard quanti-
were collected by a Bruker Senterra R200 spectrome- tative calculation method was adopted, with Al cho-
ter coupled with a 532 nm laser for identifying various sen as the normalizing element. Calibration
inclusions. The resolution was set at 5 cm–1 with a 20 standards of NIST 610, BCR-2G, BHVO-2G, and BIR-
second integration time, 2 accumulations, and 20 mV 1G were used as external references (Liu et al., 2008).
laser energy. To explore the characteristics of water in Three laser spots for each sample were applied in an
the beryl, FT-NIR was performed using a Bruker V80 area that was typically clean with an even color dis-
FTIR spectrometer at 2 cm–1 resolution and 32 accu- tribution. UV-Vis-NIR spectra were collected in the
mulations. To study the color features, UV-Vis-NIR same area analyzed by LA-ICP-MS.
spectra were recorded with a PerkinElmer 650s spec-
trophotometer equipped with a 150 nm integrating RESULTS
sphere accessory at 1 nm resolution. Gemological Properties. The beryl samples had a re-
Chemical analysis was performed by LA-ICP-MS fractive index range of ne=1.568–1.579 and
using an Agilent 7500a and 7900 ICP-MS instrument no=1.573–1.586, with birefringence between 0.005 and
combined with a GeoLas 193 nm laser. The carrier 0.006, with the exception of samples BM-1 and BRA-
gas used in the laser ablation unit was He with a flow 3. Pale green MOZ-1 had the lowest RI (1.568–1.573)
rate set at approximately 650 mL/min. Laser ablation among all the samples. Burmese deep blue sample
conditions consisted of a 44 μm diameter laser spot BM-1 and Brazilian dark greenish blue BRA-3 had the
size, a fluence of 5–6 J/cm2, and a 6–8 Hz repetition highest RI (1.589–1.600) and birefringence (0.009–

COLOR CHARACTERISTICS OF BLUE TO YELLOW BERYL GEMS & GEMOLOGY SPRING 2020 55
A B

Figure 2. Two-phase in-

clusions in blue to yel-
low beryl samples. A:
Fingerprint-like two-
phase inclusions along
a a healed fissure plane
in aquamarine. B:
Hexagonal two-phase
100 μm b 100 μm
inclusions viewed
C D down the c-axis in
aquamarine, consistent
with the crystallo-
graphic symmetry. C
and D: Isolated elon-
gated rod-like two-
phase inclusions
parallel to the c-axis in
heliodor. Photomicro-
c c graphs by Yang Hu.

a a
100 μm 100 μm

0.010). All the beryl samples displayed weak to mod- Microscopic Observation. Two-phase inclusions were
erate dichroism. All bluish beryl samples showed the most common type in the beryl samples, contain-
more blue color along the e-ray than the o-ray. It ing one gas bubble (CO2) floating in at least one kind
should be noted that the dark hue of sample BRA-3 of fluid (figure 2). Minor CH4, H2S, and N2 were found
was caused by extremely tiny cloudy dark inclusions only in some of the Pakistani beryl samples. Normally
throughout nearly the whole sample. The lemon and the Pakistani samples had two liquid phases (CO2 and
yellow heliodor samples showed weak dichroism but water with minor dissolved CO2), while samples from
a relatively saturated color. other deposits hosted one liquid phase (water). Some-

TABLE 1. Blue to yellow beryl samples selected for this study.

Sample no. Variety Geographic origin Color

PK-7 Goshenite Shigar Valley, Pakistan Near-colorless
PK-5 Goshenite Shigar Valley, Pakistan Near-colorless
PK-10 Goshenite Shigar Valley, Pakistan Colorless
PK-8 Aquamarine Shigar Valley, Pakistan Near-colorless
PK-9 Aquamarine Shigar Valley, Pakistan Pale greenish blue
BRA-3 Aquamarine Brazil Dark greenish blue
RUS-8 Aquamarine Russia Greenish blue
MOZ-2 Aquamarine Mozambique Blue
BM-1 Aquamarine Mogok, Myanmar Deep blue
MOZ-1 Green beryl Mozambique Pale green
AF-3 Green beryl Africa Yellowish green
AF-2 Heliodor Africa Greenish yellow
UK-11 Heliodor Unknown Lemon yellow
UK-10 Heliodor Unknown Golden

56 COLOR CHARACTERISTICS OF BLUE TO YELLOW BERYL GEMS & GEMOLOGY SPRING 2020
 RAMAN SPECTRA
Figure 3. This multi-
223 (S)
fluid II
phase inclusion had
two fluid phases, one
S O
crystal
gas gaseous phase, and one
O 474 (S) crystal phase. Abbrevi-
O 686 (B)
fluid I 3606 (H1) ations: B—beryl, S—
INTENSITY (a.u.)

1388 (C1) 100 μm

sulfur, O—orpiment,
1243 (C2) C1—CO2 in the inclu-
crystal: sulfur+orpiment
1284 (C1) sion, C2—CO2 in the
3200–3500 (H2)
fluid I: water beryl structure channel,
H1—water in the beryl
gas: CO2+CH4 2913 (CH4)
structure channel, and
fluid II: CO2 H2—water in the inclu-
B
B B B B B
B sion. Spectra are offset
aquamarine (E⊥c)
vertically for clarity.
0 500 1000 1500 2000 2500 3000 3500
Photomicrograph by
Yang Hu.
RAMAN SHIFT (cm–1)

times a mineral phase was hosted in the fluid, such as disappeared after heating above 400°C. Raman analy-
native sulfur and orpiment in Pakistani samples (see sis of this two-phase inclusion detected no gas or fluid.
figure 3). Two-phase inclusions had hexagonal, elon- We concluded that micro-cracks occurred after heat-
gated, round, oval, angular, needle-like, or other irreg- ing above 400°C, allowing the gaseous and fluid phase
ular shapes. They were generally distributed along to escape and leaving the empty two-phase inclusion
healed fissure planes forming “fingerprint-like” inclu- (recognizable by its original shape and distribution).
sions or isolated (again, see figure 2). Yellowish and greenish beryl are usually heated be-
Our heating experiments indicated that two-phase tween 400° and 500°C to obtain a blue color, so this
inclusions transformed after heat treatment at 400°C empty two-phase inclusion could potentially be used
or higher. Before heat treatment, a round bubble was as evidence of heat treatment of aquamarine.
floating on the fluid phase (figure 4, left). While cooling Various mineral inclusions could be detected by
the sample to room temperature after heating to 300° Raman spectroscopy. Tourmaline, albite, muscovite,
and 400°C, we photographed the same two-phase in- garnet, zircon, and tantalite-columbite were typical
clusion (figure 4, middle and right). The round bubble in our samples (figure 5). Argentojarosite was identi-

Figure 4. A two-phase inclusion before and after heat treatment at 300° and 400°C. The gaseous and fluid phases
disappeared after heat treatment above 400°C. Left: Unheated. Center: After heating to 300°C and cooling to room
temperature. Right: After heating to 400°C and cooling to room temperature. Photomicrographs by Yang Hu.

100 μm 100 μm 100 μm

COLOR CHARACTERISTICS OF BLUE TO YELLOW BERYL GEMS & GEMOLOGY SPRING 2020 57
A B
Figure 5. Various min-
eral inclusions in the
beryl samples. A: Yel-
low argentojarosite in-
side needle-like growth
tubes along the c-axis of
a Pakistani aquama-
rine. White albite was
present at the end of
growth tubes. B: Brown
1 mm 1 mm
butterfly-like tantalite-
columbite in a Pak-
C D istani aquamarine. C:
Hexagonal brownish
tantalite-columbite in a
Brazilian aquamarine.
D: Bertrandite (red rec-
tangle) and siderite
(yellow rectangle) inclu-
sions in an aquamarine
from Mozambique.
Photomicrographs by
Yang Hu.
100 μm 100 μm

fied only in Pakistani aquamarine, while bertrandite Burmese aquamarine BM-1, K content (72–219 ppmw)
and siderite were recognized only in an aquamarine was the lowest rather than Rb (1375–1459 ppmw). The
sample from Mozambique. concentration of Li was less than Na in all of the sam-
ples. Nearly all samples had total alkali content (in-
Trace Element Analysis. The alkali element and tran- cluding Li, Na, K, Rb, and Cs) below approximately
sition element contents (in ppmw) of all beryl samples 6000 ppmw. Aquamarine MOZ-2 and goshenite PK-
from multiple localities are reported in table 2. 10 had the lowest total alkali content (1250–1350
ppmw on average). The alkali content of deep blue
Alkali Elements. Among the alkali elements, Cs and Burmese aquamarine BM-1 and dark blue Brazilian
Na concentrations were relatively enriched and Rb aquamarine BRA-3 differed noticeably from those of
content was lowest in nearly all samples (table 2). In other samples. Li, Rb, and Cs contents were more than

TABLE 2. LA-ICP-MS chemical composition analysis of beryl samples.

Trace Goshenite Goshenite Goshenite Aquamarine Aquamarine Aquamarine Aquamarine Aquamarine Aquamarine
elements PK-7 PK-5 PK-10 PK-8 PK-9 BRA-3 RUS-8 MOZ-2 BM-1
(ppmw)
Pakistan Pakistan Pakistan Pakistan Pakistan Brazil Russia Mozambique Myanmar
Li 290–357 1168–1239 68–75 292–328 73–76 5633–5745 270–272 136–145 6244–6497
Na 1103–1295 3845–3970 827–843 1099–1147 1860–2039 8988–9073 1295–1425 1031–1098 7857–8102
K 94–114 118–137 54–69 83–86 434–452 1036–1067 16–21 37–44 72–219
V 1–5 1–2 0 4–5 30–33 17–18 0 4–5 82–84
Cr 0–14 0–1 0–3 0–4 1–6 3–5 0 0–1 0
Mn 28–30 52–56 17–18 19–51 19–23 287–292 23–25 7–9 568–854
Fe 1745–1826 1459–1500 1355–1414 1802–1950 4498–5080 9306–9407 5466–5703 5542–5554 6709–7668
Rb 37–42 42–44 9–10 34–37 52–63 662–671 7–9 10–11 1375–1459
Cs 2261–2861 622–676 313–333 2229–3216 1366–1414 17789–18005 184–186 133–135 46510–47496
Data analyzed by LA-ICP-MS from three spots per sample. Data reported in minimum and maximum values. Detection limits (ppmw): Li = 0.36, Na = 7.43,

58 COLOR CHARACTERISTICS OF BLUE TO YELLOW BERYL GEMS & GEMOLOGY SPRING 2020
 FT-NIR SPECTRA
Figure 6. FT-NIR spec-
tra of beryl wafers par-
allel to the c-axis. The
ABSORPTION COEFFICIENT (cm–1)

15 5275-II/5271-I peak intensity of type II

water at 7097, 7077,
7142-I
and 5275 cm–1 became
stronger (from top to
5109-I 5436-I 6816-I 8699-I
10
PK-10: bottom) with increas-
827–843 ing sodium content.
7097-II
The spectra are dis-
placed vertically by
PK-5:
3845–3970 y+4.5 cm–1 (PK-5) and
5
7077-II y+9 cm–1 (PK-10) for
clarity. Note: Na con-
BM-1: tent analyzed by LA-
7857–8102 ICP-MS is listed in
5000 6000 7000 8000 9000 ppmw.
WAVENUMBER (cm–1)

10 times higher, and Na and K were 3–10 times higher Mn content was much lower than Fe in all sam-
(table 2). BM-1 had the richest Cs content (46510– ples (table 2). The concentration of Mn varied be-
47496 ppmw) reported for blue aquamarine. tween 4 and 854 ppmw, and it was below 200 ppmw
in most samples. Burmese deep blue beryl BM-1 and
Transition Metals. Concentration of Fe varied from Brazilian dark greenish blue sample BRA-3 had rela-
1355 to 9407 ppmw for all colorless and blue to yel- tively rich Mn (187–854 ppmw). The concentration
low beryl samples (table 2). Among the samples, the of V and Cr in bluish and yellowish beryl samples
colorless and near-colorless beryl (PK-10, PK-7, PK- was quite low to even below detection limits (table
8, and PK-5) had relatively low Fe (1355–1950 ppmw). 2). But these concentrations were relatively rich in
Other saturated bluish beryl contained higher Fe con- green beryl MOZ-1 (which averaged 262 ppmw V and
279 ppmw Cr). Besides alkali and transition ele-
tent, even up to 9407 ppmw. The lemon yellow and
ments, traces of Mg, Sc, Zn, Ga, Sn, and Ta were de-
golden-colored heliodor (UK-10 and UK-11) had Fe
tected in some samples.
content between 2030 and 3812 ppmw, lower than
bluish beryl samples of similar saturation. Spectroscopy Analysis. FT-NIR Spectroscopy. The type
of water and its relative content in the channel was re-
vealed quantitatively by its near-infrared absorption
spectrum. Peaks in the NIR region represented the
Green beryl Green beryl Heliodor Heliodor Heliodor overtones and combinations of “type I” and/or “type
MOZ-1 AF-3 AF-2 UK-11 UK-10 II” water absorption. Most beryl samples contain “type
Mozambique Africa Africa Unknown Unknown I” water with peaks at 8699, 7142, 6816, 5436, 5271,
70–76 143–149 134–140 162–174 135–162
and 5109 cm–1 (Wood and Nassau, 1967), such as PK-
10 and PK-5 in figure 6. But Brazilian aquamarine BRA-
508–627 2507–2611 1625–1797 804–855 1204–1443
3 and Burmese aquamarine BM-1 were dominated by
2–36 292–297 162–195 51–66 153–175 strong “type II” water peaks at 7097, 7077, and 5275
257–268 9–10 7–14 0 0 cm–1 (Wood and Nassau, 1967) (blue trace in figure 6).
267–292 0 0–2 0–4 2–4 Also, the peak intensity of type II water became
4–7 95–96 116–160 36–43 49–57 stronger with increasing sodium content.
2698–2841 6899–7055 6292–8217 2030–2203 3225–3812
11–14 169–175 124–145 39–46 34–61
UV-Vis-NIR Spectroscopy. Quantitative UV-Vis-NIR
spectra for the e-ray and o-ray were acquired for the
300–370 1606–1630 1791–2305 1952–2092 510–689
beryl samples. The color circles of the beryl samples
K = 11.22, V = 0.07, Cr = 0.88, Mn = 0.46, Fe = 12.65, Rb = 1.66, Cs = 8.22
in figures 7–10 were calculated from visible spectra

COLOR CHARACTERISTICS OF BLUE TO YELLOW BERYL GEMS & GEMOLOGY SPRING 2020 59
 UV-VIS-NIR SPECTRA

ABSORPTION COEFFICIENT (cm–1)

between 380 and 780 nm using GRAMS software,
14
when the sample thickness was normalized to 4 mm.
In order to discuss the impact of Fe on coloration, we 12

converted ppmw to ppma in the UV-Vis-NIR region 10

AF-2 UK-11 UK-10
by the following formulation: ppma = [(atomic weight 8 Fe: 2088–2727 Fe: 674–731 Fe: 1070–1265
of beryl/number of atoms for beryl)/atomic weight of
6
Fe] ∙ ppmw value = [(537.49/29)/55.85] ∙ ppmw value.
That is, ppma value = 0.3318 ppmw value. 4

All blue to yellow beryl samples showed the typ- 2

ical iron absorption in their UV-Vis-NIR spectra. 0

Peaks at 372 and 427 nm related to Fe3+ ions (Wood 300 400 500 600 700 800

and Nassau, 1968) were observed in all samples (fig- WAVELENGTH (nm)
ures 7–9). These 372 and 427 nm peaks were more or
less covered by the absorption edge in the violet-blue Figure 7. Unoriented UV-Vis-NIR spectra of greenish
region in the yellowish beryl samples. They were yellow (AF-2), lemon yellow (UK-11), and golden
even invisible in lemon yellow and golden heliodor (UK-10) heliodor samples. When the absorption
(UK-10 and UK-11) (figure 7). These 372 and 427 nm edge shifted to a longer wavelength from the violet-
peaks did not contribute to color because of their blue to the blue region, the color turned more yel-
rather weak absorption. lowish. The spectra are displaced vertically by y+2
The absorption edge in the violet-blue region leads cm–1 (UK-11) and y+4 cm–1 (UK-10) for clarity. Note:
to yellowish color in beryl. There was no obvious dif- Fe content analyzed by LA-ICP-MS is listed in ppma.
ference between the o-ray and the e-ray. This absorp-
tion edge was assigned to Fe3+ ions (Wood and Nassau, low beryl AF-2 had the richest total Fe content (2088–
1968). The absorption edge was present in all beryl sam- 2727 ppma) of the three yellowish samples, but the in-
ples, and it was in the ultraviolet region for beryl sam- tensity of the absorption edge was weak compared with
ples without yellowish color. When this absorption other yellow heliodor samples (see figure 7).
edge shifted to a longer wavelength from the violet-blue Bluish color in the beryl samples resulted from a
to the blue region, the yellow color occurred in beryl. distinct broad absorption around 600 nm in the e-ray.
Also, the intensity of this absorption edge was not pro- The 600 nm broad absorption band was present in
portional to total Fe content. For example, greenish yel- the e-ray but quite weak in the o-ray (see figures 8

UV-VIS-NIR SPECTRA
e-ray
14

Figure 8. UV-Vis-NIR
ABSORPTION COEFFICIENT (cm–1)

12 spectra of aquamarine
and goshenite samples
10
with distinguishing Fe
content in e-ray. With
MOZ-2 BM-1
Fe: 1839–1843 Fe: 2227–2545
increasing Fe content,
8
the absorption band at
600 nm became
PK-10 PK-9 600
6
Fe: 450–469 Fe: 1493–1686 stronger and the blue
color became saturated
4 (from bottom to top).
Note: Fe content ana-
2
lyzed by LA-ICP-MS is
listed in ppma.

300 400 500 600 700 800

WAVELENGTH (nm)

60 COLOR CHARACTERISTICS OF BLUE TO YELLOW BERYL GEMS & GEMOLOGY SPRING 2020
 UV-VIS-NIR SPECTRA

o-ray
14
ABSORPTION COEFFICIENT (cm–1)

12 Figure 9. UV-Vis-NIR
spectra of aquamarine
MOZ-2
10 Fe: 1839–1843 820 and goshenite samples
with distinguishing Fe
PK-9
8
Fe: 1493–1686
content along the o-ray.
372 The band intensity at
6 PK-10 820 nm was not propor-
Fe: 450–469
tionate to the Fe con-
4
tent. Note: Fe content
427
BM-1 analyzed by LA-ICP-MS
Fe: 2227–2545
2
is listed in ppma.

300 400 500 600 700 800

WAVELENGTH (nm)

and 9), which was responsible for more blue color aquamarine with similar saturation should contain
along the e-ray than the o-ray. This absorption was at least twice the Fe content.
associated with Fe2+ or Fe2+-Fe3+ ion pairs (Wood and A broad absorption band centered at about 820 nm
Nassau, 1968; Lin et al., 2013). With increasing Fe was stronger in the o-ray than the e-ray (see figures 8
content, the 600 nm absorption band became and 9). This 820 nm broad band overlapped with the
stronger and the color became more saturated, from 600 nm band in the e-ray. This band in the o-ray was
colorless, pale greenish blue, and medium blue to related to Fe2+ (Wood and Nassau, 1968), and its in-
dark blue (from bottom to top in figure 8). Compared tensity had nothing to do with total Fe content. For
with Fe content in yellow heliodor with similar sat- example, the 820 nm band intensity of Burmese aqua-
uration, Fe ions were less efficient for coloring beryl marine BM-1 was quite weak, contrary to the abun-
blue than yellow: 600 ppma Fe in beryl could produce dant Fe content (see figure 9). However, the intensity
a relatively saturated yellow color. However, blue of the 820 nm band in the e-ray was not in propor-
tional to that in the o-ray. And the intensity of the
820 nm band in the e-ray was nearly positive to the
Figure 10. The UV-Vis-NIR spectra of green beryl total Fe content (see figure 8). So we supposed that
MOZ-1 showed typical absorption of Cr3+ and minor the origin of the 820 nm band was different between
V3+ ions, as well as Fe ions at 372 and 820 nm. the e-ray and o-ray, maybe Fe2+ in different occupancy.
The tail of the 820 nm band reached the visible
UV-VIS-NIR SPECTRA range, but its intensity in the visible range was quite
weak. So this 820 nm broad absorption could not lead
ABSORPTION COEFFICIENT (cm–1)

to any obvious color.

2.5 Pale green beryl MOZ-1 showed typical absorp-
820 tion of Cr3+ at 426 and 600 nm for the o-ray and 630
2.0
and 660 nm for the e-ray, as well as sharp R lines at
1.5 683 nm (Wood and Nassau, 1968). A weak V3+ absorp-
e-ray o-ray
1.0
tion shoulder was also observed at about 400 nm (see
426
600 683
figure 10). So this green color originated from Cr3+
372
0.5 and minor V3+ ions in beryl, similar to emerald. This
absorption was totally different from yellowish green
300 400 500 600 700 800 beryl AF-3 (no absorption spectra shown for this sam-
WAVELENGTH (nm) ple), which showed the absorption features of Fe ions,
though they both had greenish color.

COLOR CHARACTERISTICS OF BLUE TO YELLOW BERYL GEMS & GEMOLOGY SPRING 2020 61
 A B
0.28 nm

O (Si–O)
O (Al–O/Be–O)
Si
Al
Be

0.46 nm
2a Be
Al
Si
Na
H2O

2b

c
a
c

a b

0.51 nm

Figure 11. Diagrams of the unique peanut-shaped channel structure in beryl, and the internal arrangement of alkali
ions and water, viewed as “ball-wire” down into the a-axis (with a 120° angle from the b-axis, which is projected onto
the page from behind as a dashed line) (A) and as “polyhedra” down into the c-axis (B). Smaller rings of SiO4 tetrahe-
dra stack over larger rings of alternating BeO4 tetrahedra and AlO6 octahedra forming peanut-shaped enclosures. Large
alkali ion Cs+ or water would only be localized at the larger 2a site and unable to migrate along the channel. The small-
est ion Na+ fits suitably at the 2b site, interacting with “type II” water. The hydrogen-hydrogen direction is parallel to
the c-axis in “type I” water and perpendicular to the c-axis in “type II” water. Selected atoms in figure A were omitted
to better show internal details. Relevant parameters are from Artioli et al. (1993) and Wood and Nassau (1968).

DISCUSSION the 2b site, Na+, are stable because of Coulomb inter-

Crystal Structural Consideration. The crystal structure actions between Na+ and adjacent type II water mole-
of beryl, with an ideal chemical formula Be3Al2(Si6O18), cules (see again figure 11A).
consists of smaller rings of six SiO4 tetrahedra and Substitutions commonly occur in Be-O tetrahedra
larger rings of alternating BeO4 tetrahedra and AlO6 oc- by Li+ ions, and in Al-O octahedra by transition metal
tahedra (figure 11B). These two rings stack over each ions and other ions (such as Mg, Ca, and Ti). With Li+
other along the c-axis, forming a peanut-shaped struc- substituting for Be2+, slight deformation of Be-O tetra-
tural “channel” along the c-axis (figure 11A). Large ions hedra and charge deficiency arise. Substitution of di-
such as Cs or water molecules would only fit within valent cations in Al octahedra or trivalent cations in
the large ring and be therefore localized at the 2a site Be tetrahedra could also lead to charge deficiency. Al-
and unable to migrate along the channel, because the kali ions (mainly Cs+ and Na+) in the channel serve to
ionic diameter of Cs+ (0.376 nm) and the dimension of charge balance (e.g., Li+ + transition metal ions + other
a water molecule (0.28 × 0.32 × 0.37 nm) are too large substituted ions = Cs+ + Na+). Moreover, the substitu-
to fit within the 2b site. Water molecules in beryl occur tion of transition metal ions mainly included
in two configurations: “type I” with H-H direction par- Fe2+/Fe3+/Mn2+/Mn3+/Cr3+/V3+<=>Al3+, or in some cases
allel to the c-axis, and “type II” with H-H direction per- Fe2+/Fe3+<=>Be2+.
pendicular to the c-axis (Wood and Nassau, 1967).
Water molecules are only free to rotate around the c- Alkali Elements. The relative concentrations of each
axis according to the hexagonal system (Gorshunov et alkali element were nearly proportional in all beryl
al., 2016). Due to similar ionic radii and chemical prop- samples, showing similar variation (figure 12). This
erties, minor K+ and Rb+ ions in the channel behave was in agreement with the charge balance principle, in
like Cs+ ions. The smallest alkali ions coordinated in which the charge deficiency arising from the substitu-

62 COLOR CHARACTERISTICS OF BLUE TO YELLOW BERYL GEMS & GEMOLOGY SPRING 2020
 Goshenite, PK-7, Pakistan Heliodor, AF-2, Africa
Aquamarine, PK-8, Pakistan Heliodor, UK-11, Unknown
Aquamarine, BM-1, Myanmar Heliodor, UK-10, Unknown
100000 Aquamarine, MOZ-2, Mozambique Green beryl, AF-3, Africa
Aquamarine, BRA-3, Brazil Green beryl, MOZ-1, Mozambique
Aquamarine, RUS-8, Russia
Figure 12. The concen-
tration of five alkali
CONTENT (ppmw)

10000
elements for all beryl
samples from different
locations showed a
1000 similar proportional
tendency. A log scale
was used for the y-
100 axis.

10

Li Na K Rb Cs

tion of Li are compensated by a proportional amount in figure 13) and the “high-alkali beryl” area (blue cir-
of Na and Cs plus minor K and Rb in the channel. But cle in figure 13). Low-alkali beryl was dominated by
there were also some inconsistencies, especially the type I water and high-alkali beryl by type II water an-
unexpectedly low K content in Burmese aquamarine alyzed by FT-NIR spectra. This was in agreement
BM-1. This proportional tendency may result from the with the fact that Na+ coordinates with type II water
process of crystallization in the mineralizing fluid with molecules in the channel (see figure 11A) and each al-
abundant Li, Na, K, Rb, and Cs at the same time. kali was nearly proportional. Beryl samples from
The chemical fingerprint diagram of Li versus Cs most localities were in the low-alkali beryl area in-
content showed two concentrated areas, which could cluding Pakistan, Russia, Africa, Mozambique, Italy
be considered the “low-alkali beryl” area (red circle (Bocchio et al., 2009), Vietnam (Huong et al., 2011),

Li VS. Cs

Shigar Valley, Pakistan Masino-Bregaglia Massif, Italy (Bocchio et al., 2009)

Mozambique Thuong Xuan, Vietnam (Huong et al., 2011)
100000 Africa Minas Gerais, Brazil (Viana et al., 2002)
Russia Figure 13. The chemical
Brazil fingerprint diagram of
Cs CONTENT (ppmw)

Mogok, Myanmar the log-log plot of Li ver-

10000
sus Cs content in beryl
samples from various
localities shows two
concentrated areas con-
1000
sidered the “low-alkali”
(red circle) and “high-
alkali” (blue circle)
areas.
100

1 10 100 1000 10000

Li CONTENT (ppmw)/Log10

COLOR CHARACTERISTICS OF BLUE TO YELLOW BERYL GEMS & GEMOLOGY SPRING 2020 63
 Fe VS. Mn
tent showed no obvious absorption of Mn ions in UV-
10000
Vis-NIR spectra and a corresponding pink, orange, or
purple color. For example, the Mn content of deep
blue aquamarine BM-1 (568–854 ppmw) was nearly
Fe CONTENT (ppmw)

8000

half that of red beryl. And the Mn content in Brazilian

6000
dark greenish blue aquamarine BRA-3 (287–292
ppmw) and greenish yellow sample AF-2 (116–160
4000
ppmw) was close to morganite. This indicated that
Mn ions in blue to yellow beryl do not play a role in
2000
coloration, and the valence or occupancy of Mn ions
should differ from that in morganite or red beryl. The
0
1 10 100 1000
valence or occupancy of Mn ions may be affected by
Mn CONTENT (ppmw)/Log10
high Fe content (the Fe/Mn ratio was at least close to
an order of magnitude greater). The valence, occupa-
Figure 14. Fe and Mn concentrations showed an ap- tion, and coloration of Mn ions in beryl will be dis-
proximately positive relationship in all 14 blue to yel- cussed in detail in a future study.
low samples, indicating similar substitution behavior.
Blue to Yellow Color Characteristics. We determined
that Fe ions in beryl could lead to yellow and blue col-
and Minas Gerais, Brazil (Viana et al., 2002). Although ors. Blue color has been attributed to Fe2+ or Fe2+-Fe3+
this Li-Cs plot could not be used to distinguish the pairs, and yellow color to Fe3+ (Wood and Nassau, 1968;
exact geographic origin, it showed the possibility of Lin et al., 2013), though their accurate occupancy re-
separating blue to yellow beryl from locations world- mains controversial. Recently, an alternative mecha-
wide by various chemical fingerprint diagrams. nism involving trapped electrons in the creation and
Chromophores Fe and Mn. Concentrations of Fe and the decay of the yellow color was proposed by Ander-
Mn showed an approximately positive relationship in sson (2013). But the exact configurations of the electron
blue to yellow beryl samples, though they were not trap and the tetrahedral Fe3+ color center could not be
quite proportional (figure 14). This positive relationship determined. Due to different proportions of Fe ions
could be explained by Fe and Mn’s similar substitution with various valences and occupancies in beryl, color
behavior according to similar radii, ionic valence, and varied from yellow to greenish yellow, yellowish green,
coordination. Mn and Fe ions were normally present as bluish green, greenish blue, and blue. The greenish
two possible valence states: +2 and +3 in beryl. Radii of color was normally modified by a yellowish or bluish
Mn ions (approximate measurements of Mn2+: 0.066– component, such as yellowish green sample AF-3 and
0.083 nm; Mn3+: 0.058–0.064 nm) are quite similar to bluish green sample RUS-8. Green color could also be
Fe ions (Fe2+: 0.063–0.078 nm; Fe3+: 0.049–0.054 nm) caused by other distinct coloration mechanisms, Cr3+
(Shannon, 1976). Both ions were more likely to substi- and/or V3+, the same as emerald (e.g., green beryl MOZ-
tute for Al3+ in the octahedral site due to suitable radii 1). But the green beryl had a lower color saturation than
in beryl. Moreover, Fe and Mn ions could enrich to- emerald due to inadequate Cr and/or V content.
gether in mineralized fluid due to similar geochemical
features. Therefore, blue to yellow beryl showed similar Alkali and Color. The alkali ions showed no obvious
substitution behavior during crystallization. impact on blue to yellow color of beryl. The alkali ions
serve for charge deficiency arising from the substitu-
The Chromophore Mn and Color. Mn ions in blue to tion of some transition metals, but no corresponding
yellow beryl did not impart an obvious red modifica- relationship occurred between the alkali content and
tion to the hue, and no obvious pink, purple, or or- Fe/Mn content for blue to yellow beryl samples. This
ange hues were seen. Red beryl and morganite are could be because at least some Fe and Mn ions substi-
considered to be colored by Mn2+ or Mn3+ (Nassau and tuted without charge deficiency (i.e., Fe3+/Mn3+<=>Al3+,
Wood, 1968; Platonov et al., 1989; Fridrichová et al., or Fe2+/Mn2+<=>Be2+), or other substituted ions (such as
2017). Red beryl generally contains 2000 ppmw Mn Zn2+, Mg2+) participated in charge balance. Alkali ions,
(Shigley et al., 1984), and morganite has less than 200 therefore, had no direct impact on the chromophores
ppmw Mn (authors’ unpublished data). However, our and their coloration of blue to yellow beryl. Neverthe-
blue to yellow beryl samples with abundant Mn con- less, our high-alkali beryl appeared to show deeper

64 COLOR CHARACTERISTICS OF BLUE TO YELLOW BERYL GEMS & GEMOLOGY SPRING 2020
blue color. The additional dark hue of BRA-3 was the chromophores and color characteristics. Blue color
caused by dark cloudy inclusions. It was worth noting was caused by a broad absorption band at 600 nm, and
that bluish samples could be “low-alkali” or “high-al- yellow color was attributed to an absorption edge in
kali” beryl, but all yellowish samples were “low-al- the violet-blue region. Both were related to Fe ions in
kali” beryl. This manifested that the alkali may different valences and occupancies. Green beryl was
influence the valence or occupancy of Fe ions, but it colored by a combination of yellow and blue, or by
required further confirmation. Cr3+ and/or minor V3+. The concentration of Mn was
somewhat positively related to Fe in blue to yellow
Water and Color. Water type and its relative content beryl, but abundant Mn ions had no impact on color.
played little role in the blue to yellow color of beryl. This study discussed the arrangement of alkali
All blue to yellow samples contained both type I and ions and water molecules within the structural chan-
type II water in the channel. Nor was there any rela- nel, facilitating charge compensation among the
tionship between the type I/type II water ratio and channel species and aliovalent substitutions in the
color hue or saturation. Water molecules were related surrounding rings. Large Cs+, Rb+, K+ ions and water
to chromophores in beryl through Na ions, but Na ions molecules were localized at the 2a site and were un-
showed no obvious impact on the coloration of blue to able to migrate along the channel. The smallest ions
yellow beryl. Therefore, it was unlikely that water (Na+) stably positioned at the 2b site, interacting with
molecules would affect beryl’s blue to yellow color. “type II” water. Non-nominal (not in chemical for-
mula) alkali elements (Li, Na, K, Rb, and Cs) and
CONCLUSIONS water were relevant and played complex roles in
Blue to yellow gem beryl was studied by quantitative beryl. However, we found the influence on blue to
chemistry and spectroscopic properties to understand yellow color to be indirect and possibly weak.

ABOUT THE AUTHORS ACKNOWLEDGMENTS

Mr. Hu is a PhD student at the Gemological Institute, China Uni- The authors would like to thank the editorial improvement on this
versity of Geosciences in Wuhan. Dr. Lu ([email protected]) is a manuscript. We also thank anonymous reviewers for their construc-
distinguished professor at the Gemological Institute, China Uni- tive comments. This study was supported by the Center for Innova-
versity of Geosciences in Wuhan. tive Gem Testing Technology, China University of Geosciences.

REFERENCES
Andersson L.O. (2013) The yellow color center and trapped elec- Lin J., Chen N., Huang D., Pan Y. (2013) Iron pairs in beryl: New
trons in beryl. Canadian Mineralogist, Vol. 51, No. 1, pp. 15– insights from electron paramagnetic resonance, synchrotron X-
25, http://dx.doi.org/10.3749/canmin.51.1.15 ray absorption spectroscopy, and ab initio calculations. Amer-
Artioli G., Rinaldi R., Stahl K., Zanazzi P.F. (1993) Structure re- ican Mineralogist, Vol. 98, No. 10, pp. 1745–1753,
finements of beryl by single-crystal neutron and X-ray diffrac- http://dx.doi.org/10.2138/am.2013.4472
tion. American Mineralogist, Vol. 78, pp. 762–768. Liu Y.S., Hu Z.C., Gao S., Günther D., Xu J., Gao C.G., Chen H.H.
Belakovskiy D., Bradshaw J., Cairncross B., Falster A.U., et al. (2005) (2008) In situ analysis of major and trace elements of anhydrous
Beryl and Its Color Varieties - Aquamarine, Heliodor, Morgan- minerals by LA-ICP-MS without applying an internal standard.
ite, Goshenite, Emerald, and Red Beryl, English Edition. Lapis Chemical Geology, Vol. 257, No. 1-2, pp. 34–43, http://dx.doi.org/
International LLC, East Hampton, Connecticut. 10.1016/j.chemgeo.2008.08.004
Bocchio R., Adamo I., Caucia F. (2009) Aquamarine from the Nassau K., Wood D.L. (1968) An examination of red beryl from
Masino-Bregaglia Massif, Central Alps, Italy. G&G, Vol. 45, Utah. American Mineralogist, Vol. 53, No. 5/6, pp. 801–806.
No. 3, pp. 204–207, http://dx.doi.org/10.5741/GEMS.45.3.204 Platonov A.N., Taran M.N., Klyakhin V.A. (1989) On two colour
Fridrichová J., Bacík P., Ertl A., Wildner M., Dekan J., Miglierini M. types of Mn3+-bearing beryls. Zeitschrift der Deutschen Gem-
(2017) Jahn-Teller distortion of Mn3+ -occupied octahedra in red mologischen Gesellschaft, Vol. 38, pp. 147–154.
beryl from Utah indicated by optical spectroscopy. Journal of Mo- Viana R.R., Jordt-Evangelista H., Costa G.M., Stern W.B. (2002)
lecular Structure, Vol. 1152, pp. 79–86, https://doi.org/10.1016/ Characterization of beryl (aquamarine variety) from pegmatites
j.molstruc.2017.09.081 of Minas Gerais, Brazil. Physics and Chemistry of Minerals,
Gorshunov B.P., Torgashev V.I., Zhukova E.S., Thomas V.G., et al. Vol. 29, No. 10, pp. 668–679, http://dx.doi.org/10.1007/s00269-
(2016) Incipient ferroelectricity of water molecules confined to 002-0278-y
nano-channels of beryl. Nature Communications, Vol. 7, No. Wood D.L., Nassau K. (1967) Infrared spectra of foreign molecules
1284, http://dx.doi.org/10.1038/ncomms12842 in beryl. Journal of Chemical Physics, Vol. 47, No. 7, pp. 2220–
Huong L.T.-T., Hofmeister W., Häger T., Khoi N.N., Nhung N.T., 2228, http://dx.doi.org/10.1063/1.1703295
Atichat W., Pisutha-Arnond V. (2011) Aquamarine from the ——— (1968) The characterization of beryl and emerald by visible
Thuong Xuan District, Thanh Hoa Province, Vietnam. G&G, and infrared absorption spectroscopy. American Mineralogist,
Vol. 47, No. 1, pp. 42–48, http://dx.doi.org/10.5741/GEMS.47.1.42 Vol. 53, No. 5, pp. 777–800.

COLOR CHARACTERISTICS OF BLUE TO YELLOW BERYL GEMS & GEMOLOGY SPRING 2020 65
FEATURE ARTICLES

HISTORY OF THE CHIVOR EMERALD MINE,

PART I (1880–1925): FROM REDISCOVERY
TO EARLY PRODUCTION
Karl Schmetzer, Gérard Martayan, and Jose Guillermo Ortiz

The history of the Chivor emerald mine in Colombia is rife with legend and adventure. The tale traces from early ex-
ploitation by indigenous people, to work by the Spanish in the sixteenth and seventeenth centuries, to 200 years of
abandonment and jungle overgrowth. The story then picks up with rediscovery near the turn of the twentieth century
by the Colombian mining engineer Francisco Restrepo using clues from a historical manuscript. Still the saga con-
tinued, with repeated shortages of investment funds driving multiple ownership changes and little progress toward
mining the largely inaccessible deposit. The German gem merchant Fritz Klein, in cooperation with Restrepo, pursued
limited mining activities with a small number of workers for a few months prior to the outbreak of World War I.
After the war, the American company Colombian Emerald Syndicate, Ltd., took ownership, and mining operations
resumed under the new leadership. Ownership changed yet again in the 1920s, followed by multiple cycles of ex-
panding and shrinking mining activity, interrupted by completely unproductive periods.

T
he foregoing outline of the story, as presently eral, the events described by Klein, Rainier, or Ander-
known, has been drawn largely from three ton have been accepted as facts in the literature, and
books authored by individuals who led the min- only Klein’s description and dating of the rediscovery
ing activities at Chivor during different eras: German
gem merchant Fritz Klein, South African mining en-
gineer Peter W. Rainier, and American gem hunter In Brief
and buyer Russell W. Anderton. All three men wrote
• After 200 years of abandonment, the modern history of
in a style to highlight the adventurous nature of the the Chivor emerald mine began with rediscovery by
work. Klein’s memoir covered the period from ap- the Colombian mining engineer Francisco Restrepo in
proximately 1911 to 1923 and was initially published the 1880s.
in 1941, with a slightly updated version released in • After several trials to sell the mine by the Colombian
1951. Rainier’s narrative chronicled from the second owners, in 1912 German gem merchant Fritz Klein then
half of the 1920s to the early 1930s and was printed joined Restrepo in further efforts, but Klein’s attempts to
in 1942. Anderton’s work, coming to press in 1953 in complete a purchase with a group of German investors
the United States and 1954 in the United Kingdom, were thwarted by the outbreak of World War I.
recounted activities of the late 1940s and early 1950s. • After the war, an American company, the Colombian
The events presented in the three books have, Emerald Syndicate, Ltd., took ownership in late 1919.
since their respective publications, found their way in • Although operations resumed under the new leader-
numerous variations into historical articles or descrip- ship, lasting success was not to be found, and the
tions as well as gemological, mineralogical, or geolog- Colombian Emerald Syndicate entered bankruptcy pro-
ceedings in 1923.
ical papers.1 The rediscovery of Chivor by Francisco
Restrepo (figure 1) and the clues that motivated his
search have been a particularly popular topic. In gen- by Restrepo have on occasion been questioned or dis-
cussed.2 It has been noted that Klein’s time frame for
the rediscovery clearly differs from that presented in
See end of article for About the Authors and Acknowledgments.
GEMS & GEMOLOGY, Vol. 56, No. 1, pp. 66–109,
1
http://dx.doi.org/10.5741/GEMS.56.1.66 See, e.g., Weldon et al., 2016, for a recent example.
2
© 2020 Gemological Institute of America See, e.g., Sinkankas, 1981; Moore and Wilson, 2016.

66 HISTORY OF CHIVOR, PART I: 1880–1925 GEMS & GEMOLOGY SPRING 2020

 SOURCES
The search for contemporaneous materials spanned
Europe, South America, and North America, as key
figures in the story hailed from or worked in Ger-
many, the United Kingdom, Colombia, and the
United States. These primary documents were com-
pared with the known literature on the topic. Full ci-
tations for the principal existing publications
consulted are summarized in the reference list. Other
primary documents, such as governmental and judi-
cial reports and decrees that may appear in broader
compendiums, archived contracts, business records,
newspaper articles, personal correspondence, etc., are
identified to the extent feasible in the footnotes,
along with brief citations to publications included in
the reference list. Direct quotations from original
sources in German or Spanish were translated for
purposes of this publication.
Investigations began with the German Federal
Archive (Bundesarchiv) and the Historical Archive of
the German Ministry of Foreign Affairs (Auswärtiges
Amt), both in Berlin. Stored in these repositories
Figure 1. Francisco Daniel Restrepo Escobar, circa
1906. Restrepo began his career at gold and silver
mines in the Colombian states of Antioquia and
Figure 2. Portrait of the gem merchant Fritz Klein by
Tolima in the 1870s and 1880s. He is credited with re-
the artist Rudolf Wild, Idar-Oberstein, Germany, in
discovery of the Chivor emerald mines, an event that
the mid-1930s. At that time, Klein was engaged in
evidence also places in the 1880s. The early 1900s
writing the 1941 book describing his earlier adven-
found him at gold mines in Ecuador and then leading
tures searching for new gem deposits in South Amer-
operations at the Muzo emerald mines. In the 1910s,
ica; a highlight is the saga of his engagement with
he returned to Chivor, where he would cooperate with
the Chivor emerald mine in Colombia. Courtesy of
German miner Fritz Klein to find investment funding
Wolf D. Unger.
for the project. Courtesy of Eduardo Restrepo Ortega.

other publications. This discrepancy, in turn, provided

the impetus for the present project, aiming not only
to clarify this inconsistency but also, more broadly, to
probe the trustworthiness of the purported historical
“facts” recounted by Klein, Rainier, and Anderton.
The focus was on locating contemporaneous pri-
mary documents, such as original letters and con-
tracts, that could elucidate events underlying the
oft-repeated descriptions. Beginning with an empha-
sis on the period covered by Klein (figure 2), the
search was subsequently augmented to consider the
record from the 1880s to the 1970s. In summary, the
project endeavored to present a comprehensive sur-
vey of Chivor’s history from its rediscovery to the
second half of the twentieth century, based princi-
pally upon primary contemporaneous materials, di-
vided into two parts. The first installment of this
work covers the period from the 1880s to 1925.

HISTORY OF CHIVOR, PART I: 1880–1925 GEMS & GEMOLOGY SPRING 2020 67

were original documents or copies thereof involving frames. Similarly insightful were digital archives of
the emerald mines in Colombia, largely in materials newspapers across the United States, which enabled
generated during relevant periods by the German view of events as presented to the public in the early
Ministry of Foreign Affairs in Berlin and the German twentieth century.
Residency (essentially equivalent to an embassy) in
Bogotá.3 Of a similar nature and also held in Europe LOCATION
were correspondence and related materials in the col- The Chivor mining area is located northeast of
lections of the Kew National Archives in London, de- Colombia’s capital Bogotá, in proximity to the pres-
rived from the United Kingdom’s Foreign Office. ent-day small municipalities of Somondoco and
Principal Colombian materials were found in the Almeida to the north and Gachalá and Ubalá to the
form of government publications printed, typically southwest (figure 3). In the early decades of twentieth
anonymously, in the “Diario Oficial (Colombia),” in century, the nearest community of sufficient size to
the “Gaceta Judicial (Colombia)” of the Corte host a telegraph office was Guateque, accessible via
Suprema de Justicia (Supreme Court), and in the re- a one-day journey by horse.
ports of ministries or other government offices. The main emerald mines are situated south of the
These included legislative texts, decrees, governmen- town of Chivor, founded in 1930, between the valleys
tal or judicial pronouncements, and statements on of the Río Sinai and the Río Rucio, the latter of which
petitions and complaints. The Archivo General de la flows into the Río Guavio. (The two valleys are al-
Nación (Colombia) in Bogotá further yielded notarial ternatively referred to as Quebrada Sinai and Que-
acts (escrituras) and attachments reflecting mine title brada de las Minas, quebrada meaning “gorge” or
transfers, sales of shares between various sharehold- “breaking reef” in Spanish.) North of the Chivor area
ers, and agreements shedding light on the ongoing are the Río Somondoco and the Río Garagoa. Chivor
fluctuations in Chivor ownership and related events.4 is now positioned between two large dams that were
Europe and the Americas also proved to be sources built in the second half of the twentieth century, the
of collections of a more personal nature, with notable Embalse de Chivor (Chivor Dam) and the Embalse
materials being found both in the University Library del Guavio (Guavio Dam). The Río Garagoa runs to-
Freiburg (Universitätsbibliothek Freiburg) archives5 ward the Embalse de Chivor and south of the dam is
and in the private holdings of family members. For also called the Río Bata, which in turn flows into the
example, Francisco Restrepo’s grandson Eduardo Re- Río Guavio.
strepo Ortega held family documents from the late
1880s to the early 1920s describing many aspects of
his grandfather’s role with Chivor. Likewise, descen- 3
In the German Federal Archive (Bundesarchiv) in Berlin, pertinent
material was found in file R901, archive numbers 1303, 1304, and
dants of Fritz Klein and his brother August, who 1305, all covering the years 1911 to 1920. In addition, file R901,
worked together at Chivor between 1912 and 1914, archive number 81190, contained a mining report from 1913 by a
were in possession of original photos from Colombia member of the German Residency in Bogotá. In the Historical Archive
of the German Ministry of Foreign Affairs (Auswärtiges Amt), also in
and copies of letters written by family members be- Berlin, principal relevant files were designated RAV Bogotá Volume
tween 1907 and 1920. Unfortunately, the original 50 (now 136) and R Volume 91809, covering the years 1914 to 1936.
manuscript for Klein’s 1941 book was not located by 4
Most relevant at the Archivo General de la Nación (Colombia) in Bo-
the present authors. Contributions of this nature from gotá were several notarial acts (escrituras) designated herein by escritura
number and date. Also useful was a file entitled “Minas de Esmeraldas”
relatives of many key players were instrumental, and incorporating materials from 1905 to 1917, collected by the Ministerio
their importance cannot be overstated.6 de Relaciones Exteriores. Similarly instructive was the file “Joaquin
Even online resources added to the available con- Daza B.,” Volume “Propuestas Minas 99,” Ministerio de Industrias, De-
partamento de Minas y Petróleos, which contained documents covering
temporaneous data. Noteworthy in this regard were 1929 to 1954 and detailing mine boundaries and history.
digitized compilations of records detailing transit of 5
Held at the University Library Freiburg ((Universitätsbibliothek
persons, such as passenger lists maintained by the Freiburg) was a collection of documents related to Prof. Robert
United States Immigration Office in New York and Scheibe, a geologist involved in scientific investigations at and evalu-
ation of Chivor, as detailed below.
by offices in Hamburg and Bremen, Germany, and in 6
See the acknowledgments for additional indication of the scope of
Tilbury, London, United Kingdom.7 Those records such contributions.
detailed information that could incorporate depar- 7
MyHeritage.com, Ancestry.com, and Ancestry.de databases. The
tures and arrivals through national borders, starting New York passenger lists record the arrival of United States or foreign
citizens, the Hamburg and Bremen passenger lists record departures
points, destinations, and, most importantly, specific from Germany, and the London passenger lists record departures of
dates to help establish or corroborate event time British citizens from the port of Tilbury.

68 HISTORY OF CHIVOR, PART I: 1880–1925 GEMS & GEMOLOGY SPRING 2020

 Río Garagoa
N Guateque
Achiote Agua Blanca
o
oc
nd Figure 3. Map showing
omo Somondoco Macanal
Rí oS the location of the
Almeida Chivor emerald mines,
south of the towns of
Quebrada Negra
Embalse Somondoco and
de Chivor Almeida and northeast
Palo Arañado of Gachalá and Ubalá,
Río in the Departamento de
Chivor Bata
Río Ruc Gualí Boyacá, Colombia. The
io
area is now located be-
Minas de
tween two dams, Em-

ai
Chivor

Si n
Mundo Nuevo Santa Maria balse de Chivor and

Rí o
Embalse del Guavio,
Buenavista
that were built in the

io
Ru c
second half of the

Rí o
Montecristo Río twentieth century. The
G uav
io other emerald mines in
vio
the departments of
Boyacá and Cundina-
ua
oG

Mambita
marca were discovered
Rí

Ubalá La Vega de Río Guavio and exploited in the

Hacienda
San Juan Matecaña CU BOY
Las Cascadas ND A twentieth century, sub-
INA CÁ
MA
RC
A
sequent to the rediscov-
Las Cruces
ery of Chivor.
Gachalá Embalse
0 5 km
del Guavio

The Chivor mining area currently consists of sada’s camp in late May or early June 1538 and pro-
many tunnels and galleries worked employing indus- ceeded as reported in a letter from 1539:
trial methods, with production yielding large quan- Valenzuela departed with several others, and after six
tities of gem-quality emeralds.8 This scenario days they arrived at the said mines, where he and the
presents a marked contrast to the harsh working con- other Spaniards with him, watched the Indians extract
ditions and difficulties encountered in the first half the emeralds from below the ground, and they wit-
of the twentieth century. nessed such strange new things. The mines are located
roughly fifteen leagues from the Valley of the trumpet,
in a very high and sparse mountain range. It appears
PRE-NINETEENTH CENTURY HISTORY OF THE that the emeralds are extracted from an area about one
SOMONDOCO (CHIVOR) EMERALD MINES league in size. The lord of the mines is a very principal
Indian by the name of Somyndoco [Somondoco]. He is
Beginning in early April 1536, a Spanish military ex- sovereign over many vassals and settlements, and his
pedition was conducted under the leadership of con- private residence is located three leagues’ distance from
quistador Gonzalo Jiménez de Quesada into the the mines…. Those who went on the discovery of the
region that later became Nueva Granada (the New emerald mines said that from the mines they could see
some great planes [llanos], so marvelous that nothing
Kingdom of Granada, encompassing approximately like them ever had been seen. On hearing this news,
the territory of modern Colombia).9 The journey em- Jiménez moved the camp closer to the mines….10
barked from the Atlantic coastal city of Santa Marta,
and by December 1536, the expeditionary group had A similar account of the discovery of the ancient
begun to cross the Andes Mountains. They arrived Somondoco emerald mines appeared, with minor
in the territory of the indigenous Muisca in March
8
1537. As they traveled south, the Spanish soldiers en- Fortaleché et al., 2017.
9
Francis, 2007.
countered emeralds in the region, and Quesada or- 10
Letter from two Spanish captains, July 8, 1539, Archivo General de
dered the captain Pedro Fernández Valenzuela to Indias, Seville, Spain, as translated and published by Francis, 2007.
search for the mines. Valenzuela departed from Que- One league was equivalent to 5.57 km. See Lane, 2010.

HISTORY OF CHIVOR, PART I: 1880–1925 GEMS & GEMOLOGY SPRING 2020 69

variations, in all historical descriptions of the con- to 1887, numerous new deposits, mostly gold and sil-
quest of Nueva Granada penned by sixteenth- and ver mines, were discovered and registered with the
seventeenth-century Spanish authors. Examples in- government (to establish claims and valid mining ti-
clude those by Fray Pedro de Aguado, Fray Pedro tles) by both Restrepo and Correa, either alone or to-
Simón, and Lucas Fernández de Piedrahita. gether.20 Evidence of Restrepo’s mining activities in
In the following decades, the mines referred to as Tolima extends to the end of the decade, with records
Somondoco in contemporaneous Spanish texts were showing mine tax payments and sales transactions
worked with limited success, although a portion of for several gold and silver mines as late as 1889.21
the emerald mining activity was diverted to the Conversely, records for the following two decades are
Muzo deposits after their discovery to the northwest silent in that regard, but those from 1910 reflect that
in the 1560s.11 Exploitation at Somondoco continued Restrepo was still holding mining titles in the area,
under various owners and lessees12 into the 1670s, together with several partners.22 Meanwhile, as of
with the last known documentary evidence of activ- 1891 Restrepo had apparently resumed efforts back
ity dated from 1672 to 1675.13 in Antioquia as well.23
Beyond that time, there exists no known record
of further activity at the mines, nor is there any re-
port detailing why the mines were abandoned. While 11
Lane, 2010.
the literature refers to an order to close the mines is- 12
Canova, 1921.
13
sued by the Spanish monarch Carlos II (reigning Slave labor at the mines was described in a document from 1672
(Kris Lane, pers. comm., 2018). Litigation in 1672 involving Somon-
1665–1700),14 such an order has never been proven doco mining rights has also been reported (Canova, 1921). Several
and could not be located despite recent extensive re- additional years of activity are supported by archival material from
search.15 By 1772 it was noted that the mines of the second half of the seventeenth century in the file “Minas-Boyacá,”
SC 38, 1, D10, Archivo General de la Nación (Colombia).
Somondoco were deserted and that nothing more 14
Martin de Retana, 1990.
than some vague remains of the old workings could 15
Kris Lane, pers. comm., 2018.
be found.16 Thus, over the course of more than 200 16
Moreno, 1772, published 1870.
years beginning in the 1670s, any detailed knowledge 17
Commonly referred to as Francisco Restrepo, the names Francisco
about the location of the Somondoco emerald mines Restrepo Escobar or Francisco Restrepo E. also appear in official doc-
uments, with his signature on notarial acts taking the latter form. His
was essentially lost. year of birth is similarly ambiguous. A death certificate indicates
1853, while passenger lists reflecting his departure from Hamburg
and subsequent arrival in New York in November 1913 give an age of
REDISCOVERY OF CHIVOR 62, implying that he was born in 1851. File F. Restrepo, Hamburger
The rediscovery of the historical mines of Somon- Passagierlisten (list of passengers departing from Hamburg), Novem-
ber 1913, Ancestry.de; File F. Restrepo, List or Manifest of Alien Pas-
doco, henceforth known primarily as the Chivor sengers for the United States Immigration Officer at Port of Arrival,
mines, has long been attributed to the Colombian New York, November 1913, Ancestry.com. However, the date 1955
mining engineer Francisco Daniel Restrepo Escobar17 was provided by his grandson. Eduardo Restrepo Ortega, pers.
comms., 2018, 2019.
(1855–1914). Restrepo was born in Rionegro, Antio- 18
Aguilar, 1884.
quia, Colombia, and was involved in the mines of 19
Martínez Covaleda and Martínez Covaleda, 1996; Moreno
Antioquia during his younger years.18 By the early Calderón, 2011.
1880s, he had transferred his attention to the neigh- 20
Diario Oficial (Colombia), 23, No. 7208 (1887), pp. 1194–1195; Es-
boring state of Tolima, where he worked as a prospec- quivel Triana, 2001; Quintero and Centeno, 2007.
21
Informe presentado al Congreso de la República en sus sesiones or-
tor and mine production manager. At the time, dinarias de 1890 por el Ministro de Hacienda, Segunda Parte – In-
mineral deposits in Antioquia were considered formes de los administradores departamentales de Hacienda, pp.
nearly exhausted, while Tolima was viewed as hav- 219–320; Skinner, 1891–1892; Escritura 556, May 25, 1899, Notaria
3, Archivo General de la Nación (Colombia).
ing strong potential for alluvial gold and silver mines. 22
The Compañía de Minas de Los Alpes holding these mining titles
Restrepo’s activities in Tolima involved searching for was formed in 1910 and reorganized in 1916. Escritura 148, February
lost mines from the colonial era as well as discover- 3, 1910, Notaria 2, Escrituras 944 and 955, May 2, 1916, Notaria 2,
Archivo General de la Nación (Colombia). Restrepo was also in-
ing new mines. During the first half of the 1880s, at volved in the late 1800s and early 1900s in ventures unrelated to
least a portion of his work was done in association mining. For example, through an administrator he operated two ha-
with the Compañía Minera del Tolima (not to be ciendas in the district of Paime, south of Muzo, named Santa Teresa
and San Luis, which were abandoned during the Colombian civil
mistaken with the British Tolima Mining Company, war. After the war, in 1902, he was active in the coffee trade. Gaceta
Limited),19 but from the mid-1880s Restrepo also Judicial (Colombia), 20 (1911), pp. 49–57.
worked in partnership with Luis Correa. From 1885 23
Brissón, 1899.

70 HISTORY OF CHIVOR, PART I: 1880–1925 GEMS & GEMOLOGY SPRING 2020

 As to the date, Klein (1941) estimates 1903 to
1904. Conversely, Spanish texts place this event in
1896 or even as early as 1888 to 1889,29 and English-
language texts mostly quote 1896.30 The English
texts primarily refer to Latham (1911), but that work
offers no corroborative reference to support the 1896
date given. Domínguez31 (1965), the most detailed
Spanish textbook on the history of Colombian emer-
alds, cites a limited number of geological studies but
does not associate historical events with any pri-
mary source. The Spanish publication by Martin de
Retana (1990) chronicles historical events, but the
individual facts are neither consistent nor correlated
with the references cited. German authors, in most
cases, follow the version and dates given by Klein
(1941).
No direct written communication by Restrepo de-
scribes his search; the only known tangible evidence
of his efforts consists of three rough hand-sketched
maps, all undated, found in his personal papers (e.g.,
figure 4). These merely depict selected areas between
Río Somondoco and Río Guavio and mark the loca-
tions of already discovered mines and promising sites
Figure 4. Undated hand-sketched map, from a collec- for search. Nonetheless, a report authored by Klein
tion of original documents in the possession of Fran- in 1913 and notes by Prof. Robert Scheibe from 1914
cisco Restrepo’s grandson, of the region south of the offer verification from contemporaneous primary
town of Guateque and the Río Somondoco in Colom- sources. As will be detailed below, both Klein and the
bia, incorporating notations regarding already located geologist Scheibe spent time with Restrepo at
mines and promising areas for search. Courtesy of Ed- Chivor. In the report by Klein, which was an “ex-
uardo Restrepo Ortega.

24
Escritura 201, February 13, 1899, and attachments, Notaria 3,
The foregoing regions, however, did not consti- Archivo General de la Nación (Colombia); letter from a department of
tute Restrepo’s sole focus. The emerald mines in Boy- the Ministerio de Hacienda in Tunja, Colombia, to the Ministerio de
acá designated Chivor 1 and Chivor 2 were registered Industrias, 1930, file “Joaquin Daza B.,” Volume “Propuestas Minas
99,” Ministerio de Industrias, Departamento de Minas y Petróleos,
with the government in 1889, and Restrepo was Archivo General de la Nación (Colombia).
clearly involved in the process, which comprised 25
Enrique González served as director of operations at the Muzo and
multiple steps.24 The first step in applying for the ti- Coscuez emerald mines in Colombia from 1886 to 1901, first on be-
half of Lorenzo Merino and later on behalf of the English Mining Syn-
tles (acta de aviso) was undertaken by Miguel Perilla dicate, Limited. Gaceta Judicial (Colombia), 15 (1901), pp. 169–173;
in the names of Restrepo, Enrique González,25 González, 1911.
Bernardo Escovar,26 and Núñez & Compañía27 in Jan- 26
Also referred to as Bernardo Escobar in several documents, with his
uary 1889. The application was then granted and signature on notarial acts generally taking the form Escovar (used
hereinafter).
published in March 1889. Thereafter, at a date and 27
The background of Núñez & Compañía is vague; subsequent to the
under circumstances not otherwise elucidated, title reference in the 1889 application, an entity by that name is identi-
ownership apparently transferred from the original fied as having been founded in 1895 by four brothers. Escritura
1286, November 30, 1895, Notaria 1, Archivo General de la Nación
four applicants to Restrepo and Correa, who were (Colombia).
recognized as the owners in documentation from 28
Escritura 201, February 13, 1899, and attachments, Notaria 3,
1896, seven years later.28 Archivo General de la Nación (Colombia).
Nonetheless, two critical aspects regarding Re- 29
Domínguez, 1965; Martin de Retana, 1990; Calvo Perez, 1992.
30
strepo’s underlying rediscovery of the Chivor mines Latham, 1911; Pogue, 1917; Canova, 1921.
31
are fraught with discrepancy and ambiguity: the date Rafael A. Domínguez was head of the financial administration (In-
terventor de Salinas y Esmeraldas) of the Ministerio de Minas y
of the rediscovery and the source of the information Petróleos in Bogotá. Memoria del Ministro de Minas y Petróleos al
on which he relied. Congreso de 1963 (1963), p. 26.

HISTORY OF CHIVOR, PART I: 1880–1925 GEMS & GEMOLOGY SPRING 2020 71

posé”32 likely written for potential investors, 1882 Antwerp in 1688, was reprinted in Bogotá in 1881
was given as the date when indications of the mine’s and, as a consequence, could have been readily avail-
location were found, and 1885 was noted as when the able to Restrepo at the beginning of the 1880s.
old tunnels worked centuries ago were rediscovered. The historical treatise by Fray Pedro Simón, one
Scheibe’s personal notes33 discussed the boundaries of the most prominent authors of the Spanish con-
of the two mines Chivor 1 and 2 in connection with quest, has frequently been mentioned as the original
“the work that Restrepo did in the 80s of the last cen- source for Restrepo’s information. Only the first vol-
tury.” Thus, primary and secondary documentation ume of this five-volume series was published in the
places the search for the lost mines and their redis- seventeenth century, with the complete set edited
covery in the 1880s. and printed in Bogotá from 1882 to 1892. The second
The nature of the collaboration between Restrepo volume, which contained the description of Valen-
and Correa in the rediscovery remains obscure,34 but zuela’s 1538 expedition to Somondoco, was pub-
a 1911 letter from Correa to Restrepo lends a degree lished in 1891 from a handwritten manuscript
of corroboration to the 1880s timing. Although the preserved in the Biblioteca Nacional de Colombia
correspondence makes no direct mention of Chivor, (National Library of Colombia) in Bogotá. The origi-
it does discuss possible future cooperative ventures nal manuscript had been prepared by Simón in the
and refers back to an event in 1885 when they were 1620s, and that manuscript or a copy of it had been
“extremely lucky.”35 available and accessible at the national library since
Turning to the question of the source of the clues 1852, when it came into the collection from the pri-
on which Restrepo relied, it is generally reported that vate library of Joaquin Acosta.37
he was guided by hints found in historical manu- Fray Pedro de Aguado, the sixteenth-century chron-
scripts, but further details differ widely. In nine- icler of the Spanish conquest of Colombia and
teenth-century descriptions of Colombian history, Venezuela, has also been cited as a possible source for
Acosta (1848), Plaza (1850), and Groot (1869) re- Restrepo’s information. Aguado’s treatise “Recopi-
counted the expedition of Valenzuela in 1538 and in- lación Historial” was never published during the colo-
dicated that from the Somondoco mines, or on the nial period.38 The handwritten first volume,
journey thereto, great plains (llanos) were visible to concerning Colombia, was discovered in Madrid by
the east. Other nineteenth-century publications of- Acosta in 1845 and briefly mentioned in his history of
fered even somewhat misleading information such as Nueva Granada published in 1848. The handwritten
“nine leguas [leagues] south-south-east of Guateque, manuscript itself, copied in the early years of the twen-
close to the cascade of Nagar.”36 tieth century, was published in Bogotá in 1906. These
With respect to Spanish authors from the colonial facts indicate that Aguado’s treatise on Colombia was
period, the description of Colombia’s history by most likely not available to Restrepo in the 1880s.
Lucas Fernández de Piedrahita, first published in In summary, the only known information avail-
able to guide Restrepo’s journey was the vague hint
32
about the llanos being visible from the Chivor min-
Exposé der Smaragdmine “El Chivor” by Fritz Klein, undated, 16
pp., received by the German Ministry of Foreign Affairs September ing region (figure 5) and perhaps some approximate
13, 1913, and preserved in the German Federal Archive. A portion of indication of distance from the town of Somondoco.
this document was translated and published in English by Canova This information would have been accessible to him
(1921).
33
Prof. Robert Scheibe collection, University Library Freiburg. from several sixteenth- and seventeenth-century au-
34
Likewise minimal is information about Luis Correa’s life after the thors and as repeated in nineteenth-century histori-
Chivor rediscovery. He did, however, remain involved in the emerald cal texts. The present authors are aware of neither
business, serving for example as Inspector de las minas de esmeraldas further handwritten manuscripts from sixteenth- or
de Muzo & Coscuez. Diario Oficial (Colombia), 42, No. 12693
(1906), p. 630. seventeenth-century historians nor any other de-
35
Letter from Luis Correa, March 15, 1911, to Restrepo. tailed accounts or maps that might have identified
36
Ancizar, 1853; Schuhmacher, 1875. The cascade of Nagar was lo- the location of the ancient Somondoco mines.
cated close to Macanal but is now covered by the Chivor Dam. Enci-
clopedia Universal Ilustrada Europeo-Americana, 1931; Geografía
económica de Colombia, 1936. This locality would be about 15 km MINING TITLE TRANSFERS, MINING COMPANY
too far northeast from the present Chivor mines; see also figure 3.
37
FORMATION, AND EARLY YEARS (1896–1910)
Robinson López, Biblioteca Nacional de Colombia, Bogotá, pers.
comm., 2017. As previously noted, titles to mines designated Chivor
38
Fals-Borda, 1955; Friede, 1964. 1 and Chivor 2 were registered in 1889, and by 1896

72 HISTORY OF CHIVOR, PART I: 1880–1925 GEMS & GEMOLOGY SPRING 2020

 Figure 5. In all the six-
teenth- and seven-
teenth-century texts
that describe the
Spaniards’ discovery of
the Chivor emerald
mines—named Somon-
doco at that time—it
was mentioned that
great plains (llanos)
were seen from the
mines or on the way to
the mines in a view to
the east. Courtesy of
Kris Lane, Tulane
University.

Restrepo and Correa were recognized as the owners. for Restrepo, 20 for González, 20 for Escovar, and 20
The registration was made in accordance with the for Núñez & Compañía. The latter two were to act
Colombian mining law then in effect, which stipu- as investors, while Restrepo and González, both of
lated that a gem mine could not exceed one square whom had mining experience, were to continue ex-
kilometer in size (Law No. 38 of 1887, Article 2).39 The ploration and to start practical operations. The com-
Chivor emerald mines thus consisted of two contigu- pany later became known as the Compañía de las
ous areas of one square kilometer each and were not Minas de Esmeraldas de Chivor,44 or, in certain ma-
otherwise split or subdivided during the relevant pe- terials, Sociedad Ordinaria de las Minas de Chivor.
riod. Any mining in the region employed open-pit For clarity, the different companies involved in the
methods, as was generally the case in Colombia prior Chivor emerald business are listed in table 1.
to 1930. Although the Chibcha and Spanish in prior In 1898 and likely into the next year, Restrepo
centuries may have attempted to build tunnels, even spent time at the mine, mainly supervising restora-
those extended only a few meters into the mountains. tion and construction of a water supply and irrigation
At the Chivor claims, no mining activity took system. These efforts aimed to evaluate the potential
place for several years after registration. The 1891 an- of the deposit and to prepare the mine for visits by
nual taxes for the two mines were paid by Correa,40 possible investors, looking for veins and samples in
but no further payments are known. As a result,
Chivor 1 and 2 were declared “abandoned” on August
23, 1896, on account of an absence of operations in the 39
Bullman, 1892.
seven years since the registration and a failure to pay 40
Diario Oficial (Colombia), 28, No. 8842 (1892), p. 819.
taxes on the mines since 1891.41 Attendant legal for- 41
Escritura 201, February 13, 1899, and attachments, Notaria 3,
malities led to the titles being transferred from Correa Archivo General de la Nación (Colombia).
42
and Restrepo to González on February 16, 1898.42 Escritura 201, February 13, 1899, and attachments, Notaria 3,
Archivo General de la Nación (Colombia); Report by Dr. Antonio José
Later in 1898, a mining entity was formed as a Cadavid, March 18, 1913, Historical Archive of the German Ministry
joint stock company between González, Restrepo, of Foreign Affairs.
43
Escovar, and Núñez & Compañía, but without Cor- Escritura 1337, August 20, 1898, Notaria 3, Archivo General de la
Nación (Colombia). No name for the company was specified at that
rea, and the Chivor 1 and 2 mining titles were trans- time.
ferred from González to that company.43 Share 44
The name appeared in documents and reports of the Ministerio de
distribution at formation was as follows: 40 shares Hacienda and the German Resident in Bogotá.

HISTORY OF CHIVOR, PART I: 1880–1925 GEMS & GEMOLOGY SPRING 2020 73

TABLE 1. Companies related to the Chivor 1 and 2 emerald mines in Colombia.

Formation Date Incorporators and Agents in

Company Name References
Registration Date Colombia

Compañía de las Minas de

Esmeraldas de Chivor; August 20, 1898 Francisco Restrepo, Enrique González, Escritura 1337, August 20, 1898,
Sociedad Ordinaria de las Minas (Formation in Colombia) Bernardo Escovar, Nuñez & Cía. Notaria 3
de Chivora

February 14, 1914 August Stauch, Hjalmar Schacht, Historical Archive of the German
Syndikat El Chivor
(Formation in Germany) Rudolf Hahn & Sons, Fritz Klein Ministry of Foreign Affairs

June 22, 1918 State of Delaware, Department of

S.J. Nathan, Edgar W. Ward, J.E.
Colombian Emerald Syndicate, (Formation in U.S.) State, Division of Corporations
Cochrane, M.F. O’Dell, F. A. Daly
Ltd. May 7, 1920 Escritura 1402, May 7, 1920,
Agent: Miguel S. Uribe Holguin
(Registration in Colombia) Notaria 1

Colombian Chivor Emerald

Wilson E. Griffiths, Carl K. MacFadden
Company
Compañía Colombiana de July 6, 1920 Directors: Carl K. MacFadden, Escritura 2041, July 6, 1920,
Esmeraldas de Chivor (Formation in Colombia) Nathaniel W. Ross, William M. Schaill Notaria 1
(Operating company of the
Agent: Miguel S. Uribe Holguin
Colombian Emerald Syndicate)

March 6, 1924 State of Delaware, Department of

Samuel C. Wood, Harry C. Hand,
(Formation in U.S.) State, Division of Corporations
Chivor Emerald Corporation Raymond J. Gorman
February 16, 1925 Escritura 344, February 16, 1925,
Agent: Miguel S. Uribe Holguin
(Registration in Colombia) Notaria 1

Colombia Emerald Development November 7, 1924 State of Delaware, Department of

Corporation Edmund J. MacNamara, Ernest W.
(Formation in U.S.) State, Division of Corporations
(Operating company in U.S. and Brown, William B. Anderson
Colombia, connected to the July 11, 1925 Escritura 1607, July 11,1925,
Agent: Miguel S. Uribe Holguin
Chivor Emerald Corporation) (Registration in Colombia) Notaria 1

September 28, 1933

Chivor Emerald Mines, Inc. Agents (1920–1970): State of Delaware, Department of
(Change of name in U.S.)
(Change of name from the Miguel S. Uribe Holguin State, Division of Corporations
Colombia Emerald Development November 17, 1934 Hernando Uribe Cualla Escritura 2045, November 17,
Corporation) (Change of name Eduardo Torres R. 1934, Notaria 1
registered in Colombia)
a
Both names were used interchangeably in various documents between 1908 and 1921.

the process. Emerald samples were sent to Europe to rate of 5 pesos per mine per year. Taxes for the mines
attract investors (e.g., major European banks), and an were then paid for 20 years in advance, at the rate of
estimated price of £40,000 was suggested in early ne- 100 pesos per mine, by Antonio Núñez of Núñez &
gotiations.45 However, no deal culminated before out- Compañía on behalf of the shareholders.47
break of the Colombian civil war (October
1899–November 1902) stopped all activities at the
45
mine. Meanwhile, controversy had also developed Letter from Núñez & Compañía, December 16, 1898, to Restrepo.
Lorenzo Merino, who had recently completed a ten-year lease of the
over the water supply, as the irrigation canal ran across Muzo emerald mines (1886–1896), was involved in the discussions.
private land owned by the local Acosta family, and 46
Letters by Núñez & Compañía, June 18 and June 23, 1898, to
available water was also used for a nearby hacienda.46 Restrepo.
47
In the following era, activities pertaining to Escritura 1242, December 3, 1901, Notaria 3, Archivo General de
la Nación (Colombia). See also Reports by the German Resident in
Chivor tended to be more financial than operational Bogotá, Dorotheus Kracker von Schwartzenfeldt, November 16,
in nature. Tax payments, share transfers, and invest- 1912, and by a member of the German Residency in Bogotá, Wil-
ment pursuits featured heavily, while mining was helm Gerlach, November 15, 1912, German Federal Archive; Diario
Oficial (Colombia), 48, No. 14720 (1912), pp. 812–813; Report by
relatively minimal. Annual taxes for Chivor 1 and 2 Dr. Antonio José Cadavid, March 18, 1913, Historical Archive of the
were paid by González in 1898, 1899, and 1900 at the German Ministry of Foreign Affairs; Domínguez, 1965.

74 HISTORY OF CHIVOR, PART I: 1880–1925 GEMS & GEMOLOGY SPRING 2020

TABLE 2. Transfers of shares of the Compañía de las Minas de Esmeraldas de Chivor between 1899 and 1909.

Escritura 1337, August 20, 1898, Notaria 3: 100 shares are issued and distributed among 4 shareholders as follows: Restrepo 40, González
20, Escovar 20, Núñez & Compañía 20

Approximate
Number Escritura (notarial act), Archivo General
Vendor Buyer Total Price price per share
of shares de la Nación (Colombia), Bogotá
in pesos oro a,c

Francisco Restrepo Carlos Uribe 15 1,000 £ ≙ 333 Escritura 189, February 13, 1899, Notaria 3
5,000 pesos oro
Francisco Restrepo Carlos Uribe 2 4,000 pesos paper 600 Escritura 868, June 13, 1899, Notaria 3
Francisco Restrepo Álvaro Uribe 2 4,000 pesos paper 600 Escritura 908, June 16, 1899, Notaria 3
Francisco Restrepo Diódoro Sánchez 1 2,000 pesos paper 200 Escritura 100, March 27, 1900, Notaria 5
Enrique González Christopher Dixon 1 5,000 pesos paper 500 Escritura 915, December 7, 1900, Notaria 3
Enrique González Christopher Dixon 1 25,000 pesos paper 500 Escritura 1201, November 21, 1901,
Notaria 3
Álvaro Uribe Christopher Dixon 1 48,000 pesos paper 320 Escritura 108, March 9, 1902, Notaria 3
Francisco Restrepo Eduardo Sayer Vega 1 25,000 pesos paper 167 Escritura 109, March 10, 1902, Notaria 3
Carlos Uribe Fould & Compagnie 17 2,000 £ ≙ 588 Escritura 1848, October 10, 1905, Notaria 2
10,000 pesos oro
Francisco Restrepo Emiliano Isaza 3 10,000 pesos paper 33 Escritura 758, October 28, 1905, Notaria 3
Eduardo Sayer Vega Lisandro Moreno 1 30,000 pesos paper 300 Escritura 851, December 4, 1905, Notaria 3
Lisandro Moreno Eduardo Sayer Vega 1 30,000 pesos paper 300 Escritura 134, March 7, 1906, Notaria 3
Núñez & Compañía Fould & Compagnie 20 400 £ ≙ 100 Escritura 1834, December 17, 1907,
2,000 pesos orob Notaria 2
Eduardo Sayer Vega Luis Francisco Moreno 1 64,000 pesos paper 640 Escritura 749, October 15, 1908, Notaria 3
Luis Francisco Emiliano Isaza 1/2 22,500 pesos paper 450 Escritura 174, March 20, 1909, Notaria 3
Moreno
Luis Francisco Francisco Restrepo 1/2 22,500 pesos paper 450 Escritura 174, March 20, 1909, Notaria 3
Moreno
a
Conversion factors used for pesos oro:pesos paper are 1:3.33 (1899), 1:10 (1900), 1:50 (1901), 1:150 (1902), 1:100 (1905 and later). With a total of
100 shares outstanding, 1 share represents 1% ownership of the company (see footnote 48 in text).
b
Núñez & Compañía sold multiple assets, not only shares in Compañía de las Minas de Esmeraldas de Chivor, to Fould & Compagnie.
c
The price for 90% of the mine offered to F. Klein in 1914 was £51,000 ≙ 255,000 pesos oro; the value for 1% of the mine (equivalent to 1 share) would
be calculated at 2,833 pesos oro.
The sale price for the mine and the land to W.E. Griffiths and C.K. MacFadden in 1919 was £46,000 ≙ 230,000 pesos oro; the value for 1% of the mine
(equivalent to 1 share) would be calculated at 2,300 pesos oro.

A lively trade in shares of the Compañía de las

48
Minas de Esmeraldas de Chivor also took place dur- Conversion factors reflected in certain documents indicate £1 ster-
ling ≙ 5 pesos oro. At the time of the civil war (1899–1902), inflation
ing the first decade from 1899 to 1909, and examples caused rapid change in the exchange rate of the gold peso (peso oro)
of these transactions are summarized in table 2.48 A to paper currency, for example: 1:10 (1900), 1:50 (1901), 1:189 (Oc-
notable series of transfers began in 1899 when Re- tober 1902), and 1:260 (1902/1903). After the war, the value of the
paper currency stabilized, and in 1904, the exchange rate was fixed
strepo sold 15 shares to Carlos Uribe,49 who in turn by the government at 1:100. Mitteilungen des Deutsch-
sold 17 shares in 1905 to Fould & Compagnie, a südamerikanischen und Iberischen Instituts, 3 (1915), p. 198; Bürger,
Parisian bank.50 Two years later, in 1907, Fould & 1919; Eder, 2017.
50
Escritura 1848, October 10, 1905, Notaria 2, Archivo General de la
Compagnie became the majority shareholder upon Nación (Colombia).
acquiring an additional 20 shares from Núñez & 51
Escritura 1834, December 17, 1907, Notaria 2, Archivo General de la
Compañía.51 Other individuals who would play a role Nación (Colombia).

HISTORY OF CHIVOR, PART I: 1880–1925 GEMS & GEMOLOGY SPRING 2020 75

 TABLE 3. Currency equivalencies circa 1910–1914.a

£ Sterling Peso Oro $US Gold Mark

£ Sterling 1 5 5 20
Peso Oro 0.2 1 1 4
$US Gold 0.2 1 1 4
Mark 0.05 0.25 0.25 1
a
As used in various publications and contracts of the period.

then taken in March 1904, when Álvaro Uribe was

authorized for a nine-month term to search for a for-
eign entity, especially in France and England, to lease
Chivor for an estimated price of £10,000.57 As to the
apparent lack of success of such efforts, the United
States Consul-General in Bogotá, Snyder, wrote in
1904: “The prohibitively high price asked by the own-
ers of these five mines, called the ‘Chivor group’ is
the reason, I am informed, that they are not being
worked.”58
Meanwhile, a few months before the end of the
civil war, Restrepo had left Colombia. He arrived in
Quito, Ecuador, in June 1902, accompanied by his
older son Anibal and two other Colombians. In
Ecuador, Restrepo worked on government projects
Figure 6. Dr. Emiliano Isaza Gutiérrez—lawyer, scien-
such as road and railway construction, and he was
tist, politician, diplomat, and minister—was one of
the most prominent shareholders in the Compañía de
involved in gold mining in the Oriente region as
las Minas de Esmeraldas de Chivor. He cooperated owner and/or supervisor of several gold mines.59 In
with Klein in the 1910s but as president sold the com- early 1904, Restrepo was asked to return to his home
pany in 1919 to American interests. Courtesy of Ban- country because there was urgent need of an experi-
repcultural, Bogotá.

52
Escritura 915, December 7, 1900, Notaria 3, Escritura 1201, No-
vember 21, 1901, Notaria 3, and Escritura 108, March 9, 1902, No-
in Chivor’s history likewise became involved as taria 3, Archivo General de la Nación (Colombia).
shareholders during this period, such as the English 53
Dr. Emiliano Isaza Gutiérrez (1850–1930) studied law and political
science, becoming a professor at the University of Antioquia in 1875.
businessman and miner Christopher Ernest Dixon In 1876 he moved to Bogotá. Throughout his career, he authored a
(1868–1961; see box A), who purchased three shares number of academic publications, including a classic Spanish gram-
between 1900 and 1902,52 and Dr. Emiliano Isaza mar textbook Gramática Práctica de la Lengua Castellana (1880), and
he served Colombia in various diplomatic capacities, such as
Gutiérrez (figure 6),53 who acquired three shares in plenipotentiary minister in Ecuador and as minister of public instruc-
190554 and would later become president of the Com- tion in the homeland (1908–1909). See Martínez, 1950.
pañía de las Minas de Esmeraldas de Chivor. Prices 54
Escritura 758, October 28, 1905, Notaria 3, Archivo General de la
for one share in paper currency increased from 2,000 Nación (Colombia).
55
pesos in 1900 to 64,000 pesos in 1908, but the infla- For comparison, monthly salaries at the Muzo emerald mine in
1904 were 30,000 pesos paper (300 pesos oro) for the director and
tion factor impacting paper currency changed 3,000–4,000 pesos paper (30–40 pesos oro) for the foreman or other
markedly during the Colombian civil war (again, see staff. Junta Nacional de Amortización, Informe al Congreso y al Gob-
table 2).55 The values of different Colombian and for- ierno Ejecutivo 1904, p. 62.
56
Escritura 1172, November 12, 1901, Notaria 3, Archivo General de
eign currencies are compared in table 3. la Nación (Colombia).
Throughout the same period, the mine owners 57
Escritura 172, March 5, 1904, Notaria 3, and Escritura 193, March
continued seeking foreign investment or participa- 16, 1905, Notaria 3, Archivo General de la Nación (Colombia).
tion. In November 1901, Dixon was authorized for a 58
Bulletin of the International Bureau of the American Republics, Vol.
18 (1904), pp. 858–859.
term of six months to attempt to form a company in 59
Restrepo’s activities in Ecuador between 1902 and 1904 are well
Europe that would be responsible for organizing and documented in newspaper articles, contracts, reports, and letters re-
operating the emerald mines.56 A different tack was maining in the possession of his family.

76 HISTORY OF CHIVOR, PART I: 1880–1925 GEMS & GEMOLOGY SPRING 2020

 Chivor

Río
Sinai

Río Río
Rucio Sinai

Río
Rucio

0 500 m

Figure 8. Satellite photo of the region between the val-

leys of the Río Rucio and the Río Sinai where the
Figure 7. Map published in 1921 showing the location areas covered by the mining titles for Chivor 1,
of the Chivor 1 and 2 emerald mines between the val- Chivor 2, and Triunfo de Chivor are located. The area
leys of the Río Rucio and the Río Sinai in the Depar- is situated in the Departamento de Boyacá, near the
tamento de Boyacá, near the boundary with the boundary with the Departamento de Cundinamarca
Departamento de Cundinamarca. The mining titles (yellow line), which runs partly along the Río Rucio
for Chivor 1 and 2, measuring one square kilometer valley. Modified from Google Earth; the town of
each, were registered in 1889 to Francisco Restrepo Chivor and the labeling were added.
and partners; three additional mining titles were reg-
istered in 1904 for the Triunfo de Chivor, Paz de
Chivor, and Esperanza de Chivor claims. After las Minas de Esmeraldas de Chivor was treated as the
Canova (1921), representing the mine boundaries as owner of the Chivor 1 and 2 mining titles and of land
assumed by the Colombian Emerald Syndicate, Ltd.; in the area. Through this purchase, the mine owners
the town of Chivor was added. Note: In a map pre-
resolved with the Acosta family the controversy over
pared by the Ministerio de Minas y Petróleos in 1941,
the water supply that had plagued the site since the
Paz de Chivor and Esperanza de Chivor are located
much closer to Chivor 1 and 2, at a distance of ap- late 1890s.
proximately 1 km. On December 13, 1904, three additional mining
titles were registered in the vicinity of Chivor 1 and
2: “Triunfo de Chivor,” “Paz de Chivor,” and “Esper-
enced miner to direct the Muzo emerald mines. Re- anza de Chivor” (figures 7 and 8), with Restrepo,
strepo began in that position in March 1904.60 Henrique Campo, and Dixon named as the co-own-
The year 1904 also saw a widening of the focus ers.62 From that time, the five mining titles in the
for investment activities pertaining to Chivor, be-
yond the existing two claims. On August 25, 1904,
Restrepo, González, Escovar, Carlos Uribe, and 60
El Ecuador. Guía Comercial, Agrícola, e Industrial de la República
Núñez & Compañía purchased land in the Chivor (1909), Guayaquil, Compañía “Guía del Ecuador,” pp. 243–251;
area, between Río Rucio and Río Sinai, from Aurelio Ochoa Ortiz, 1921.
61
Escritura 1479, August 25, 1904, Notaria 2, and Escritura 3084,
Ruedo Acosta.61 The purchase price was 100,000 December 27, 1919, Notaria 1, Archivo General de la Nación
pesos, and 20,000 pesos were paid by each of the five (Colombia).
principal shareholders. Thereafter the Compañía de 62
Diario Oficial (Colombia), 48, No. 14720 (1912), pp. 812–813.

HISTORY OF CHIVOR, PART I: 1880–1925 GEMS & GEMOLOGY SPRING 2020 77

 BOX A: CHRISTOPHER DIXON’S LEGACY IN COLOMBIAN EMERALD MINING
stay in Colombia, George might also have been involved
in emerald prospecting and mining at Muzo, as suggested
by photos from the family album (figures A-2 and A-3).5
Although Christopher Dixon would become most
known for his activities with emerald, he was always
looking for new commercial opportunities and was ac-
tive across a variety of sectors throughout his career. In
1891, he became associated with the “Magdalena Estates
Company,” a real estate development venture.6 In the
1892 to 1893 period, he and his brother Frank obtained
grants to parcels of public land in northern Tolima and
Caldas.7 In 1910, he was named administrator of the “H.
and U. Rubber and Coffee Estates, Limited.”8 Then, in
1916, Dixon turned to cattle breeding, already an impor-
tant industry in Colombia, and planned to build a new
meat-packing plant on the Caribbean coast.9
The earliest known reference to emeralds in Dixon’s
career is from 1896, when he became an agent of the Eng-
lish Mining Syndicate, Limited, a British company oper-
ating the Muzo mines, a position he held until October
1900.10 The early 1900s then found Dixon buying shares
in the Compañía de las Minas de Esmeraldas de Chivor
(see the main text) and being involved in various other
activities around Chivor.
Meanwhile, in August 1898, preliminary applica-
tions for three additional mining claims in the Chivor
Figure A-1. Christopher Dixon worked on behalf of the
area had been filed by Francisco Restrepo and various
English Mining Syndicate, Limited, when the company
partners.11 Such action was taken contemporaneously
operated the Muzo mine during the 1890s and early
with when Restrepo was involved at Chivor in efforts to
1900s. Dixon later became one of the shareholders in
the Compañía de las Minas de Esmeraldas de Chivor
restore the water supply. The three mining claims later
that owned the mining titles for Chivor 1 and 2. After became known as El Triunfo de Chivor, La Paz de
the mine was sold in 1919 to the U.S.-based Colom- Chivor, and La Esperanza de Chivor.
bian Emerald Syndicate, Ltd., he served as its official
representative in Colombia in the early 1920s. Photo 1
circa 1900, public domain. Transactions of the Institution of Mining and Metallurgy, Vol. 21
(1912), p. 723.
2
The arrival of three members of the Dixon family in Colombia in
1886 is documented in the Diario Oficial (Colombia), 22, No. 6836
Christopher Ernest Dixon was born in 1868 in Watling- (1886), p. 1114, but further information as to identities is unavailable.
ton, Oxfordshire, England, the eleventh of Dr. Henry and 3
Informe presentado al Congreso de la Republica en sus sesiones or-
Helen Dixon’s twelve children. In the mid to late 1880s, dinarias de 1890 por el Ministro de Hacienda, Casa Editorial de J.J.
Pérez, Bogotá, 1890.
he traveled to South America along with his brother Frank 4
Skinner, 1889–1890, 1891–1892.
(1865–1956) to join an elder brother George (1861–1911), 5
Simon Hamilton (great-grandson of George Dixon), pers. comm.,
who had been prospecting for gold and silver in the state 2018.
of Tolima, in central Colombia, since 1884.1 6
The Mining Journal, Railway and Commercial Gazette, Vol. 61
Although the activities and precise whereabouts of the (1891), p. 851.
two younger Dixon brothers during their early years in 7
Informe del Ministro de Hacienda de la Republica de Colombia al
Colombia2 remain obscure, by 1889, at the age of 21, Congreso Constitucional de 1894, p. 101; LeGrand, 1988.
Christopher (figure A-1) was already operating at least 24 8
India-Rubber Journal, Vol. 43 (1912), p. 21.
gold and silver mines in the Mariquita area, in northern 9
Memoria del Ministro de Agricultura y Comercio al Congreso de
Tolima.3 During this period, George worked for British 1918, pp. 173–177.
10
mining companies at the Silencio mine, in central Tolima Atuesta, 1899; Escritura 314, October 31, 1900, Notaria 3, Archivo
General de la Nación (Colombia); Gaceta Judicial (Colombia) 15 (1901)
and elsewhere.4 Over the course of the 1880s, Tolima had pp. 169–173 and 16 (1903) pp. 78–79; Céspedes Cubides, 2017.
become a mining bonanza for gold and silver seekers, at- 11
Diligencia No. 2, Municipality of Macanal, August 8, 1898; Aviso
tracting prospectors and investors from neighboring An- No. 81, Municipality of Ubalá, August 21, 1898; Aviso No. 82, Au-
tioquia, as well as many foreigners. During his seven-year gust 21, 1898, Municipality of Ubalá.

78 HISTORY OF CHIVOR, PART I: 1880–1925 GEMS & GEMOLOGY SPRING 2020

Figure A-2. View of the landscape of the Muzo area, from Figure A-3. A bridge in the Muzo area, from the family
the family album of George Dixon, taken before 1890, album of George Dixon, taken before 1890, with the
with the handwritten comment reading: “A picturesque handwritten comment reading: “Another rustic bridge,
native bridge across stream. Road leading to Muzo close to Muzo mines. Indian miners bidding adieu to
mines.” Courtesy of Simon Hamilton. their wives, no women are permitted to enter the estab-
lishment or to pass beyond a given boundary.” Courtesy
of Simon Hamilton.
As shown in a map by Canova (1921), El Triunfo de
Chivor adjoined directly to Chivor 1 and 2, situated be-
tween the valleys of the Río Rucio and the Río Sinai (see periods in particular saw flurries of activity in this regard,
figure 7 of the main text). La Paz de Chivor and La Es- with Dixon engaged in extensive correspondence and
peranza de Chivor were located northeast and southeast, even travel to England17 in the search for investors. Par-
respectively, of the current town of Chivor. No indica- alleling these efforts was Dixon’s long-running battle
tion exists as to why these three particular locations against the change in Colombian mining laws that had
were chosen for claims—it is possible that emerald sam- taken effect in 1905 (Law No. 40 of 1905), restricting ap-
ples and/or former native or Spanish workings (tunnels) plications for new emerald mines and increasing the tax
had been discovered by prospecting or other activities in burden on existing mines. A complaint was lodged
the areas and that the claims were defined so as to cover against the Colombian government, and Dixon, sup-
mining rights over those regions. Regardless, and for rea- ported by British officials in Bogotá, embarked on a multi-
sons otherwise not elucidated but perhaps impacted by year quest to establish that he had been harmed by the
the outbreak of the Colombian civil war (The Thousand change and was entitled to recompense from Colombia.18
Days’ War, 1899–1902), the registration process was not Dixon’s allegations centered on concerns that the
completed, and no mining claims were officially granted. higher taxes would prevent profitable mining and should
After the war and Restrepo’s return to Colombia from not be applied to titles granted before 1905 and that the
a sojourn in Ecuador (see main text), the registration uncertain and detrimental situation for mine owners had
process was restarted. A February 1904 agreement be-
tween Henrique Campo (a Swiss citizen),12 Dixon, and Re-
12
strepo set forth the conditions under which the they Also referred to as Enrique Campo in several documents, with his
signature on notarial acts generally taking the form of Henrique
would join in new applications for the three claims, Campo (used hereinafter).
namely, that ownership would be divided equally among 13
Letter from Christopher Dixon, December 3, 1904, to Restrepo.
the three partners (figure A-4). The applications were then 14
Diario Oficial (Colombia), 48, No. 14720 (1912), pp. 812–813.
filed in September 1904,13 and the three mining titles were 15
Foreign Office, Political Departments, General Correspondence
granted on December 13 of that year,14 with annual taxes from 1906 to 1966, Colombia: Code 11, Files 47 - 6847 and 904,
being paid in advance for the period through 1910.15 Ap- The National Archives, Kew, London.
proximately one week later, on December 19, 1904, the 16
Escritura 2326, December 19, 1904, Notaria 5, Archivo General de
three owners formally granted Dixon a power of attorney la Nación (Colombia).
17
covering all future matters related to the three claims.16 File Christopher Dixon, Names and Descriptions of British Passen-
gers Embarked at the Port of Tilbury, January 13, 1915,
Consistent with what had generally transpired to that MyHeritage.com.
point with respect to Chivor 1 and 2, focus for the newly 18
Foreign Office, Political Departments, General Correspondence
acquired claims centered on investment pursuits rather from 1906-1966, Colombia: Code 11, Files 47 - 6847 and 904, The
than practical mining. The 1905 to 1906 and 1915 to 1916 National Archives, Kew, London.

HISTORY OF CHIVOR, PART I: 1880–1925 GEMS & GEMOLOGY SPRING 2020 79

 of the so-called Chivor group is incomplete regarding
specifics of any pertinent ownership transfers, and the
few references are disjointed, if not contradictory, al-
though all five mines eventually came to be considered
in part or in full under the purview of the American
company.23
Following his involvement with the Colombian
Emerald Syndicate, Ltd., Dixon left the mine for a period
but returned in the early 1930s to help Peter W. Rainier,
then the administrator at the mine, fight attacking ban-
dits.24 Later, he also cooperated with Rainier during field
work at the Coscuez emerald deposit.25
Dixon remained in Colombia with his wife and chil-
dren. His last known assignment was in the mid-1940s,
as one of the participants in geological investigations car-
ried out in Colombia by the Foreign Economic Admin-
istration, in cooperation with the United States
Geological Survey.26 In the last years of his life, Dixon
was celebrated in the U.S. as an authority on emerald
mining.27 He died in Bogotá in 1961, and some of his de-
scendants still live in Colombia. He will be remembered
for the legacy he left in the history of Chivor and in the
modern history of Colombian mining at large.

19
Ibid.
20
Ibid.
21
Figure A-4. A letter dated February 23, 1904, from File “Minas de Esmeraldas,” Ministerio de Relaciones Exteriores,
Henrique Campo and Christopher Dixon to Restrepo Archivo General de la Nación (Colombia).
22
memorialized an agreement under which the three Diario Oficial (Colombia), 50, No. 15279 (1914), p. 396; Diario
Oficial (Colombia), 51, No. 15557 (1915), p. 448; Diario Oficial
would join in applying for the three additional
(Colombia), 54, No. 16306 (1918), p. 121; Diario Oficial (Colombia),
claims in the Chivor area, sharing ownership equally 58, No. 18425–18426 (1922), p. 243.
and with all expenses paid by Dixon and Campo. 23
For example, a 1920 listing of Restrepo’s assets at the time of his
Courtesy of Eduardo Restrepo Ortega. death in 1914 included a one-third share in each of the three newer
mines. Escritura 188, January 27, 1920, Notaria 3, Archivo General
de la Nación (Colombia). In 1921, it was represented that the titles to
all five mines of the Chivor group were under the ownership or con-
caused potentially interested investors to turn away.19 trol of the Colombian Emerald Syndicate, Ltd., or its operating entity
His assertions also apparently became increasingly ex- the Colombia Chivor Emerald Company. Canova, 1921. Concomitant
aggerated as the contest went on, with documents pre- with the bankruptcy of the Colombian Emerald Syndicate in 1923, its
pared by British officials even giving Dixon credit for assets were stated to incorporate complete ownership of Chivor 1 and
2 and one-third of the three newer claims. Jewelers’ Circular, Vol. 87,
having discovered the three mines.20 In negotiations he No. 22 (1923), p. 91. Annual payments of taxes for the three claims
offered to settle by selling the mines to the Colombian were made from 1929 to 1931 by Pedro Ignacio Uribe, a Colombian
government for £200,000 but was asked to prove the entrepreneur and public figure (as well as founder of the “Compañía
value of the claims through an international expert.21 He Colombiana del Petróleo” in 1918 along with Wilson E. Griffiths and
Carl K. MacFadden). Diario Oficial (Colombia), 54, No. 16364
failed to do so and instead continued paying annual taxes
(1918), p. 56; Diario Oficial (Colombia), 65, No. 21072 (1929), p.
for the three mines up through 1922.22 191; Diario Oficial (Colombia), 65, No. 21275 (1929), p. 780; Diario
By 1920, Dixon had been named as the Colombian Oficial (Colombia), 67, No. 21674 (1931), p. 189.
manager for the Colombian Emerald Syndicate, Ltd., the 24
See main text, Part II, and Rainier, 1942.
American company that had just taken over the Chivor 25
Rainier, 1933.
1 and Chivor 2 mines (see main text). From this date on- 26
Singewald, 1950.
wards, the historical record about the three other claims 27
Lancaster Eagle-Gazette, July 2, 1951, p. 14.

80 HISTORY OF CHIVOR, PART I: 1880–1925 GEMS & GEMOLOGY SPRING 2020

 ACTIVITIES AT MUZO
As alluded to previously, the year 1904 also wit-
nessed relevant activity connected to the Muzo
mines and to the broader Colombian emerald story
that included Chivor. The important mines at Muzo,
as well as at Coscuez, had been owned by the Colom-
bian government since 1886. Nonetheless, the mines
were often leased to private operators against prede-
termined charges, and those tenants might even sub-
lease to another individual or company. There were
also repeated cycles of inactive periods.68
After the civil war, an entity known as the Sindi-
cato de Muzo (Syndicate of Muzo, used hereinafter)
Figure 9. Handwritten register reflecting the owner- was founded and took over exploitation of the Muzo
ship distribution of shares in the Chivor 1 and 2 mines from 1904 to 1909.69 Restrepo was named as
emerald mines before and after a stock split between director, and his son Anibal joined the mine’s payroll
1909 and 1912 that increased the capital from 100 to as well.70 Toward the end of that period, in December
250 shares. The participants and their holdings re- 1908, it was agreed between the government of
mained unchanged until the mines were sold in De- Colombia, the Syndicate of Muzo, and the London-
cember 1919 to an American group. Courtesy of
based Colombian Emerald Company that the right
Eduardo Restrepo Ortega.
to mine at Muzo from March 1909 would pass to the
Colombian Emerald Company.71 Restrepo remained
in his position until resigning in December 1909, ex-
Chivor area were occasionally referred to collectively
pressing in a letter to the Colombian president that
as the Chivor group (see also box B),63 but ownership
he did not agree with the new company’s policies.72
of the three mining claims registered in 1904 and
The offered justification was that such policies were
that of Chivor 1 and 2 remained, at least before 1920,
contrary to the interests of the Colombian nation.
formally separate (see also box A).
The contract with the Colombian Emerald Com-
Yet, in stark contrast to the ongoing financial and
pany was to last for more than 20 years, but it was
investment maneuvers, practical mining activities
terminated by the government after just 14 months,
were nearly nonexistent. As noted above, prior to the
in May 1910.73 Mining operations were taken over by
onset of the civil war, initial steps to clean old infra-
structure and restore the water supply were under-
taken,64 but endeavors were curtailed by the conflict.
63
The decade after the war then saw Restrepo, one of For other mines that have been erroneously associated with Chivor,
see also box B.
the key participants with practical mining experi- 64
See also Bulletin of the International Bureau of the American Re-
ence, either out of the country (in Ecuador) or occu- publics, Vol. 28 (1909), pp. 1027–1039.
pied with other ventures (Muzo). With Restrepo thus 65
Escritura 3084, December 27, 1919, Notaria 1, Archivo General de
unavailable and the majority of the other sharehold- la Nación (Colombia).
66
ers focused on selling the mine, there was little in- Diario Oficial (Colombia), 48, No. 14720 (1912), 812–813; Escrit-
ura 3084, December 27, 1919, Notaria 1, Archivo General de la
centive to invest further in operations. Nación (Colombia).
By 1909, the shares were distributed among Fould 67
Escritura 3084, December 27, 1919, Notaria 1, Archivo General de
& Compagnie, six Colombians (mostly lawyers or la Nación (Colombia).
68
engineers), and Dixon.65 Then, between March 1909 Informe del Ministro de Hacienda al Congreso de 1922, pp. 24–29.
69
Diario Oficial (Colombia), 49, No. 15043 (1913), p. 3221.
and October 1912, a stock split increased the capital 70
Junta nacional de Amortización, Informe al Congreso y al Gobierno
from 100 to 250 shares (figure 9).66 By the time of the Ejecutivo 1904, p. 60; Revista del Ministerio de Obras Públicas, 4
1919 sale discussed below, ownership was distrib- (1909), p. 49; Domínguez, 1965.
uted as follows: 92½ shares for Fould & Compagnie, 71
Bulletin of the International Union of the American Republics, Vol.
50 for descendants of Escovar, 45 for González, 41¼ 29 (1909), pp. 1120–1121; Montaña, 1915.
72
Letter from Restrepo, December 22, 1909, to the Colombian presi-
for the widow and descendants of Restrepo, 8¾ for dent Ramon González Valencia.
Isaza, 7½ for Dixon, 2½ for Álvaro Uribe, and 2½ for 73
Report by a member of the German Residency in Bogotá, Wilhelm
Diódoro Sánchez (figure 9).67 Gerlach, November 15, 1912, German Federal Archive.

HISTORY OF CHIVOR, PART I: 1880–1925 GEMS & GEMOLOGY SPRING 2020 81

 BOX B: MINING COMPANIES ERRONEOUSLY ASSOCIATED WITH CHIVOR
Two additional mining enterprises were associated with ing in 1900. The corporate reconstruction had brought
Chivor by Sinkankas (1981): Somondoco Emeralds, Lim- sufficient funds to pay off liabilities, but practical work
ited, a successor entity to the Somondoco Company, had suffered due to the ongoing civil war (1899–1902). Al-
Limited; and the Emerald Mining Company of Colom- though the superintendent Mr. Gage and the consulting
bia.1 The first company Sinkankas assumed worked at engineer Mr. Russell had been working at the deposit, and
Chivor following the rediscovery by Restrepo but before some stones were ready to be sent overseas, export was
Restrepo and Klein began their joint mining efforts there. impossible given the hostilities and political disruptions.5
The second company he realized did not fit with avail- The situation at the 1901 annual meeting was simi-
able references concerning the rediscovery and subse- lar. Mr. Gage and Mr. Russell reported that they “had
quent work. Evidence indicates, however, that each was not been able to obtain sufficient emeralds to put a prof-
actually involved with properties unrelated to the his- itable consignment upon the market,” but they still
torical Spanish mines. hoped to make the mine productive: “[W]e have passed
through a period of disturbance in the veins, but that is
SOMONDOCO COMPANY, LIMITED, AND a matter, it appears, we get full compensation for, be-
SOMONDOCO EMERALDS, LIMITED cause a disturbance of the veins is generally a precursor
of good and valuable stones when the veins reunite.”6
Somondoco Company, Limited, was registered in Lon- Such expectations fell short, however, as a court order
don under chairman H.S. Sankey on July 1, 1895, with to wind up the company was issued on December 15,
the stated purpose of acquiring the Somondoco emerald 1904, on the ground that the company was unable to pay
mines in the Republic of Colombia.2 Somondoco Emer- its creditors.7
alds, Limited, was then registered, also in London and The foregoing publications tracking the company’s
under chairman H.S. Sankey, on May 13, 1899, as a re- progress never clearly indicated the location of the prop-
construction of the Somondoco Company, Limited, and
with property reported to comprise the Somondoco
emerald mines in the Republic of Colombia.3 1
Sinkankas, 1981, p. 411.
Industry publications track the company’s progress 2
Skinner, 1896, 1897.
through the end of the nineteenth century and the first 3
Skinner, 1900.
few years of the twentieth century. As of 1899, the situ- 4
The Mining Journal, Vol. 69 (1899), p. 936.
ation was not promising: “Up to the present it has 5
The Mining Journal, Vol. 70 (1900), p. 1482.
achieved nothing of importance, and the valuable stones 6
The Mining Journal, Vol. 71, (1901), pp. 1400–1401.
it has found are only to be counted by the half-dozen. It 7
Register of defunct companies, 1990. The International Stock Ex-
now finds itself in sore straits for money….”4 Some im- change of the United Kingdom and the Republic of Ireland, 2nd ed.,
provement was reported at the annual shareholders meet- Macmillan Publishers, London, 552 pp.

the Colombian government from June 1910, and the stones, mostly red and green tourmaline and aqua-
government then appointed Restrepo as director on marine, in Brazil in the last decade of the nineteenth
its behalf from January to May 1911.74 In early Janu- century. August Senior was supported in Brazil by his
ary 1913, production at Muzo was discontinued for sons Gustav (1876–1931), Fritz (1882–1953), and Au-
several years due to lack of profitability75 and only gust Junior (1886–1971) from 1902, 1904, and 1906,
resumed in 1923.76 Meanwhile, the termination of respectively.77 Although the Klein brothers were suc-
the contract in 1910 had resulted in lengthy negoti- cessful in their endeavors as gem merchants in
ations on claims for damages involving both the Brazil, competition from other Idar traders in pur-
Colombian Emerald Company and the Syndicate of chasing rough gemstones in Minas Gerais increased.
Muzo.
74
Diario Oficial (Colombia), 47, No. 12241 (1911), p. 404; Diario
FRITZ KLEIN AND HIS EARLY ACTIVITIES IN Oficial (Colombia), 49, No. 15043 (1913), p. 3221.
COLOMBIA (1905–1913) 75
Pogue, 1917.
76
Diario Oficial (Colombia), 59, No. 19011–19014 (1923), p. 572;
Klein’s family hailed from Idar, Germany, and was Diario Oficial (Colombia), 59, No. 18849–18850 (1923), pp. 658–
engaged in the gem trade. His father August Klein 659.
(1852–1927) began mining and purchasing gem- 77
Falz, 1939.

82 HISTORY OF CHIVOR, PART I: 1880–1925 GEMS & GEMOLOGY SPRING 2020

 erty involved. Nonetheless, other materials offered fur- EMERALD MINES OF COLOMBIA, LIMITED, AND
ther insight. A 1902 report by the British Vice-Consul S.S. EMERALD MINING COMPANY OF COLOMBIA
Dixon gave the Somondoco district as the location of the Emerald Mines of Colombia, Limited, was registered in
mine worked by Somondoco Emeralds, Limited, but it London on September 28, 1888.10 The company pur-
distinguished those workings from the historical Chivor chased the Yacopi mines, located in the Western Colom-
mines. After discussing both Somondoco Emeralds, Lim- bian Emerald Belt, where the Muzo and Coscuez mines
ited, and the Muzo deposits, S.S. Dixon continued: “I am are also situated. The following decade, a company called
informed that an emerald-producing formation of great the Emerald Mining Company of Colombia was men-
importance has been discovered by the aid of old Spanish tioned in a report by George Frederick Kunz incorporated
parchments in the Somondoco district, but as yet has not in a publication from the United States Geological Sur-
been worked by the discoverers. At present only the old vey.11 Kunz noted that a property had been acquired by
Spanish tunnels and workings indicating that in past cen- the company that was expected to produce emeralds
turies great mining operations had been carried on “quite as fine as those from the famous Muzo mine.”
there….”8 This report is the oldest known published ref- Based on this report, the Colombia Mining Company of
erence related to the rediscovery of Chivor. Colombia was later potentially associated with Chivor
With similar import, a map published in 1913 by Sinkankas, as previously stated. The identity of the
showed emerald mines in the valley of the Río Somon- property had not been specified by Kunz, leading to spec-
doco near the small municipalities of Somondoco and ulation connecting the company with the likewise fa-
Guazeque [Guateque]9, and Pogue (1917) expressly dif- mous Chivor mines. However, research indicates that the
ferentiated the historical Chivor workings from those Emerald Mining Company of Colombia referred to by
along the river: “Some confusion of terms has arisen. Kunz was most likely Emerald Mines of Colombia, Lim-
These deposits [Somondoco or Chivor] are near the val- ited, and therefore was engaged at Yacopi, not Chivor.
ley of Chivor, and have been called the Chivor deposits.
Under Spanish dominion they came to be known as the
Somondoco deposits… and by this name they are gener- 8
Dixon, 1902. A nearly identical text was published anonymously in
ally called in the literature. They do not, however, occur
Monthly Bulletin of the International Bureau of the American Re-
on Río Somondoco, but other less important deposits are publics, 13 (1902), pp. 1349–1350.
known on that stream.” Hence, all evidence points to 9
Gamba, 1913.
one of these less important deposits in the Somondoco 10
Skinner, 1889–1890, 1891–1892.
valley being the locality mined without great success by 11
Mineral Resources of the United States: Calendar Year 1893. United
the British “Somondoco” company. States Geological Survey, 1894, pp. 696–697.

The growing competition prompted Fritz to travel Given these incidents, returning to Bogotá would
through various South American countries such as have been advisable only under protection of the
Argentina, Bolivia, and Peru, always searching for German Resident.80 Nonetheless, Klein continued to
new gem materials, sources, and trade opportunities. pursue his goal of establishing himself in the Colom-
Fritz Klein’s first trip to Colombia occurred in bian emerald trade. In December 1907, for instance,
1905 and nearly ended in disaster. Purchase and sale
of gemstones at that time was strictly controlled
78
through government-authorized channels, and Klein Prior to 1914, the German empire designated ambassadors to only
a small number of important nations, with such individuals serving
sought to buy and export rough emeralds from a pri- in cities such as Paris, London, Rome, Madrid, Washington, and St.
vate Colombian middleman in violation of applica- Petersburg. For other countries, including Colombia, the German
Empire designated officials termed residents, who held rights for
ble regulations. He was accused of this infraction by purposes of international law essentially equivalent to those of am-
the Syndicate of Muzo and arrested in Honda, in the bassadors.
department of Tolima. The emeralds he had acquired 79
Report by the former translator of the German Resident, Septem-
were confiscated. Only through intervention of the ber 1925, Historical Archive of the German Ministry of Foreign Af-
fairs.
German Resident78 in Bogotá was he later released 80
Letter from Fritz Klein, January 19, 1909, to his brother-in-law
and permitted to leave the country.79 August Hahn in Idar, Germany.

HISTORY OF CHIVOR, PART I: 1880–1925 GEMS & GEMOLOGY SPRING 2020 83

Figure 10. Dr. Ernst Jäckh, the brother-in-law of Fritz Figure 11. The German Resident in Bogotá,
Klein, had access to the elite of the German Empire. In Dorotheus Kracker von Schwartzenfeldt, was in-
the 1910s, he served as an unofficial consultant with structed by the German Ministry of Foreign Affairs to
the Ministry of Foreign Affairs, Berlin, focusing on deal- assist Fritz Klein in Colombia, especially as regarded
ings with the Ottoman Empire (Turkey). Jäckh used his purchasing the Chivor emerald mines. Kracker von
influence to further Klein’s efforts both in Colombia Schwartzenfeldt later worked to clarify the legal situ-
and in finding investors for the Chivor project. Photo of ation in response to inquiries from the Syndicate El
Jäckh (center) in Istanbul in 1917 with German Em- Chivor. Photo circa 1910, courtesy of the Historical
peror Wilhelm II (to the right of Jäckh); courtesy of the Archive of the German Ministry of Foreign Affairs.
Ernst Jäckh collection, Columbia University.

Klein spent a short period in Barranquilla, a port city Colombia and stay in Bogotá.83 An order was issued
in northern Colombia, for a meeting with a local to the German Resident in Bogotá, Dorotheus
partner.81 By 1911, Klein’s brother-in-law Dr. Ernst Kracker von Schwartzenfeldt (1869–1935, figure 11),
Jäckh82 (1875–1959, figure 10) had asserted his influ- in support of Klein’s activities.
ence at the Ministry of Foreign Affairs in Berlin to Klein returned to Bogotá in September 1911.84 He
secure the protection needed for Klein to re-enter initially negotiated with the Colombian government

81
Letter from Fritz Klein, December 23, 1907, to his father August April 1933, he realized that he would not be able to keep his positions
Klein in Idar, Germany. as the leader of various institutions. In May 1933 Jäckh emigrated to-
82
Dr. Ernst Jäckh married Fritz Klein’s oldest sister, Bertha (1874–1928), gether with his second wife Martha to Great Britain. He remained there
in 1899. From 1902 to 1912 he was editor-in-chief of a daily newspa- until 1940, working for the British government in multiple functions.
per in Heilbronn, Germany. Beginning in 1908, Jäckh also made nu- He moved in 1940 to the United States and taught as a professor at Co-
merous trips to Turkey, where he came into contact with high-ranking lumbia University in New York. See Jäckh, 1954.
83
representatives of the Ottoman Empire and became a friend of the Ger- Letter from the German Minister of Foreign Affairs, Alfred von
man diplomat Alfred von Kiderlen-Waechter, who was appointed Min- Kiderlen-Waechter, November 25, 1911, to Fritz Klein’s brother-in-
ister of Foreign Affairs in 1910. Jäckh moved to Berlin in 1912 and law Dr. Ernst Jäckh.
84
served as a “consultant” with the Ministry of Foreign Affairs. In 1920, Fritz Klein traveled via New York, where he arrived in June 1911.
Jäckh founded the German Academy for Political Sciences in Berlin, File Fritz Klein, List or Manifest of Alien Passengers for the United
and during the 1920s, he was a member of the German delegation to States Immigration Officer at Port of Arrival, New York, June 1911,
several international conferences. Through a conversation with Hitler in MyHeritage.com.

84 HISTORY OF CHIVOR, PART I: 1880–1925 GEMS & GEMOLOGY SPRING 2020

 and was valid for a one-year term.87 Span and Klein
were also given the right to mine samples in Chivor
during the stated pendency of the option.
Klein was not alone in his interest in Chivor dur-
ing this period. Dixon (see again box A), previously
active in Muzo88 and already both a shareholder in
the Compañía de las Minas de Esmeraldas de Chivor
and a co-owner of the three newer Chivor mining ti-
tles, was engaged anew in exploration in the region,
commencing additional mining ventures. As re-
ported in 1911: “Mr. Christopher E. Dixon, formerly
engineer in charge of Muzo mines, personally ex-
tracted over a pound of emeralds by scratching about
in one place and another…. Somondoco (sometimes
called ‘Chivor’ from the valley near it) will soon yield
gems to delight the world.”89
On June 11, 1912, Klein set out from Bogotá for
Chivor to obtain samples for presentation to poten-
tial investors.90 He was accompanied by Restrepo and
remained at the mine until January 1913. During
that time, from late October to early November
1912, an official from the German Residency in Bo-
gotá visited the site and reported91 that within a short
period of four to five weeks, the work undertaken by
15 to 20 indigenous people had produced quite favor-

85
Report by the German Resident in Bogotá, Dorotheus Kracker von
Schwartzenfeldt, December 21, 1911, German Federal Archive; letter
Figure 12. Opening page of a report from the German from Fritz Klein, July 24, 1912, to Kracker von Schwartzenfeldt, Ger-
Resident in Bogotá, Dorotheus Kracker von man Federal Archive.
86
Schwartzenfeldt, to the German Ministry of Foreign Exposé der Smaragdmine “El Chivor” by Fritz Klein, undated, 16
pp., received by the German Ministry of Foreign Affairs September
Affairs, dated December 21, 1911. The report de- 13, 1913, German Federal Archive. Hermann Span (1870–1943?) and
scribed Fritz Klein’s negotiations with the Colombian his brother Emil (1869–1944), from Reutlingen, Württemberg, Ger-
government in seeking permission to establish an many, moved in the 1890s to Central America and were active in
emerald-cutting enterprise. Historical Archive of the Guatemala, Costa Rica, and El Salvador. Hermann Span was also in-
volved in Colombia by 1910. Hauptstaatsarchiv Stuttgart, E40-76 Bü
German Ministry of Foreign Affairs. 1370; Diario Oficial (Colombia), 48, No. 14678 (1912), pp. 411–
413. Additionally, multiple notarial acts concerning Span and sundry
business partners are found in the Archivo General de la Nación
(Colombia).
in hopes of establishing an emerald-cutting enterprise 87
Escritura 1096, June 5, 1912, Notaria 2, Archivo General de la
(figure 12), but these efforts were unsuccessful.85 Nación (Colombia); Report by a member of the German Residency in
In November 1911, Klein became aware of the Bogotá, Wilhelm Gerlach, November 15, 1912, German Federal
Archive.
Chivor mine through another German, Hermann 88
During the 1890s and early 1900s, Christopher Dixon worked on
Span.86 A preliminary option contract for the mine, behalf of the English Mining Syndicate, Limited, which operated the
initiated by Span, was executed on February 13, Muzo mine prior to the contract with the Syndicate of Muzo. Atuesta,
1912, and on June 5, 1912, after Klein conducted pre- 1899; Monthly Bulletin of the International Union of the American
Republics, 8 (1900), pp. 581–582; Gaceta Judicial (Colombia), 15
liminary exploration of the mining area, terms were (1901), pp. 169–173; Gaceta Judicial (Colombia), 16 (1903), pp. 78–
further formalized in a notarized option contract 79; Domínguez, 1965.
signed between Span and Klein on one hand and the 89
Latham, 1911.
90
mine owners Compañía de las Minas de Esmeraldas Exposé der Smaragdmine “El Chivor” by Fritz Klein, undated, 16
pp., received by the German Ministry of Foreign Affairs September
de Chivor on the other. The latter contract provided 13, 1913, German Federal Archive.
an exclusive right to purchase the mine for £51,000 91
Report by a member of the German Residency in Bogotá, Wilhelm
(equivalent to 255,000 pesos oro; see again table 3) Gerlach, November 15, 1912, German Federal Archive.

HISTORY OF CHIVOR, PART I: 1880–1925 GEMS & GEMOLOGY SPRING 2020 85

Figure 13. Export license issued to Fritz Klain [sic] by
the Ministerio de Hacienda, March 5, 1913, for 157
“gangas” (emeralds in the host rock) and 5,450 carats
of rough emerald crystals. Courtesy of Eduardo Re-
strepo Ortega.

Figure 14. Dr. Antonio José Cadavid, a lawyer, law pro-

able results. According to Klein,92 they focused for fessor, and politician, filed a lawsuit on behalf of three
the first four months on poor targets, but when he fi- shareholders of the Compañía de las Minas de Esmeral-
nally understood “the direction of the veins and the das de Chivor to challenge new laws and decrees affect-
character of the mine, the production of emeralds ing the independence of the “free” emerald mines in
began in surprising abundance. With 20 workers, I Colombia. Cadavid also worked with Fritz Klein and
unearthed in 3½ months under the most primitive the German Resident Dorotheus Kracker von
conditions conceivable about 6,000 carats of emer- Schwartzenfeldt in 1913 and 1914. Courtesy of Banrep-
cultural, Bogotá.
alds, which have a minimum value of 150,000 marks
[then equivalent to US $37,500].” At the end of 1912,
Klein’s younger brother August arrived in Colombia
José Cadavid96 (figure 14), representing Klein.97 Word
for support.93 He remained at the mine, together with
had already come to the German Residency in Bo-
Restrepo’s younger son Alejandro, when Klein and
Restrepo subsequently departed for Germany.
Klein left Bogotá in March 1913 to search for in-
92
vestors in Germany. With the approval of the local Exposé der Smaragdmine “El Chivor” by Fritz Klein, undated, 16
pp., received by the German Ministry of Foreign Affairs September
government, he carried a sample of emeralds pro- 13, 1913, German Federal Archive.
duced during his approximately six months at the 93
August Klein traveled from Hamburg via New York, where he ar-
mine. The set consisted of 157 “gangas” (emeralds rived in October 1912. File A. Klein, List or Manifest of Alien Passen-
gers for the United States Immigration Officer at Port of Arrival, New
still embedded in the original host rock) and over York, October 1912, MyHeritage.com.
5,000 carats of crystals in varying sizes and shapes.94 94
Report by a member of the German Residency in Bogotá, Josef Wip-
Having received a license for exportation (figure 13), perfeld, March 27, 1913, German Federal Archive.
Klein wanted to use these samples to interest poten- 95
Hermann Span left Colombia for Germany in August 1913 to form a
banana company, apparently ending his pursuit of the Chivor emer-
tial European investors. In furtherance of this mis- ald business. He returned to Colombia in April 1920. Diario Oficial
sion, Klein arranged before leaving to have the term (Colombia), 58, No. 18319–18320 (1922), p. 478.
of the option contract extended, this time exclusively 96
The lawyer, law professor, and politician Dr. Antonio José Cadavid
in his own name and without Span.95 That extension, (1866–1919) served in various positions as professor at the National
University, Minister of the Treasury (1909–1910), and Minister of War
valid for nine months, was signed in July 1913 by (1915–1916). See Bermúdez (1966).
several of the shareholders together with Escovar, 97
Escritura 1474, July 15, 1913, Notaria 2, Archivo General de la
representing Restrepo, and the attorney Dr. Antonio Nación (Colombia).

86 HISTORY OF CHIVOR, PART I: 1880–1925 GEMS & GEMOLOGY SPRING 2020

 doco, Ltd. The stated aim of the entity was “to ac-
quire and work the Emerald Mines called El Chivor
No. 1 and El Chivor No. 2” (figure 16). The capital
was to consist of £90,000, with 90,000 shares to be
issued at £1 each.104 Nonetheless, no further activity
pertaining to this company is known, so any under-
lying rival or alternate plan targeting Chivor was ap-
parently abandoned.
Soon after Restrepo’s return to Bogotá, another ex-
tension to the option contract, granting Klein exclu-
sive rights to purchase the mine until December
1914, was signed by Cadavid, representing Klein.105

LITIGATION BETWEEN MINE OWNERS AND

THE COLOMBIAN GOVERNMENT (1912–1913)
The commercial success of a gem mine depends pri-
marily on the value of mined stones and the cost of
mining operations. Also significant are the amount
of duties and taxes imposed and the ability to sell the

98
Restrepo traveled from Hamburg via New York, where he arrived in
Figure 15. Dr. Hjalmar Schacht served as a vice direc- November 1913. File F. Restrepo, Hamburger Passagierlisten,
November 1913, Ancestry.de; File F. Restrepo, List or Manifest of
tor of Dresdner Bank, one of the larger German finan- Alien Passengers for the United States Immigration Officer at Port of
cial institutions, from 1908 to 1915. Separate from his Arrival, New York, November 1913, Ancestry.com.
role at the bank, he opted in 1914 to privately invest 99
Dr. Hjalmar Schacht (1877–1970) was one of the vice directors of
in the Chivor mine with Fritz Klein, August Stauch, Dresdner Bank in Berlin from 1908 to 1915. An ascendant career
and the company Rudolph Hahn & Sons, London, in thereafter led him to the presidency of the German National Bank
(Reichsbank), with two terms of office from 1923 to 1930 and from
pursuing the Chivor mine. Photo circa 1900, courtesy
1933 to 1939. He also served concurrently as Minister of Economics
of Cordula Schacht. from 1934 to 1937. Between 1933 and 1939 Schacht was instrumen-
tal in the financing of Hitler’s rearmament, which in 1945 resulted in
his indictment at the Nuremberg Trials. He was ultimately acquitted.
See Kopper, 2006.
gotá that Dresdner Bank, one of the larger financial In 1909, Schacht and Klein’s brother-in law Dr. Ernst Jäckh traveled
institutions in Germany, might be interested in fund- together to Turkey. Following that trip and Jäckh’s move to Berlin in
1912, they cooperated on many political projects. After Jäckh’s emi-
ing the purchase of Chivor. gration in 1933 and Schacht’s new engagement, however, their ways
Klein traveled to Germany with Restrepo, who re- parted.
mained there until his return in November 1913.98 100
Karlsbader Kurliste, No. 135, May 23, 1913, 2 pp.
Restrepo’s stay in Germany included time in Idar, 101
Letters by Fritz Klein, November 27, 1913, and January 26, 1914,
to Restrepo; Letter from August Klein Senior, December 27, 1913, to
where he met members of Klein’s family, and in Restrepo.
Berlin, where he met Jäckh, Dr. Hjalmar Schacht (fig- 102
The company Rudolf Hahn & Sons, London, dated back to 1893
ure 15),99 and presumably other possible investors. and in 1914 was managed by the brothers Eugen and Paul Hahn. The
Restrepo’s sojourn incorporated a period in May 1913 Hahn brothers had roots in Idar, Germany, and were cousins of Au-
gust Hahn, brother-in-law of Fritz and August Klein. August Hahn
at the renowned sanatorium of Karlsbad100 to restore married Klein’s younger sister Luise (1884–1976).
his health. Restrepo also traveled to Meran, Tyrol, 103
Letter from Rudolf Hahn & Sons, London, November 7, 1913, to
Austria, to see Klein’s father August.101 Klein, in turn, the mine owners.
104
traveled to England in October 1913 to show the Prospectus, Emerald Mines of Somondoco, Ltd., Undated. Al-
though the prospectus and attendant documents are undated, refer-
emerald samples to members of the company Rudolf ences therein to several specific individuals (e.g., Diódoro Sánchez as
Hahn & Sons, London,102 dealers in high-end gem president of the Sociedad Colombiana de Ingenieros and Francisco
materials, and other potential backers.103 Restrepo Plata as Ministro de Hacienda, Colombia) enable prepara-
tion and publication of the material to be placed within the 1913 to
Meanwhile, initial steps were taken by unknown 1914 timeframe.
parties between 1913 and 1914 to start a company in 105
Escritura 2759, December 24, 1913, Notaria 2, Archivo General de
London under the name Emerald Mines of Somon- la Nación (Colombia).

HISTORY OF CHIVOR, PART I: 1880–1925 GEMS & GEMOLOGY SPRING 2020 87

 Thereafter, these so-called free mines were to be for-
ever excluded from any taxes and fees and could no
longer be registered by other persons (Law No. 292
of 1875, Article 45, adapted in Law No. 38 of 1887).106
As noted previously, the taxes for Chivor 1 and 2 had
been paid in advance for 20 years.
After the law changed, the nature of the rights af-
forded to the free mines became a subject of repeated
controversy and litigation between the mine owners
and the Colombian government. The disputes cen-
tered on three issues:
1. Taxes and fees payable by the owners of free
mines
2. Control of mining operations by the government
3. Control of emerald sale and export by the
government
An initial legal action challenging several laws
and decrees was filed by Francisco de P. Matéus in
his name and on behalf of several other owners of free
mines. The eventual decision of the Corte Suprema
de Justicia on August 2, 1912, was a partial victory
for the mine owners, ruling that governmental con-
trol of the mining operations was illegal but that
state control of emerald exportation could con-
tinue.107 A copy of the decision was mailed to
Guateque to inform Restrepo and Klein, who were
working at Chivor at the time.
In October 1912, the government rejected a
Figure 16. Prospectus drafted in 1913–1914 to form a complaint by Dixon about restrictive rules circum-
company in London under the name Emerald Mines scribing the emerald mining business after a seven-
of Somondoco, Ltd., “to acquire and work the Emer- year dispute that had begun in 1905 (again, see box
ald mines called El Chivor No. 1 and El Chivor No. A).108
2.” Courtesy of Eduardo Restrepo Ortega. A new decree by the Colombian government was
then issued in December 1912, followed by a supple-
ment in January 1913, which sought again to in-
crease control by governmental institutions over free
gems on the open market. During relevant periods in mines. Opposition to the new regulations began im-
Colombia, the applicable framework of various laws mediately, with a petition by Isaza, shareholder in
and decrees, issued at different times and subse- the Compañía de las Minas de Esmeraldas de Chivor,
quently incorporating multiple addendums and
amendments, was rife with complexity. The major
emerald mines of Muzo and Coscuez had belonged
to the government since 1886, but a number of 106
Bullman, 1892.
mostly smaller mines existed under private owner- 107
Corte Suprema de Justicia, Bogotá, agosto 2 de 1912, “Sobre la in-
ship. After 1905, however, registration of new emer- exequibilidad de algunas disposiciones relacionadas con la ex-
plotación de minas de esmeraldas [About the unenforceability of
ald mines by private owners was no longer permitted some decrees related to the exploitation of emerald mines],” 21 pp.;
and additional taxes were imposed upon the existing Gaceta Judicial (Colombia), 22 (1913), pp. 2–9.
mines (Law No. 40 of 1905). 108
Letter from Christopher Dixon, November 10, 1905, to the Ministe-
Prior to the 1905 change in regulations, a provi- rio de Relaciones Exteriores, Ministerio de Industrias, Archivo, Corre-
spondencia 1905, Archivo General de la Nación (Colombia); Diario
sion allowed for the annual taxes on privately owned Oficial (Colombia), 48, No. 14720 (1912), pp. 812–813; Annual re-
mines to be paid in advance for a period of 20 years. port on Colombia for the year 1913.

88 HISTORY OF CHIVOR, PART I: 1880–1925 GEMS & GEMOLOGY SPRING 2020

 Affairs in Berlin by the German Resident in Bogotá
on September 7, 1913, and Klein was promptly in-
formed via his brother-in-law Jäckh.

ACTIVITIES IN 1914
After the favorable resolution of the legal situation in
Colombia, Klein was able to continue searching for
and negotiating with potential investors. In this en-
deavor, Jäckh and his contacts with political and busi-
ness elites once again proved instrumental. Restrepo,
as indicated above, had already returned to Colombia.
The search culminated on February 14, 1914,
with the founding of the Syndikat El Chivor (Syndi-
cate El Chivor, used hereinafter) between four mem-
bers, with initial capital contributions as follows:113
August Stauch (figure 18),114 Berlin-Zehlendorf,
25,000 marks; Dr. Hjalmar Schacht, Berlin-Zehlen-
dorf, 10,000 marks; Rudolf Hahn & Sons, London,
10,000 marks; and Fritz Klein, Idar, 5,000 marks. The
founding contract further contained provisions ad-
dressing usage of the initial capital, investigation of
the property by an expert, and terms of any future
purchase of the mine. The services of an expert were
considered necessary because the quality of Klein’s
samples was inadequate to convince all investors to
formalize the purchase, absent independent exami-
nation of the mine (figure 19).115 In these endeavors,
the four members agreed that the company should
be privately financed, without institutional funding
Figure 17. Report from a member of the German Resi- from Dresdner Bank or any other German banking
dency in Bogotá, Josef Wipperfeld, to the German firm.
Ministry of Foreign Affairs, Berlin, dated September 6,
1913. The report briefly described the August 22 deci-
sion by the Colombian Corte Suprema de Justicia ad- 109
Diario Oficial (Colombia), 49, No. 14823 (1913), pp. 341–342.
dressing privately owned “free” mines. Historical 110
Ibid., pp. 546–548.
Archive of the German Ministry of Foreign Affairs. 111
All four Colombians involved were well-known lawyers who had
held high-level positions in Colombia as university professors, diplo-
mats, and ministers. As previously noted, Enrique González also for-
merly served as director of operations at the Muzo emerald mine.
Diaz, 1967.
to the Minister of Finance in January 1913109 and a 112
Gaceta Judicial (Colombia), 22 (1913), pp. 242–244; Cadavid,
similar petition by Klein in February 1913.110 An ad- 1913.
ditional legal action challenging several of the older 113
Historical Archive of the German Ministry of Foreign Affairs; Prof.
laws and decrees and the two new decrees was filed Robert Scheibe collection, University Library Freiburg.
114
August Stauch (1878–1947) has been recognized as the discoverer
on February 22, 1913, by the attorney Cadavid in the of the diamond deposits in Namibia, then known as German South-
names of three Compañía de las Minas de Esmeral- West Africa, in 1908. Diamond mining made him and those involved
das de Chivor shareholders (the Colombian citizens in his Colonial Mining Company extremely wealthy. Stauch bought
farms in southwest Africa and a villa in Berlin-Zehlendorf, Germany.
Isaza, Escovar, and González).111 The decision of the In the 1920s, he was active as an investor not only in a variety of min-
Corte Suprema de Justicia, issued August 22, 1913, ing companies but also in technology-focused enterprises. See Levin-
nullified several of the contested laws and decrees, son, 2007. The Schacht and Stauch families lived near each other in
Berlin-Zehlendorf, Germany, and might thus have interacted on a per-
ruling that the free emerald mines could work with- sonal level as well.
out any further governmental control.112 That result 115
Letter from Rudolf Hahn & Sons, London, November 7, 1913, to
(figure 17) was telegraphed to the Ministry of Foreign the mine owners.

HISTORY OF CHIVOR, PART I: 1880–1925 GEMS & GEMOLOGY SPRING 2020 89

EMERALDS FROM THE CHIVOR MINING DISTRICT, COLOMBIA
Emeralds from the Chivor region are found as individual single crystals (A–C) or as crystal aggregates (D–F).
These aggregates consist of parallel or subparallel oriented crystals or even of radiating, strongly inclined indi-
viduals. The habit of the emerald crystals is formed by six or twelve hexagonal prism faces in combination with
two basal pinacoids (A–C, G–J). Occasionally, different hexagonal dipyramids are also seen (I and M). The basal
pinacoid might show incomplete growth surfaces (K), and, in extreme rare cases, cavities parallel to the hexag-
onal crystal axis are seen (L). The mineral assemblage within the emerald-bearing veins at Chivor consists of
pyrite (A–C, N–R) frequently associated with albite (var. cleavelandite), calcite, dolomite (O–R), occasionally
quartz, and limonite (mainly goethite) formed by the alteration of pyrite. The host rocks of the Chivor deposit
are breccia, layered limestones, albitites, and black and gray shales (C–K, M–S). Photos by G. Martayan, Paris.

Emerald crystal: 19 × 7.5 mm

Habit: hexagonal prism
faces terminated by flat
basal pinacoids; the large
beryl crystal is intergrown
at different angles with
smaller crystals
Associated minerals: dodeca-
hedral pyrite crystals
 B C

Specimen: 21 × 19 × 14 mm Emerald crystal: length 7 mm

Habit: hexagonal prism faces terminated by flat basal
pinacoids and hexagonal dipyramids
Associated minerals: gray dolomite, albite, pyrite
Matrix: gray shale

Emerald crystal: length 33 mm Cluster of intergrown emerald crystals, specimen: 43 × 38

Habit: hexagonal prism faces terminated by flat basal × 36 mm; several crystals exhibit striations (growth pat-
pinacoids terns) on basal pinacoids or on prism faces; some crys-
tals of the aggregate contain carbonaceous inclusions
Associated minerals: albite, pyrite, calcite
Associated minerals: limonite
 E F

Specimen with emerald clusters: 45 × 33 × 23 mm, Emerald cluster, specimen: 20 mm across; the crystals of
largest cluster: 18 × 16 mm the aggregate are inclined to each other
Associated minerals: dolomite, albite, pyrite Associated minerals: albite
Matrix: gray shale

G H

Specimen with emerald crystal Crystal: length 8 mm Specimen with emerald crystal Crystal: 11 × 4.5 mm
Habit: hexagonal prism faces terminated by flat basal Habit: hexagonal prism faces terminated by flat basal
pinacoids and hexagonal dipyramids pinacoids and hexagonal dipyramids
Associated minerals: pyrite, dolomite Associated minerals: white albite, pyrite, dolomite
Matrix: breccia Matrix: gray limestone
 I J

Specimen with emerald crystal Crystal: length 13 mm

Habit: hexagonal prism faces terminated by flat basal
pinacoids
Associated minerals: pyrite, gray dolomite rhombs
Matrix: gray shale

L
Specimen with emerald crystal Crystal: length 8 mm
Habit: hexagonal prism faces terminated by flat basal
pinacoids and hexagonal dipyramid
Associated minerals: pyrite, dolomite
Matrix: breccia

Emerald crystal with channels parallel to the c-axis

Specimen: 17 × 11 × 11 mm
Associated minerals: gray dolomite

Specimen with emerald crystal Crystal: length 5 mm Specimen with emerald crystal Crystal: length 6 mm
Habit: hexagonal prism faces terminated by flat basal Habit: hexagonal prism faces terminated by flat basal
pinacoids, indentations in the basal pinacoid pinacoids and hexagonal dipyramids
Associated minerals: gray calcite rhombs, white albite, Associated minerals: white albite
pyrite
Matrix: black shale
 N

Specimen: 60 × 45 × 55 mm Emerald crystal: length 6 mm

Habit: hexagonal prism faces terminated by flat basal pinacoids and hexagonal dipyramids
Associated minerals: pyrite, dolomite
Matrix: black shale

O P

Specimen: 42 mm across Emerald crystals: up to 14 mm Specimen: 32 mm across Emerald crystal: length 9 mm

in length Habit: hexagonal prism faces terminated by flat basal
Habit: hexagonal prism faces terminated by flat basal pinacoids
pinacoids, prism faces with striations parallel to the c- Associated minerals: pyrite, gray dolomite rhombs,
axis white albite
Associated minerals: dolomite, pyrite, embedded in a Matrix: black shale
white dolomitic matrix
 Q

Specimen: 15 × 12 × 5 mm
Largest emerald crystal:
length 11 mm
Habit: hexagonal prism
faces terminated by flat
basal pinacoids and hexag-
onal dipyramids
Associated minerals: white
albite
Matrix: gray shale

R S

Specimen: 38 × 35 × 18 mm Emerald crystal: length 20 Specimen: 67 × 50 × 50 mm Emerald crystal: length 18

mm mm
Habit: hexagonal prism faces terminated by flat basal Habit: hexagonal prism faces terminated by flat basal
pinacoids and hexagonal dipyramids, pinacoids and hexagonal dipyramids
Associated minerals: white albite, small pyrite nodules Associated minerals: albite, dolomite, small pyrite nodules
Matrix: black shale Matrix: layered gray limestone
 Figure 19. In a letter dated November 7, 1913, Lon-
don-based Rudolf Hahn & Sons advised owners of the
Chivor mines that the quality of the emerald samples
shown by Fritz Klein was inadequate to justify pur-
chase of the property. Courtesy of Eduardo Restrepo
Ortega.

few days later, Klein and Scheibe embarked together

in Hamburg, bound for Puerto Colombia.118
Following the just-described formalities, letters
were sent from the Syndicate El Chivor to the Ger-
man Ministry of Foreign Affairs in Berlin and to the
German Resident in Bogotá.119 Therein, Schacht and
Stauch sought to obtain clarification regarding the
situation in Colombia, inquiring as to the legality of
the mining titles, any obligations to pay taxes or ad-
ditional fees, the parameters for authorized export of
the emeralds and so forth. The German Resident in
Bogotá, Dorotheus Kracker von Schwartzenfeldt, in
Figure 18. August Stauch is credited with discovering turn contacted González as a representative of the
the diamond deposits in Namibia in 1908. Stauch mine owners, the attorney Cadavid, and the Ministro
was a key investor in the group assembled by Fritz de Hacienda for details before responding in a letter
Klein in 1914 to pursue the Chivor mine. Photo 1921, to Schacht dated June 15, 1914.120
courtesy of Dr. Gabriele Schneider.

116
Letter from the Syndicate El Chivor, February 16, 1914, to Prof.
The initial paid-in capital was to be used solely to Robert Scheibe, Historical Archive of the German Ministry of Foreign
finance the expert evaluation. Klein was to assist the Affairs.
117
scientific expert in his activities. Any subsequent pur- Prof. Robert Scheibe (1859–1923) studied mathematics and natural
sciences at the Martin Luther University of Halle-Wittenberg, Ger-
chase of the mine would proceed through exercise of many, where he received his PhD in 1882. In 1885, he became a
Klein’s option by paying the price of £51,000 (equiva- member of the Royal Mining Academy (Königliche Bergakademie) in
lent to 1,020,000 marks and identical to the 255,000 Berlin, where he was appointed professor in 1895. In the summer of
1908, Scheibe traveled to Namibia to study primary kimberlites. He
pesos oro specified in the option contract). Klein was met August Stauch, the discoverer of the secondary diamond deposits
to receive 1,250,000 marks cash from the investors to in the area of Lüderitzbucht. From late December 1908 to early Janu-
purchase the mine and would be entitled to 20 percent ary 1909, the two men made an excursion to the area south of Lüder-
itzbucht, where they found Pomona, potentially the richest diamond
of the shares of a future mining company. deposit ever discovered. Scheibe returned to Germany at the end of
As communicated in a letter from the Syndicate 1909, but he and Stauch remained friends for life. See Rauff, 1926;
El Chivor dated February 16, 1914,116 Prof. Robert Strunz, 1970; Levinson, 2007.
118
Scheibe (figure 20),117 professor of geology at the Files Robert Scheibe and Fritz Klein, Hamburger Passagierlisten,
February 1914, Ancestry.de.
Royal Mining Academy (Königliche Bergakademie) 119
Letters by the Syndicate El Chivor, March 14, 1914, and April 9,
in Berlin and close friend of Stauch, was selected to 1914, Historical Archive of the German Ministry of Foreign Affairs.
perform the expert evaluation of the Chivor mine. A 120
Historical Archive of the German Ministry of Foreign Affairs.

96 HISTORY OF CHIVOR, PART I: 1880–1925 GEMS & GEMOLOGY SPRING 2020

Figure 20. Prof. Robert Scheibe of the Royal Mining Figure 21. Restrepo’s death certificate, dated April 24,
Academy (Königliche Bergakademie) in Berlin joined 1914. File Inheritance F. Restrepo, Archivo General de
August Stauch in late 1908 to early 1909 in finding la Nación (Colombia), Bogotá.
the major Pomona diamond deposit in Namibia. He
was hired in 1914 by the Syndicate El Chivor to eval-
uate the Chivor mine. Photo circa 1910, Prof. Robert
Scheibe collection, University Library Freiburg. Mr. Restrepo’s son, who was working as foreman here,
accompanied the transport. On arrival at Guateque late
in the evening, Mr. Restrepo died and was buried the
day after.
Scheibe arrived with Klein at Chivor on April 4, A Colombian certificate of death dated April 24,
1914. Over the course of the next several months, he 1914, was issued documenting Restrepo’s passing
prepared four interim communications121 and a final (figure 21).
report,122 offering a window into circumstances at the When conditions improved, three weeks were
mine during that time. At the outset of his visit, a spent repairing the mine’s aqueduct, after which
period of heavy rain made geological fieldwork al- the first clean but small emeralds were found in
most impossible. Progress was also impacted when,
on April 20, 1914, Restrepo became seriously ill and
lost consciousness. As Scheibe wrote:123 121
Letters by Prof. Robert Scheibe, Historical Archive of the German
We believed it a crisis in encephalitis but he did not Ministry of Foreign Affairs.
wake up again, he lay motionless. On April 23rd we de- 122
The final report, dated September 3, 1914, was provided by the
cided to transport him to Guateque despite the rain. grandchildren of Prof. Robert Scheibe.
The workers did as much as humanly as possible on 123
Letter from Prof. Robert Scheibe, May 1, 1914, Historical Archive
the muddy and unstable mountain path. Mr. Klein and of the German Ministry of Foreign Affairs.

HISTORY OF CHIVOR, PART I: 1880–1925 GEMS & GEMOLOGY SPRING 2020 97

May and June and commercial material was mined time there and the purchase price for the mine,
in July. A representative of the German Residency Scheibe saw a realistic chance for a successful ven-
in Bogotá also visited the mine from July 3 to 16, ture but also an extremely high “risk” in buying the
1914, and reported on his observations.124 Between mine under the existing conditions. Hence, he was
late July and early August, Scheibe finished his in- unable to offer an unqualified recommendation to
vestigations, planning to return to Bogotá together buy Chivor for the contract price.
with August Klein on August 2.125 The outbreak
of war in early August, however, curtailed such WORLD WAR I
plans and suspended Klein’s pursuit of the Chivor
project. The international conflict that became World War I
Scheibe’s final report,126 dated September 3, 1914, began in August 1914 and had far-reaching conse-
and written in Bogotá, offered his conclusions as to quences for Chivor. Germans who found themselves
the viability of Chivor. The report began with a de- overseas faced an ongoing series of complications and
tailed description of the geological conditions, then decisions. Early on, there was hope that hostilities
covered both the expected yield and the anticipated would end quickly and normal business pursuits
costs. As to the yield, he reported: could be resumed after waiting a few months. As that
hope faded, the desire to return home became
In the 4 months of my presence at the mine [April 4 stronger. However, a simple trip to Germany was
to August 2, 1914] with a labor force of 25 men gem
quality emeralds of a value of about 40,000 marks hindered by the fact that British warships stopped
were mined…. I do not doubt that emeralds of a good commercial or passenger ships, even vessels of neu-
kind are to be found elsewhere in the above men- tral states, and transported any German passengers
tioned area and beyond…. I consider the pit itself to to internment camps.
be exploitable, especially since the prices of emeralds Klein waited, using the time to reprise his efforts
are high—also in Chivor stones were found that are
estimated to 800 marks per carat. The average value to obtain permission for an emerald-cutting enter-
of the usable stones found during my time there (1,267 prise in Colombia. As before, his petition dated No-
ct = 39,420 marks) is about 31 marks for one carat of vember 12, 1914, was swiftly rejected.127 Klein then
raw stones. left Colombia in December 1914 or January 1915, re-
The start-up investment for mining equipment turning to Germany, possibly under false papers.128
was calculated at 150,000 marks. Annual operating He entered a cavalry regiment stationed in Ludwigs-
costs for 150 workers, six supervisors, one commer- burg, southern Germany, as a volunteer on February
cial manager, one mining engineer, and one head of 20, 1915.129 According to Klein, he felt that this serv-
operational activities was estimated at 142,000 ice was a duty he owed to his country.130 The equip-
marks. Taking into account those parameters, in ment used at Chivor was held by a custodian in
combination with the material recovered during his Klein’s absence, presumably in hopes of resuming op-
erations in the future.131
Scheibe deemed an attempt to return to Germany,
124
Report by a member of the German Residency in Bogotá, Josef combined possibly with years of internment, too
Wipperfeld, July 20, 1914, Historical Archive of the German Ministry risky. He therefore decided to stay in Colombia,
of Foreign Affairs. working for the government as a geologist. Becoming
125
Letter from Fritz Klein, August 1, 1914, to the German Resident in
Bogotá, German Federal Archive.
head of the Comisión Científica Nacional (National
126
Courtesy of Andreas Scheibe (grandson of Prof. Robert Scheibe). Geological Commission), he devoted himself to ex-
127
Diario Oficial (Colombia), 50, No. 15355 (1914), pp. 1127–1128. ploring the country’s important mineral resources.
128
Klein, 1941, 1951. For instance, he prepared reports on the Muzo emer-
129
Military files of the 20th lancer regiment, Ludwigsburg, Germany, ald deposit, published in 1916 and 1922, for the Min-
Hauptstaatsarchiv Stuttgart.
130
istro de Hacienda.132 His work on behalf of the
Klein, 1941, 1951.
131
government continued until his return to Germany
”Inventory of the tools of the Chivor mine which are held by Anato-
lio Morales in El Rosal,” January 5, 1916, 2 pp. in the spring of 1920133 and was recognized with ap-
132
The scientific studies on Muzo by Prof. Robert Scheibe were pub- pointment as an honorary professor at the University
lished initially in Spanish and later edited by his son, Dr. Ernst Al- of Bogotá. Upon request of the government, Scheibe
brecht Scheibe, in German. See Scheibe, 1916, 1922, 1926. The
younger Scheibe, like his father, traveled to Colombia and worked as subsequently returned to Colombia in 1921 under a
a geologist from 1924 to 1929. three-year contract, again as the head of the
133
Rauff, 1926; Strunz, 1970. Comisión Científica Nacional.

98 HISTORY OF CHIVOR, PART I: 1880–1925 GEMS & GEMOLOGY SPRING 2020

 Figure 22. In 1917, a
group of American and
Colombian business-
men and geologists as-
sociated with the Carib
Syndicate, Ltd., investi-
gated oil fields located
at the border between
Colombia and
Venezuela. Carl K.
MacFadden (left) and
Wilson E. Griffiths par-
ticipated in the excur-
sion, which resulted in
the 1918 transfer of
property referred to as
the “Barco concession”
to an American oil
company. Photo 1917,
public domain.

PURCHASE OF CHIVOR BY AN AMERICAN been founded in the United States on June 22,
COMPANY (1919) 1918,137 by a group of American bankers and specu-
Following World War I, both Klein and his brother lators from the oil industry. Those involved were
August were interested in continuing with the principally associated with the Carib Syndicate,
Chivor project. Nonetheless, several developments Ltd. and the Colombian Petroleum Co., which pur-
during the interim changed their previous plans sued commercial ventures in the oil business in
and expectations. From a personal standpoint, each Colombia and Venezuela.138 Key figures included
of the Klein brothers had married in 1919, Fritz to Wilson E. Griffiths (1851–1929) and Carl K. MacFad-
Claire Cullmann and August to Liese Brill, so fu- den (1872–1952).139 Griffiths and MacFadden (figure
ture work would need to take their new wives into 22) made several trips to Colombia between 1917
consideration.
From a business perspective, war altered the eco-
nomic landscape for the Syndicate El Chivor. The ex- 134
Eugen and Paul Hahn, as well as their brother Rudolf, had been in-
change rate of the mark decreased dramatically, and terned in the United Kingdom from 1915 to 1918. Assets of the com-
the financial situation for at least one investor, pany in the United States worth $225,000 and in Britain worth
£250,000 (equivalent to US $1,250,000) had been confiscated as
Rudolf Hahn & Sons, London, deteriorated substan- enemy property. See Hahn v. Public Trustee, [1925] Ch. 715 (Eng.);
tially.134 To accommodate this scenario, a new in- [1926] 95 Law Journal Reports 9 (Ch.), pp. 9–22 (Eng.).
vestor in the Syndicate El Chivor was found, the 135
Letter from August and Liese Klein, January 4, 1920, to Fritz and
Claire Klein. No further relevant documentation was found in
Dutch shipping firm Wm H. Müller & Co., based in archived records of Wm H. Müller & Co., held at the Rotterdam City
The Hague and Rotterdam.135 In addition to its pri- Archive (Stadsarchief Rotterdam).
mary shipping and transport business, Wm H. Müller 136
Letter from August and Liese Klein, September 26, 1919, to Fritz
& Co. participated in various mines and other indus- and Claire Klein.
137
trial companies. https://opencorporates.com/companies/us_de/74916.
138
The Wall Street Journal, May 26, 1919, p. 8; The Honolulu Adver-
After the foregoing preparations were initiated, tiser, December 7, 1919, p. 29; The Wall Street Journal, May 21,
Klein and his wife Claire left Germany for Bogotá in 1920, p. 8; Encyclopedia of American Biography, New Series, 9
late July or early August 1919.136 (1938), pp. 401–402.
139
Meanwhile, a new competitor for Chivor had en- The Pittsburgh Press, December 29, 1895, p. 12; Escritura 3084,
December 27, 1919, Notaria 1, Archivo General de la Nación
tered the fray after Klein’s option expired during the (Colombia); Rausch, 2014; Torres del Río, 2015. MacFadden’s name
war. The Colombian Emerald Syndicate, Ltd. had is also written as Mac Fadden or McFadden.

HISTORY OF CHIVOR, PART I: 1880–1925 GEMS & GEMOLOGY SPRING 2020 99

 drawn to the Chivor emerald mine and its history
by Dixon.144
Upon his arrival at the Colombian port of Barran-
quilla, Klein was informed of a legally valid option
contract held by the U.S.-based corporation.145
Hence, Klein could only wait to see if the Americans
exercised the option, as expressed by his brother:
“Now, December 8 is getting close, and thus the final
decision. We are of course all very excited about the
result and would be very happy if it turns out in your
favor.”146
In a last-ditch effort to thwart the signing of a deal
by the Americans, Klein on December 22, 1919, filed
a complaint with the Colombian government alleg-
ing that the legal status of Chivor 1 and 2 was
unclear.147 Such a move, which would likely have ren-
dered the property unavailable for sale pending any
attendant investigation, resulted only in directives
that Klein support his assertions with documentary
evidence.148 When he failed to do so, the complaint
was dismissed in April 1921, having had no effect.149
On December 27, 1919, Chivor 1 and 2 were pur-
chased from the Sociedad Ordinaria de las Minas de

140
This first option was mentioned in the sales contract for Chivor, Es-
critura 3084, December 27, 1919, Notaria 1, Archivo General de la
Figure 23. Although the American companies in- Nación (Colombia).
volved in Colombian oil and emerald were formally 141
These two events coincided with the travels of Griffiths and Mac-
separate, they were led by the same individuals and Fadden, who were together in Colombia in early 1918 before leav-
were even marketed in joint advertisements. Image ing the country on April 20 and arriving in New York on May 3,
1918. Files Wilson E. Griffiths and Carl K. MacFadden, List of United
from “The Colombian Review,” published by the States Citizens, Arriving at Port of New York, May 3, 1918, Ances-
Government Information Bureau of the Republic of try.com. The option was thus signed while they were in Colombia,
Colombia, New York, 1920. and the company was incorporated shortly after their return to the
United States. Although the actual April 3, 1918, option contract for
Chivor has not been found, an option contract with that date and
signed by Griffiths was located for the transfer of an oil field conces-
sion from Virgilio Barco to the Compañía Colombiana de Petróleo.
Escritura 331, April 3, 1918, Notaria 4, Archivo General de la
and 1920, focused generally on establishing Ameri- Nación (Colombia).
can interests in the oil business there and particu- 142
This second option was also mentioned in the sales contract for
larly on obtaining control of an oil field at the Chivor, Escritura 3084, December 27, 1919, Notaria 1, Archivo Gen-
eral de la Nación (Colombia).
border between Colombia and Venezuela, referred 143
Letter from August and Liese Klein, December 5, 1919, to Fritz and
to as the “Barco concession.” The emerald business Claire Klein; Canova, 1921.
was merely an adjunct to their activities in the oil 144
Rainier, 1931, p. 223 (comment by Carl K. MacFadden).
industry. The companies operating in the two sec- 145
Letter from August and Liese Klein, September 26, 1919, to Fritz
tors were formally separate but were led by the and Claire Klein.
146
same officials and were even marketed jointly in ad- Letter by August and Liese Klein, December 5, 1919, to Fritz and
Claire Klein. The more dramatic situation suggested by Klein (1941,
vertisements (figure 23). 1951) and by Canova (1921) appears to have been exaggerated. The
An option contract to purchase Chivor had al- Historical Archive of the German Ministry of Foreign Affairs contains
ready been obtained on April 3, 1918,140 prior to es- only a brief note, dated December 29, 1919, concerning the sale of
Chivor to an American company.
tablishing the Colombian Emerald Syndicate in June 147
Diario Oficial (Colombia), 56, No. 17080–17081 (1920), p. 273.
1918.141 A second option contract was executed on 148
Diario Oficial (Colombia), 56, No. 17102–17103 (1920), pp. 361–
August 13, 1919,142 with a term running to Decem- 362.
ber 8, 1919.143 MacFadden’s attention had been 149
Diario Oficial (Colombia), 57, No. 17687–17688 (1921), p. 258.

100 HISTORY OF CHIVOR, PART I: 1880–1925 GEMS & GEMOLOGY SPRING 2020
 In July 1920, a local mining company was formed
in Colombia by Griffiths and MacFadden to serve as
the operating entity for the Colombian Emerald Syn-
dicate.157 Known as Colombian Chivor Emerald
Company (Compañía Colombiana de Esmeraldas de
Chivor), stated capital was to consist of 20,000 shares
at $5 each (figure 24; again, see table 2).
In September 1920, Ross retired from business
and was thereafter succeeded by MacFadden as pres-
ident of the Colombian Emerald Syndicate.158 Grif-
fiths served as one of the company’s directors, and
Dixon acted as the Colombian manager.159
By 1921, interests in the titles to the other three
Chivor mines, Triunfo de Chivor, Paz de Chivor, and
Esperanza de Chivor were represented to be under
Figure 24. Share certificate for the Colombian Chivor the ownership or control of the Colombian Emerald
Emerald Company (Compañía Colombiana de Es- Syndicate or the Colombian Chivor Emerald Com-
meraldas de Chivor), formed in 1920 in Colombia as pany,160 but the nature of any relevant transactions
the operating company for the U.S.-based Colombian is not otherwise elucidated (again, see box A).
Emerald Syndicate, Ltd., which had purchased the
Chivor mines in December 1919. Courtesy of Eduardo
KLEIN’S ACTIVITIES FOR THE COLOMBIAN
Restrepo Ortega.
EMERALD SYNDICATE, LTD. (1920–1921 AND
1922–1923)
After the sale of Chivor to the American corporation
150 at the end of 1919, Klein and his wife remained in
Chivor by the American group. Pursuant to the
Colombia. Although the idea of opening a jewelry
sale contract, the mine titles for Chivor 1 and 2 and
shop in Bogotá was apparently broached by the Klein
the land holdings of the Colombian company were
transferred for a purchase price of £46,000 (equivalent
to US $230,000) to Griffiths and MacFadden. Grif-
fiths and MacFadden were represented in the trans- 150
Escritura 3084, December 27, 1919, Notaria 1, Archivo General de
action by Leon J. Canova, acting under a power of la Nación (Colombia).
151
Ibid.
attorney dated October 31, 1919.151 152
A share split was performed in late October to early November
In retrospect, circumstances rendered it almost a 1919, with the value of 50 new shares being equivalent to that of one
foregone conclusion that the Colombian Emerald old share. The Washington Post, December 11, 1919, p. 11.
Syndicate would not turn away from the opportu- 153
The Buffalo Enquirer, May 27, 1919, p. 8; New York Tribune, Octo-
nity to purchase Chivor. Between the corporation’s ber 26, 1919, p. 26; Quad-City Times, October 27, 1919, p. 12;
Springfield Missouri Republican, November 7, 1919, p. 6; The Salt
founding and the time the option was exercised, the Lake Tribune, December 3, 1919, p. 18; The Washington Post, De-
price of the 4,000 issued shares had increased from cember 11, 1919, p.11.
154
$25 to approximately $1,100,152 with continued ap- The Wall Street Journal, October 29, 1919, p. 15.
155
preciation to $4,000 or $5,000 expected.153 Moreover, Schwarz, 1953.
156
American Society of Civil Engineers, Year Book 1920, p. 312; Oil-
the company had already started mining operations, dom, 12 (1921), pp. 28–30.
and president Nathaniel W. Ross (1874–1938) was by 157
Escritura 2041, July 6, 1920, Notaria 1, and Escritura 2047, July 7,
October 1919 referring to Chivor as “the company’s 1920, Notaria 1, Archivo General de la Nación (Colombia). Diario
mine.”154 Civil mining engineer Harold Case Will- Oficial (Colombia) 56, No. 17238-17239 (1920), pp. 99–100.
158
cox had been hired in December 1919 to map the MacFadden, 1920; National Petroleum News, 12, December 1,
1920, pp. 75–76; Pittsburgh Daily Post, April 27, 1921, p. 11;
area for a deeper understanding of the geological set- Canova, 1921; Moore and Wilson, 2016. MacFadden had also be-
ting, and he continued in that position until March come president of the Carib Syndicate, Ltd., in September 1920. The
1920.155 Later, Willcox worked for the Carib Syndi- New York Times, September 10, 1920, p. 24.
159
Moody’s Manual of Investments: American and Foreign, 11, Part 2
cate, Ltd., in Colombia, underscoring the intercon- (1920), p. 1189; Moody’s Manual of Investments: American and For-
nectedness of the companies involved with the eign, 12, Part 2 (1921), p. 697; Canova, 1921.
American investors.156 160
Canova, 1921.

HISTORY OF CHIVOR, PART I: 1880–1925 GEMS & GEMOLOGY SPRING 2020 101
 to obtain a profitable supply of gem-quality emeralds.
However, during Klein’s first engagement, Dixon con-
tinued to be involved in the background, still working
for the Colombian Emerald Syndicate.
At that time, the mine was worked exclusively by
the traditional open-cut method of preparing terraces
or steps in the extremely steep environment (figure
26). During the initial contract period of approxi-
mately six months from the fall of 1920 to the spring
of 1921, Klein was assisted by several American min-
ing engineers, including Eugene Brossard.163 A notable
discovery occurred in December 1920 when a 632 ct
emerald crystal (figure 27) was found at Chivor by
miner Justo Daza.164 Klein (1941, 1951) recounted that
Daza located the cavity and that Klein himself care-
fully removed the crystal. The emerald was personally
transported to New York by Brossard in January
1921.165 The extraordinary stone generated great inter-
est in the United States, and its arrival in New York
was widely reported in the nation’s press.166
The crystal was subsequently named the “Patriz-
ius” or “Patricia” emerald after the patron saint of
Ireland, the green island, according to Klein.167 Spec-
ulation that it might have been named after Klein’s

161
Klein, 1941, 1951. No known contracts have been preserved to
offer specific details, but general parameters and time frames can also
be drawn from other contemporaneous materials such as newspaper
reports and travel documents.
162
An example of Christopher Dixon’s leadership can be found in the
form of a certificate he signed on September 27, 1920, at the time of
Anibal Restrepo’s departure from employment at Chivor and recog-
nizing nine months of exemplary service for the Colombian Emerald
Syndicate, Ltd.
163
Eugene Brossard’s work in Colombia was recognized in an alumni
Figure 25. Fritz Klein was unable to purchase the publication of his alma mater: “Eugene Brossard Madison, is in the
Chivor mining titles in 1919 on account of interven- employ of the Colombian Emerald Syndicate… as assistant to the su-
ing developments during World War I. The U.S.-based perintendent of the Chivor emerald mines.” The Wisconsin Alumni
Colombian Emerald Syndicate, Ltd., obtained option Magazine, 21, No. 6 (1920), p. 171. Another assistant at Chivor was
F. Stewart Turneaure. Wisconsin State Journal, September 4, 1921, p.
contracts in the interim and subsequently completed 7; Proceedings of the Geological Society of America for 1945, pub-
the acquisition. Klein remained in Colombia and was lished 1946, p. 106. Yet another American involved at Chivor was the
hired in the fall of 1920 by the new mine owners to mining engineer William Burns, who was first employed by the
lead operations at Chivor. Photo circa 1920 to 1921, Colombian Emerald Syndicate, Ltd., returning to New York in May
courtesy of Carola Kroll. 1922, and who later worked for the Colombia Emerald Development
Corporation at Chivor in 1925. File William Burns, List of United
States Citizens, Arriving at Port of New York, May 14, 1922, MyHer-
itage.com; Brock (1929).
164
brothers, such a plan was never implemented. In- Canova, 1921; Klein 1941, 1951; Johnson, 1959, 1961; Weldon et
al., 2016.
stead, Klein (figure 25) was hired by the Colombian 165
File Eugene Brossard, List of United States Citizens, Arriving at Port
Emerald Syndicate for two terms as head of mining of New York, January 14, 1921, MyHeritage.com.
operations at Chivor, from 1920 to 1921 and 1922 to 166
The Brooklyn Daily Eagle, February 2, 1921, p. 1; The Montana
1923.161 Standard, February 7, 1921, p. 3; Sikeston Standard, February 8,
1921, p. 3; The Tennessean, February 15, 1921, p. 9. Inconsistently,
At the time of Klein’s hiring, Dixon had been over- multiple articles or reports in American newspapers from February or
seeing Chivor’s administration.162 According to Klein March 1921 provided a weight of 630 ct.
(1941, 1951), he was hired because Dixon had failed 167
Klein, 1941.

102 HISTORY OF CHIVOR, PART I: 1880–1925 GEMS & GEMOLOGY SPRING 2020
Figure 26. Mining at Chivor in the 1920s was performed exclusively by the open-cut method, preparing terraces or
narrow steps in the extremely steep environment. Indigenous miners worked with long iron bars and hoes to re-
move the overburden and locate emerald-bearing veins or cavities. Photos of open-cast mining at Banco Klein,
circa 1920 to 1921, courtesy of Carola Kroll.

daughter168 (whose name was Susanne)169 is un- 632 ct emerald crystal is now in the collection of the
founded. Whether to cut the stone was initially de- American Museum of Natural History, New York,
bated, but the crystal was preserved as such.170 The inventory number 42027, and was donated to the
museum in 1953 by Mr. and Mrs. J. Robert Rubin.171
Figure 27. Left: In 1920, while Fritz Klein was leading
Rubin (1882–1958), a New York lawyer, served as one
mining operations at Chivor, an extraordinarily large of the trustees and/or directors of the Colombian
emerald crystal weighing 632 ct was discovered in the Emerald Syndicate in the 1920s.172 Prior to its dona-
open-cast mine. The crystal was transported to New tion, the emerald was transferred internally among
York by mining engineer Eugene Brossard and widely company officials after financial pressures in March
reported in the American press. The emerald was do- 1923 led the company to borrow $3,000 from two of
nated to the American Museum of Natural History, the directors, pledging the emerald as security.173
New York, in 1953. Courtesy of the American Mu- After Klein’s first contractual commitment with
seum of Natural History. Right: Klein (1941) also de- the Colombian Emerald Syndicate ended, he and his
picted the crystal, named the “Patricia” emerald. wife returned to Germany via New York. They ar-
rived at the New York port in August 1921 and may
have spent a period of several months before contin-
uing their journey.174

168
https://www.internetstones.com/patricia-emerald-632-carat-uncut-
colombian-origin-chivor-mine.html.
169
Carola Kroll (granddaughter of Fritz Klein), pers. comm., 2017.
170
Jewelers’ Circular, 82, No. 6 (1921), p. 79.
171
Press release, November 9, 1953, American Museum of Natural
History, New York; Jamie Newman, American Museum of Natural
History, New York, pers. comm., 2016.
172
Encyclopedia of American Biography, New Series, 9 (1938), pp.
401–402.
173
Jewelers’ Circular, 86, No. 10 (1923), p. 83.
174
The documentation cites the Colombian Emerald Syndicate, Ltd.,
as Fritz Klein’s employer and gives four months as the intended length
of stay in the United States. File Fritz Klein, List or Manifest of Alien
Passengers for the United States Immigration Officer at Port of Arrival,
New York, August 1921, MyHeritage.com.

HISTORY OF CHIVOR, PART I: 1880–1925 GEMS & GEMOLOGY SPRING 2020 103
 Figure 28. Hand-drawn
map showing the topog-
raphy of the Chivor
mine and related geo-
logical details, signed
R.S. [Robert Scheibe]
1922. The drawing also
identified the locations
principally worked by
Fritz Klein and Christo-
pher Dixon, labeled
“Banco Klein” and
“Banco Dixon,” respec-
tively. Prof. Robert
Scheibe collection,
University Library
Freiburg.

Following Klein’s initial engagement at the helm shrouded in history. On the corporate side, Paris re-
in Chivor, Dixon resumed leadership of mining oper- placed Arroyo as president in October 1922.185 Min-
ations.175 Dixon reportedly claimed that during the six
months from June to December 1921, emeralds worth
$176,000 were recovered under his management.176 175
Moody’s Manual of Investments: American and Foreign, 11, Part 2
Meanwhile, the corporate leadership in New York had (1920), p. 1189; Moody’s Manual of Investments: American and For-
undergone a transition. MacFadden resigned in Sep- eign, 12, Part 2 (1921), p. 697; Canova, 1921.
176
tember 1921 from his role at the head of the Colom- Gilles, 1930.
177
New York Tribune, September 13, 1921, p. 6; Engineering and Min-
bian Emerald Syndicate and was succeeded by Julian ing Journal 112 (1921), p. 466; The New York Herald, September 13,
A. Arroyo.177 In summer of 1922, the company’s vice 1921; The New York Times, September 13, 1921. Some months later
president, Joseph J. Paris, visited Colombia, returning Carl K. MacFadden was also dismissed from his role at the Carib Syn-
dicate, Ltd., oil company. New York Times, April 18, 1922, p. 32.
to New York on June 15.178 About one week later, 178
File Joseph Paris, List or Manifest of Alien Passengers for the United
Dixon arrived in New York,179 and he spent time in States Immigration Officer at Port of Arrival, New York, June 1922,
the city during June and July,180 presumably meeting MyHeritage.com; Investor & Trader, July 1, 1922, p. 6.
with Colombian Emerald Syndicate officials to dis- 179
File Christopher Dixon, List or Manifest of Alien Passengers for the
United States Immigration Officer at Port of Arrival, New York, June
cuss further activities at Chivor. Because no documen- 1922, MyHeritage.com.
tation after 1922 mentions Dixon in a management 180
The Washington Times, July 2, 1922, p. 3.
role at Chivor, the implication is that no agreement 181
Moody’s Manual of Investments: American and Foreign, 13, Part 2
for his continued employment was reached.181 That, (1922), p. 874; Investor & Trader, June 24, 1922, p. 6.
182
in turn, might have led to the rehiring of Klein. Claire Klein traveled alone from Hamburg to Puerto Colombia. File
Claire Klein, Hamburger Passagierlisten, August 1922, Ancestry.de.
The precise dates of Klein’s second contractual 183
Investor & Trader, August 19, 1922, p. 2, September 9, 1922, p. 2,
commitment with the Colombian Emerald Syndi- and September 30, 1922, p. 12; Klein, 1941, 1951.
cate are unknown. He first traveled alone to Colom- 184
Prof. Robert Scheibe fell ill in Colombia in February 1923 and died
bia, and his wife followed in August 1922.182 Good on March 3, following appendicitis and peritonitis. As a result, the
Chivor study was never completed. A decree by the Colombian Presi-
production was reported during this time,183 and dent dated March 8, 1923, was issued in honor of the deceased, and
Scheibe returned to conduct further scientific study numerous Colombian newspapers carried his obituary. The congress
at Chivor (figure 28).184 Nevertheless, the financial in Bogotá honored his work and his selfless service to Colombia in
June 1923 with an official state ceremony. Historical Archive of the
support for work at the mine was reduced, and, ulti- German Ministry of Foreign Affairs. Diario Oficial (Colombia) 59, No.
mately, mining activities were terminated. Even 18841-18842 (1923), p. 629.
now, the reasons that precipitated the decline remain 185
Investor & Trader, October 14, 1922, p. 8.

104 HISTORY OF CHIVOR, PART I: 1880–1925 GEMS & GEMOLOGY SPRING 2020
ing engineer Brossard, with whom Klein had worked With his employer’s demise, Klein returned to
at Chivor, returned to the United States in July Germany and over the following months established
1923.186 his own company in Idar in 1924, focused on the
emerald trade.199 He decided to travel to Colombia
BANKRUPTCY OF THE COLOMBIAN EMERALD once again in March 1925.200 At that time, the sale of
SYNDICATE, LTD. (1923) Colombian emeralds was still strictly controlled by
the government, with the only authorized channel
As noted above, financial pressures at the company
being through the firm Rosenthal in Paris. Klein’s
had become severe by March 1923 on account of ex-
aim was to obtain a similar authorization under
penses that exceeded the income generated from
which emeralds could also be sold to him directly or
emerald sales. By April 4, 1923, an involuntary peti-
to a partner company. After intensive negotiations
tion in bankruptcy had been filed against the Colom-
with Colombian officials, Klein was hired by the gov-
bian Emerald Syndicate in the United States District
ernment as an evaluator for rough emeralds in Paris,
Court in New York.187 At that time, Paris was still
in cooperation with the Colombian government’s
serving as the president, and Wolcott H. Pitkin was representative in the city, J.M. Arango.201 Klein was
appointed receiver for purposes of the bankruptcy never again active as a miner in Colombia, but re-
proceeding.188 Liabilities were initially estimated at mained at the center of the emerald trade. Several
between $45,000 and $48,000.189 Schedules detailing trips to Colombia in the second half of the 1920s and
the assets of $314,039 and liabilities of $37,816 were the early 1930s reflect his ongoing involvement.
then filed with the court on December 17, 1923.190
At that time, the Colombian Emerald Syndicate
owned the Chivor 1 and 2 mine titles and a one-third
interest in the titles to the three other mines of the 186
File Eugene Brossard, List of United States Citizens, Arriving at Port
Chivor group.191 Outstanding debts were owed to, of New York, July 1923, MyHeritage.com.
187
among others, present or former presidents, vice Jewelers’ Circular, Vol. 86, No. 10 (1923), p. 83; The Wall Street
Journal, April 5, 1923, p. 14; The Cumulative Daily Digest of Corpora-
presidents, and/or directors of the corporation, such tion News, No. 2 (1923), pp. 142–143.
as MacFadden, Griffiths, Arroyo, and Rubin. Other 188
The Wall Street Journal, April 5, 1923, p. 14; The Cumulative Daily
listed creditors included Klein, mining engineer Digest of Corporation News, No. 2 (1923), p. 142; Jewelers’ Circular,
Vol. 86, No. 10 (1923), p. 83.
Brossard, the corporation’s attorney in Colombia 189
The Wall Street Journal, April 5, 1923, p. 14; The Cumulative Daily
Miguel S. Uribe Holguin, the Carib Syndicate, and Digest of Corporation News, No. 2 (1923), pp. 142–143; Jewelers’
the German firm Hahn & Klein.192 Circular, Vol. 87, No. 22 (1923) p. 91.
190
Meanwhile, a stockholders’ committee was estab- Jewelers’ Circular, Vol. 87, No. 22 (1923), p. 91; The Cumulative
Daily Digest of Corporation News, No. 4 (1923), p. 118.
lished to devise a method of raising funds to pay off 191
Jewelers’ Circular, Vol. 86, No. 10 (1923), p. 83; Jewelers’ Circular,
the liabilities, and it was proposed that a new com- Vol. 87, No. 22 (1923), p. 91.
pany would be formed to bid for the assets in an an- 192
A company based in Idar, Germany, owned by August Klein and
ticipated bankruptcy auction.193 A special meeting of his brother-in-law August Hahn.
193
creditors was held in January 1924 to consider and The Cumulative Daily Digest of Corporation News, No. 2 (1923), p.
143; No. 3 (1923), p. 125.
act upon offers to purchase the property.194 194
The Cumulative Daily Digest of Corporation News, No. 1 (1924), p.
The mining titles for Chivor 1 and 2 and the re- 164.
lated land ownership were sold on or about August 195
The Cumulative Daily Digest of Corporation News, No. 3 (1924), p.
23, 1924, to the recently formed U.S. company 150, 161.
196
The Cumulative Daily Digest of Corporation News, No. 4 (1924),
Chivor Emerald Corporation.195 Another new entity, pp. 209, 222; Winkler, 1928; Rainier, 1929, 1931.
the Colombia Emerald Development Corporation, 197
The Buffalo Enquirer, June 30, 1922, p. 5; The Baltimore Sun, De-
was then established on November 7, and the Chivor cember 30, 1922, p. 11; Democrat and Chronicle, April 27, 1923;
claims and land were transferred to that company in The Tampa Tribune, December 14, 1923, p. 15. Additional details
charting stock values can be traced through issues of Investor &
late 1924.196 Throughout these machinations, the Trader published in 1922 and 1923.
value of stock in the Colombian Emerald Syndicate 198
Engineering and Mining Journal-Press, 21 (1925), p. 999.
fell precipitously, dropping from 65 cents per share 199
Carola Kroll (granddaughter of Fritz Klein), pers. comm., 2017.
200
in June 1922 and 40 cents in December 1922 to 5 Klein traveled via the port of Hull in Great Britain to Santa Marta in
Colombia. https://www.findmypast.co.uk/Colombia. Findmypast.co.uk
cents in April 1923 and 1 cent in December 1923.197 201
J.M. Arango was nominated General Consul of Colombia in Paris in
The downward spiral continued thereafter, and the March 1925 and was responsible for the selling of emeralds to Rosen-
value was reported in 1925 at $1 for 600 shares.198 thal. Diario Oficial (Colombia), 61, No. 19862 (1925), p. 519.

HISTORY OF CHIVOR, PART I: 1880–1925 GEMS & GEMOLOGY SPRING 2020 105
TABLE 4. History of Chivor, 1880–1925.

Date Events in Colombia, Germany, or the United States

1880s Search for and rediscovery of old native and Spanish tunnels by F. Restrepo
March 1889 Mining titles for Chivor 1 and 2 are granted to F. Restrepo, E. González, B. Escobar, and Núñez & Compañía
August 1896 F. Restrepo and L. Correa are referred to as the mine owners; E. González declares the Chivor mines “abandoned”
February 1898 Transfer of the mining titles to E. González
August 1898 Formation of a company by F. Restrepo and three companions; transfer of the mining titles to the company
1898–1899 Cleaning of former tunnels and repair of water channels at the mine by F. Restrepo
1899–1909 Numerous sales of shares involving different shareholders take place
1902 First publications by the British Vice-Consul S. Dixon mentioning the rediscovery of Chivor
August 1904 The five principal shareholders buy land from A. Acosta
December 1904 Three additional claims near Chivor 1 and 2 are registered for the Triunfo de Chivor, Paz de Chivor, and
Esperanza de Chivor mines, with F. Restrepo, C. Dixon, and H. Campo named as co-owners of the three titles
1905–1906 C. Dixon attempts but fails to sell the three additional mines
February 1912 Preliminary option contract between F. Klein, H. Span, and the mine owners
June 1912 Formalized option contract with the mine owners; F. Klein and F. Restrepo start mining at Chivor
End of 1912 A. Klein joins the miners at Chivor
March 1913 F. Klein and F. Restrepo travel to Germany; A. Klein remains at Chivor
July 1913 Option contract with F. Klein is extended by Klein’s lawyer
November 1913 F. Restrepo leaves Germany for Colombia
December 1913 F. Klein’s option contract is extended again
February 1914 Formation of the “Syndikat El Chivor”; F. Klein and R. Scheibe leave Germany for Colombia
April 1914 R. Scheibe begins evaluation of Chivor; Restrepo becomes ill and dies at Guateque
August 1914 R. Scheibe finishes the evaluation of Chivor; outbreak of World War I
End of 1914 or early 1915 F. Klein leaves Colombia for Germany
1915–1916 C. Dixon again fails to sell the three additional mines
April 1918 Option contract between W.E. Griffiths, C.K. MacFadden, and the mine owners
June 1918 Formation of the “Colombian Emerald Syndicate, Ltd.”; N.W. Ross becomes president
July or August 1919 F. Klein and his wife leave Germany for Colombia
August 1919 Second option contract between W.E. Griffiths, C.K. MacFadden, and the mine owners
December 1919 Purchase of Chivor 1 and 2 by W.E. Griffiths and C.K. MacFadden
By early 1920 C. Dixon serves as Colombian manager for the “Colombian Emerald Syndicate, Ltd.”
By September 1920 C.K. MacFadden becomes president of the Colombian Emerald Syndicate, Ltd., succeeding N.W. Ross
Fall 1920–Spring 1921 F. Klein serves as head of mining operations at Chivor
December 1920 632 ct “Patricia” emerald is found and sent to New York
June–December 1921 C. Dixon serves as head of mining operations at Chivor
September 1921 C.K. MacFadden resigns as president and is succeeded by J.A. Arroyo
June–July 1922 C. Dixon travels to New York and presumably meets with company officials to discuss further activities at Chivor
Summer 1922–1923 F. Klein again serves as head of the mining operations at Chivor
October 1922 J.A. Arroyo is succeeded by J.J. Paris
By April 1923 An involuntary bankruptcy petition is filed in New York against the “Colombian Emerald Syndicate, Ltd.”
August 1924 Properties are sold to the “Chivor Emerald Corporation” in a bankruptcy sale
End of 1924 Properties are transferred to the “Colombia Emerald Development Corporation”

106 HISTORY OF CHIVOR, PART I: 1880–1925 GEMS & GEMOLOGY SPRING 2020
LEGENDS AND REALITY: KLEIN AND RESTREPO Nonetheless, Klein did recount a large number of
The present authors have endeavored to describe the details that have been verified by contemporaneous
events between 1880 and 1925 to the extent corrob- materials, particularly concerning travel arrange-
orated by contemporaneous primary documents, ments and individuals involved. The names offered
without filling gaps with dramatizations and appar- for ships by which he journeyed between Germany,
ent fabrications, some of which have been repeated Colombia, and the United States have been corrobo-
so often as to become accepted “facts.” Also, no ef- rated by manifests, contracts, or letters, and many
fort has been made to explain the origin of these dis- persons are accurately named and characterized,
credited or unverified views. Consequently, the story aside from minor spelling errors (e.g., the Colombian
given here leaves various points open, waiting for fu- Emerald Syndicate’s attorney Uribe-Holguin and
ture closure through discovery of further records, Klein’s lawyer Cadavid in Bogotá). On the other
whether in Colombian archives or elsewhere. A few hand, some key figures are inexplicably absent or
of these lingering ambiguities and the milieu in given a much diminished role. Stauch, a principal in-
which they arise may be highlighted as follows. vestor in the German group, is missing, and Klein’s
Klein’s 1941 memoir describes in a lively way his brother-in-law Jäckh is given little credit for his cru-
travels in South America, always connected with the cial influence in obtaining protection for Klein to re-
pursuit of new gem deposits. It went to press in the turn to Colombia in 1911 and in facilitating contact
middle of World War II and was geared toward a cir- with potential investors such as Schacht. The name
cle of readers interested in gemstones and the “ad- of his own brother August, who spent nearly two
ventures” associated with their discovery, mining, years at Chivor, is never mentioned.
and trading. The book was not intended to be a his- Further complicating the situation created by the
torical document, consistent in all aspects with the dramatized version of events is the problem of gaps
rather complex and vacillating legal situation in in the historical record. Regarding Restrepo, for ex-
Colombia, but was written for the entertainment of ample, certain aspects of his activities are clear,
the interested reader. while others are unexplained. Materials found to date
Notably, the book was penned after an interval of are silent as to his partner Correa’s involvement in
almost 30 years, in terms of events before 1915, and the search for the forgotten Chivor mines in the
nearly 20 years, based on the events after 1918. It is 1880s; the reasons why the mining titles were twice
not known what contemporaneous documents and transferred, first from Restrepo and three partners to
records were available to Klein when drafting the Restrepo and Correa between 1889 and 1896, and
original manuscript or to what extent such manu- later from Restrepo and Correa to González between
script was edited by the publisher or elsewhere, 1896 and 1898; and the rationale for González shortly
much less the expertise or awareness of those in- thereafter in 1898 to form a new entity in partnership
volved in the editorial process. A degree of liberality with Restrepo and the other two original title appli-
with facts is evidenced by the statement on the back cants. Restrepo’s activities during several largely
cover of the 1941 edition referring to “drill holes that silent periods and his involvement in a range of other
produced surprising results,” although no under- ventures, such as farming and the coffee trade,202 are
ground mining activities or their results were men- equally obscure.
tioned in the pages. The record is likewise deficient as to the dealings
Looking at the chronology of events (see table 4), of the Colombian Emerald Syndicate, including
it can be seen that actual developments in Germany Dixon’s mining activities for the company, the cir-
and Colombia, especially regarding the acquisition of cumstances that led to Klein’s employment simulta-
Chivor, formed a rough framework for the story told neously with Dixon during one interval, and the
by Klein. This background was supplemented and em- reasons precipitating the 1923 bankruptcy. Thus the
bellished in some places and even substantially dram- story, and the search, continues.
atized in others. For instance, the purported
one-and-a-half-year search for the first emeralds in
Chivor, Restrepo’s death in 1914, and the “fight” for
the mine with the American competitors in 1919 are
all illustrative of the tendency to sensationalize. 202
Gaceta Judicial (Colombia), 20 (1911), pp. 49–57.

HISTORY OF CHIVOR, PART I: 1880–1925 GEMS & GEMOLOGY SPRING 2020 107
 ABOUT THE AUTHORS (grandson of Prof. Robert Scheibe). Literature on the history of the
Dr. Schmetzer is an independent researcher living in Peters- gem trade in Idar-Oberstein, as well as historical and biographical
hausen, near Munich, Germany. Mr. Martayan is a senior geo- data, were shared from the private archives of Mr. Dieter
physicist and longtime emerald aficionado residing in Paris, Jerusalem and Mr. Klaus Eberhard Wild; biographical data from
France. Mr. Ortiz is a mechanical engineer and emerald dealer the Idar-Oberstein city archive was supplied by Mr. Manfred
who owns the Colombia Emerald Co., based in Bogotá, and Los Rauscher. In the search for the relevant files on German activities
Angeles, California. in Colombia and related biographical data, Ms. Annegret Wilke
and Ms. Lucia van der Linde, Historical Archive of the German
ACKNOWLEDGMENTS Ministry of Foreign Affairs, and Mr. Ralf Engel, German Federal
Documents, pictures, biographical data, and other information Archive, assisted. Prof. Kris Lane, Tulane University, New Orleans,
concerning the events of the years 1880 to 1925 described here generously discussed his extensive research on Colombian his-
have been provided by relatives of the persons involved, who have tory. The Ernst Jäckh collection at Columbia University in New York
willingly reviewed their private records. Thanks are given especially was reviewed by Ms. Andrea Blake, and files of Wm H. Müller &
to Ms. Iris Böhmer (granddaughter of Dorotheus Kracker von Co. in the Stadsarchief Rotterdam were accessed by Mr. Martijn
Schwartzenfeldt), Mr. Simon Hamilton (great-grandson of George Verbon. Victor Castañeda and Yovanny Alba accompanied one of
Dixon), Ms. Anette Klein (granddaughter of August Klein), Mr. the authors (GM) on various expeditions to the Chivor mining area.
Michael Krafft (grandson of August Stauch), Ms. Carola Kroll Finally, countless other unnamed individuals and staff members at
(granddaughter of Fritz Klein), Mr. Eduardo Restrepo Ortega various archives, libraries, and local administrations should also be
(grandson of Francisco Restrepo Escobar), Ms. Cordula Schacht thanked here for numerous details and clues that helped in finding
(daughter of Dr. Hjalmar Schacht), and Mr. Andreas Scheibe descendants of individuals involved and in completing the puzzle.

REFERENCES
Acosta J. (1848) Compendio histórico del descubrimiento y colo- dad extracontractual del Estado, con posterioridad a la Guerra
nización de la Nueva Granada en el siglo décimo sexto. Im- Civil de 1899 a 1902. Universidad del Rosario, Facultad de Ju-
prenta de Beau, Paris. risprudencia, Maestria en derecho administrative, Bogotá.
Aguado P. de, Fr. (1906) Recopilación Historial. Imprenta Na- Díaz C.A. (1967) Páginas de Historia Colombiana. Recuerdo de
cional, Bogotá. Antonio José Cadavid. Imprenta del Departamento, Bucara-
Aguilar F.C. (1884) Ligero estudio sobre las minas de Méjico y manga.
Colombia. Anales de la Instrucción Pública en los Estados Dixon S.S. (1902) Mining in Colombia in 1901. The Mining Jour-
Unidos de Colombia, Vol.7, pp. 493–528. nal, Vol. 72, p. 429.
Ancizar M. (1853) Peregrinación de Alpha: por las provincias del Domínguez R.A. (1965) Historia de las esmeraldas de Colombia.
norte de la Nueva Granada en 1850 y 51. Imprenta de Echev- Gráficas Ducal, Bogotá.
erria Hermanos, Bogotá. Eder P.J. (2017) A banker invites other bankers to make money in
Anderton R. (1953) Tic-Polonga. Doubleday, New York. Colombia. In Farnsworth-Alvear A., Palacios M., and Gómez
——— (1954) Tic-Polonga. Museum Press Limited, London. López A.M., Eds., The Colombia Reader: History, Culture, Pol-
Annual report on Colombia for the year 1913. In: Philip G., Ed., itics. Duke University Press, Durham and London, pp. 457–
British documents on foreign affairs – reports and papers from 463.
the Foreign Office Confidential print. Part I, from the mid-nine- Enciclopedia universal ilustrada europeo-americana: Apéndice 2
teenth century to the First World War. Series D, Latin America, (BED-CEQ). (1931) Espasa-Calpe, Madrid.
1845–1914, published 1992, pp. 182–193. Esquivel Triana R. (2001) Fuentes para la historia agraria e indus-
Atuesta D. (1899) Informe relativo al levantamiento de los planos trial de la provincia de Neiva 1778-1938. Memoria y Sociedad:
y mensura de los terrenos de las minas de esmeraldas de Revista del Departamento de Historia y Geografía, Vol. 5, No.
propiedad de la nación. Imprenta Nacional, Bogotá. 9, pp. 61–88.
Bermúdez J.A. (1966) Antonio José Cadavid. Revistas Universidad Fals-Borda O. (1955) Odyssey of a sixteenth-century document –
Pontificia Bolivariana, Vol. 28, No.100, pp. 347–354. Fray Pedro de Aguado’s “Recopilacion Historial.” The Hispanic
Brisson J. (1899) Viajes por Colombia en los años de 1891 á 1897. American Historical Review, Vol. 35, No.2, pp. 203–220.
Imprenta Nacional, Bogotá. Falz E. (1939) Von Menschen und Edlen Steinen. Selbstverlag, Idar.
Brock F.W. (1929) Investment pitfalls. Oral Hygiene, Vol. 19, pp. Fernández Piedrahita L. (1881) Historia general de las conquistas
789–795. del nuevo reino de Granada. A la S.C.R.M. de D. Carlos Se-
Bullman C. (1892) The Mining Laws of the Republic of Colombia. gundo Rey de las Españas y de las Indias. Imprenta de Medardo
The Scientific Publishing Company, New York. Rivas, Bogotá.
Bürger O. (1919) Reisen eines Naturforschers im tropischen Süd- Fortaleché D., Lucas A., Muyal J., Hsu T., Padua P. (2017) The
amerika. Dieterich´sche Verlagsbuchhandlung, Leipzig. Colombian emerald industry: Winds of change. G&G, Vol. 53,
Cadavid A.J. (1913) En defensa del derecho de propiedad. Imprenta No. 3, pp. 332–358, http://dx.doi.org/10.5741/GEMS.53.3.332
de La Luz, Bogotá. Francis J.M. (2007) Invading Colombia. The Pennsylvania State
Calvo Perez B. (1992) Minas y minerales de Iberoamérica. IGME, University Press, University Park, PA.
Madrid. Friede J. (1964) Fray Pedro Aguado, con Ocasión del 450o Aniver-
Canova L.J. (1921) Legend of the Discovery of Emeralds in the sario de su Nacimiento. The Hispanic American Historical Re-
Colombian Andes. Colombian Emerald Syndicate, Ltd., New view, Vol. 44, No. 3, pp. 382–389.
York. Gamba F.P. (1913) Emerald fields of Colombia. Mining and Scien-
Céspedes Cubides J.E. (2017) Reminiscencias de la Guerra de los tific Press, Vol. 107, p. 345.
Mil Días. Descripción y análisis del régimen de responsabili- Geografía económica de Colombia, Tomo III: Boyacá, 1936. Bo-

108 HISTORY OF CHIVOR, PART I: 1880–1925 GEMS & GEMOLOGY SPRING 2020
 gotá, Contraloría General de la República. Quintero A., Centeno R. (2007) Hacendados, comerciantes y ne-
Gilles V.A. (1930) Geology of the Chivor emerald mine. In Buis, gociantes de Neiva a finales del siglo XIX: actores, capitales,
O. and Bronkie W.F., Eds., Guide Book to the Geology of the sociedades y negocios. Universidad Surcolombiana, Neiva.
Chivor Emerald Mine. Colombian Society of Petroleum Geol- Rainier P.W. (1929) The Chivor-Somondoco emerald mines of
ogists and Geophysicists, published 1966, pp. 5–18. Colombia. The American Institute of Mining and Metallurgical
González E. (1911) Esmeraldas de las minas de Muzo y Coscuez. Engineers, Technical Publication No. 258.
Imprenta Moderna, Bogotá. ——— (1931) The Chivor-Somondoco emerald mines of Colom-
Groot J.M. (1869) Nueva Granada, escrita sobre documentos bia. Transactions of the American Institute of Mining and Met-
auténticos. Imprenta a cargo de Foción Mantilla, Bogotá. allurgical Engineers AIME, Vol. 96, pp. 204–223.
Jäckh E. (1954) Der Goldene Pflug. Deutsche Verlags-Anstalt, ——— (1933) Las minas de esmeraldas de Muzo y Coscuez. Inter-
Stuttgart. nal Report printed 1958 in Boletín de Minas Vol. 5, Nos. 56–
Johnson P.W. (1959) The emeralds of Chivor today. Lapidary Jour- 57, pp. 10–19.
nal, Vol. 13, No. 4, pp. 516–524. ——— (1942) Green Fire. Random House, New York.
——— (1961) The Chivor emerald mine. Journal of Gemmology, Rauff H. (1926) Robert Scheibe †. Gedächtnisreede. Jahrbuch der
Vol. 8, No. 4, pp. 126–152. Preußischen Geologischen Landesanstalt zu Berlin, Vol. 47,
Klein F. (1941) Smaragde unter dem Urwald. Oswald Arnold Ver- No. 2, pp. 63–71.
lag, Berlin. Rausch J.M. (2014) Colombia and World War I: The Experience
——— (1951) Smaragde unter dem Urwald. Selbstverlag, Idar- of a Neutral Latin American Nation during the Great War
Oberstein. and its Aftermath, 1914-1921. Lexington Books, Lanham,
Kopper C. (2006) Hjalmar Schacht: Aufstieg und Fall von Maryland.
Hitlers mächtigstem Bankier. Deutscher Taschenbuch Ver- Scheibe R. (1916) Minas de Muzo. Informe del Ministro de Ha-
lag, Munich. cienda (Colombia) 1916, pp.181–198.
Lane K. (2010) Colour of Paradise: The Emerald in the Age of ——— (1922) Informe del doctor Robert Scheibe sobre la mina de
Gunpowder Empires. Yale University Press, New Haven and Muzo. Informe del Ministro de Hacienda al Congreso de 1922,
London. pp. 99–112.
Latham E.B. (1911) The newly discovered emerald mines of ——— (1926) Die Smaragdlagerstätte von Muzo (Kolumbien) und
“Somondoco.” The School of Mines Quarterly, Vol. 32, No. 3, ihre nähere Umgebung. Neues Jahrbuch für Mineralogie, Beila-
pp. 210–214. genband 54, Abt. B, pp. 417–447.
Legrand C. (1988) Colonización y protesta campesina en Colom- Schuhmacher H.A. (1875) Ueber die Columbischen Smaragden.
bia (1850-1950). Universidad Nacional de Colombia, Bogotá. Zeitschrift der Gesellschaft für Erdkunde zu Berlin, Vol. 10,
Levinson O. (2007) Diamanten im Sand: Das wechselvolle Leben pp. 38–62.
von August Stauch. Kuiseb Verlag, Windhoek. Schwarz H.F. (1953) In memory. The Explorers Journal, Vol. 31,
MacFadden C.K. (1920) Petroleum in Colombia. The Colombian No. 4, pp. 54–55, 64.
Review, Vol. 1, pp. 36–38. Simón P., Fr. (1891) Noticias historiales de las conquistas de
Martin de Retana J.M., Ed. (1990) El gran libro de la esmeralda. Tierra Firme en las Indias Occidentales, Segunda Parte. Casa
Editorial la Gran Enciclopedia. Vasca, Bilbao. Editorial de Medardo Rivas, Bogotá.
Martínez F.A. (1950) Don Emiliano Isaza. Thesaurus: Boletín del Singewald Q.D. (1950) Mineral resources of Colombia (other
Instituto Caro y Cuervo, Vol. 6, No. 2, pp. 271–278. than petroleum). Geologic Investigations in the American
Martínez Covaleda H.J., Martínez Covaleda R. (1996) Economía y Republics, 1949. Geological Survey Bulletin 964-B, pp. 53–
región: aproximación a la historia del siglo XIX en el Huila. In 61.
Tovar Zambrano, B., Ed., Historia general de Huila, Vol. 3, Ac- Sinkankas J. (1981) Emerald and Other Beryls. Chilton Book
ademia Huilense de Historia, pp. 15–76. Company, Radnor, Pennsylvania.
Montaña F. (1915) Renta de esmeraldas: Apuntes fiscales. The Skinner W.R. (1889-1890) The Mining Manual, Vol. 3, pp. 116–117,
Salesian Press, London. 333.
Moore T.P., Wilson W.E. (2016) The emerald mines of Colombia. ——— (1891–1892) The Mining Manual, Vol. 4, pp. 75, 121, 341.
The Mineralogical Record, Vol. 47, No. 1, pp. 5–68. ——— (1896) The Mining Manual, Vol. 8, p. 1151.
Moreno F.A. (1772) Estado del Virreinato de Santafé, Nuevo Reino ——— (1897) The Mining Manual, Vol. 9, pp. 1279–1280.
de Granada. Anales de la Universidad Nacional de los Estados ——— (1900) The Mining Manual, Vol. 12, p. 1190.
Unidos de Colombia, Vol. 4, No. 19, 1870, pp. 34–96. Strunz H. (1970) Von der Bergakademie zur Technischen Univer-
Moreno Calderón D. (2011) Cronología de la historia del Estado sität Berlin 1770-1970. Förderer der Berliner Fakultät für Berg-
Soberano del Tolima. HUILA, Vol. 14, No. 62, pp. 29–67. bau und Hüttenwesen e.V., Berlin.
Ochoa Ortiz F. (1921) Immigration into Ecuador. The Republic of Torres del Río C.M. (2015) Colombia siglo XX: desde la Guerra
Ecuador, Vol. 1, No. 1, pp. 24–25. de los Mil Días hasta la elección de Álvaro Uribe. Editorial
Plaza J.A. (1850) Memorias para la historia de la Nueva Granada Pontificia Universidad Javeriana, Bogotá.
desde su descubrimiento hasta el 20 de julio de 1810. Imprenta Weldon R., Ortiz J.G., Ottaway T. (2016) In Rainier’s footsteps:
del Neo-Granadino, Bogotá. Journey to the Chivor emerald mine. G&G, Vol. 52, No. 2, pp.
Pogue J.E. de (1917) The emerald deposits of Muzo, Colombia. 168–187, http://dx.doi.org/10.5741/GEMS.52.2.168
Transactions of the American Institute of Mining Engineers, Winkler M. (1928) Investments of United States Capital in Latin
Vol. 55, pp. 910–934. America. World Peace Foundation, Boston.

HISTORY OF CHIVOR, PART I: 1880–1925 GEMS & GEMOLOGY SPRING 2020 109
FEATURE ARTICLES

RHODONITE-PYROXMANGITE FROM
TANATZ ALP, SWITZERLAND
Franca Caucia, Luigi Marinoni, Maria Pia Riccardi, Omar Bartoli, and Maurizio Scacchetti

The chemical, physical, and gemological properties of attractive rhodonite-rich rocks from Tanatz Alp (Switzer-
land) were investigated using classical gemological methods, petrographic observation, and analytical techniques
such as X-ray powder diffraction, Raman spectroscopy, and scanning electron microscopy. The color of the sam-
ples varied from pale pink to purplish pink; they were opaque and contained braunite, rhodochrosite, and spes-
sartine. The gem-quality samples have an attractive color and are compact with no fractures. Tanatz Alp
rhodonites are characterized by the presence of other minerals such as kutnohorite, spessartine, ankerite,
rhodochrosite, khristovite-(Ce), pyroxmangite, and chlorite. The presence of the polymorph pyroxmangite was
also confirmed by micro-Raman analyses. The samples had a chemical composition of MnO (45.97–48.64 wt.%)
and SiO2 (46.70–47.92 wt.%), with very low amounts of MgO (0.64–1.65 wt.%) and CaO (<5 wt.%). Rhodonites
from Tanatz Alp have some economic importance because of their use in jewelry and ornamental objects.

R
hodonite is the Mn-rich member of the pyrox- state gem in 1979. Rhodonite is collected as a lap-
enoid group; it is triclinic with a structure idary material and ornamental stone and is used to
made up of silicate chains parallel to the c-axis, make cabochons, beads, small sculptures, tumbled
in turn composed of a repeating sequence of five stones, and other objects. High-quality crystals of
tetrahedral units. The chains alternate with M octa-
hedral sites, designated M1 though M5, where diva-
lent cations reside (Deer et al., 1997, and references
therein). The ideal chemical formula is MnSiO3.
In Brief
Rhodonite of this composition has been synthesized • Rhodonite from Tanatz Alp in Switzerland is associ-
but has never been found in nature, where Ca, Mg, ated with kutnohorite, spessartine, ankerite, and
rhodochrosite.
and Fe2+ replace Mn (Ito, 1972).
Rhodonite was first discovered in 1790 in the Ural • Classical gemological methods, petrographic obser-
vation, and analytical techniques demonstrated the
Mountains near Sidelnikovo. The mineral was
presence of the two polymorphs of MnSiO3:
named in 1819 by the German naturalist Christoph rhodonite and pyroxmangite.
Friedrich Jasche. The name is derived from the Greek
• Fashioned rhodonites from Tanatz Alp display an at-
rhodo, which means “rose,” because of its character- tractive pale pink purplish pink color.
istic pink color (figure 1). The Hermitage Museum in
Saint Petersburg preserves many precious and deco-
rative objects made with rhodonite that belonged to
the Russian aristocracy, especially the czars. Even this mineral can be very expensive. Rhodonite is gen-
today, Russian children exchange rhodonite eggs for erally found in massive translucent to opaque aggre-
Easter as a gesture of friendship. In the United States gates, and the best-quality gems are transparent. The
there are important rhodonite mines in New Jersey crystals have perfect cleavage in two directions and
and in Massachusetts, which declared it the official low hardness (5.5–6 Mohs), making it one of the most
difficult gemstones to cut. For this reason, faceted
rhodonites are typically sold as collectible gems
See end of article for About the Authors and Acknowledgments.
GEMS & GEMOLOGY, Vol. 56, No. 1, pp. 110–123,
rather than for jewelry use.
http://dx.doi.org/10.5741/GEMS.56.1.110 Rhodonite is very similar to its polymorph, pyrox-
© 2020 Gemological Institute of America mangite, and to rhodochrosite (MnCO3). Pyroxman-

110 SWISS RHODONITE FROM TANATZ ALP GEMS & GEMOLOGY SPRING 2020
 Figure 1. Rhodonite
bead necklace and ear-
rings. Photo by Emily
Lane; courtesy of Dona
Dirlam.

gite has a structure very similar to that of rhodonite, acids, whereas rhodonite is resistant to acids and usu-
with a repeating sequence of seven tetrahedral units; ally associated with manganese oxides.
the distinction from rhodonite therefore requires Rhodonite and pyroxmangite contain small
more accurate investigations with techniques such amounts of Ca, Mg, and Fe2+ substituting for Mn2+;
as X-ray powder diffraction (XRPD) and electron mi- the P-T compositional stability limits have been in-
croprobe. Rhodochrosite differs in that it frequently vestigated by many authors, but there are still uncer-
shows white streaks of calcite and is reactive to hot tainties (Ito, 1972; Peters et al., 1973; Ohashi et al.,

SWISS RHODONITE FROM TANATZ ALP GEMS & GEMOLOGY SPRING 2020 111
1975; Maresch and Mottana, 1976; Brown et al., 1980; 1999), but those in the Tanatz Alp (figure 2) are less
Akaogy and Navrotsky, 1985; Abrecht, 1988; Taka- well known. This work investigates the chemical,
hashi and Hariya, 1995; Zanazzi et al., 2008; Diella et physical, and gemological properties of rhodonite
al., 2014, and references therein). The two minerals from Tanatz Alp.
can generally be distinguished on the basis of their
chemical composition, as rhodonite contains a higher DESCRIPTION OF THE DEPOSIT
amount of Ca (more than 0.05 wt.%) than pyroxman-
Tanatz Alp is located about 2.5 kilometers south of
gite. Rhodonite and pyroxmangite often occur to-
the village of Splügen in the Swiss canton of Grisons.
gether as a bladed intergrowth, in various types of ore
The geological setting is similar to that of the more
deposits and Mn-rich lithologies (Ohashi et al., 1975;
famous Val Ferrera. Tanatz Alp lies between the crys-
Jefferson et al., 1980; Pinckney and Burnham, 1988;
talline basement and the sedimentary cover of the
Millsteed et al., 2005; Michailidis and Sofianska,
Tambo nappe in the so-called Spluga zone; a short dis-
2010).
tance to the east is the nappe of Suretta. Both nappes
Deposits of rhodonite are found all over the world.
belong to the Pennidic domain and are formed by
The most important are those at Broken Hill in New
rocks of the crystalline basement, metamorphosed
South Wales, Australia, where the best single-crystal
gems are found (Millsteed et al., 2005; Millsteed, during the Alpine orogeny, and by overlying shallow-
2006); Franklin, New Jersey, in the United States basin sedimentary sequences, which were subse-
(Nelson and Griffen, 2005); many areas of the Ural quently metamorphosed into Mn- and Fe-rich schists
Mountains (Bukanov, 2006; Brusnitsyn, 2010); and quartzites (Biggioggero and Montrasio, 1990; fig-
British Columbia, Canada (Simandl et al., 2001); ure 2). The first description of the minerals from
Huánuco, Peru (Wilson,1989); Minas Gerais, Brazil Tanatz Alps was by Kenngott (1866), who reported
(Quinn, 2004; Leverett et al., 2008); Pajnsberg, Swe- the discovery of rhodonite in Mount Splügen.
den (Lee, 1958); and several localities in Japan At the beginning of the twentieth century, interest
(Ohashi et al., 1975). A variety of rhodonite contain- in manganese minerals increased considerably, and
ing high levels of zinc (up to 10 wt.% ZnO), called this Swiss resort became an economically important
fowlerite, was found at Franklin, New Jersey (Nelson source (Grenouillet, 1920). Rhodonite in Tanatz Alp
and Griffen, 2005). occurs in large isolated boulders between half a meter
In the Italian Alps, rhodonite has been found in and a meter long (figure 3), distributed along an ex-
the Praborna mine near San Marcel in Valle d'Aosta tension of about 100 meters on the slope to the west
(Mottana, 1986), Feglierec near Alagna Valsesia in of the switchbacks ascending the pass at an elevation
Piedmont (Peters et al., 1978), Monte Forno in Ma- of 1,720 m. Between 1918 and 1920, excavations were
lenco Valley, Scerscen in Lombardy (Diella et al., carried out to identify the original outcrop of this
2014), and Monte Civillina in Veneto (Schiavinato, mineral, but without success. Since 1970, stonecut-
1953). Rhodonite is also found in the Apennines ters and artisans have rediscovered an appreciation for
chain, in the mines of Gambatesa and Molinello in the manganese minerals of Tanatz Alp, which are
Graveglia Valley in Liguria (Marchesini and Pagano, multi-colored, hard, and free of cavities and cracks.
2001), Alpe Ravinella in Strona Valley (Bertolani, Many collectors around the world also became inter-
1967), Scortico in the Apuan Alps (Di Sabatino, 1967; ested in this area, where several very rare and attrac-
Mancini, 1997), and Campiglia Marittima in Tuscany tive minerals can be found (gasparite, khristovite,
(Capitani et al., 2003). Italian rhodonites are translu- franklinfilite, tiragalloite, and manganberzeliite; Roth
cent or opaque and microcrystalline with a pale to and Meisser, 2011).
deep pink color; only those from Monte Forno in Val The source region of these rhodonite-bearing
Malenco (Diella et al., 2014) and Scortico (Apuan boulders has never been identified, and its mineralog-
Alps) can be truly considered gem-quality since they ical composition is still the subject of investigation.
are transparent enough to be faceted. It is believed that the manganese-rich deposits
The manganese deposits in the Rhetic Alps of formed during the Tertiary, under zeolite and green-
Switzerland that contain rhodonite, such as Falotta schists facies metamorphic conditions (Roth and
and Alpe Parsettens in the Oberhalbstein or Fianel Meisser, 2011). It was assumed that the mineraliza-
and Starlera in Val Ferrera, have been widely investi- tion took place in the Triassic carbonate sediments,
gated (Trommsdorff et al., 1970; Peters et al., 1973, similar to the famous deposits of Fianel and Starlera
1978; Wenk and Maurizio, 1978; Brugger and Giere, in nearby Val Ferrera (Brugger and Giere, 1999), but

112 SWISS RHODONITE FROM TANATZ ALP GEMS & GEMOLOGY SPRING 2020
 Innerferrera
Innerferrera

P. Suretta
P. Suretta

Splügen Pass
Spluga Pass

P.
P. Tambò
Tambò

Montespluga
Lake
lake

P. Emet
Emet
Chur 0 1 km

SWITZERLAND

P.Bernina
Bormio
Chiavenna
Calcschist Marbles and quarzites Amphibolites Orthogneiss Roffna porphyroids Garnet micaschists
Bellinzona ITALY Tirano and paragneiss
Sondrio 0 30 km Fault Village Mountain

Figure 2. Geological map of the Tanatz Alp area in Switzerland. Modified from Bedogné et al. (1995).

the original outcrop is likely located in the crys- The mineralogy of the rhodonite-bearing rocks
talline intercalation of the Splügen area, which is tec- from Tanatz Alp shows similarities with that of de-
tonically very complex (figure 2). posits from Scerscen in Val Malenco, located about

Figure 3. A loose assort-

ment of blocks contain-
ing the rhodonite
mineralizations investi-
gated in this work. The
largest is about 1 m
long, while the overly-
ing blocks are about
20–25 cm long. Photo
by F. Vanini.

SWISS RHODONITE FROM TANATZ ALP GEMS & GEMOLOGY SPRING 2020 113
 Sample 1 Sample 2

1 cm 1 cm

Sample 3 Sample 4

Figure 4. Rough sam-

ples of rhodonite from
Tanatz Alp investigated
in this work (samples
1–6). Photos by Maur-
izio Scacchetti.

1 cm 1 cm

Sample 5 Sample 6

1 cm 1 cm

40 km away (Bedogné et al., 1993, 1995; Diella et al., and brown or blackish portions made up of braunite
2014). On the surface, the manganese boulders are and manganese oxides such as rancieite. All the sam-
black. The alteration layer varies from a few millime- ples were compact.
ters to one centimeter and is composed of rancieite. Standard gemological analyses were carried out
Rhodonite is present in small, needle-like crystals on seven fashioned stones (figure 5), cut from the
with a characteristic and very pleasing pink color, rough samples in figure 4, to describe the optical
often in contact with spessartine. properties, specific gravity, and ultraviolet fluores-
cence. Density was measured using a Presidium
MATERIALS AND METHODS PCS100 Sensible hydrostatic balance, and the color
Analyses were carried out on six samples of rock con- was evaluated with an RGB (red, green, blue) color
taining rhodonite (figure 4), collected from Maurizio table. Refractive index was measured by the distant
Scacchetti in the summer of 2017 along the main vision method using a Kruss refractometer (1.45–1.80
road leading to the Splügen Pass. The samples were range) and a contact liquid with an RI of 1.80. Ultra-
in centimeter sizes, with very uneven color due to violet fluorescence was investigated with a short-
the presence of different minerals. The rhodonite was wave (254 nm) and long-wave (365 nm) UV lamp.
present in these samples as fine needle-like crystals. X-ray powder analyses to determine the miner-
There were also millimeter-sized and rounded orange alogical composition were carried out using a Philips
crystals of spessartine and light pink rhodochrosite PW1800 powder diffractometer, with CuKa radiation

114 SWISS RHODONITE FROM TANATZ ALP GEMS & GEMOLOGY SPRING 2020
 Sample 1 Sample 2 Sample 3 Sample 4

Sample 5 Sample 6 Sample 7

Figure 5. Cut gems of

rhodonite from Tanatz
Alp, ranging from 0.43 to
12.70 ct. Photos by Fed-
erica Montanelli.

(λ = 1.5418 Å) and a scan speed of 1°/min, in the 2– silicon Raman peak at 520.5 cm–1. The spectral reso-
65° 2θ range. The samples for the analyses (about 3 g) lution was ~2 cm–1, and the instrumental accuracy
were previously ground in an agate mortar and re- in determining the peak positions was approximately
duced to a very fine powder (about 5 microns). We 0.56 cm–1. Raman spectra were collected in the spec-
ground the rough samples shown in figure 4. The ma- tral range 100–1300 cm–1 for 5 seconds averaging over
terial was selected based on the pink color before the 40 scans accumulated.
grinding. Unfortunately, many phases of rhodonite-
bearing rock show a similar color (spessartine orange- RESULTS
pink, rhodochrosite reddish pink, and kutnohorite Microscopic Observations. Figure 6A shows a macro-
pale pink). Therefore, in the Results section rhodonite scopic section of a rhodonite sample (no. 2) that
represents less than 50% of all samples except sample contains a pink and a dark part. Thin-section investi-
2. Qualitative and semi-quantitative evaluation was gation revealed that the pink portion (figure 6B) con-
performed with X’Pert HighScore software, which is sists of an intimate association of rhodonite (or
designed to obtain all phase information from loose pyroxmangite) and kutnohorite (see chemical analyses
and pressed powders and other polycrystalline sam- in table 5). Here, the size of rhodonite/pyroxmangite
ples (Markvardsen et al., 2008; Degen et al., 2014). crystals ranges from tens of micrometers to a few mil-
Major element analysis (Si, Mn, Ca, Mg, Fe, Zn, limeters. Conversely, the dark part is composed
Ti, and Al) was carried out on select crystals from all mainly of spessartine and Mn oxides (figure 6C).
the rhodonite samples with an electron microscope,
model TESCAN series Mira XMU, coupled with an Gemological and Physical Properties. Seven cabo-
EDAX system with energy dispersion and no internal chons and faceted stones were cut with various
standard (analytical error of about 3%). The acceler- shapes (figure 5) from the six rough samples; table 1
ating voltage was 20 KV, while the analysis area was presents the gemological data.
100 × 100 μ2. The accuracy of the analyses is related The cabochons and faceted stones had a massive
to the calibration of the detector and the maintaining and opaque aspect with an intense and variable pink
of standard conditions during the work. color (pink, pale violet red, and Indian red according
Micro-Raman scattering measurements were con- to the RGB color table). The rhodonites were strongly
ducted with a Horiba Jobin Yvon Explora Plus single inhomogeneous due to the presence of various asso-
monochromator spectrometer (grating of 2400 ciated minerals in spots that were dark to black and
groove/mm) equipped with an Olympus BX41 micro- yellow-orange. The measured specific gravity ranged
scope. Raman spectra were recorded with 532 nm ex- from 3.40 to 3.63, in agreement with data for other
citation. The spectrometer was calibrated to the alpine rhodonites (Deer et al., 1997; Diella et al.,

SWISS RHODONITE FROM TANATZ ALP GEMS & GEMOLOGY SPRING 2020 115
 A B C

Kutnohorite

c Spessartine
Rhodonite/Pyroxmangite

0.5 cm 200 μm 100 μm

Figure 6. A: In this slice obtained from rough sample 2, red squares indicate the points of microscopic observation
shown in figures 4B and 4C. B: A crossed polarizer image reveals the coexistence of kutnohorite and rhodonite/
pyroxmangite. C: A plane-polarized light image shows the presence of Mn-rich garnet. Photos by Omar Bartoli.

2014). The refractive indices measured with the dis- cause of the two polymorphs’ overlapping peaks at
tant vision method ranged from approximately 1.72 26.6, 29.1, 30.2, 34.4, and 36.5° 2θ. The rhodonite was
to 1.74 (in agreement with Deer et al., 1997; Diella et identified by the characteristic peaks at 21.5, 25.0,
al., 2014). The absorption spectra were characterized 27.4, 28.5, 30.5, 32.2, and 35.4° 2θ, while the sole py-
by a broad band at 550 nm, a strong narrow line at roxmangite presented diffraction effects at 25.5, 28.1,
500 nm, and a diffuse weak band at 450 nm. All the 29.8, 31.3, 40.8, and 44.3° 2θ.
samples were inert to short-wave (254 nm) and long- The samples showed variable amounts of
wave (365 nm) UV. rhodonite-pyroxmangite (22–66%) and kutnohorite
(8–62%); samples 1, 4, and 6 also contained ankerite.
XRPD. Figure 7 shows the X-ray patterns of some in- Other detected minerals were spessartine,
vestigated samples, while the mineralogical compo- rhodochrosite, calcite, braunite, chlorite, and traces of
sition is reported in table 2. The samples showed a quartz. In samples 1, 2, 4, and 6, the X’Pert software
variable occurrence of rhodonite-pyroxmangite, but also suggested traces of the rare mineral khristovite,
distinguishing the two phases is quite difficult be- identified by the peaks at 25.3, 30.6, 32.8, and 34° 2θ.

TABLE 1. Gemological data of the rhodonite cabochons and faceted stones investigated.

Color Dimensions Weight Specific

Sample Cut RIa
(RGB color table) (mm) (ct) gravity
1 Pink; black spots 7×4×3 Triangular cabochon 0.43 3.40 1.72
2 Pink/pink1; black spots 12 × 7 × 6 Oval cabochon 5.30 3.41 1.73
3 Pale violet/”Indian red 7×5×5 Oval cabochon 1.08 3.42 1.73
2”; pink veins
4 Pale violet/“Indian red 10 × 12 × 6 Rectangular cabochon 12.01 3.60 1.74
2” veins
5 Pale violet/“Indian red 12 × 13 × 5 Oval cabochon 12.51 3.56 1.74
2” veins; black spots
6 Pale violet/“Indian red 13 × 10 × 7 Rectangular cabochon 12.70 3.58 1.74
2”; black spots
7 “Indian red”; black 12 × 6 × 4 Rectangular emerald-cut 3.70 3.63 1.74
spots
a
Determined using the distant vision method

116 SWISS RHODONITE FROM TANATZ ALP GEMS & GEMOLOGY SPRING 2020
 XRPD SPECTRA

Sample 1
3600

Ank

1600

Ank
Rh Sp
Rh Ank
Pyr
400 Pyr
Rh Rhc Pyr
Rh
Chl Chl Kr
Rhc

0
10 20 30 40 50 60

3600 Sample 2

Figure 7. XRPD patterns of

the rhodonite samples 1, 2,
Ank + Rh
and 4 from figure 4. The
1600 most important minerals
COUNTS

are reported, close to their

Kut
diagnostic peaks. Ank =
Chl Rh ankerite; Pyr = pyroxman-
Chl Kut
Chl Kut RhPyr Sp gite; Rh = rhodonite; Kut =
Ank
400 Kr Pyr kutnohorite; Sp= spessar-
Kr Kut tine; Rhc = rhodochrosite;
Chl = chlorite.

0
10 20 30 40 50 60

Sample 4
1600
Ank
Rh

Kut
Chl

900
Kut
Rh
Chl Pyr
Chl
Chl
400 Pyr
Sp Ank Kut
Kut Rh Kr Ank
Kr Sp
Kut

100

0
10 20 30 40 50 60
POSITION (°2Θ)

SWISS RHODONITE FROM TANATZ ALP GEMS & GEMOLOGY SPRING 2020 117
TABLE 2. Mineralogical composition of the analyzed samples (from figure 4) obtained
through XRPD analysis.
Sample no. Mineral assemblages (wt.%)
1 Ankerite (60%), rhodonite-pyroxmangite (20%), rhodochrosite (20%), khristovite (traces),
chlorite (traces)
2 Rhodonite-pyroxmangite (65%), kutnohorite (25%), spessartine (5%), chlorite (5%),
quartz (traces), khristovite (traces)
3 Kutnohorite (60%), rhodonite-pyroxmangite (30%), calcite (10%)
4 Rhodonite-pyroxmangite (35%), spessartine (30%), ankerite (15%), kutnohorite (10%),
chlorite (10%), khristovite (traces)
5 Rhodonite-pyroxmangite (45%), braunite (20%), ankerite (10%), kutnohorite (10%),
spessartine (10%), quartz (5%)
6 Rhodonite-pyroxmangite (40%), spessartine (30%), kutnohorite (20%), ankerite (10%),
khristovite (traces)

Mineral Chemistry. SEM analyses performed on se- ric (Si = 1.00–1.01 apfu) while that of Mn deviated
lected crystals of rhodonite-pyroxmangite are shown (0.82–0.88 apfu) from the idealized formula (1.00
in table 3 (error of analyses about 3%). Table 4 reports apfu), due to the occurrence of other cations such as
the chemical composition of rhodonites from the Ca and Mg in the M-sites.
Alps and other worldwide occurrences reported in the The Ca and Mn contents in our samples were
literature, for comparison. In our samples the most similar to those reported for other samples from Val
abundant oxides were MnO (46–48.64 wt.%) and SiO2 Scerscen, Val Sesia Val d’Ayas, and Val Malenco (Pe-
(46.70–47.92 wt.%), with very low amounts of MgO ters et al., 1978; Mottana, 1986; Diella et al., 2014;
(0.64–1.65 wt.%) and CaO (<5 wt.%). see table 4).
Figure 8 and table 5 provide the backscattered CaO contents also showed large variability in the
electron images and corresponding analyses of the as- same areas investigated by energy-dispersive spec-
sociation of rhodonite (points 3, 4, and 5) with Ca- troscopy (EDS) analyses. It varied from 0.97% to
Mn-carbonate kutnohorite (CaMn2+(CO3)2; points 1 3.90% in samples 1 and 2, from 0.99% to 3.97% in
and 2). We also observed the presence of Mn oxides sample 4, and from 1.02% to 4.64% in sample 5; see
and spessartine as accessory phases. The rhodonite table 3. The values of Ca vs. Mn are plotted in figure
in the samples appeared to be composed of micro- 9 with the different fields for rhodonite and pyrox-
crystals, and the chemical composition was not ho- mangite (see also Diella et al., 2014).
mogeneous. Similar to what was reported for Fe was absent in our samples, as in rhodonite
rhodonites from Val Malenco (Diella et al., 2014), we from other localities in the Alps such as Praborna and
observed that the Si content was almost stoichiomet- Val Sesia and from Japan; see table 4. In the

TABLE 3. Chemical composition (wt.%) of the rhodonite/pyroxmangite samples investigated, obtained through SEM analyses.

Oxides 1/1a 1/2 2/1 2/2 3/1 3/2 4/1 4/2 5/1 5/2 6/1 6/2
SiO2 46.91 46.62 47.02 46.63 46.70 47.03 47.43 46.92 47.66 46.53 47.62 47.54
MgO 0.67 1.44 0.70 1.45 0.64 0.68 0.57 1.65 0.75 1.46 0.84 0.85
MnO 48.52 50.75 48.42 50.95 48.64 48.49 48.33 50.44 46.95 50.98 46.85 47.41
CaO 3.90 0.98 3.86 0.97 4.01 3.80 3.67 0.99 4.64 1.02 4.70 4.20
Total 100 99.99 100 100 99.99 100 100 100 100 99.99 100.01 100
apfu (3 oxygens)
Si 1.00 1.00 1.01 1.00 1.00 1.01 1.01 1.01 1.01 1.00 1.01 1.01
Mg 0.02 0.05 0.02 0.05 0.02 0.02 0.02 0.06 0.02 0.05 0.03 0.03
Mn 0.88 0.93 0.88 0.93 0.88 0.88 0.87 0.90 0.85 0.93 0.84 0.85
Ca 0.09 0.02 0.09 0.02 0.09 0.09 0.08 0.02 0.11 0.02 0.11 0.10
a
Sample/crystal
Fe2O3 was below detection limits (0.1 wt.%).

118 SWISS RHODONITE FROM TANATZ ALP GEMS & GEMOLOGY SPRING 2020
 2 Ca VS. Mn
0.20

0.16

Rhodonites

Ca (apfu)
3 0.12

0.08

1 Pyroxmangites
4 0.04

0
5 0.80 0.82 0.84 0.86 0.88 0.90 0.92 0.94 0.96 0.98 1.0

Mn (apfu)

Figure 9. Compositional variation of Ca vs. Mn of co-

existing rhodonite-pyroxmangite from Tanatz Alp in
Switzerland.
Figure 8. Backscattered electron image of rhodonite/
pyroxmangite (points 3, 4, and 5) and associated kut-
nohorite (points 1 and 2) in rough sample 2. Micro-Raman Spectroscopy. Mineral polymorphs are
distinguishable by Raman spectroscopy (Mazzoleni
rhodonites from the famous deposits of Broken Hill et al., 2015), and this analysis was carried out to con-
(Australia) and Franklin (New Jersey), Fe is an impor- firm the presence of the polymorph pyroxmangite
tant chromophore element; again, see table 4. with rhodonite.

TABLE 4. Chemical composition (wt.%) of rhodonite samples reported in the literature: Italy (Peters et al., 1978; Mottana,
1986; Diella et al., 2014); United States and Australia (Nelson and Griffin, 2005); and Japan (Aikawa, 1984).
United United
States States United United
Italy Italy Italy Italy Australia Australia Australia
Italy (Franklin (Franklin States States Japan Japan
Oxides (Val (Val (Val (Monte (Broken (Broken (Broken
(Prabornaa) Mine 1, Mine 2, (New (Plainfield, (Sankei) (Sankei)
Malenco 1) Malenco 2) Sesiaa) Alpea) Hill) Hill) Hill)
New New York) Maine)
Jersey) Jersey)
SiO2 46.82 46.76 46.40 46.91 46.70 46.35 46.46 45.71 45.69 46.15 45.94 45.83 46.88 47.34
TiO2 n.d. n.d. 0.07 n.d. n.d. 0.04 n.d. 0.01 0.01 n.d. 0.04 0.02 n.d. n.d.
Al2O3 n.d. 0.02 n.d. n.d. n.d. 0.01 0.02 0.01 0.01 0.93 0.03 0.04 0.07 0.59
MgO n.d. 0.66 0.65 0.95 1.71 1.54 0.88 0.13 0.08 1.41 0.3 0.25 1.50 0.87
FeO n.d. 3.27 3.20 n.d. 5.21 2.79 1.91 12.34 1.93 10.51 11.76 11.92 0.45 0.31
MnO 47.07 41.80 41.68 46.74 40.0 35.71 38.04 40.21 48.13 38.62 37.35 37.91 50.69 44.18
ZnO n.d. n.d. n.d. n.d. n.d. 6.96 5.09 0.16 0.03 n.d. 0.33 0.35 n.d. n.d.
CaO 6.11 6.41 6.79 5.40 6.38 6.59 7.61 1.73 3.65 2.01 4.32 3.94 1.37 7.11
Total 100.00 100.02 99.99 100.00 100.00 99.99 100.01 100.30 99.53 99.63 100.07 100.26 100.06 100.41
Cations based on 3 oxygens apfu
Si 1.00 1.006 1.000 1.00 0.99 0.994 0.996 0.994 0.994 0.992 0.993 0.991 0.998 0.998
Ti n.d. n.d. 0.001 n.d. n.d. 0.001 n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d.
Al n.d. n.d. n.d. n.d. n.d. n.d. 0.001 n.d. n.d. 0.024 0.001 0.001 0.002 0.015
Mg n.d. 0.021 0.020 0.03 0.05 0.049 0.028 0.004 0.003 0.04 0.010 0.008 0.048 0.027
Fe n.d. 0.057 0.058 n.d. 0.09 0.050 0.034 0.224 0.035 0.188 0.213 0.216 0.008 0.031
Mn 0.85 0.761 0.761 0.85 0.72 0.649 0.690 0.740 0.887 0.702 0.684 0.694 0.914 0.806
Zn n.d. n.d. n.d. n.d. n.d. n.d. 0.081 0.003 n.d. n.d. 0.005 0.006 n.d. n.d.
Ca 0.12 0.148 0.157 0.12 0.15 0.151 0.175 0.034 0.072 0.040 0.099 0.090 0.032 0.164
a
EDS analyses
n.d. = not detected

SWISS RHODONITE FROM TANATZ ALP GEMS & GEMOLOGY SPRING 2020 119
TABLE 5. EDS analyses of the five backscattered electron image spots from figure 8.

3 4 5
Oxides 1 2 (rhodonite/ (rhodonite/ (rhodonite/
(wt.%) (kutnohorite) (kutnohorite) pyroxmangite) pyroxmangite) pyroxmangite)
MgO 1.79 1.81 1.37 1.28 1.53
SiO2 — — 47.87 47.83 47.92
CaO 26.03 25.31 3.79 3.89 4.58
MnO 30.73 31.50 46.97 47.01 45.97
CO2calc 41.45 41.38 — — —
Total 100 100 99.99 100.01 100
apfu (6 oxygens) apfu (3 oxygens)
Mg 0.09 0.10 0.04 0.04 0.05
Si — — 1.01 1.01 1.01
Ca 0.99 0.96 0.09 0.09 0.10
Mn 0.92 0.94 0.85 0.86 0.82
C 2 2 — — —

The Raman spectrum of the sample shown in fig- The Raman spectra of pyroxmangite (from the
ure 6A (area b) is presented in figure 10. The spectrum RRUFF database, ID R070212 at rruff.info) present
is characterized by an intense band at 1000 cm–1 as- many bands that overlap with those of rhodonite (fig-
signed to the 1 symmetric stretching mode and two ure 10). However, the small bands at 185, 252, and
bands at 972 and 944 cm–1 assigned to the ν3 asymmet- 299 cm–1 are characteristic of pyroxmangite and con-
ric stretching modes. The band at 670 cm–1 has been firm the presence of this polymorph in our samples.
assigned to the ν4 Si-O-Si bending mode (Mills et al.,
2005). The low wavenumber region of rhodonite is DISCUSSION AND CONCLUDING REMARKS
complex (Mills et al., 2005). There are two small bands The present study, performed on both cut and rough
at 399 and 424 cm–1 assigned to the ν2 bending mode. samples of rhodonite-rich rocks from Tanatz Alp,
The spectra reported in literature are very similar to provided new data to better understand the charac-
ours; the small differences observed can be attributed teristics of this attractive gemstone.
to cationic substitution of Mn by Ca, Fe, and Mg The color of the investigated rhodonite rocks var-
(Mills et al., 2005; Diella et al., 2014). ied from pale pink to purplish pink. The mineralogi-

RAMAN SPECTRUM
12000

10000 1000
INTENSITY (a.u.)

8000 466 1170

972
Figure 10. Raman spec-
676 1053
6000 trum of the rhodonite
from figure 6A (area b).
129 944
4000 870
218 424
399
2000 185 252
299

0
0 200 400 600 800 1000 1200 1400

RAMAN SHIFT (cm–1)

120 SWISS RHODONITE FROM TANATZ ALP GEMS & GEMOLOGY SPRING 2020
cal composition is an assemblage of rhodonite and py- Hill (up to about 12%; Millsteed et al., 2005; Mill-
roxmangite and other Mn-rich minerals such as steed, 2006), Franklin (up to about 2.8%; Nelson and
kutnohorite, spessartine, ankerite, and rhodochrosite. Griffen, 2005), and Val Malenco (up to 3.27%; Diella
Low amounts of chlorite, calcite, and quartz were also et al., 2014). Together with Mn and Ca, Fe substitutes
detected. We also identified the presence, as an acces- for Mg in the octahedral sites of the mineral’s struc-
sory phase, of khristovite, a Ce-Fe-Mn2+-rich epidote ture. The presence of Fe modifies the color from pink
group mineral. This finding is very interesting owing to deep pinkish red, as reported for rhodonites from
to the rarity of this mineral; the presence of khris- Broken Hill and the Franklin mine, which are Fe-rich
tovite in this locality was also reported by Roth and and have a deep pinkish red color (Millsteed, 2005;
Meisser (2011). Nelson and Griffen, 2005). Since Fe is absent in our
The chemical composition of our rhodonite sam- samples, the color is softer; the Praborna material,
ples highlights a negative correlation between Ca and which also lacks Fe, is similar in color to our samples
Mn, which was previously observed in rhodonites from Tanatz Alp.
from Tanatz Alp and the Rhetic Alps (Peters et al., The Ca content in our rhodonites ranged be-
1978; Diella et al., 2014). The amount of Mg ions sub- tween 0.97% and 4.70 wt.%, lower than that of
stituting for Mn in the samples we analyzed showed other rhodonites from the Alps (table 4). The vari-
large variations that reflect the type of carbonate min- ability of Ca content in deposits worldwide can be
eral present, the bulk composition of the mother rock related to the coexistence of pyroxmangite and
and the availability of the element (Diella et al., 2014). rhodonite.
It is interesting to observe that our samples did not XRPD patterns, micro-Raman analyses, and low
show the presence of FeO, which has been reported Ca contents indicate that the pyroxmangite should
in rhodonites from other occurrences such as Broken coexist with rhodonite in the Mn-rich rocks from

Figure 11. Handmade

jewels produced with
rhodonite from Tanatz
Alp by Lutezia Gioielli
(Stradella, Italy). Left:
necklace of freeform
rhodonite beads. The
knot in the center of
the necklace contains a
pink rhodonite cabo-
chon. Right: Vintage
pendant with a drop-
shaped rhodonite.
Photos by Federica
Montanelli.

SWISS RHODONITE FROM TANATZ ALP GEMS & GEMOLOGY SPRING 2020 121
Tanatz Alp. The pyroxmangite crystals probably in- derstanding the genesis of the manganese mineral-
tergrow with rhodonite crystals, but the two phases ization of Tanatz Alp.
are visually undetectable. The intergrowth of the two Regarding gemological properties, the rhodonite-
phases is also confirmed by the large variability of Ca bearing rocks investigated in this study were compact
in the same areas of EDS analyses, ranging from 0.02 with no fractures. They also showed a pink color,
to 0.10 apfu. Both polymorphs usually form through though it was not homogeneous and contained orange
processes of regional or contact metamorphism, as and purple tints due to spessartine and rhodochrosite,
well as metasomatic processes from carbonate rocks respectively. Since these rhodonites are translucent
that underwent low-grade regional or very high-grade and microcrystalline, their commercial value is far
contact-metamorphic conditions (Abrecht, 1988). lower than that of the transparent gems from Broken
Several studies on the coexistence of pyroxmangite Hill (Millsteed, 2005) or Franklin (Nelson and Griffen,
and rhodonite have been performed on synthetic crys- 2005). Although some Tanatz Alp rhodonites were
tals of MnSiO3; synthetic pyroxmangite is usually sta- carved and traded in the past, their marketing did not
ble at temperatures below about 400°C and pressures begin until the end of the twentieth century. This
< 2 kbar, while rhodonite is stable at higher tempera- charming pink stone with a fine-grained structure is
tures (Abrecht and Peters, 1975; Maresch and Mot- now available for pendants, necklaces, brooches, and
tana, 1976; Pinckney and Burnham,1988).Yet the ornamental objects (figure 11).
application of these studies to natural rocks is quite At present the deposit is only mined occasionally
difficult because of the great variety and the extent of with hand tools and not for commercial purposes.
cation substitutions observed in natural pyroxenoids Despite the difficulty of access, the geological char-
(Ca, Mg, Fe2+→ Mn). acteristics of the locality suggest the possibility of
Furthermore, the inability to identify the original substantial future production if a small commercial
outcropping represents a significant challenge in un- mine were to be established.

ABOUT THE AUTHORS ACKNOWLEDGMENTS

Dr. Caucia is associate professor, Dr. Marinoni is a geologist, and The authors are grateful to Antonio Langone for help during the
Dr. Riccardi is associate professor with the Department of Earth traces analyses, Mattia Gilio for micro-Raman analyses,
Sciences and Environment, University of Pavia, Italy. Dr. Bartoli is Francesco Vanini for supplying the rhodonite samples, Federica
associate professor with the Department of Geosciences, Univer- Montanelli for the gem photography, and Riccardo Fantini and
sity of Padova, Italy. Dr. Scacchetti is a geologist with Reggiana Francesco Leoni for the creating the illustrations.
Society of Natural Sciences, Reggio Emilia, Italy.

REFERENCES
Abrecht J. (1988) Experimental evaluation of the MnCO3 Bedogné F., Maurizio R., Montrasio A., Sciesa E. (1995) I minerali
+SiO2=MnSiO3+CO2 equilibrium at 1 kbar. American Miner- della Provincia di Sondrio e della Bregaglia Grigionese. Val
alogist, Vol. 73, No. 11-12, pp. 1285–1291. Bregaglia, Val Masino, Val Codera, Bettini Editore, Sondrio.
Abrecht J., Peters T. (1975) Hydrothermal synthesis of pyroxenoid Bertolani M. (1967) Rocce manganesifere tra le granuliti della Valle
in the system MnSiO3–CaSiO3 at Pf = 2kb. Contributions to Strona (Novara). Periodico di Mineralogia, Vol. 36, pp. 1011–
Mineralogy and Petrology, Vol. 50, No. 4, pp. 241–246, 1032.
http://dx.doi.org/10.1007/BF00394851 Biggioggero B., Montrasio A. (1990) Val Chiavenna e Valle Spluga.
Aikawa N. (1984) Lamellar structure of rhodonite and pyroxman- Itinerario n. 9. In Guide Geologiche Regionali. 1. Alpi e Prealpi
gite intergrowths. American Mineralogist, Vol. 69, No. 3-4, pp. Lombarde. Società Geologica Italiana, BE-MA Editore Milano.
270–276. Brown P.E., Essene E.J., Peacor D.R. (1980) Phase relations inferred
Akaogy M., Navrotsky A. (1985) Calorimetric study of high-pressure from field data for Mn pyroxenes and pyroxenoids. Contribu-
polymorph of MnSiO3. Physics and Chemistry of Minerals, Vol. tions to Mineralogy and Petrology, Vol. 74, No. 4, pp. 417–425,
12, No. 6, pp. 317–323, http://dx.doi.org/10.1007/BF00654341 http://dx.doi.org/10.1007/BF00518121
Armbruster T., Bonazzi P., Akasaka M., Bermanec V., Chopin C., Brugger J., Giere R. (1999) As, Sb and Ce enrichment in minerals
Gieré R., Heuss-Assbichler S., Liebscher A., Menchetti S., Pan from a metamorphic Fe-Mn deposit, Val Ferrera, Eastern Swiss
Y., Pasero M. (2006) Recommended nomenclature of epidote- Alps. Canadian Mineralogist, Vol. 37, pp. 37–52.
group minerals. European Journal of Mineralogy, Vol. 18, No. Brusnitsyn A.I. (2010) Mineralogy of metamorphosed manganese de-
5, pp. 551–567, http://dx.doi.org/10.1127/0935-1221/2006/ posits of the South Urals. Geology of Ore Deposits, Vol. 52, No.
0018-0551 7, pp. 551–565, http://dx.doi.org/10.1134/S1075701510070044
Bedogné F., Montrasio A., Sciesa E. (1993) I minerali della Provin- Bukanov V.V. (2006) Russian Gemstones Encyclopedia. Granit Ed-
cia di Sondrio. Bettini Editore, Sondrio. itor, Saint Petersburg.

122 SWISS RHODONITE FROM TANATZ ALP GEMS & GEMOLOGY SPRING 2020
Capitani G.C., Grobéty H.B., Mellini M. (2003) Reaction se- Vol. 62, No. 1-3, pp. 171–175, http://dx.doi.org/10.1016/
quences, polysomatic faults and chemical compositions of j.saa.2004.12.022
manganese pyroxenoids from Campiglia Marittima skarn. Eu- Millsteed P.W. (2006) Faceting transparent rhodonite from Broken
ropean Journal of Mineralogy, Vol. 15, No. 2, pp. 381–391, Hill, New South Wales, Australia. G&G, Vol. 42, No. 2, pp.
http://dx.doi.org/10.1127/0935-1221/2003/0015-0381 151–158, http://dx.doi.org/10.5741/GEMS.42.2.151
Deer W.A., Howie R.A., Zussman J. (1997) Rock Forming Miner- Millsteed P.W., Mernagh T.P., Otieno-Alego V., Creagh D.C. (2005)
als, Vol. 2A, 2nd ed. Single Chain Silicates. The Geological So- Inclusions in transparent gem rhodonite from Broken Hill,
ciety, London. New South Wales, Australia. G&G, Vol. 41, No. 3, pp. 246–
Degen T., Sadki M., Bron E., König U., Nénert G. (2014) The High- 254, http://dx.doi.org/10.5741/GEMS.41.3.246
Score suite. Powder Diffraction, Vol. 29, No. S2, pp. S13–S18, Mottana A. (1986) Blueschist-facies metamorphism of manganif-
http://dx.doi.org/10.1017/S0885715614000840 erous cherts: A review of the alpine occurrences. In B.W. Evans
Di Sabatino B. (1967) Su una paragenesi del giacimento mangane- and E.H. Brown, Eds., GSA Memoirs: Blueschists and Eclogites,
sifero di Scortico (Alpi Apuane). Periodico di Mineralogia, Vol. Vol. 164, pp. 267–300, http://dx.doi.org/10.1130/MEM164-p267
36, pp. 965–992. Nelson W.R., Griffen D.T. (2005) Crystal chemistry of Zn-rich
Diella V., Adamo I., Bocchio R. (2014) Gem-quality rhodonite from rhodonite (“fowlerite”). American Mineralogist, Vol. 90, No.
Val Malenco (Central Alps, Italy). Periodico di Mineralogia, 5-6, pp. 969–983, http://dx.doi.org/10.2138/am.2005.1694
Vol. 83, pp. 207–221. Ohashi Y., Kato A., Matsubara S. (1975) Pyroxenoids: A variation
Grenouillet W. (1920) Geologische Untersuchungen am Splügen- in chemistry of natural rhodonite and pyroxmangite. In
pass und Mte di San Bernardino. Jahresbericht der Natur- Carnegie Institution of Washington Year Book 75.
forschenden Gesellschaft Graubundens, N.F. 60, pp. 31–60. Peters T., Schwander H., Trommsdorff V. (1973) Assemblages
Ito J. (1972) Rhodonite-pyroxmangite peritectic along the join among tephroite, pyroxmangite, rhodochrosite, quartz: Exper-
MnSiO3-MgSiO3 in air. American Mineralogist, Vol. 5, pp. 865– imental data and occurrences in the Rhetic Alps. Contributions
876. to Mineralogy and Petrology, Vol. 42, No. 4, pp. 325–332,
Jefferson D.A., Pugh N.J., Alario-Franco M., Mallison L.G., Mill- http://dx.doi.org/10.1007/BF00372610
ward G.R., Thomas J.M. (1980) The ultrastructure of pyrox- Peters T., Trommsdorff V., Sommerauer J. (1978) Manganese py-
enoid chain silicates. I. Variation of the chain configuration in roxenoids and carbonates: Critical phase relations in metamor-
rhodonite. Acta Crystallographica Section A, Vol. 36, No. 6, phic assemblages from the Alps. Contributions to Mineralogy
pp. 1058–1065, http://dx.doi.org/10.1107/S0567739480002136 and Petrology, Vol. 66, No. 4, pp. 383–388, http://dx.doi.org/
Kenngott A. (1866) Die Minerale der Schweiz. Engelmann Editor, 10.1007/BF00403423
Leipzig. Pinckney L.R., Burnham C.W. (1988) Effects of compositional vari-
Lee D.E. (1958) An andradite-spessartine garnet from Pajsberg, ation on the crystal structures of pyroxmangite and rhodonite.
Sweden. American Mineralogist, Vol. 43, pp. 208–215. American Mineralogist, Vol. 73, No. 7-8, pp. 798–808.
Leverett P., Williams P.A., Hibbs D.E. (2008) Ca-Mg-Fe-rich Quinn E.P. (2004) Gem News International: Rhodonite of facet and
rhodonite from the Morro da Mina mine, Comselheiro Lafaiete, cabochon quality from Brazil. G&G, Vol. 40, No. 3, pp. 260–
Minas Gerais, Brazil. Mineralogical Record, Vol. 39, No. 2, pp. 261.
125–130. Roth P., Meisser N. (2011) I minerali dell'Alpe Tanatz. Passo dello
Mancini S. (1997) Le miniere di manganese di Scortico (Fivizzano, Spluga (Grigioni, Svizzera). Rivista Mineralogica Italiana, Vol.
Massa-Carrara). Rivista Mineralogica Italiana, Vol. 3, pp. 295– 17, pp. 90–99.
296. Schiavinato G. (1953) Sulla johannsenite dei giacimenti a silicati
Marchesini M., Pagano R. (2001) The Val Graveglia manganese manganesiferi del Monte Civillina presso Recoaro (Vicenza).
district, Liguria, Italy. Mineralogical Record, Vol. 32, No. 5, pp. Rendiconti Società Mineralogica Italiana, Vol. 9, pp. 210–218.
349–379. Simandl G.J., Paradis S., Nelson J. (2001) Jade and rhodonite de-
Maresch W.V., Mottana A. (1976) The pyroxmangite-rhodonite posits, British Columbia, Canada. In L. Bon, Ed., Proceedings
transformation for the MnSiO3 composition. Contributions to of the 35th Forum on the Geology of Industrial Minerals: The
Mineralogy and Petrology, Vol. 55, No. 1, pp. 69–79, Intermountain West Forum, 1999. Utah Geological Survey,
http://dx.doi.org/10.1007/BF00372755 Salt Lake City.
Markvardsen A.J., Shankland K., David W.I.F., Johnston J.C., Ibber- Takahashi K., Hariya Y. (1995) Phase relations of pyroxenoid in
son R.M., Tucker M., Nowell H., Griffin T. (2008) ExtSym: a the system MnSiO3 – MgSiO3 – CaSiO3. In J. Pašava et al., Eds.,
program to aid space-group determination from powder diffrac- Mineral Deposits: From Their Origin to Their Environmental
tion data. Journal of Applied Crystallography, Vol. 41, No. 6, Impacts. CRC Press, Prague, pp. 1011–1014.
pp. 1177–1181, http://dx.doi.org/10.1107/S0021889808031087 Trommsdorff V., Schwander H., Peters T. (1970) Mangasilikate der
Mazzoleni P., Barone G., Aquilia E., Bersani D., Cirrincione R., alpine metamorphose in Radiolariten des Julier-Bernina-Gabi-
Raneri S. (2015) Application of micro-Raman spectroscopy for etes. Schweizerische Mineralische und Petrographische Mit-
the identification of unclassified minerals preserved in old mu- teilungen, Vol. 50 pp. 539–545.
seum collections. Plinius, Vol. 42, pp. 112–120, http://dx.doi.org/ Wenk H.R., Maurizio R. (1978) Kutnahorite, a rare Mn mineral
10.19276/plinius.2016.02015 from Piz Cam (Bergell Alps). Schweizerische Mineralische und
Michailidis K., Sofianska E. (2010) Spheroidal and radiating aggre- Petrographische Mitteilungen, Vol. 58, pp. 97–100.
gates of Mn-pyroxenoids in the Olympias carbonate-hosted Wilson W.E. (1989) What’s new in minerals: Springfield Show
polymetallic sulphide ore deposit, E. Chalkidiki peninsula, N. 1989. Mineralogical Record, Vol. 20, No. 1, p. 481.
Greece. Review of the Bulgarian Geological Society, Vol. 71, Zanazzi P.F., Nestola F., Nazzareni S., Comodi P. (2008) Pyroxman-
pp. 59–68. gite: A high-pressure single crystal study. American Mineralo-
Mills S.J., Frost R.L., Kloprogge J.T., Weier M.L. (2005) Raman gist, Vol. 93, No. 11-12, pp. 1921–1928, http://dx.doi.org/
spectroscopy of the mineral rhodonite. Spectrochimica Acta, 10.2138/am.2008.2963

SWISS RHODONITE FROM TANATZ ALP GEMS & GEMOLOGY SPRING 2020 123
 The following 25 questions are from the four 2019 issues of G&G. Please refer to the articles in those issues to
find the single best answer for each question.
Mark your choice on the response card provided in this issue or visit gia.edu/gems-gemology to take the
Challenge online. Entries must be received no later than Friday, October 9, 2020. All entries will be acknowledged
with an e-mail, so please remember to include your name and e-mail address (and write clearly).
Score 75% or better, and you will receive a certificate of completion (PDF file). Earn a perfect score, and your
name also will be listed in the Fall 2020 issue of Gems & Gemology.

1. Platelet inclusions are generally C. owe their “hazy” appearance mary deposits.
diagnostic of rubies from to strong blue fluorescence. B. Mugloto-type rubies have
A. Cambodia. D. exhibit unusual light scattering stronger fluorescence.
B. Madagascar. from inclusions or dislocations. C. Maninge Nice–type rubies
C. Afghanistan. have a slight orangy tint.
4. Coarse silk, both long and short, D. Mugloto-type rubies tend to be
D. Mozambique.
is seen in sapphires from flatter and hexagonally
2. The perceived color of faceted A. Thailand. shaped.
alexandrite B. Ethiopia.
A. shows less noticeable differ- C. Australia. 7. With increasing iron content in
ence with orientation for step- D. All of the above nephrite,
cut stones. A. the refractive index increases.
B. is essentially the same in all 5. Which material is not found in B. the specific gravity increases.
orientations with brilliant-cut Ethiopia? C. the color deepens.
stones. A. “Zebra sapphire” D. All of the above
C. is strongly dependent upon B. Star sapphire
C. Garnet 8. Which of these statements
orientation regardless of cut.
D. Gold regarding Malipo emeralds is
D. is more greenish in incandes-
true?
cent light.
6. Mugloto-type and Maninge Nice- A. They have extremely high Cr3+
3. Fancy white diamonds type rubies from Mozambique levels.
A. always have a “D” color grade. differ in that B. They display distinct dichroism.
B. always show blue fluorescence A. Maninge Nice–type rubies C. They display high UV fluores-
with long-wave UV light. tend to be associated with pri- cence.

124 G&G CHALLENGE GEMS & GEMOLOGY SPRING 2020

 D. They display no color zoning. 15. What variety of amber is most the United States
prized in the Chinese market? A. contain higher levels of cal-
9. The center gemstone of the
A. Chicken-fat yellow beeswax cium (Ca) on average than Chi-
Talisman of Charlemagne was
B. Roasted old beeswax nese freshwater cultured pearls.
once described as
C. Golden amber B. contain higher levels of man-
A. surrounded only by pearls.
D. Transparent ganese (Mn) on average than
B. originating from Egypt.
natural freshwater pearls from
C. being set in silver. 16. What do natural rhodochrosite
the United States.
D. emerald. and imitation rhodochrosite made
C. are indistinguishable from Chi-
from pressed gibbsite and calcite
10. Which process starts with an nese freshwater cultured pearls.
have in common?
artist drawing out a pattern with D. contain unusually high levels
A. Within each, the composition of lead (Pb).
henna dye on a marble surface?
and properties of their red and
A. Parchin kari 22. Chalky blue fluorescence was
white areas are very similar
B. Pietra dura observed in Madagascar
B. Both contain calcium
C. Intarsia sapphires
C. Refractive index
D. Inlay
D. Luster A. prior to any heat treatment.
11. In general, a freshwater non-bead B. only if they contained yellow
17. Pearls originating from bivalves of
cultured (NBC) pearl would exhibit colored regions.
the Tridacnidae family
A. lower manganese (Mn) levels C. with short-wave UV light after
A. likely rotate during growth. heating to 1300°C.
than saltwater pearls.
B. always display a visible flame D. with long-wave UV light after
B. lower strontium (Sr) levels than
structure. heating to 1000°C.
saltwater pearls.
C. rarely have a white bodycolor.
C. no reaction to optical X-ray flu- 23. What difference is seen between
D. are always nacreous.
orescence. pearls produced by small and
D. pronounced voids. 18. Which mineral inclusion is most large Pteria penguin mollusks?
often found in ruby and purple
12. Which location produces A. Pink hue for pearls from small
sapphire from the Bo Welu mine
emeralds with few inclusions? mollusks
in Chanthaburi, Thailand?
A. Afghanistan B. Rate of nacre deposition
A. High-alumina diopside C. Pearl quality
B. Ethiopia
B. Monazite D. Pearl shape
C. Russia
C. Sapphirine
D. Brazil 24. Alexandrite from Brazil
D. Pyrope garnet
13. Dating zircon inclusions in A. demonstrates a more distinct
19. What is one reason emerald
sapphire color change than material
occurrences are rare?
A. always gives an accurate age of from Russia.
A. They can only form in one type B. may show chatoyancy, unlike
the sapphire host.
of geological environment. alexandrite from all other
B. is straightforward, as there is
B. All emerald crystals are more countries.
no zoning in zircon.
than nine million years old. C. can have strong color change
C. is done using Raman spec-
C. Aluminum is not very preva- with good clarity and size.
troscopy.
lent in the continental crust. D. is a byproduct of emerald
D. None of the above
D. Beryllium is very rare in the mining.
14. Laser ablation–inductively upper continental crust.
coupled plasma–mass spectro- 25. Graphitization in diamond
20. Which of the following methods is
metry (LA-ICP-MS) analysis of A. is rarely observed in cut dia-
most accurate in positively identi-
tourmaline monds with no treatment.
fying an inclusion species in a
A. is more accurate and precise B. can indicate HPHT treatment.
gem?
than electron microprobe C. can happen naturally.
A. Energy dispersive X-ray fluores- D. All of the above
analysis (EMPA).
cence spectroscopy
B. can separate hydroxy- from
B. UV-Vis-NIR spectroscopy
oxy-tourmaline species.
C. Raman spectroscopy To take the Challenge online
C. can separate liddicoatite from
D. Optical microscopy
dravite. please scan the QR code to
D. reliably measures lithium (Li). 21. Freshwater cultured pearls from the right.

G&G CHALLENGE GEMS & GEMOLOGY SPRING 2020 125

 Lab Notes
Editors
Thomas M. Moses | Shane F. McClure

DIAMOND as NV–, NV0, and H4 (a B aggregate

Diamond with Cavities Showing with an additional vacancy) within
Radiation Evidence these cavities (figures 3 and 4). Al-
though this diamond was type IIa,
The Carlsbad laboratory recently ex-
there was sufficient nitrogen to form
amined a 0.70 ct, E-color round bril-
related defects that could be observed
liant. Infrared spectroscopy showed
at the lower level of detection of PL
this to be a type IIa diamond, so we
spectra.
performed a variety of additional
A number of other peaks often
spectroscopy and imaging to verify its
Figure 1. This 0.70 ct natural di- seen in natural diamonds but not yet
natural origin. This diamond also had
amond contained an inclusion identified, such as those at 490.7, 498,
I1 clarity due to a large inclusion
under the table (figure 1). Raman of wollastonite and CaSiO3-
analysis of the inclusions verified that breyite with a large disk-like Figure 2. This DiamondView flu-
this crystal was a metastable compos- graphitic fracture due to exhu- orescence image of the table and
ite of the minerals wollastonite mation from the mantle. Field of crown facets shows green fluores-
(CaSiO3) and CaSiO3-breyite (E.M. view 1.15 mm. cence halos around two cavity
Smith et al., “The very deep origin of features (as indicated by the red
the world’s biggest diamonds,” Win- arrows). The green fluorescence,
ter 2017 G&G, pp. 388–403), which 360–750 km before being transported caused by the H4 defect, was
indicates a sublithospheric origin. to near the surface. caused by localized radiation
These minerals are believed to be the But even after such a tumultuous within those cavities while the
lower-pressure phases of CaSiO3-per- journey, the earth was not finished diamond resided within the
ovskite. Around these minerals were writing the story of this diamond. earth’s crust and was exposed to
large disk-like graphitic fractures in- While in the earth’s crust, the dia- radioactive fluids.
dicating inclusion expansion as pres- mond was exposed to radioactive flu-
sures on the diamond reduced during ids, particularly in etch channels now
exhumation from the mantle. The in the form of cavities on the table
other inclusion present was unidenti- and crown facets. DiamondView im-
fiable due to its graphitic casing. Re- aging revealed fluorescing green halos
cent research of inclusions in other around these cavities (figure 2). Al-
type II diamonds shows that many, if though these isolated areas of fluores-
not most, have a superdeep origin cence around the cavities indicated
(again, see Smith et al., 2017). This they had likely been filled with a ra-
stone is one more example of dia- dioactive fluid, there was none of the
monds forming at incredible depths of greenish color or radiation staining
that would likely accompany higher
radiation doses.
Editors’ note: All items were written by staff Photoluminescence (PL) mapping
members of GIA laboratories. with 532 nm (figure 3) and 455 nm
(figure 4) excitation shows a pro-
GEMS & GEMOLOGY, Vol. 56, No. 1, pp. 126–139.
nounced increase in radiation-related
© 2020 Gemological Institute of America defects such as GR1, 3H, and TR12
and other vacancy-related peaks such

126 LAB NOTES GEMS & GEMOLOGY SPRING 2020

 servations (I.A. Dobrinets et al.,
HPHT-Treated Diamonds: Diamonds

CALCULATED CONCENTRATIONS
GR1 Forever, 2013, Springer). The width of
NV0 the fluorescence halos in figure 2 and

(FROM PEAK AREA)

NV– the elevated GR1 shown in figure 3 is
approximately 30 μm, consistent with
the penetration depth of alpha radia-
tion (S. Eaton-Magaña and K.S. Moe,
“Temperature effects on radiation
stains in natural diamonds,” Dia-
0 50 100 150 200 250 300
mond and Related Materials, Vol. 64,
2016, pp. 130–142). We know this ra-
LATERAL DIMENSION (microns)
dioactive fluid exposure must have
occurred in the crustal region because
Figure 3. Left: A 532 nm PL map of the top cavity pictured in figure 2 shows many of the features, such as 3H,
a false-color image of GR1 intensity, revealing much higher concentrations would not survive if this exposure oc-
along the perimeter of the cavity. Right: A plot of calculated concentrations curred at greater depths and, there-
for several vacancy-related centers from the PL map at left and collected at 3 fore, higher temperatures.
μm intervals along the red line in the left image. The plot shows the elevated These localized effects of radiation
concentrations along the cavity edges of GR1 (neutral vacancy; zero-phonon are particularly interesting because
line [ZPL] at 741.2 nm), NV0 (ZPL at 575 nm), and NV– (ZPL at 637 nm). the dose was low enough to not im-
part green coloration or green to
brown radiation stains. Nevertheless,
and 505 nm, also showed elevated enhanced presence due to natural ra- the radioactive fluids left their mark
concentrations in these cavities; their diation was consistent with prior ob- on this diamond with the green fluo-
rescence halos around the cavities,
providing an excellent example of
Figure 4. A photomicrograph (top left) of the lower cavity pictured in figure
how many peaks are enhanced by ra-
2 is also shown in a 455 nm PL map (top right) displaying a false-color
diation compared with the unaffected
image of the H4 intensity. The corresponding spectra (bottom) for two
portions within the remainder of the
points from the red and blue color regions are shown. Although the spectra
diamond. It is also interesting that
were collected at only ~50 μm distance, there are pronounced differences in
two distinctly different geological
peak intensities. The diamond Raman peaks within the two spectra are
conditions and depths within the
scaled as equal.
earth both contributed to the gemo-
logical characteristics of this dia-
mond. The sublithospheric depths
imparted some notable minerals as
evidence of its superdeep formation,
and radioactive fluids within the shal-
low, crustal region of the earth created
the green-fluorescing halos and other
100 µm
radiation features.
Sally Eaton-Magaña and
PL SPECTRA Garrett McElhenny
PL INTENSITY (COUNTS)

H4
3000
Formation of the “Matryoshka”
Diamond from Siberia
2000
Raman A freely moving diamond trapped in-
3H side another diamond was discovered
TR12
1000 in Siberia by Alrosa in 2019. The un-
505 usual diamond, nicknamed the
H3
490.7 498
0 “Matryoshka” after the traditional
470 475 480 485 490 495 500 505 510 515
Russian nesting dolls, attracted wide-
WAVELENGTH (nm) spread interest in how this feature
formed.

LAB NOTES GEMS & GEMOLOGY SPRING 2020 127

 https://www.gia.edu/gems-gemology/
spring-2020-labnotes-matryoshka).
The surface of the small diamond was
covered with groups of straight paral-
lel striations following the diamond
crystal symmetry. No etched trigons
were observed on the surfaces of the
small diamond (figure 6). Except for
tiny foreign-material contaminations
at the two entrances of the open chan-
nels, no other inclusions were ob-
served in this crystal (figure 7).
Figure 5. The 0.62 ct green “Ma- Figure 6. The small, flat diamond Absorption spectroscopy in the in-
tryoshka” diamond (4.8 × 4.9 × crystal enclosed in the cavity frared region indicated a type Ia dia-
2.8 mm) from Siberia has an in- showed a hexagonal outline. The mond, with high concentration of
ternal open cavity connected by surface of the small diamond aggregated nitrogen. A moderately
two small channels to the out- was covered with groups of paral- strong absorption at 3107 cm–1 from
side. The surface is covered by lel straight striations following the N3VH defect was also observed.
etched trigons and striations. A the diamond crystal symmetry. More detailed analysis of selected
small diamond crystal can move areas with and without the small in-
freely inside the cavity. ternal crystal showed almost identical
spectral features, confirming that the
The 0.62 ct flat octahedral dia- edges had rectangular openings about small crystal is a diamond with nearly
mond, a twinned macle, was recently 0.2 mm in width. The channels ex- identical trace element chemistry as
tended into the internal enclosed cav- the host. The UV-Vis absorption spec-
examined by the New York labora-
trum collected at liquid nitrogen tem-
tory. Flat-bottom trigon etch pits were ity. These features made this diamond
perature (–196°C) showed clear
well developed on the face {111} (fig- unique. Trapped in the cavity was a
absorptions from the N3 defect (ZPL
ure 5). The crystal showed a clear small, flat diamond crystal with a
at 415 nm) and GR1 (ZPL at 741 nm).
green bodycolor, with small dark hexagonal outline. The small dia-
These spectral features are typical for
green radiation stains in shallow frac- mond, covered with some green radi-
a natural type Ia diamond, except for
tures along the edges when viewed ation stains on the surface, is entirely
the strong GR1 absorption, which is
from the top of the crystal. Two etch detached from its host crystal and can
attributed to irradiation. Occurrence
channels on opposite sides of the move freely inside (see video 1 at
of the small dark green radiation
stains revealed that this diamond was
naturally irradiated. Under various
Figure 7. An open channel about 200 μm in width connected the internal laser excitations at liquid nitrogen
cavity with the outside. temperature, photoluminescence
spectroscopy showed emissions at
911, 787, 741 (GR1), 700, 535, 503.5
(3H), 489, and 468 nm. Absence of
emission from the H3 defect (N2V,
ZPL 503.2 nm) indicated that the dia-
mond crystal was not annealed to any
elevated temperature after being nat-
urally irradiated.
X-ray computed microtomography
(μ-CT) scanning and analysis revealed
some very interesting observations
(video 2 at https://www.gia.edu/gems-
gemology/spring-2020-labnotes-
matryoshka). First, the small internal
crystal showed the same intensity of
X-ray absorption as the host diamond
supporting the conclusion that the
small crystal is a diamond. Second, CT
technology can accurately compile the

128 LAB NOTES GEMS & GEMOLOGY SPRING 2020

Figure 8. The laser cavities on the star facet (left) and in reflected light (right). Field of view 1.5 mm.

morphology of the internal cavity lectively dissolved during interactions were not the typical oscillating lines
through stacked imaging. The cavity with special type(s) of melt/fluid after of manufacturing remnants. These
showed a flattened octahedral shape, its crystallization. About 0.11 carats remnants on extra facets or natural
composed of two octahedron crystals of diamond dissolved through the two surfaces are usually left behind for a
of parallel growth. The crystal habit of small open channels that created the variety of reasons such as saving
the void, an extremely important ob- pathway for the internal cavity. The weight. Neither was there an associ-
servation of this study, strongly indi- host diamond and the small internal ated internal inclusion, which one
cated that the open cavity was diamond crystals were inactive or less would expect if this were a treatment,
originally occupied by another dia- active to the melt/fluid and conse- though a shallow inclusion could
mond crystal instead of other mantle quently survived. Fluid with radioac- have been completely removed by
minerals. Once crystallized in the tive elements responsible for the this lasering. The only other internal
earth’s mantle under very high pres- green bodycolor and formation of ra- inclusion was a dark brown crystal on
sure, it is not possible for a diamond diation stains would have been the the opposite side of the table, which
crystal to have any internal open last step in decorating this crystal. had not been treated by lasering.
space. Other mantle minerals such as Special chemistries and interactions Reflected light (figure 8, right) re-
garnet, olivine, or pyroxenes will show with multiple types of melt/fluid sub- vealed a small “island” of the original
entirely different crystal habits. Fi- sequent to its initial crystallization facet left behind, evidence that many
nally, volumes of different parts of this led to the formation of this unique di- craters in close proximity joined to-
crystal were calculated. The external amond in Siberia. gether to form one larger cavity. This
crystal was 33.16 mm3 in volume (0.58 is also the reason it is suspected that
ct), while the internal small crystal Wuyi Wang, Emiko Yazawa,
Stephanie Persaud, Elina Myagkaya, this cavity system was caused by a
had a volume of 1.51 mm3 (0.03 ct). laser as opposed to another thermal
The total calculated weight of 0.61 ct Ulrika D’Haenens-Johansson, and
process, such as a jeweler’s torch.
is very close to the actual weight of Thomas M. Moses
Something like a torch is unlikely to
0.62 ct (excluding the contaminations have left an intact section of the facet
at entrances of the etching channels). in the center of a burnt-out cavity. The
The volume of the open cavity was Graphitic Cavities on Diamond
fact that there were many connected
5.99 mm3, equal to 0.11 ct of diamond An atypical series of graphitized cavi- craters does make these cavities seem
originally occupying that space. Based ties was recently observed on a star intentional, though the purpose is not
on these calculations, the total initial facet (figure 8, left) of a 0.45 ct near- obvious.
weight of this crystal would have been colorless round brilliant diamond.
about 0.72 ct. The presence of graphite was con- Troy Ardon
In summary, this diamond crystal firmed with Raman spectroscopy. The
was initially a solid diamond without observed residue surrounding the
the cavity when it formed in the craters could have been the result of a Corundum Inclusions in
earth’s mantle. Due to chemical het- lasering process. Gem Diamond
erogeneity in trace element chemistry Laser sawing is now the standard Inclusions can tell us a great deal
or sub-micro inclusions/structure way to shape rough into blocked about a diamond’s formation history.
(such as those fibrous diamonds), the shapes for brillianteering and polish- Inclusions such as olivine, garnet, and
middle part of the diamond crystal ing, and laser drilling is used to treat chromite are more common, while
(now represented by the void) was se- internal inclusions. The laser marks others such as kyanite, zircon, and

LAB NOTES GEMS & GEMOLOGY SPRING 2020 129

Figure 9. Left: Face-up image of the 0.13 ct round brilliant diamond. Right: Photomicrograph of the largest corun-
dum inclusion exhibiting a slight pinkish color when viewed through a microscope with a diffuser plate. The triple
image of the inclusion is due to the diamond’s faceting. Field of view 1.76 mm.

corundum (Al2O3) can be quite rare. surface-reaching ruby crystal. The ex- Here we report the second identi-
Regardless of their rarity, diamond in- posed inclusion was analyzed using fication of chromium-rich corundum
clusions are often quite fascinating as single-crystal X-ray diffraction, and in a natural gem diamond. This dia-
they trap a small bit of the deep earth chemical analysis was conducted mond was discovered by Michael
that cannot otherwise be sampled. using a reference ruby to calculate Turner at Stuller during routine
To our knowledge, the only pre- oxide percentages. At least two occur- screening for undisclosed laboratory-
viously recorded occurrence of rences of blue corundum (i.e., sapphire) grown diamonds. Mr. Turner deter-
chromium-bearing corundum (i.e., have also been reported (Summer 2006 mined that it was type IIa using FTIR
ruby or pink sapphire) as an inclusion Lab Notes, pp. 165–166; M.T. Hutchi- spectroscopy and then noted a pecu-
in a gem diamond was nearly 40 years son et al., “Corundum inclusions in di- liar chromium doublet emission in
ago (H.O.A. Meyer and E. Gübelin, amonds—discriminatory criteria and a PL. The diamond was sent to the
“Ruby in diamond,” Fall 1981 G&G, corundum compositional dataset,” Carlsbad laboratory for further test-
pp. 153–156). In that study, the 0.06 ct Lithos, Vol. 77, No. 1-4, 2004, pp. 273– ing. Upon examination, the 0.13 ct
round brilliant diamond contained a 286). round brilliant (figure 9, left) con-
tained at least five internal inclusions
that were positively identified as
Figure 10. Raman spectra identifying the corundum inclusions. The red chromium-rich corundum (ruby or
trace shows the figure 9 (right) inclusion, while the blue trace is a known pink sapphire). The largest inclusion
corundum reference. Spectra vertically offset for clarity. (figure 9, right) measured 0.18 mm
long; the additional inclusions were
0.07 mm or smaller.
RAMAN SPECTRA When viewed through the micro-
scope with a diffuser plate, the largest
of the inclusions displayed a very
Large central inclusion light pinkish color, while the rest ex-
hibited no visible coloration. A
DXR2xi Raman imaging microscope
was used to create a map of the indi-
INTENSITY

vidual inclusions. These maps, con-

sisting of thousands of individual
Raman spectra, were analyzed and
Corundum reference spectra compared to Raman spectra from
known corundum samples to deter-
mine that at least five of the inclu-
sions in the diamond were in fact
200 400 600 800 1000 corundum (figure 10). A chromium
RAMAN SHIFT (cm–1) doublet centered around 1369 and
1399 cm–1 (~693 and 694 nm) was also

130 LAB NOTES GEMS & GEMOLOGY SPRING 2020

 PL SPECTRUM
Previous research by G.R. Watt et
al. (“A high-chromium corundum
694 (ruby) inclusion in diamond from the
São Luiz alluvial mine, Brazil,” Min-
eralogical Magazine, Vol. 58, No. 392,
1994, pp. 490–493) and Hutchison et
INTENSITY

al. (2004) found that inclusions of

693 corundum (colorless, ruby, and sap-
phire) in their non-gem type II dia-
monds contained large amounts of
704 chromium and nickel. Both studies
declared the diamonds to be sublithos-
691 – 695 nm pheric in origin with an eclogitic par-
200 400 600 800 1000 agenesis due to the association of the
WAVELENGTH (nm) corundum inclusions with other deep-
mantle inclusions. We did not perform
Figure 11. Left: Photoluminescence spectrum showing the detection of chemical analysis on the inclusions
chromium with 532 nm laser excitation. Top right: Photo of the table with described here because they were en-
the corundum inclusions circled in red. Bottom right: False-color PL map tirely encased within the diamond.
generated by the Raman imaging microscope collected from the table in While we cannot determine the depth
the same orientation showing the 693–694 nm chromium peaks and peak of formation for this diamond, it can
intensities. be reasonably inferred, based on previ-
ous studies, to have a sublithospheric
origin and eclogitic paragenesis, as
detected in the Raman spectra, fur- largest one displayed a beautiful deep Al2O3 is rare in peridotitic mantle. Op-
ther confirming that the corundum red color (figure 12). Ruby and pink portunities to examine diamonds con-
taining rare inclusions provide an
contained significant chromium. Also sapphire typically exhibit a weak to
intriguing glimpse into the deep earth.
observed in the inclusions were addi- strong red color when exposed to
tional luminescence peaks that are long- and short-wave UV due to fluo- It is with great sadness that we note
not typical of diamond (figure 11). In rescence from abundant chromium the passing of Michael Turner. Michael
particular, a 704 nm peak was local- impurities; the same was observed for was a well-respected contributor to
ized around the corundum inclusions. some of the corundum inclusions GIA, and he will be dearly missed.
After identifying the inclusions, within this diamond. Other corun- Garrett McElhenny, Michael Turner,
we tested their reaction to UV light. dum inclusions were either too weak and Christopher M. Breeding
Under illumination from a long-wave to photograph or did not display any
ultraviolet light source (365 nm), the visible reaction to UV light.

Figure 12. The largest of the corundum inclusions shown with ordinary fiber-optic lighting (left) and long-wave
UV illumination (right). The right image displays the corundum’s deep red fluorescence color. Field of view
2.34 mm.

LAB NOTES GEMS & GEMOLOGY SPRING 2020 131

Figure 13. This carved Fancy Figure 14. Needle clusters in the carved Fancy Dark gray diamond ring.
Dark gray diamond ring weighs
13.15 ct. The whitish areas are
clear viewing windows.
lions of years ago under high pressure. impurities were detected by photolu-
During regional metamorphism, H3 minescence spectroscopy. Infrared
and NiN complexes were formed spectroscopy was used to detect the B
Solid Carved Dark Gray within the crystal lattice (K. Smit et aggregates and hydrogen complexes
Diamond Ring al., “Black diamonds from Marange within the diamond lattice, categoriz-
A ring made entirely of natural dia- (Zimbabwe): A result of natural irradi- ing this diamond as type IaB. Fluores-
mond (figure 13) was recently submit- ation and graphite inclusions,” Sum- cence imaging of the ring using the
ted to GIA’s New York laboratory. A mer 2018 G&G, pp. 132–148). These DiamondView showed bundled dislo-
large rough weighing approximately
20 ct was used to create this unusual
13.15 ct carved piece. It is common to Figure 15. Fluorescence imaging using the DiamondView shows bundled
find solid carved rings of jade and dislocation networks.
wood, which were traditionally worn
as symbols of status or wedding rings
signifying eternity. This is the first ex-
ample of a solid diamond ring submit-
ted to GIA for identification.
This solid type IaB diamond ring
owes its Fancy Dark gray color to the
graphite needles (figure 14) trapped
during formation deep below the
earth’s surface. Like famous gem-
quality black diamonds such as the
Black Orlov, the Korloff Noir, the
Black Star of Africa, and the Spirit of
de Grisogono, this diamond has clear
spaces creating viewing windows that
capture a snapshot of the ring’s forma-
tion history. The ring was fashioned
from the Beaufort diamond from
Canada’s Northwest territories and
named the “Beaufort Ring.”
This diamond rough from which
the ring was carved likely formed bil-

132 LAB NOTES GEMS & GEMOLOGY SPRING 2020

cation networks indicative of natural
diamond growth (figure 15). While this
diamond ring may not be a traditional
piece, it carries many unique aspects
capturing a snapshot of Earth’s history
within a true infinity band.
Stephanie Persaud, Paul Johnson,
and John King

Clarity-Enhanced GLASS Imitating

Emerald
The Carlsbad lab received a transpar-
ent green octagonal step cut for iden-
tification. This stone visually
resembled emerald due to its rich
green color and large fractures. How-
ever, its single refractive index of
1.510 was not consistent with emer- Figure 16. Round gas bubbles in host material and flat gas bubbles within
ald’s double refractive index of 1.577 the break of fractures. Field of view 1.99 mm.
to 1.583.
Examination with a standard
translucent stone weighing approxi- 480 and 490 nm seen in the handheld
gemological microscope revealed no
mately 763.5 ct and measuring 50.32 spectroscope. Examination with a
natural inclusions. Instead, it showed
× 48.15 × 36.17 mm (figure 18). Stan- standard gemological microscope re-
rounded gas bubbles in the body as
dard gemological examination re- vealed the stone to be heavily in-
well as flattened gas bubbles in sur-
vealed a refractive index of cluded with fractures, fine particulate
face-reaching fractures (figure 16).
1.621–1.581, weak green-blue to green clouds, and miscellaneous whitish
This stone has also been examined
pleochroism, and absorption lines at minerals. The results from Raman
with FTIR and long-wave UV. When
observed with long-wave UV, the frac-
tures emitted a weak white fluores- Figure 17. Shown are the infrared spectrum for the octagonal step cut (red
cence while the body of the stone line) and the typical reference spectrum for manufactured glass (dotted
showed a very weak blue fluores- black line). The spectra have been offset for clarity.
cence. The presence of these features
was enough to conclude it had been
clarity enhanced. IR SPECTRA
Additionally, rounded gas bubbles
separate from fractures, an RI of 1.51,
and the sample’s FTIR spectrum (fig-
ure 17) were consistent with manu-
factured glass.
These gemological properties and
observations identified the material as
ABSORBANCE

clarity-enhanced manufactured glass.

Due to emerald’s high value, it is
common for imitations to show up in
the market. Items such as this
demonstrate the need to always be
cautious when purchasing gemstones.
Michaela Stephan

Unusually Large GRANDIDIERITE 6500 6000 5500 5000 4500 4000 3500 3000 2500 2000
The Carlsbad laboratory received a WAVENUMBER (cm–1)
green-blue translucent to semi-

LAB NOTES GEMS & GEMOLOGY SPRING 2020 133

 a Pinctada margaritifera mollusk. Ad-
vanced testing supported this initial
observation; a 700 nm feature in the
UV-Vis spectrum (Summer 2016 Gem
News International, pp. 207–208), a
“bumpy” Raman spectrum, and a se-
ries of peaks at around 620, 650, and
680 nm in the photoluminescence (PL)
spectrum confirmed the mollusk’s
identification (S. Elen, “Identification
Figure 18. A green-blue translu- of yellow cultured pearls from the
cent to semi-translucent grandi- black-lipped oyster Pinctada margari-
dierite weighing 763.5 ct. tifera,” Spring 2002 G&G, pp. 66–72).
Preliminary examination with Figure 19. The dark gray round
the unaided eye and subsequent mi- nacreous Tahitian cultured pearl
croscopic observation revealed char- mounted in a ring that was sub-
spectroscopy, microscopic observa- mitted for identification.
acteristic features of an untreated
tions, and standard gemological ex-
nacreous pearl: soft luster; slightly
amination were consistent with the
smoothened platy structure with a
rare mineral grandidierite.
few negligible blemishes, pits, and tured pearls from Australian Pinc-
Named after French naturalist Al-
scratches; and no indications of color tada maxima,” GIA Research News,
fred Grandidier (1836–1912), grandi-
concentration or coatings. Although February 13, 2017) with a variety of
dierite is a very rare mineral first
the mounting prevented clear obser- untraditional nuclei. Further testing
discovered in 1902 at the cliffs of An-
vation of the drill hole, some greenish might have assisted in identifying the
drahomana on the southern coast of blue material was observed and did greenish blue material in the submit-
Madagascar (D. Bruyere et al., “A new not conform to the features expected ted pearl, but that would have re-
deposit of gem-quality grandidierite in for a traditional shell bead nucleus. quired the pearl’s removal.
Madagascar,” Fall 2016 G&G, pp. 266– This therefore appeared to be an atyp- Despite the typical Tahitian bead
275). Gem-quality material of facetable ical bead cultured pearl. While not cultured pearl appearance, real-time
size was not found in the market until routinely encountered in the trade, microradiography (RTX) quickly re-
after the summer of 2015 (Winter 2015 this is not the first example GIA’s lab- vealed a rather uncharacteristic inter-
Gem News International, pp. 449– oratories have examined. Indeed, nal structure. Instead of a traditional
450). Grandidierite is found in various some branded products even exist in shell bead nucleus (figure 20, left), a
localities as an accessory mineral in the market, such as Galatea’s carved “nucleus” containing a probable elec-
aluminous boron-rich pegmatites and atypical bead cultured pearls (“Atyp- tronic “capsule-like” device was ex-
in rocks subjected to local high-tem- ical ‘beading’ in the production of cul- posed (figure 20, right). Greater detail
perature, low-pressure metamorphism
such as contact aureoles and xenoliths
(again, see D. Bruyere et al., 2016). It Figure 20. Real-time X-ray (RTX) microradiographs showing a traditional
very seldom reaches large sizes, and freshwater shell bead nucleus in a typical bead cultured pearl (left) and
this is the largest grandidierite GIA has the unusual internal scene within the client’s pearl (right). The majority of
seen to date. the “bead” material within the client’s pearl looks more radio-translucent
Michaela Stephan than the shell bead, and the various thicknesses of the metal components
are visible as more radio-opaque areas.

PEARL
“Electronic Device” in an Atypical
Bead Cultured Pearl
GIA’s Hong Kong laboratory recently
received a white metal ring adorned
with a round partially drilled dark gray
nacreous pearl measuring 16.30 mm in
diameter (figure 19). Externally, the
pearl appeared to be a typical bead cul-
tured pearl, routinely referred to in the
trade as “Tahitian,” that had formed in

134 LAB NOTES GEMS & GEMOLOGY SPRING 2020

Figure 21. Greater detail of the coiled wire around one end of the metal
capsule component (left) and the two contact points at the opposite end
(right).

of the device’s structure showed an radio-frequency identification (RFID)

elongated metal component with to pearl culturing. As a result, GIA’s Figure 23. A registered smart phone
coiled wire around one end and two Hong Kong team obtained some pearl app (connected to an RFID reader)
contact points at the opposite end (fig- samples containing the RFID chips will display the scan result after an
ure 21). The other notable feature ob- directly from the local supplier RFID chip is detected. The result
served was that the radio-opacity of (Fukui Shell Nucleus Factory) for shows the information uploaded to a
the remaining constituent of the nu- comparison. particular chip. The optional infor-
cleus was not in keeping with that ex- The pearls were found to contain mation (e.g., product info, nucleation
pected of typical freshwater shell. The RFID chips in each nucleus (figure 22, date, harvest date) has not been en-
white outer layer represents the A and B), consistent with previous re- tered for this sample.
nacreous layers that overgrew the in- ports (H.A. Hänni and L.E. Cartier,
serted materials, and the nacre thick- “Tracing cultured pearls from farm to
ness also seemed rather thin in consumer: A review of potential
comparison to that typically observed methods and solutions,” Journal of plier stated that the chips are placed
in Tahitian bead cultured pearls rou- Gemmology, Vol. 33, No. 7-8, 2013, slightly off-center to prevent damage
tinely examined. pp. 239–246). The RTX images clearly from drilling and that the beads are
The RTX results reminded the au- revealed an RFID chip positioned fashioned from freshwater shell, like
thor of a recently patented pearl iden- within a small recess inside the shell the vast majority of nuclei used in the
tification technology that introduced bead nucleus (figure 22C). The sup- cultured pearl industry. The freshwa-
ter environment of an intact bead nu-
cleus (before RFID chip insertion)
Figure 22. Externally, this specimen with an RFID chip embedded (A) provided by the same company was
looks no different from bead cultured pearls routinely seen in the market. confirmed by energy-dispersive X-ray
However, RTX examination revealed a square chip positioned inside a re- fluorescence (EDXRF) analysis. An-
cess within a sawn and reconstructed shell bead nucleus (B and C). The other characteristic feature of these
arrow in image C indicates the sawn plane. RFID bead cultured pearls revealed by
the RTX work was that the beads
A B C were sawn in half prior to the chip’s
insertion (indicated by the black line
seen in figure 22C when aligned cor-
rectly) and then bonded back together.
By connecting a registered smart
phone app to an RFID reader, the in-
formation saved in the chips may be
retrieved (figure 23). The purpose of
introducing such technology is to en-
able consumers to identify, track, and
trace the sources of the pearls en-
countered (“Fukui unveils South Sea

LAB NOTES GEMS & GEMOLOGY SPRING 2020 135

pearls with Metakaku nuclei,” Jew-
ellery News Asia, No. 392, 2017, pp.
66–67).
Obviously the RFID beads re-
vealed in the pearls we sourced in no
way match the internal structure ob-
served in the client-submitted pearl
described in this note. The RFID
reader also failed to read the informa-
tion (if any) stored in the capsule-like
electronic device. Thus, we still can- Figure 24. Treated pearls (left) with a similar range of colors as naturally
not be certain of the identity of the colored pearls (right).
materials incorporated in the unusual
cultured pearl, but we will continue
to investigate. 340 nm. This fluorescence feature the response in the 300 to 700 nm
Pearl culturing technologies are may be attributable to the organic range.
forever adapting in order to meet the compounds contained within the A set of 12 pearl samples was se-
perpetually high demand. Besides nacreous layers (J. Hiramatsu et al., lected for evaluation. Figure 24 shows
RFID, near-field communication “Non-destructive assessment of the six treated-color pearls (left) and six
(NFC) technology has also been effects of heat and sunlight on akoya naturally colored pearls (right). The
adopted by some pearl brands (e.g., pearl quality,” Seibutsu Kogaku, Vol. samples included both freshwater and
Galatea and Gyso) to gain market 88, No. 8, 2010, pp. 378–383; F.W.J. saltwater cultured pearls of colors fre-
share. One of their selling points is Teale, “The ultraviolet fluorescence quently encountered in the market.
that by simply scanning the pearl of proteins in neutral solution,” Jour- Figure 25 shows the experimental
with an NFC-enabled smart phone nal of Biochemistry, Vol. 76, No. 2, results of this fluorescence measure-
with the registered app installed, dig- 1960, pp. 381–388). It can be identified ment prototype. The horizontal axis
ital information such as audio, videos, in all untreated pearls, corresponding indicates the fluorescence wave-
and photos can be retrieved. It is too to the UV absorption band around 280 lengths from 300 to 550 nm while the
soon to tell whether these tracking nm (J. Yan et al., “Origin of the com- vertical axis shows the normalized
and marketing technologies for pearls mon UV absorption feature in cul- detector counts, which is the relative
will be widely accepted by the buying tured pearls and shells,” Journal of intensity of the signal normalized to
public. Nonetheless, GIA will remain Materials Science, Vol. 52, No. 14, the spectrometer’s integration time
alert to these developments. 2017, pp. 8362–8369). per millisecond. Based on the results,
Commonly applied color treat- the six naturally colored pearls
Cheryl Ying Wai Au ments such as dyeing and irradiation showed fluorescence signals in the
tend to damage or mask the conchi- UV region at least 2.5 times stronger
olin in the nacre, significantly reduc- than the six treated pearls. This char-
Fluorescence Spectroscopy for ing the fluorescence intensity. By acteristic feature may be a useful and
Colored Pearl Treatment Screening evaluating the intensity of a pearl’s rapid screening technique for gemo-
Color is one of the most important fluorescence in the UV region, it is logical laboratories to detect color
value factors for pearls. Commonly possible to rapidly detect potential treatments in pearls.
applied treatment methods used to color treatments on pearls in a nonde-
alter their color to increase commer- structive manner. Tsung-Han Tsai and Chunhui Zhou
cial value include dyeing, irradiation, A prototype fluorescence spec-
and bleaching. Unfortunately, the troscopy system was designed to
identification of some color treat- measure the fluorescence signal in A Fossilized Shell Blister and
ments is challenging and time con- order to detect potential color treat- Blister Pearl
suming. In this study, we tested the ments. A 275 nm UV light-emitting Fossils are the remains or impressions
use of a fluorescence spectroscopy diode (LED) was used as the excitation of prehistoric organisms preserved
system to nondestructively inspect source. The excitation light was from past geological ages. Some of the
pearls by measuring their near ultra- guided by a bi-fabricated fiber probe to most common fossils are shells of var-
violet (UV) response to visible fluores- generate the fluorescence signal from ious marine mollusks and, occasion-
cence under mid-UV excitation. the pearl sample, and the fluorescence ally, their associated pearls (Winter
When excited by deep to mid-UV signal was relayed by the same fiber 2015 Lab Notes, pp. 432–434). Re-
(200 to 300 nm) light, naturally col- probe to the detector. Finally, a spec- cently two interesting fossilized spec-
ored pearls emit a fluorescence band trometer was used to disperse the flu- imens were studied by staff of GIA’s
between 320 and 400 nm centered at orescence light emitted and monitor New York laboratory. One resembled

136 LAB NOTES GEMS & GEMOLOGY SPRING 2020

 FLUORESCENCE SPECTRA

Natural color golden

Natural color strong purplish pink
Natural color strong pinkish purple
Natural color brown
Natural color dark gray
NORMALIZED INTENSITY

Natural color pistachio

Treated color golden
Treated color strong pinkish purple
Treated color strong pinkish purple
Treated color dark bluish gray
Treated color chocolate
Treated color pistachio

Figure 26. Two fossilized speci-

mens, a shell blister (left) and a
blister pearl (right), both with
their associated shell. Courtesy of
300 350 400 450 500 550
Gina Latendresse, American
WAVELENGTH (nm) Pearl Company, Inc.

Figure 25. Fluorescence spectra of naturally colored and treated-color the related winged pearly oysters of the
pearls. The horizontal axis represents the wavelength of the signal, and extant genus Pteria). Fossils of the In-
the vertical axis displays the signal level in normalized intensity. oceramus are commonly found in the
Pierre Shale of the Western Interior
Seaway in North America, where the
a shell blister and the other a blister the difference?” GIA Research News, state of Kansas lies today (figure 29).
pearl, both with associated shell (fig- https://www.gia.edu/gia-news- The specimen is reportedly estimated
ure 26). research/natural-shell-blisters-and- to be as much as 86 million years old.
The smaller sample appeared to be blister-pearls). No additional informa- EDXRF analysis detected low con-
a shell fragment measuring approxi- tion was provided about this specimen. centrations of manganese and high
mately 24.4 × 18.6 × 6.5 mm and The second sample measured ap- concentrations of strontium, typical
weighing 12.62 ct. The back of the proximately 37 × 25 mm and weighed of marine mollusks. Iron was also de-
shell showed even striations and a cel- 32.63 ct. Its form resembled a concave tected in both specimens, probably in-
lular crystal growth pattern, while the shell fragment, and a near-round blis- troduced through infiltration from the
cross section showed a columnar ter pearl was clearly attached to its surrounding environment over mil-
structure (figure 27). A shallow inner surface (figure 28). According to lions of years. Microradiography did
“bump” noted on the face side could our source, this specimen was found in not reveal any clear internal growth
be best described as a shell blister, ac- Ellis County, Kansas, and formed
cording to a recent study (“Natural within an Inoceramus mollusk genus
shell blisters and blister pearls: What’s (an extinct marine bivalve resembling Figure 28. A near-round blister
pearl seen lightly attached to its
host shell.
Figure 27. Left: Clear striations are visible on the back of the shell blister
(measuring approximately 24.4 × 18.6 × 6.5 mm). Center: Cellular crystal
growth patterns are evident on the back of the shell blister. Right:
Columnar growth structures are prominent with the cross section of the
shell fragment.

LAB NOTES GEMS & GEMOLOGY SPRING 2020 137

 16.00 mm (figure 30) was submitted
to GIA’s Bangkok laboratory for iden-
tification, the gemologists were in-
trigued by what the X-rays revealed.
Externally, the pearl showed char-
acteristics typical of bead cultured
pearls originating from the silver-
lipped variety of Pinctada maxima:
large size, symmetrical shape, satiny
luster, and white color. Preliminary
observation with a loupe and micro-
scope revealed faint and smooth over-
lapping nacre platelets with some Figure 31. Real-time microradiogra-
surface blemishes and polishing lines. phy revealed two planes and two re-
Energy-dispersive X-ray fluorescence cesses (one in the center of each
(EDXRF) analysis failed to detect any plane) within a large shell bead nu-
manganese (Mn) but did reveal high cleus. In the optimal direction, one
Figure 29. Map of North America levels of strontium (Sr), while X-ray lu- plane (red arrow) shows as a distinct
highlighting the shallow inland minescence revealed a moderate dark line owing to the perfect align-
seaways present during the mid- greenish yellow reaction. These results ment with the X-ray source relative
Cretaceous period. The arrow in- are consistent with those expected for to the detector, while the other plane
dicates the present-day state of saltwater cultured pearls containing a (yellow arrow) is less well aligned
Kansas. © Colorado Plateau freshwater shell bead nucleus (H.A. and hence appears more like a dis-
Geosystems Inc. Hänni et al., “X-ray luminescence, a coid feature.
valuable test in pearl identification,”
Journal of Gemmology, Vol. 29, No.
structures. These two intriguing spec- 5/6, 2005, pp. 325–329). with RTX do not show such defined
imens created by Mother Nature and However, several interesting fea- bands. (These are not to be confused
preserved by her until their discovery tures were revealed by examination with the mostly weak and more dif-
provide unique insights into ancient with real-time microradiography fused bands sometimes seen that re-
mollusks and their associated pearl or (RTX) (figure 31). The bead nucleus late to the shell’s layered growth
pearl-like companions. was clearly visible in the RTX images structure.) The parallel lines matched
and measured approximately 10.21 structures observed in GIA’s previous
Chunhui Zhou and Tao Z. Hsu research on known samples of bead
mm in diameter. This uncommon
size was the first indication that this cultured pearls in which laminated
was not a standard bead. Moreover, shell bead nuclei were used. In keep-
Saltwater Bead Cultured Pearl with ing with samples from that research,
two sharp planes were clearly visible,
Laminated Nucleus another interesting observation since the most defined line appears when
The nuclei used for bead cultured most shell bead nuclei examined the nucleus is in the optimal direc-
pearl production are usually spheres tion (figure 31, red arrow) relative to
fashioned from nacreous shell (J. Tay- the X-ray source and detector, while
lor and E. Strack, “Pearl production,” the less defined line (figure 31, yellow
in P.C. Southgate and J.S. Lucas, The Figure 30. A white round nacre- arrow) appears more like a discoid
Pearl Oyster, 2008, Elsevier, Oxford, ous pearl weighing 28.89 ct and feature when the alignment is not so
UK, pp. 273–302). The majority of this measuring 16.00 mm. perfect. Additionally, a shallow and
shell is sourced from freshwater mus- circular-looking recessed feature in
sels found in the Mississippi and Ten- the center of each plane aroused fur-
nessee River systems in the United ther curiosity. Such structures are not
States (L.E. Cartier and M.S. Krzem- typical of other laminated beads pre-
nicki, “New developments in cul- viously examined by GIA.
tured pearl production: Use of organic Laminated shell bead nuclei orig-
and baroque shell nuclei,” Australian inated a few decades ago when some
Gemmologist, Vol. 25, No. 1, 2013, Australian pearl farms attempted to
pp. 6–13), and the resulting nuclei typ- produce large pearls to meet demand.
ically range between 6.0 and 8.0 mm. However, the high price of the larger
When a white round nacreous pearl shell bead nuclei needed for this
weighing 28.89 ct and measuring forced them to experiment with other

138 LAB NOTES GEMS & GEMOLOGY SPRING 2020

Figure 32. A 5.35 ct purplish red Figure 33. Left: Blue flashes and trapped gas bubbles of various sizes and
cabochon displaying asterism. shapes along surface-reaching fractures and cavities. Field of view 7.64
mm. Right: The bismuth glass-filled fractures and cavities were easily vis-
ible in the X-ray image.
materials such as dolomite (Summer
1998 Lab Notes, pp. 130–131; Winter
2001 Gem News International, pp. ray luminescence reaction prove that diography indicated heavy element
332) to produce the large nuclei. Un- the bead was fashioned from freshwa- depositions along these fractures and
fortunately, none of the materials ter shell. Therefore, even though the cavities (figure 33, right). The Fourier-
were suitable from a commercial per- nucleus is laminated, the pearl is still transform infrared (FTIR) spectrum
spective. Some farms started bonding classified as a “bead cultured pearl.” showed two broad absorptions cen-
strips of shell together to fashion The only mystery left is the reason tered at 3500 and 2670 cm–1 associated
larger bead nuclei of around 10–10.9 for the two shallow circular recesses with manufactured glass. Qualitative
mm. The majority of such pearls in the surfaces of two of the shell analysis using energy-dispersive X-ray
were produced from second opera- pieces used to form the laminated fluorescence (EDXRF) spectroscopy re-
tions. Since this experiment proved bead nucleus. vealed the presence of bismuth,
more successful, a number of saltwa- whereas no lead was detected. This
Areeya Manustrong and
ter bead cultured pearls possessing a confirmed that the foreign material
Kwanreun Lawanwong
laminated shell bead nucleus are still was actually a bismuth-based glass.
encountered in the market today. Based on internal features and ad-
GIA labs globally have encoun- vanced testing results, the stone was
tered numerous laminated shell bead Bismuth Glass-Filled Burmese
positively identified as a manufactured
nuclei. However, this is the first ex- Star RUBY product consisting of bismuth glass
ample with the unusual recessed fea- The Hong Kong laboratory recently and ruby. Although bismuth-based
tures, and the reason for their examined a 5.35 ct purplish red oval glass has occasionally been applied to
existence is unclear. They are un- cabochon displaying asterism (figure corundum as a filling material (Spring
likely to be related to damage caused 32). Standard gemological tests 2017 Lab Notes, p. 94), it is rare to de-
during the manufacturing of the bead. yielded a spot refractive index (RI) of tect it in star ruby.
The almost identical positions and ap- 1.76, medium red fluorescence under
long-wave ultraviolet (UV) light, Xiaodan Jia and Mei Mei Sit
pearances suggest that they were cre-
ated for some purpose. Since the pearl weak red fluorescence under short-
is undrilled, damage resulting from wave UV, and a diagnostic spectrum
any drilling process (Summer 1995 in the handheld spectroscope, all of
Lab Notes, pp. 125) may also be elim- which were consistent with ruby.
PHOTO CREDITS
inated. Another possibility is that Magnification revealed iridescent
Garrett McElhenny—1, 9 (right), 12; Sally
they may be related to placement silk and arrowhead-like inclusions
Eaton-Magaña—4 (top left); Jianxin (Jae)
areas for radio-frequency identifica- similar to those frequently found in Liao—5, 13; Towfiq Ahmed—6; Elina
tion (RFID) chips (see Spring 2020 Lab Burmese rubies. Apart from the natu- Myagkaya—7; Troy Ardon—8; Mitchell Lyn—9
Notes, pp. 134–136 of this issue). But ral inclusions, easily observable foreign (left), 11 (top right); Stephanie Persaud—14,
since it is customary to only use one substances for clarity enhancement 15; Michaela Stephan—16; Diego Sanchez—
chip per pearl and the recesses are rel- were found along surface-reaching 18; Tony Leung—19, 22; Sood Oil (Judy)
atively thin with rounded outlines, fractures and cavities. Numerous Chia—24, 26, 27 (left), 28; Chunhui Zhou—27
which differ from the usual deeper rounded to flattened gas bubbles and (center and right); Nuttapol Kitdee—30; Kwan-
square recesses encountered for RFID blue flashes were visible within filled reun Lawanwong—31; Johnny Leung—32;
pearls, this seems unlikely. fractures (figure 33, left). The filler also Xiaodan Jia—33 (left); Sze Ling Wong—33
(right).
The presence of faint banding exhibited a different luster from that of
within the nucleus and the pearl’s X- the ruby under reflected light. X-ray ra-

LAB NOTES GEMS & GEMOLOGY SPRING 2020 139

 Editor
Nathan Renfro

Contributing Editors
Elise A. Skalwold and John I. Koivula

Actinolite in Spinel
Lotus Gemology received a vivid cobalt blue spinel from
Vietnam for laboratory testing. Despite its small size of well
under a carat, its stunning color stood out immediately. The
stone proved to be exceptionally clean. Most cobalt spinels
we test contain fissures, small crystals, or other inclusions.
After careful examination, we found just one tiny inclusion,
a transparent teardrop-shaped crystal (figure 1).
Raman microscopy revealed the crystal to be actinolite,
a mineral in the amphibole group. As far as we are aware,
this is the first reported inclusion of actinolite in spinel.
E. Billie Hughes
Lotus Gemology, Bangkok

Figure 1. This small transparent colorless crystal in spinel

Bavenite in Quartz was identified by Raman analysis as actinolite. Photo-
The author recently had the opportunity to examine a micrograph by E. Billie Hughes; field of view 1.7 mm.
233.60 ct faceted quartz containing numerous white radial
inclusions, courtesy of Mike Bowers (figure 2). The unusual found. The mineral bavenite, Ca4Be2Al2Si9O26(OH)2, is an
inclusions were identified as the mineral bavenite by uncommon orthorhombic mineral that occurs as drusy in
Raman and laser ablation–inductively coupled plasma–
mass spectrometry (LA-ICP-MS) analysis.
South America claims a long history of gemstone pro- Figure 2. Clusters of white radial needles of the min-
duction and arguably produces some of the finest gem eral bavenite were seen in this 233.60 ct quartz. Pho-
specimens in the world. Of the South American gem-pro- tomicrograph by Nathan Renfro; field of view 13.43
ducing areas, Brazil is by far the most prolific source of im- mm. Courtesy of Mike Bowers.
portant gemstones. Due to the array of different geologic
environments, a number of uncommon minerals can be

About the banner: Green crystals of tangeite are perched on bright blue
needles of papagoite in this quartz from the Messina mine in Transvaal,
South Africa. Photomicrograph by Nathan Renfro; field of view 3.35 mm.
Editors’ note: Interested contributors should contact Nathan Renfro at
[email protected] and Stuart Overlin at [email protected] for submission
information.

GEMS & GEMOLOGY, VOL. 56, NO. 1, pp. 140–147.

© 2020 Gemological Institute of America

140 MICRO-WORLD GEMS & GEMOLOGY SPRING 2020

miarolitic cavities in granite and associated pegmatite,
formed by a low-temperature alteration of beryl and other
beryllium-bearing minerals, and also in hydrothermal
veins and skarns. It generally forms in more alkali envi-
ronments and is often associated with bertrandite, titanite,
danalite, and zeolites (J.W. Anthony et al., Eds., Handbook
of Mineralogy, Mineralogical Society of America,
http://www.handbookofmineralogy.org). This is believed
to the first time GIA has encountered bavenite inclusions
in quartz.
Maxwell Hain
GIA, Carlsbad
Figure 3. This negative crystal in a diamond is col-
ored green by radiation staining and contains a loose
Diamond with Mobile Green Diamond Inclusion
green diamond crystal. Photomicrograph by Nathan
Diamond inclusions are relatively common guests in dia- Renfro; field of view 4.23 mm.
mond crystals. They generally go unnoticed as the refrac-
tive index of the host and guest are the same. So it is often
the case that the only visual clues to such an inclusion lie
in the interface between the host and guest, where the Presumably this diamond inclusion was completely en-
slight mismatch traps fluids from the growth environment cased in its diamond host, but dissolution occurred as the
and leaves a delicate optical irregularity with the appear- host made its way to the earth’s surface, causing the inter-
ance of a ghost-like framework of diamond. Inclusions like face between the two crystals to widen until the small en-
this can be further explored by examining stones in polar- trapped crystal was liberated enough to become mobile (J.I.
ized light, which often reveals significant strain around the Koivula, The MicroWorld of Diamonds, Gemworld Inter-
guest diamond. national, Northbrook, Illinois, 2000, 157 pp.). Subse-
One of the most interesting examples of a diamond in- quently, the diamond must have been exposed to fluids
clusion in diamond the authors have encountered is a 0.87 carrying radioactive materials that entered through the
ct sawn crystal section containing a negative crystal cavity small opening, causing green radiation staining of the neg-
(figure 3) that is open at the surface. Inside is a tiny green ative crystal and guest diamond inclusion. While mobile
octahedral diamond crystal that is free to rattle around but diamond inclusions in diamond are always remarkable (see
too large to exit through the opening at the surface (figure Lab Notes, pp. 127–129 of this issue), this example is the
4). For a video of the green diamond inclusion moving in- only one the authors have observed of a mobile green dia-
side the negative crystal contained in the host diamond, mond inclusion trapped in a colorless diamond.
go to https://www.gia.edu/gems-gemology/spring-2020- Nathan Renfro and John I. Koivula
microworld-diamond-mobile-diamond-inclusion. GIA, Carlsbad

Figure 4. This registered pair of images shows how the green diamond contained inside the negative crystal can
change position inside the host diamond and how the green diamond is too large to fit through the opening to the
surface. Photomicrographs by Nathan Renfro; field of view 3.20 mm.

MICRO-WORLD GEMS & GEMOLOGY SPRING 2020 141

Figure 5. Neodymium pentaphosphate is a rare experimental synthetic material that shows reversible lamellar
twinning under polarized light when gentle pressure is applied perpendicular to the twinning planes. This series of
images shows vertical striations that have changed with slight pressure. Photomicrographs by Nathan Renfro;
field of view 14.10 mm.

Reversible Twinning in unusual elastic twinning of this material makes it exceed-

Neodymium Pentaphosphate ingly difficult to cut into a gem, though a 5.33 ct stone was
Unusual synthetic gem materials are often the result of ex- faceted by Art Grant of Coast-to-Coast Rare Gems in the
perimental crystal growth for industrial applications, includ- 1990s. To the authors’ knowledge, no other material displays
ing use in the laser industry. One such experimental this interesting property.
synthetic that was produced in a very limited quantity is Nathan Renfro and John I. Koivula
neodymium pentaphosphate, with the chemical formula of
NdP5O14 (Winter 1997 Gem News, pp. 307–308). Examined
using polarized light, it displays very prominent lamellar Pallasitic Peridot with Iridescent
twinning. Lamellar twinning is a regular parallel repeated Needle-Like Inclusions
reversal in growth that causes the twinning planes to appear Pallasite is a rare type of stony iron meteorite that contains
as alternating dark and light bands when observed in polar- crystalline olivine. The author recently had the opportu-
ized light (figure 5). The material’s most unusual property is nity to examine a suite of pallasitic peridot that were re-
that when slight pressure is applied perpendicular to the ported to have originated from the Jepara meteorite found
twinning planes, the twinning reverses so that the dark areas on the Indonesian island of Java in 2008 (figure 6). The
can become light, and the light areas become dark. See video stones were selected by Bradley Payne for the abundance
of this at https://www.gia.edu/gems-gemology/spring-2020- of needle-like inclusions that showed vibrant interference
microworld-reversible-twinning-ndp. Interestingly, the re- colors. To observe the inclusions, oblique fiber-optic illu-
versal is also accompanied by a weak clicking sound. The mination was used to explore the gems until the light re-

Figure 6. This suite of

pallasitic peridot (0.17–
1.40 ct) from the Jepara
meteorite contains col-
orful needle-like inclu-
sions. Photo by Robison
McMurtry; courtesy of
Bradley Payne.

142 MICRO-WORLD GEMS & GEMOLOGY SPRING 2020

 Figure 7. Oblique fiber-
optic illumination re-
vealed colorful
iridescent needle-like
inclusions in the palla-
sitic peridot from Java.
Photomicrograph by
Nathan Renfro; field of
view 1.83 mm.

flected off the inclusions at the same time, revealing their https://www.gia.edu/gems-gemology/spring-2020-
hidden colorful beauty (figure 7) and confirming the crys- microworld-play-of-color-sapphire), the sapphire displayed
tallographic alignment of the inclusions. According to this noteworthy effect, which is most commonly seen in
Payne, less than 5% of the gems from the Jepara meteorite precious opal. It occurred within a small section of a single
show these types of inclusions. These are the most beauti- fine, milky cloud with an unusual shape (figure 8, left). The
ful inclusions the author has examined in a suite of palla- cloud also possessed extensive transparent graining con-
sitic peridot from the Jepara meteorite. fined within its borders, best seen in brightfield illumina-
Nathan Renfro tion (figure 8, right).
Some regions of the cloud displayed an appearance of
rainbow graining with spectral colors confined to the linear
Phenomenon Resembling Play-of-Color in Sapphire grain lines and looked one-dimensional. Other regions, how-
A phenomenon resembling play-of-color was seen in a 4.13 ever, expressed a soft billowy broad flash of altering colors
ct unheated sapphire recently examined by the author. with diffuse edges resembling play-of-color and did not fol-
When tilted in darkfield lighting (watch the video at low linear grain lines (figure 9). Play-of-color in opal is defined

Figure 8. Left: A fine, angular milky cloud as seen in darkfield illumination. Right: Brightfield illumination of the
same cloud reveals fine transparent graining causing diffraction of light into spectral colors. Photomicrographs by
Britni LeCroy; field of view 3.57 mm.

MICRO-WORLD GEMS & GEMOLOGY SPRING 2020 143

 Figure 9. Subtle tilting
of the sapphire, as
shown in this image se-
quence viewed with
darkfield illumination,
revealed a play-of-color
phenomenon within
the milky cloud. Photo-
micrographs by Britni
LeCroy; field of view
14.52 mm.

as a display of spectral colors due to the diffraction of light as croscope are never mundane because each stone captures
it passes through organized, submicroscopic spherical parti- its own piece of history through its inclusions. Recently a
cles. It is possible that the combination of fine milky parti- client submitted one such gem for testing, a 3.04 ct speci-
cles and abundant transparent graining within the cloud were men measuring 11.98 × 7.13 × 4.21 mm that was identified
able to diffract light in a way that resembles play-of-color. as an untreated ruby from Madagascar based on inclusions.
Britni LeCroy In the microscope, a few bright brownish red crystal in-
GIA, Carlsbad clusions immediately stood out. They were transparent,
with a saturated red hue that, in darkfield illumination,
was conspicuous even against their ruby-red background
Staurolite in Ruby (figure 10). Some were also cut through on the surface, al-
We have examined thousands of rubies in Lotus Gemol- lowing us to view them with reflected light. Observation
ogy’s laboratory, yet the hours spent peering into the mi- of the surface showed that the crystals had a luster just

Figure 10. Viewed with

darkfield illumination, a
transparent brownish
red crystal inclusion
clearly stands out, even
in a ruby matrix. This is
believed to be the first
staurolite inclusion
found in corundum.
Photomicrograph by E.
Billie Hughes; field of
view 2 mm.

144 MICRO-WORLD GEMS & GEMOLOGY SPRING 2020

 Figure 11. When viewed
in reflected light, the
staurolite inclusion,
which is cut through on
the surface, displays a
slightly lower luster than
its corundum host. This
suggests that the inclu-
sion has an RI slightly
below that of corundum.
Darkfield and diffuse
fiber-optic illumination.
Photomicrograph by E.
Billie Hughes; field of
view 2 mm.

slightly lower than that of the surrounding corundum, sug- beryl came from Pakistan. As triplite is known to be found
gesting a slightly lower RI (figure 11). in Pakistan, this is certainly a possibility. The owner was
To get a better idea of the crystals’ identity, we exam- happy to learn that his gem contained a relatively rare min-
ined them using micro-Raman spectroscopy with a WITec eral.
Alpha 300R Raman imaging system with a 532 nm laser, We are excited to see these examples of inclusions
which identified them as staurolite. Staurolite, being viewed as points of interest within a gem.
Fe2Al9Si4O22(OH)2, is a red to brown mineral known to E. Billie Hughes
occur in Madagascar. Its RI values are na = 1.736–1.747, nb=
1.740–1.754, and ng = 1.745–1.762. Corundum has a slightly
higher RI range, with n𝝎 = 1.767–1.722 and ne = 1.759– Quarterly Crystal: Unknown Inclusion in
1.763. The combination of microscopic observation and Triphane Spodumene
Raman spectroscopy gave us confidence that these were When exploring in the micro-world, we occasionally run
indeed staurolite crystals. into problems we cannot solve and inclusions that cannot
It is no wonder these crystals initially stood out. To the
best of our knowledge, this is the first recorded observation
of staurolite as an inclusion in corundum. This Madagascar
ruby has encapsulated a fascinating kernel of history Figure 12. A dark orange triplite inclusion stands out
within its depths. against its colorless beryl host. Photomicrograph by
E. Billie Hughes; field of view 8 mm. Courtesy of
E. Billie Hughes Ayub Muhammad.

Triplite in Beryl
Among our Lotus Gemology lab clients and the gem trade
as a whole, inclusions have often been seen as negative fea-
tures. However, people are starting to become more inter-
ested in inclusions and see them as unique characteristics
rather than mere imperfections.
One such client brought us a colorless beryl with a
large brownish orange crystal near the culet (figure 12),
hoping to learn the inclusion’s identity. Raman microscopy
revealed it to be triplite. Triplite, (Mn,Fe)2PO4(F,OH), is a
mineral named for the Greek triplos, or triple, which refers
to its three directions of cleavage. While some triplites
have been faceted, they are quite rare both as a mineral and
as a gem material. The specimen’s owner stated that the

MICRO-WORLD GEMS & GEMOLOGY SPRING 2020 145

 Figure 13. The interior
of this 91.02 ct Afghani
spodumene plays host
to a 4 mm translucent
orange crystal. Photo
by Diego Sanchez.

be identified. Such was the case with the dominant orange dumene crystal clearly hosts a prominent 4 mm translucent
inclusion in this 91.02 ct, 39.29 × 19.34 × 12.93 mm, termi- orange crystal surrounded by a stress-related iridescent
nated light yellow triphane spodumene crystal. This spo- cleavage halo (figure 13). The spodumene crystal, from Dara-

146 MICRO-WORLD GEMS & GEMOLOGY SPRING 2020

 Figure 14. None of the
analytical techniques
at our disposal were
able to conclusively pin
down the identity of
the orange inclusion.
Photomicrograph by
Nathan Renfro; field of
view 9.31 mm.

i-Pech pegmatite field, Chapa Dara District, Konar Province, inclusion. In particular, we were looking for manganese
Afghanistan, was acquired from gem and mineral dealer which did show up in the EDXRF scan. Through the micro-
Russell E. Behnke of Meriden, Connecticut. The bodycolor scope, in polarized light, no pleochroism was detected in the
and rounded habit of the orange inclusion (figure 14), to- inclusion which pointed to the inclusion being isometric.
gether with its origin in spodumene from a pegmatite field, This was as far as we could take the analysis. In order to de-
suggested that it might be spessartine. However, laser termine for certain what the unknown inclusion was we re-
Raman microspectrometry was not able to pin down the alized that destructive analysis would be needed to get a
identification of that inclusion. This was because the crystal clear identification of the orange inclusion. Since this inclu-
faces of the spodumene were etched, which interfered with sion specimen was relatively valuable we decided that de-
the passage of the laser in the hosting spodumene. As a next structive analysis would not be used, and we would keep
step, energy-dispersive X-ray fluorescence (EDXRF) was tried the spodumene crystal as it is for future exploration.
to see if we could pick up any hints of the chemistry in the John I. Koivula and Nathan Renfro

For online access to all issues of GEMS & GEMOLOGY from 1934 to the present, visit:

gia.edu/gems-gemology

MICRO-WORLD GEMS & GEMOLOGY SPRING 2020 147

 DIAMONDS FROM THE DEEP
WINDOWS INTO SCIENTIFIC RESEARCH
Karen V. Smit and Steven B. Shirey

Diamonds Are Not Forever! Diamond Dissolution

Before cutting and polishing, diamonds have highly vari- diamond can change from an octahedron to other forms
able surface features rarely, if ever, seen by the jewelry such as dodecahedron or tetrahexahedron, and even form
wearer. These features can tell an interesting story of dia- “irregular” diamonds with no discernible shape.
mond’s geological history deep within Earth—both in the These transformations to the rough diamond’s exterior
mantle rocks where diamonds grew and during their sub- surface can be distinguished from one another. Complex
sequent volcanic transport. Our previous column showed shapes can form in the mantle prior to being picked up by
that volcanic eruptions of kimberlite are how diamonds the kimberlite. Resorption to secondary shapes and trigon
make their way from depth in the mantle to Earth’s sur- formation can occur during kimberlite eruption. Surface
face. But this violent process does not leave the rough dia- features like hillocks and frosting are also regarded as fea-
mond unscathed. tures imposed by the kimberlite. Finally, if released from
These early histories are rarely considered once the di- easily weathered kimberlite, diamond lying in streambed
amond has been faceted and set into jewelry, but they raise deposits known as “placers” and in crudely sorted sedi-
interesting and geologically important questions: mentary rocks known as “conglomerates” can take on new
surface features. Crystal rounding and percussion marks
• Why do rough diamonds look so different from each
form when a diamond is transported in high-energy fluvial
other, and what might this tell us about their geo-
environments. Green and brown radiation stains are sur-
logical history?
face features that are due to a diamond’s proximity to ra-
• What effect does the kimberlite magma have on the dioactive fluids or minerals.
diamond cargo?
• How can we see through this later stage of the dia-
mond’s history to its millions and billions of years Dissolution During Kimberlite Transport and
of mantle storage? Mantle Storage
Unless a diamond is weathered out of kimberlite, kimber-
lite eruption will produce the last visible resorption effects,
Features on Natural Rough Diamonds as it is the last event affecting the diamond before it
There is so much variety in natural diamond surfaces that, reaches Earth’s surface. Features produced by the kimber-
like snowflakes, no two diamonds are exactly alike. These lite will overprint or even remove those features produced
differences can give us useful information about how dia- in the mantle. Once kimberlite effects are studied and un-
monds react with fluids in the mantle after crystallization derstood, we can peer beyond them to a diamond’s earlier
and also reveal the dissolving properties of the kimberlite history in the mantle. But why does kimberlite magma at-
magma that brought them to Earth’s surface. tack diamonds in the first place?
External surface and internal features are related to the Successful diamond transport and delivery occurs be-
internal crystallographic structure of the diamond. Defor- cause kimberlites erupt very fast (transiting 150–200 km
mation lamellae (also known as “graining” to gemologists) in <10 hours to ~2 days; Russell et al., 2019). This rapid
can be subtle and evenly distributed. Other features relat- eruption means that any entrained mantle components—
ing to diamond structure, such as trigons, are only skin such as xenoliths and diamonds—lose pressure very
deep, microscopic, and can seem randomly distributed. quickly, are violently tumbled with upward movement,
The external form of a diamond crystal can be dissolved to and can break apart.
form secondary shapes by the partial removal of crystalline Because kimberlites are more oxidizing than diamonds,
diamond in a geological process known as dissolution or they typically are in the process of dissolving diamonds—
resorption. Left alone without dissolution, diamond will it’s just that this process has often not gone to completion.
form a perfect octahedron or a cube. But with dissolution, The chemical composition of the kimberlite and its
volatile components (such as carbon dioxide and water) are
important factors. For example, as pressure drops during
GEMS & GEMOLOGY, VOL. 56, NO. 1, pp. 148–155. ascent, the kimberlite magma is not able to dissolve as
© 2020 Gemological Institute of America much carbon dioxide, and a fluid rich in carbon dioxide and
water exsolves from the magma (figure 1; Brey and

148 DIAMONDS FROM THE DEEP GEMS & GEMOLOGY SPRING 2020
 More CO2 ± H2O fluid exsolved out of
the kimberlite as pressure drops

Pressure (GPa)
4
Thermodynamically
modeled CO2
solubility 5

Not to scale Experimental volatile

7
with pressure solubility data
given on right
8
Not to scale with pressure given 60 40 30 20 10 5 1
on the right
CO2 ± H2O wt.%

Decreasing solubility of CO2 ± H2O in

kimberlite as pressure drops

Figure 1. Volatile solubility in kimberlite versus pressure as determined by high-pressure experiments and ther-
modynamic modeling. At lower pressures, volatiles start to exsolve from the kimberlite magma, and these
volatiles often start to etch and dissolve diamonds. Modified from Foley et al. (2019), with data from Edgar et al.
(1988), Brey and Ryabchikov (1994), Brooker et al. (2011), Sharygin et al. (2017), Stamm and Schmidt (2017), and
Moussallam et al. (2016).

Ryabchikov, 1994). This fluid, which often dissolves dia- from the kimberlite melt at the near-surface will have a
mond, normally starts to exsolve at around 3–4 GPa (a giga- direct effect on the amount and style of resorption that the
pascal is a unit of pressure equal to 10,000 times diamond cargo experiences.
atmospheric pressure), with maximum exsolution once the Storage in the mantle is static for the most part and
kimberlite is at pressures below 1 GPa. does not involve such drastic pressure changes. But we
The starting composition of the kimberlite melt at know that dissolution in the mantle occurs regularly since
depth in the mantle affects the amount of carbon dioxide the internal structures of most diamonds show evidence
that can be dissolved in it, and hence the amount of carbon for multiple growth episodes often intersected by periods
dioxide that will exsolve as it reaches the near-surface. of resorption (figure 2). We know from diamond ages, and
Carbon dioxide solubility is significantly lower in water- the much younger kimberlite eruption ages, that diamonds
and silica-rich kimberlites, meaning they will exsolve reside in the mantle for millions to billions of years. Geo-
more carbon dioxide than water- and silica-poor kimber- chemical analysis shows that during this time, fluids pass
lites (Brooker et al., 2011; Russell et al., 2012; Moussallam through the mantle during a process known as mantle
et al., 2016). The amount of carbon dioxide that exsolves metasomatism. Mantle metasomatism seems to be similar

DIAMONDS FROM THE DEEP GEMS & GEMOLOGY SPRING 2020 149
 Even the best age dating techniques we can apply sim-
Ion probe
analytical spots ply do not have the precision to distinguish the small dif-
ferences (e.g., <100 million years) in age when diamonds are
Growth often billions of years old. However, microscopic imaging
Mixed-habit interruption techniques such as cathodoluminescence and Diamond-
growth
View can reveal the growth layering that occurs internally
in most gem diamonds. Based just on cross-cutting relation-
ships and the superposition of younger diamond layers on
top of older, preexisting diamond, we are sure that different
generations of diamond-forming fluids exist. We can see
what layer must have come first and what layer second, and
that dissolution of preexisting diamond is part of the story.

200 μm
Experiments Reveal Which Features Originate During
Storage Versus Transport. The only way to find out
Figure 2. Cathodoluminescence image of a plate cut
whether features originate during mantle storage or kim-
through the center of a diamond. The image reveals
berlite transport is to classify the type of dissolution fea-
growth zones (similar to tree rings) that are ubiqui-
tures seen, and then to reproduce them experimentally to
tous in the majority of natural diamonds. Diamond
match the conditions in each setting. Dissolution experi-
grows in stages, and these growth periods are often in-
ments on diamonds have been conducted for decades, but
terrupted by periods of dissolution or stasis. Courtesy
many recent advances in our understanding of how differ-
of Richard Stern (University of Alberta).
ent fluid compositions affect the resorption and features
on diamond surfaces have been made by Yana Fedortchouk
and Zhihai Zhang, working at Dalhousie University in
to the process by which infiltrating fluids lead to diamond Canada. For a detailed overview of the experimental liter-
crystallization in the first place. Thus these fluids have the ature on diamond dissolution, the reader is referred to Fe-
potential to react with the diamonds that are already there. dortchouk (2019) and Fedortchouk et al. (2019).
Over this long duration in the mantle, there is the potential Their work has produced a roadmap of the dissolution
for many different carbon-bearing fluids/melts to con- features (figure 3), allowing gemologists and research scien-
tribute to diamond growth. tists to peer into the complicated history of diamond re-
vealed by its surface features. The scientific goal of such
Timescales for Mantle Resorption/Dissolution from Radio- work is to understand the fluids that exist in the mantle
metric Dating. In rare instances, diamonds with multiple and during kimberlite transport: If these fluids can dissolve
inclusions concurrently crystallizing with different growth diamond, they are also the kind of fluids from which dia-
zones have revealed the timescales over which some dia- mond can grow.
monds grow in the mantle. A diamond from the Mir mine
in Russia contains multiple sulfide inclusions in core and
rim growth zones. The inclusions in each zone are hosted
Diamond Resorption into Rounded Shapes
by diamond with vastly different carbon isotope composi- Sometimes octahedral shapes are near perfect (figures 4 and
tions and have a core that differs by about 1 billion years 5) and show very little attack by fluids either during mantle
from the rim, which is much greater than the age uncer- storage or during kimberlite transport. In other instances,
tainty (Wiggers de Vries et al., 2013; Bulanova et al., 2014). diamonds are resorbed from their primary octahedral
An even larger age difference of around 2 billion years was shapes into secondary shapes that are often grouped to-
recorded by garnet and clinopyroxene inclusions in a dia- gether and called “dodecahedra,” or “dodecs” for short.
mond from the Letlhakane mine in Botswana (Timmerman Dodecahedra are actually (at least) two different second-
et al., 2017). These growth events are intersected by periods ary shapes, and they should have names ending in -oid to
of diamond dissolution/resorption, or otherwise by periods reflect the fact that they have curved resorption surfaces
of stasis where neither growth nor resorption is occurring. that approximate similar crystallographic directions to true
It is unclear whether such long periods between growth crystal faces.
and resorption are normal or just rare examples. For exam- Dodecahedroids have 12 curved resorption surfaces
ple, the oldest diamonds ever, dated from the Ekati mine that approximate true dodecahedra. Tetrahexahedroids
in Canada (Westerlund et al., 2006) with ages of 3.5 billion (THHs) have 24 curved resorption surfaces that approxi-
years, show older cores and younger rims that are likely mate the shape of tetrahexahedra.
separated by less than 100 million years of age. These zones Dissolution experiments show that three main factors
may even have grown with a very slight age separation, may impact the shape of a resorbed diamond (Fedortchouk,
perhaps almost contemporaneously. 2019; Fedortchouk et al., 2019). Different features can be

150 DIAMONDS FROM THE DEEP GEMS & GEMOLOGY SPRING 2020
 Figure 3. Surface fea-
tures on diamond that
Diamond resorption indicate resorption in
the kimberlite magma
versus those features
In kimberlite magma In the mantle
resulting from resorp-
tion by mantle flu-
ids/melts prior to
Ditrigonal shape of {111} face Trigonal shape of {111} face
diamond entrainment
in the kimberlite (origi-
nal figure from Fe-
dortchouk et al., 2019;
used with permission).
Complex morphologies Features originating in
the mantle can be pre-
Resorption in Resorption in volatile-
served during kimber-
C-O-H fluid undersaturated melt
lite eruption if the
diamonds are shielded
Stepped-face morphologies by mantle xenoliths, or
they can be overprinted
by features resulting
from the kimberlite
magma.
Glossy Corrosive surfaces and
smooth surfaces sharp features

produced by varying the depth of the fluid exsolution from 1. Whether a fluid phase is present or not. A fluid-free
the kimberlite, or through varying the Si content and tem- kimberlite melt will not lead to rounded resorbed
perature of melts in the mantle: THH. Instead, there is some resorption along the oc-

Figure 4. A: General se-

A Resorption quence of resorption from
primary octahedron to a sec-
Volume loss
ondary rounded shape often
3% 5% 20% 35% 45 to >99% called a dodecahedroid
≤1%
(modified from Tappert and
Tappert, 2011; McCallum et
al., 1994). B, C, and D:
Rough diamonds from the
(Rounded) Diavik mine (Canada) that
Octahedral Transitional Dodecahedroidal show various features of
habit habit habit growth and resorption. B:
Terraces or growth faces in
B C D an octahedral diamond that
has undergone very little re-
sorption. C: Partially re-
sorbed octahedral diamond
with rounded corners and
edges. D: Rounded, resorbed,
tetrahexahedroidal (THH)
diamond. Photos by Jian
Xin (Jae) Liao.

DIAMONDS FROM THE DEEP GEMS & GEMOLOGY SPRING 2020 151
 Octahedron
8 crystal faces Resorption starting along crystal edges

Primary growth shape

Tetrahexahedroid (THH) Exposed growth terraces are often

24 curved resorption surfaces shield-shaped (ditrigonal)
Figure 5. Examples of
the primary octahedral
growth shape of dia-
mond and the two
shapes that form after
resorption. Photos by
Thomas Hunn (top left),
Robert Weldon (top
Secondary shape from resorption right), and Karen Smit
mostly during kimberlite eruption (middle and bottom).

Dodecahedroid
12 curved resorption surfaces

Secondary shape from resorption

mostly during mantle storage
before kimberlite eruption

tahedral crystal edges and the development of “sharp” (Chepurov et al., 1985; Khokhyrakov and Pal’yanov, 2007).
corrosive surface features (Fedortchouk and Zhang, Tetrahexahedroid crystals have been experimentally pro-
2011). duced at 1.0–1.5 GPa by dissolution in natural kimberlite
2. Depth of fluid exsolution from the kimberlite. A with 6.2 wt.% H2O and 4.8 wt.% CO2 (Kozai and Arima,
water-rich fluid at 3 GPa will favor rapid formation 2005) and in experiments with H2O and CO2 fluids (Fe-
of THH, whereas the same fluid at 1 GPa will favor dortchouk et al., 2007). Because they form during kimber-
octahedra preservation (Zhang et al., 2015). lite eruption, THH diamond crystals are much more
common than dodecahedroids, and their formation may
3. Amount of CO2 and H2O in the fluid. Experimental
overprint any mantle resorption features or shapes.
work by Fedortchouk et al. (2007) shows that at 1
GPa, there is higher preservation of the octahedral
faces in a CO2-rich fluid. A water-rich fluid would Dodecahedroid. Dodecahedroids form in CO2-rich carbon-
favor formation of THH and elimination of octahe- ate melts that are poor in water (Khokhyrakov and
dral faces. Since kimberlite magmas change compo- Pal’yanov, 2007, 2010 and many references therein). They
sition during ascent, so does the magma’s effect on are thought to form mostly during dissolution processes in
diamond resorption. Different resorption styles can the mantle (Fedortchouk and Zhang, 2011). Dodecahe-
be expected, even during the same eruption! droidal shapes are more rare, as they can be overprinted to
THH during kimberlite eruption. These shapes have a
Tetrahexahedroid (THH). Experiments indicate that tetra- higher likelihood of survival if the kimberlite magma and
hexahedroids (THH) form in water-rich carbonate melts its exsolving fluids are water poor, and also if diamonds are

152 DIAMONDS FROM THE DEEP GEMS & GEMOLOGY SPRING 2020
 Figure 6. Trigons and
hexagons are dissolu-
tion features related to
the effect of more oxi-
dizing fluids/melts on
the diamond’s surface.
Dissolution typically
starts at a lattice dislo-
cation or imperfection.
Photos by Evan Smith
and Ulrika D’Haenens-
Johansson.

shielded from the kimberlite magma by enclosure in their that has a high proportion of diamonds with positive trigons,
mantle host rocks such as peridotite and eclogite. and their development was attributed to near-surface reac-
tion of the diamonds with late-stage, low-temperature C-O-
H fluids (Li et al., 2018). This is consistent with early
Trigons experimental work that showed that positive trigons are the
Examples of one of the most common surface details of
rough natural diamonds—trigons—are shown in figure 6.
Additional excellent examples can be found in the book Figure 7. Trigons can be either “negative” or “posi-
Diamonds in Nature: A Guide to Rough Diamonds (Tap- tive,” where a “negative trigon” is more common and
pert and Tappert, 2011). has an opposite orientation to the octahedral face of
Trigons can be either “negative” or “positive,” where the diamond.
a “negative trigon” has an opposite orientation to the oc-
tahedral face of the diamond (figure 7). In the past, there
Negative Positive
was significant debate about whether trigons were growth trigon trigon
features (e.g., Tolansky, 1955, 1960) or etch pits (e.g., Frank
and Puttick, 1958; Patel et al., 1964; Lang, 1964). These
days it is accepted that a trigon is an etch feature, produced
when oxidizing fluids/melts start to dissolve the diamond.
Dissolution typically starts at an imperfection, such as a
lattice dislocation (Wilks and Wilks, 1991).

When Do Rare Positive Trigons Form? Positive trigons are Common Rare
extremely rarely seen. Snap Lake in Canada is one locality

DIAMONDS FROM THE DEEP GEMS & GEMOLOGY SPRING 2020 153
most common etch feature at atmospheric pressure (Evans ids and their universal applicability to all monocrystalline
and Sauter, 1961). These early experiments also showed that gem diamonds needs to be better established. Since very
the orientation of a trigon (positive versus negative) could few diamonds actually contain fluids, the study of fluids
be changed by using different etchants at different tempera- resulting in surface features on a wide variety of different
tures (Frank and Puttick, 1958; Evans and Sauter, 1961). diamonds may provide a broader understanding of carbon
More recent experiments have shown that positive fluids in the deep Earth. Furthermore, there is a need to un-
trigons can be produced by dry carbonate melts, and that derstand diamond resorption styles during the economic
increasing water content changes trigon orientation to neg- evaluation of kimberlite since complete resorption can de-
ative (Khokhyrakov and Pal’yanov, 2000, 2010). Fe- stroy diamond grade by eliminating diamonds altogether.
dortchouk (2019) attributes positive trigons to resorption Identification of diamond quality in each kimberlite unit
by a rare type of mantle fluid (rare since positive trigons leads to better decisions about which kimberlite units to
are associated with other rare features such as hexagons prioritize during mining.
and transverse hillocks). It is possible that the development When does a fluid result in diamond growth, and when
of positive trigons is common in the mantle, and that they does it dissolve diamonds? We need a better understanding
have just been “erased” by later resorption of the diamond of how diamonds are resorbed and destroyed in the mantle.
into THH during kimberlite transport. Carbon-bearing fluids in the mantle seem to be largely re-
sponsible for diamond growth (e.g., Stachel et al., 2017) and
Influences on the Shape of a Trigon. Experimental work has do not contribute to diamond dissolution at high pressures
shown that the shape of trigons is influenced by the amount during mantle storage (Fedortchouk et al., 2019). Conversely,
of H2O and CO2 in the fluid, as well as temperature. Trigons carbon-bearing fluids exsolved at low pressures (figure 1; less
produced by water-rich fluids have a limited size range but a than 3 GPa) during kimberlite eruption do dissolve dia-
large depth variation; in contrast, trigons produced by CO2- monds to form rounded THH (figure 5; Fedortchouk et al.,
rich fluids can have varying sizes but always with a positive 2019). In the mantle, the main diamond dissolution agents
correlation between depth and diameter (Fedortchouk, 2015). seem to be carbonate melts that can produce several features
Large trigons are produced at higher temperatures (Kanda et through variation of silica contents and temperature (Fe-
al., 1977; Fedortchouk, 2015). Additionally, experiments at dortchouk and Zhang, 2012; Fedortchouk et al., 2019).
1 to 3 GPa showed that CO2-rich fluids mostly produced We have shown that diamond can be dissolved during
trigons with pointed and curved bottoms, whereas CO2-poor two periods of residence within the earth: where it grew in
fluids produced mostly flat-bottomed trigons (Zhang, 2016). the mantle and during its volcanic transport to the surface.
Given that diamond dissolution is common, it is fortuitous
that the mantle’s temperature, high pressure, and low oxy-
The Importance of Understanding Resorption, gen state makes it friendly for diamond growth. Similarly,
and Remaining Questions given that diamond dissolution can occur in fluid-rich kim-
Since the surface features of diamonds are heavily influ- berlites, it is a good thing that transport is fast enough to
enced by the composition and temperature of the dissolv- overcome the speed of dissolution. In the end, given that
ing fluid/melt, these features are another tool that mantle diamond dissolution is common in two settings in Earth,
geochemists can use to study carbon-bearing fluids/melts it is rather miraculous that diamonds do actually survive.
in the deep Earth. Actual fluids are preserved in fibrous di- Here we have yet another way in which natural diamonds
amonds, but these may be a specific subtype of mantle flu- are remarkable.

REFERENCES
Brey G.P., Ryabchikov I.D. (1994) Carbon-dioxide in strongly silica melts at high pressure. Doklady Akademii Nauk SSSR, Vol. 1,
undersaturated melts and origin of kimberlite magmas. Neues pp. 212–216.
Jahrbuch fur Mineralogie, Monatshefte, Vol. 10, pp. 449–463. Edgar A.D., Arima M., Baldwin D.K., Bell D.R., Shee S.R., Skinner
Brooker R.A., Sparks R.S.J., Kavanagh J.L., Field M. (2011) The E.M.W., Walker E.C. (1988) High-pressure-high-temperature
volatile content of hypabyssal kimberlite magmas: Some con- melting experiments on a SiO2-poor aphanitic kimberlite from
straints from experiments on natural rock compositions. Bul- the Wesselton mine, Kimberley, South Africa. American Min-
letin of Volcanology, Vol. 73, No. 8, pp. 959–981, eralogist, Vol. 73, No. 5, pp. 524–533.
http://dx.doi.org/10.1007/s00445-011-0523-7 Evans T., Sauter D.H. (1961) Etching of diamond surfaces with
Bulanova G.P., de Vries D.F.W., Pearson D.G., Beard A., Mikhail gases. Philosophical Magazine, Vol. 6, No. 63, pp. 429–440,
S., Smelov A.P., Davies G.R. (2014) An eclogitic diamond from http://dx.doi.org/10.1080/14786436108235896
Mir pipe (Yakutia), recording two growth events from different Fedortchouk Y. (2015) Diamond resorption features as a new
isotopic sources. Chemical Geology, Vol. 381, pp. 40–54, method for examining conditions of kimberlite emplacement.
http://dx.doi.org/10.1016/j.chemgeo.2014.05.011 Contributions to Mineralogy and Petrology, Vol. 170, No. 4, p.
Chepurov A.I., Khokhryakov A.F., Sonin V.M., Pal’yanov Y.N. 36, http://dx.doi.org/10.1007/s00410-015-1190-z
(1985) Forms of dissolution of diamond crystals in silicate ——— (2019) A new approach to understanding diamond surface

154 DIAMONDS FROM THE DEEP GEMS & GEMOLOGY SPRING 2020
 features based on a review of experimental and natural dia- and emplacement. Earth and Planetary Science Letters, Vol.
mond studies. Earth-Science Reviews, Vol. 193, pp. 45–65, 447, pp. 151–160, http://dx.doi.org/10.1016/j.epsl.2016.04.037
http://dx.doi.org/10.1016/j.earscirev.2019.02.013 Patel A.R., Goswami K.N., Desai C.C. (1964) Trigon patterns on
Fedortchouk Y., Zhang Z. (2011) Diamond resorption: Link to etched calcium fluoride cleavages. Philosophical Magazine,
metasomatic events in the mantle or record of magmatic fluid Vol. 10, No. 108, pp. 931–935, http://dx.doi.org/10.1080/
in kimberlitic magma? Canadian Mineralogist, Vol. 49, No. 3, 14786436408225401
pp. 707–719, http://dx.doi.org/10.3749/canmin.49.3.707 Russell J.K., Porritt L.A., Lavallée Y. (2012) Kimberlite ascent by
Fedortchouk Y., Canil D., Semenets E. (2007) Mechanisms of dia- assimilation-fuelled buoyancy. Nature, Vol. 481, No. 7381, pp.
mond oxidation and their bearing on the fluid composition in 352–356, http://dx.doi.org/10.1038/nature10740
kimberlite magmas. American Mineralogist, Vol. 92, No. 7, pp. Russell J.K., Sparks R.S.J., Kavanagh J.L. (2019) Kimberlite volcanol-
1200–1212, http://dx.doi.org/10.2138/am.2007.2416 ogy: Transport, ascent, and eruption. Elements, Vol. 15, No. 6,
Fedortchouk Y., Liebske C., McCammon C. (2019) Diamond de- pp. 405–410, http://dx.doi.org/10.2138/gselements.15.6.405
struction and growth during mantle metasomatism: An exper- Sharygin I.S., Litasov K.D., Shatskiy A., Safonov O.G., Golovin
imental study of diamond resorption features. Earth and A.V., Ohtani E., Pokhilenko N.P. (2017) Experimental con-
Planetary Science Letters, Vol. 506, pp. 493–506, straints on orthopyroxene dissolution in alkali-carbonate melts
http://dx.doi.org/10.1016/j.epsl.2018.11.025 in the lithospheric mantle: Implications for kimberlite melt
Foley S.F., Yaxley G.M., Kjarsgaard B.A. (2019) Kimberlites from composition and magma ascent. Chemical Geology, Vol. 455,
source to surface: Insights from experiments. Elements, Vol. pp. 44–56, http://dx.doi.org/10.1016/j.chemgeo.2016.09.030
15, No. 6, pp. 393–398, http://dx.doi.org/10.2138/gselements. Stachel T., Chacko T., Luth R.W. (2017) Carbon isotope fractiona-
15.6.393 tion during diamond growth in depleted peridotite: Counterin-
Frank F.C., Puttick K.E. (1958) Etch pits and trigons on diamond: tuitive insights from modelling water-maximum CHO fluids as
II. Philosophical Magazine, Vol. 3, No. 35, pp. 1273–1279, multi-component systems. Earth and Planetary Science Letters,
http://dx.doi.org/10.1080/14786435808233309 Vol. 473, pp. 44–51, http://dx.doi.org/10.1016/j.epsl.2017.05.037
Kanda H., Yamaoko S., Setaka N. (1977) Etching of diamond octa- Stamm N., Schmidt M.W. (2017) Asthenospheric kimberlites:
hedrons by high pressure water. Journal of Crystal Growth, Vol. Volatile contents and bulk compositions at 7 GPa. Earth and
38, No. 1, pp. 1–7, http://dx.doi.org/10.1016/0022-0248(77) Planetary Science Letters, Vol. 474, pp. 309–321,
90365-7 http://dx.doi.org/10.1016/j.epsl.2017.06.037
Khokhryakov A.F., Pal’yanov Y.N. (2000) Dissolution forms of di- Tappert R., Tappert M.C. (2011) Diamonds in Nature: A Guide to
amond crystals in CaCO3 melt at 7 GPa. Russian Geology and Rough Diamonds. Springer, Berlin.
Geophysics, Vol. 41, No. 5, pp. 682–687. Timmerman S., Koornneef J.M., Chinn I.L., Davies G.R. (2017)
——— (2007) The evolution of diamond morphology in the Dated eclogitic diamond growth zones reveal variable recycling
process of dissolution: Experimental data. American Mineral- of crustal carbon through time. Earth and Planetary Science
ogist, Vol. 92, No. 5-6, pp. 909–917, http://dx.doi.org/10.2138/ Letters, Vol. 463, pp. 178–188, http://dx.doi.org/10.1016/j.epsl.
am.2007.2342 2017.02.001
——— (2010) Influence of the fluid composition on diamond dis- Tolansky S. (1955) The Microstructures of Diamond Surfaces.
solution forms in carbonate melts. American Mineralogist, Vol. NAG Press, London.
95, No. 10, pp. 1508–1514, http://dx.doi.org/10.2138/am.2010. Tolansky S. (1960) Surface Microtopography. Longmans, London.
3451 Westerlund K., Shirey S., Richardson S., Carlson R., Gurney J.,
Kozai Y., Arima M. (2005) Experimental study on diamond disso- Harris J. (2006) A subduction wedge origin for Paleoarchean
lution in kimberlitic and lamproitic melts at 1300–1420°C and peridotitic diamonds and harzburgites from the Panda kimber-
1 GPa with controlled oxygen partial pressure. American Min- lite, Slave craton: Evidence from Re-Os isotope systematics.
eralogist, Vol. 90, No. 11-12, pp. 1759–1766, http://dx.doi.org/ Contributions to Mineralogy and Petrology, Vol. 152, No. 3,
10.2138/am.2005.1862 pp. 275–294, http://dx.doi.org/10.1007/s00410-006-0101-8
Lang A.R. (1964) Dislocations in diamond and the origin of trigons. Wiggers de Vries D.F., Bulanova G.P., de Corte K., Pearson D.G.,
Proceedings of the Royal Society of London, Series A: Mathe- Craven J.A., Davies G.R. (2013) Micron-scale coupled carbon
matical and Physical Sciences, Vol. 278, pp. 234–242. isotope and nitrogen abundance variations in diamonds: Evi-
Li Z., Fedortchouk Y., Fulop A., Chinn I.L., Forbes N. (2018) Pos- dence for episodic diamond formation beneath the Siberian
itively oriented trigons on diamonds from the Snap Lake kim- Craton. Geochimica et Cosmochimica Acta, Vol. 100, pp. 176–
berlite dike, Canada: Implications for fluids and kimberlite 199, http://dx.doi.org/10.1016/j.gca.2012.08.034
cooling rates. American Mineralogist, Vol. 103, No. 10, pp. Wilks E., Wilks J. (1991) Properties and Application of Diamond.
1634–1648, http://dx.doi.org/10.2138/am-2018-6496 Butterworth-Heinemann Ltd., London.
McCallum M.E., Huntley P.M., Falk R.W., Otter M.L. (1994) Mor- Zhang Z. (2016) Diamond resorption morphology as a fluid proxy
phological, resorption and etch feature trends of diamonds from in diamond-bearing environments: Constraints from empirical
kimberlite populations within the Colorado-Wyoming state and experimental studies. PhD thesis, Dalhousie University,
line district. In H.O.A. Meyer and O.H. Leonardos, Eds., Dia- Halifax.
monds: Characterization, genesis and exploration. Proceedings Zhang Z., Fedortchouk Y., Hanley J.J. (2015) Evolution of diamond
of the 5th International Kimberlite Conference, Volume 2, pp. resorption in a silicic aqueous fluid at 1-3 GPa: Application to
32–50. kimberlite emplacement and mantle metasomatism. Lithos,
Moussallam Y., Morizet Y., Gaillard F. (2016) H2O–CO2 solubility Vol. 227, pp. 179–193, http://dx.doi.org/10.1016/j.lithos.
in low SiO2-melts and the unique mode of kimberlite degassing 2015.04.003

DIAMONDS FROM THE DEEP GEMS & GEMOLOGY SPRING 2020 155
 Gem News International
Contributing Editors
Emmanuel Fritsch, University of Nantes, CNRS, Team 6502, Institut des Matériaux Jean Rouxel (IMN),
Nantes, France ([email protected])
Gagan Choudhary, Gem Testing Laboratory, Jaipur, India
([email protected])
Christopher M. Breeding, GIA, Carlsbad ([email protected])

TUCSON 2020

In early February, despite the early phase of the COVID-19

outbreak, the Tucson shows were well attended and sales
were brisk. Several vendors indicated that they had their
biggest single day of sales ever this year (figure 1). This an-
ecdotal evidence was supported by the strong demand for
identification and origin services reported by the GIA Show
Service Lab. This year, it was hard to identify any one par-
ticular item leading demand or any game-changing new
gem material finds. Instead, we found that vendors were
carrying high-quality material in terms of color, clarity, and
cut. In fact, many were carrying items that had been metic-
ulously carved, fantasy-cut, or recut for ideal proportions
(figure 2).
Surpassing its strong momentum from the past few
years, teal blue sapphire from Montana was a prominent
and well sought-after item (figure 3). Sapphire was doing
very well at the shows, including parti-colored, fancy-col-
ored (electric colors such as fuchsia and pinks as well as
pastels such as lavender), and slabs displaying zoning or
trapiche-like patterning (figure 4).
Emerald from around the world could be found, includ-
ing some untreated smaller sizes from Russia (figure 5),
melee and small-sized finished stones from Pakistan, at-
tractive material from Ethiopia, and an abundance of ma-
terial in a wide variety of color, clarity, and size from
Colombia, Brazil, and Zambia.

Figure 1. A busy booth on the first day of the AGTA

show. Photo by Tao Hsu.
Editors’ note: Interested contributors should send information and illustra-
tions to Stuart Overlin at [email protected] or GIA, The Robert Mouawad
Campus, 5345 Armada Drive, Carlsbad, CA 92008.
Outstanding brilliant green garnets in both deman-
GEMS & GEMOLOGY, VOL. 56, NO. 1, pp. 156–192. toid and tsavorite varieties were available, including sev-
© 2020 Gemological Institute of America eral larger-sized, ideally colored Russian demantoids
(figure 6).

156 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2020

Figure 2. A 72 ct gem silica carving by Nick Alexan- Figure 3. A suite of the popular teal blue Montana sap-
der showing a highly desirable intense shade of blue. phires from Rock Creek. The center stone is 11.03 ct, and
The material is from the Ray mine in Arizona. Photo the other 14 stones total 8.38 carats. Photo by Jennifer
by Kevin Schumacher; courtesy of Nick Alexander. Stone-Sundberg; courtesy of Pala International.

As “Classic Blue” is the Pantone color of the year and Electric-colored accent stones and melee in materials
varieties of blue gemstones are very popular in general, such as sapphire, spinel, apatite, and garnet varieties tsa-
gems prominently displaying a blue hue could be found vorite and demantoid were easy to find both loose and in
throughout the shows. Along with sapphire, deep blue finished pieces (figure 8).
aquamarine from Nigeria (figure 7) was particularly popular, As has been the case for the past couple of years, gem
and many vendors displayed blue zircon, gem silica (again, slices in a variety of materials, shapes, and price points
see figure 2), apatite, cobalt spinel, blue topaz, haüyne, were very popular in both the AGTA and GJX shows.
turquoise, blue moonstone, and lapis in their cases. These are attractive to many designers, whether they are

Figure 4. Left: This set of 110 Australian sapphire “fancies” in 3 mm sizes was collected over several years. Right:
A pair of trapiche-like Australian sapphire slices. Photos by Jennifer Stone-Sundberg; courtesy of Terry Coldham.

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2020 157

Figure 5. A 0.33 ct emerald from Malysheva, Russia, Figure 7. Deep blue aquamarine from Nasarawa,
offered by Dudley Blauwet at the AGTA show. Photo Nigeria (left to right: 1.94, 2.84, and 1.84 ct). Photo by
by Kevin Schumacher. Emily Lane; courtesy of Jeff Hapeman.

working with higher-price-point material such as parti-col- peridot, brilliant green fluorite (figure 11, left), hackmanite,
ored sapphire (figure 9), diamond, and trapiche emerald, or triphylite, grandidierite, axinite, and an array of bicolor
more moderately priced material including amethyst, stones in materials such as garnet, chrysoberyl (figure 11,
rhodochrosite, and petrified wood. right), tanzanite, and sphene.
As always, Tucson was the place to find the unusual
and phenomenal, with several vendors carrying exotic gem
materials such as fluorescent hyalite opal (figure 10), star
Figure 8. Sea star pendant by Paula Crevoshay featur-
ing 4.04 carats of ruby, 5.06 carats of lavender sap-
phire, 9.40 carats of purple sapphire, 3.87 carats of
Figure 6. A 6.73 ct round brilliant Russian demantoid fuchsia sapphire, 6.79 carats of crystal spinel, and
of top quality showing a horsetail inclusion. Photo © 0.43 carats of amethyst, set in 18K yellow gold. Photo
Christian Wild. by Emily Lane; courtesy of Paula Crevoshay.

158 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2020

Figure 9. Parti-colored sapphire slices from Montana
(left) and Australia (right). Photos by Jennifer Stone-
Sundberg; courtesy of Warren Boyd (left) and Terry
Coldham (right).
Figure 10. Fluorescent hyalite opal flower pendant by
Loretta Castoro, photographed in daylight-equivalent
(left) and lower-power daylight-equivalent light with
2020 also marked the first Tucson edition of the Ethical added short-wave UV light (right). The 12.63 ct fluo-
Gem Fair. The exhibitors showed a wide range of gemstones, rescent opal is from Zacatecas, Mexico. It is set in
prices, and origins. They were all united in their commit- 18K yellow gold and has 48 yellow diamond accents
ment to responsible sourcing, traceability, transparency, and (1.14 carats total) and 72 white diamond accents
social awareness. The variety of exhibitors reflected the dif- (1.27 carats total). Photos by Robert Weldon; cour-
ferent approaches taken to address these diverse challenges tesy of Loretta Castoro.
around the world. During its inaugural Tucson show, the
Ethical Gem Fair mainly attracted young independent jew-
elry designers looking for gems with a transparent chain of
January 2020, Tom Trozzo submitted to Stone Group Lab-
custody from the mine to their use in finished jewelry.
oratories a 20.13 ct oval cabochon peridot (figure 12, left)
Duncan Pay, Jennifer Stone-Sundberg, Tao Hsu, that exhibited a soft yet prominent four-rayed star. Com-
Wim Vertriest, Aaron Palke, and Nathan Renfro parisons were made with an in-house lab sample that also
exhibited reflection effects from tiny acicular inclusions.
Burmese star peridot. Asterism in peridot is a rare treat For purposes of this study, this comparison stone was recut
that has been reported on occasionally (S. Borg, “An un- to a 7.67 ct pear-shaped cabochon (figure 12, right) to en-
usual star peridot,” Journal of Gemmology, Vol. 17, No. 1, hance any potential asterism. However, the result appeared
1980, pp. 1–4; Summer 2009 Lab Notes, pp. 138–139). In to be more of a cat’s-eye effect, presumably due to the

Figure 11. An extremely diverse selection of unusual stones was on display at the GJX show. Left: Bright green flu-
orite offered by a European vendor; photo by Tao Hsu. Right: Bicolor chrysoberyl carried by United Colour Stone
Co. from Bangkok; photo by Jennifer Stone-Sundberg.

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2020 159

Figure 12. The 20.13 ct star peridot (left) and the 7.67 Figure 14. A top-quality 9.49 ct Russian demantoid from
ct peridot exhibiting weak asterism with more promi- the Korkodin mine. The horsetail inclusion under the
nent chatoyancy (right), presumably due to cutting table is the telltale signature of its origin. Photo by Kevin
style. Photos by Bear Williams. Schumacher; courtesy of Tsarina Jewels.

prominent keel from the original faceted pear shape. It is The spot RI reading of approximately 1.65 and hydro-
presumed that, based on the arrangement and concentra- static SG of 3.33 were consistent with peridot. The iden-
tion of inclusions, a more prominent asterism will be re- tification of both peridots was confirmed by
vealed upon further cutting. Apparent clarity was good on GemmoRaman532-SG. Their provenance was reportedly
both stones, but with a slightly diffused effect; no large eye- Myanmar (Burma), and all tests were consistent with a
visible inclusions were present. While the asterism was Burmese origin. Interestingly, all known star peridots to
more pronounced on the oval, the acicular inclusions were date have been of Burmese origin. Although Pakistani
more plentiful and of higher relief in the pear-shaped stone. peridot may often feature acicular inclusions of ludwigite,
Both stones had a yellowish green color, with the oval ex- these are presumably protogenetic and not oriented con-
hibiting a very slight grayish modifier. sistently to the crystal structure of the host peridot.
Microscopic observation of the larger oval peridot re- Bear Williams and Cara Williams
vealed evenly distributed concentrations of extremely fine, Stone Group Labs
short, brownish, acicular inclusions oriented in two direc- Jefferson City, Missouri
tions (figure 13, left). In the pear shape, the needles were
evenly dispersed throughout and oriented in three direc- Large and fine demantoid from Russia. The green to yel-
tions. The needles were also short and whitish to colorless lowish green variety of andradite garnet was first found in
(figure 13, right). Reflection effects in some materials may the Ural Mountains. It was identified by Finnish mineral-
be difficult to define. There may be reflection effects from ogist Nils von Nordensheld in 1864 and presented at the
nonacicular inclusions, creating schiller as well as aster- Ural Industrial Exhibition as a new mineral in 1887. Due
ism. This was the case with the oval peridot, although as- to the garnet’s high RI and dispersion, it was named de-
terism was clearly the dominant phenomenon. mantoid, meaning “diamond-like.”

Figure 13. Brownish, acicu-

lar inclusions in the 20.13
ct star peridot (left, field of
view 4 mm). Short, feath-
ery, whitish needle-like in-
clusions in the 7.67 ct
pear-shaped peridot (right,
field of view 3.5 mm).
Photomicrographs by Bear
Williams.

160 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2020

 Figure 15. The Korko-
din demantoid mine is
an open-pit operation.
Trucks transport the
ore from the bottom of
the pit to the nearby
washing plant. Photo
by Tao Hsu.

Russia remained the only source for demantoid until stone was also of top quality, showing high brilliance and
the mid-1990s, when the gem was discovered in Africa and fire. The owner informed the authors that the stone was cut
later elsewhere. Today, Russian demantoid is still highly from rough obtained in 2019 and possessed the best color he
desired. At the GJX show, the authors saw an 8.49 ct fine- had seen in this material in decades (see figure 6 on p. 158).
quality demantoid (figure 14) carried by Tsarina Jewels, the Since Russian demantoid often occurs as small grains
largest demantoid seen by the authors at this year’s Tucson with no well-developed crystal forms in the matrix (fig-
shows. It was from the Korkodin mine, located about 80 ure 16), any finished stone above one carat is considered
km south of Ekaterinburg, one of the two active mining rare.
operations in this area (figure 15). The Poldnevaya mine Illustrating the attention being given to this material,
lies about 7 km to the north of Korkodin. Both operations the 2019 AGTA Spectrum Award winner under the classi-
are working on primary demantoid deposits with machin- cal category featured a perfectly matched suite of 15 stan-
ery. Together the two mines supply the majority of Russian dard round brilliant cut Russian demantoid garnets. The
demantoid. bracelet by Jeffrey Bilgore included a 3.00 ct center Russian
Another impressive Russian demantoid was a 6.73 ct demantoid accented with 14 smaller Russian demantoids
stone carried by Constantin Wild at the GJX show. This gradually decreasing in size (12.20 carats total) and 14 step-

Figure 16. Demantoid gen-

erally occurs as very small
grains in the host rock
(left). It is very difficult to
find large size rough (right)
of quality. This makes
faceted stones over 1 ct
very rare. Left photo by
Tao Hsu; right photo cour-
tesy of Constantin Wild.

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2020 161

2020 Tucson Photo Gallery
A

B C

A: “Atlantis,” a malachite and chrysocolla specimen

from the Democratic Republic of the Congo measuring
70 × 25 cm. B: Lightning Ridge black opal and diamond
ring, courtesy of The Gem Garden. C: Jellyfish design by
Loretta Castoro displaying fluorescence, featuring a 49.37
ct moonstone and a Mexican hyalite opal. D: 5.74 ct
cushion-cut spinel from Vietnam, shown in daylight and
incandescent lighting, courtesy of Bryan Lichtenstein.

Photos by Robert Weldon (A, C, D), Emily Lane (B, E–H,

J), and Kevin Schumacher (I).
E F

G J

E: Bulgari multi-colored sapphire and diamond neck-

I
lace with 36 sapphires totaling 72.29 carats, courtesy of
M. Kantor & Associates. F: South Sea cultured pearl
necklace, courtesy of Tara & Sons, Inc. G: Jadeite ear-
rings, 16.77 carats total, courtesy of Jye’s International.
H: Paula Crevoshay brown-eyed Susan brooch featur-
ing an orange Montana sapphire with multi-colored
sapphires and diamonds. I: 7.35 ct Fantasy-cut heliodor
carving by Nick Alexander. J: 1940s diamond and sap-
phire link bracelet by Oscar Heyman with 27 carats of
blue sapphire, courtesy of M. Kantor & Associates.
 Figure 17. Russian de-
mantoid garnet and
diamond bracelet set in
platinum. The center
stone is 3.00 ct, while
the accent garnets total
12.20 carats and the ac-
cent diamonds 4.18
carats. Photos courtesy
of Jeffrey Bilgore.

cut (4.18 carats total) diamonds all set in platinum (figure Russian emerald. On the long list of emerald-producing
17). Mr. Bilgore said he considers Russian demantoid to be countries, Russia is one of the more mysterious to the trade
on par with top gems such as Kashmir sapphire, Burmese and consumers. Emerald was found in the Ural Mountains
ruby, and Colombian emerald. The suite came from the es- in the early nineteenth century. Malysheva was the most
tate of a collector and attracted much attention from gem famous of these deposits and the world’s largest emerald
dealers familiar with Russian demantoids before ulti- producer at the start of World War I. During the Soviet era,
mately being used in this award-winning piece of jewelry. this deposit was nationalized and mined for beryllium in-
Tao Hsu, Jennifer Stone-Sundberg, and Aaron Palke stead of emerald. Today, underground mining is going
GIA, Carlsbad strong and actively producing emeralds (figure 18).

Figure 18. The Maly-

sheva emerald mine
outside of Ekaterin-
burg, Russia. The giant
open pit was originally
mined for emerald and
then beryllium. Now
the mine operation is
underground below the
processing facility (the
green building on the
other side of the pit).
Photo by Tao Hsu.

164 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2020

Figure 19. A collection of Russian emerald offered by
Dudley Blauwet Gems. These stones range from
about 0.1 to 3 ct. Photo by Tao Hsu.

At this year’s show, the authors found two exhibitors

with Russian emerald ranging in quality from commercial
to fine without any treatment. Dudley Blauwet Gems of-
Figure 21. This bead strand is from Pakistan (9 mm
fered Russian emeralds from 0.1 to about 3 ct at the AGTA
beads), while the bead bracelet is from Siberia (20
show. The light to slightly dark green stones were offered
mm beads). Green nephrite from Pakistan has a very
as singles, pairs, and sets (see figure 19 and figure 5 on p.
similar appearance to material from British Colum-
158). Worth noting is that most of the Russian stones the
bia. The Siberian beads are much brighter. Photo by
authors saw showed bright colors and high clarity.
Kevin Schumacher.
At the GJX show, Tsarina Jewels offered Russian emer-
alds of larger sizes and fine quality. This exhibitor carried
stones as large as 8 ct or more (figure 20). These stones also
displayed a wide range of various shades of green colors. The authors noticed that Russian emeralds are still quite
rare to find on the market, while emeralds from Colombia
and Zambia dominate the market.
Figure 20. A selection of fine Russian emerald from Tao Hsu and Jennifer Stone-Sundberg
Tsarina Jewels ranging from 1.27 to 8.38 ct. All
stones are natural with no filling. Photo by Kevin
Schumacher. Nephrite jade from multiple sources. Green nephrite de-
posits are found at many locations around the globe, with
British Columbia, Siberia, and northwestern China as the
leading producers. Old and still active subduction zones
with ophiolite have the potential to produce green
nephrite. This is because green nephrite bodies are closely
associated with their serpentinite host rocks, part of the
ophiolite exposed on land when an ocean basin closes. All
major green nephrite deposits cluster in this type of con-
vergent-margin environment.
At the 2020 GJX show, Nikki Makepeace from Jade
West in British Columbia showed the authors nephrite
pieces from Pakistan, an emerging green nephrite source.
The green to dark green appearance and the quality of these
pieces (such as the bead strand in figure 21) are very similar
to the production from British Columbia. In comparison,
green nephrite from Siberia shows brighter colors (the
bracelet in figure 21 and the pendant in figure 22). Ms.

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2020 165

 Nigerian gems and jewelry. During the JCK show in Tucson,
GIA had the opportunity to meet up with Amina Egwuatu,
founder of Mina Stones. Mrs. Egwuatu, a GIA Graduate
Gemologist from Abuja, Nigeria, started the company with
the aim of promoting African culture through gems and jew-
elry. Mina Stones sources, cuts, designs, and manufactures
everything in their workshops to keep as much value as pos-
sible in Africa. Mrs. Egwuatu works directly with the miners
and has visited many of the sites where her stones are pro-
duced. This allows traceability throughout the entire supply
chain. Mina Stones helped to establish the AGJES Sparkle
Foundation, which supports miners through education
about their stones, helps them with administration, and pro-
vides tools. By working this way, Mina Stones improves the
conditions of its miners and suppliers.
Nigeria is known for tourmalines that possess a wide
range of colors, rubellite being the most famous. A more
recent rubellite discovery near Calabar, Cross River State,
has produced higher volumes (though slightly lower qual-
ity) than the original find in Oyo State, which has not pro-
duced for a long time.
Apart from the red variety, this gem is also found in
Figure 22. High-quality nephrite from Siberia can green, pink, and various combinations of these colors.
sometimes show a color similar to that of green About 20 years ago, there was even a discovery of cuprian
jadeite. Photo by Tao Hsu. tourmaline in western Nigeria that produced some fine
material for a short period. According to various sources, a
new pocket was discovered in early 2020 that made this
material available in the market.
Makepeace informed the authors that the production from
Another common stone from Nigeria is beryl. Various
Pakistan is quite substantial.
regions in the country produce fine aquamarine, morgan-
The earliest report on Pakistan nephrite was published
ite, heliodor, and green beryl (figure 23). While there are
in 1963, with the positive identification of two pebbles col-
many emerald deposits known throughout the country,
lected in 1955 in the riverbed of the Teri Toi in Kohat Dis-
facet-grade material is limited to an area in Nasarawa State.
trict of Pakistan (B.C.M. Butler, “Nephrite jade in West
Sapphires are found throughout the eastern and cen-
Pakistan,” The Journal of the Royal Asiatic Society of
tral parts of Nigeria. In several areas such as Antang,
Great Britain and Ireland, No. 3/4, 1963, pp. 130–139). The
Bauchi, Kaduna, and Gombe, the production is exclu-
geographical location of this deposit placed it in the colli-
sively from artisanal miners. The better-known Mambilla
sion belt along the Himalaya Mountains. This is the con-
vergent margin between the Eurasian and Indian Plate
when the Tethys Ocean closed about 50 million years ago,
providing an ideal geological environment for nephrite for- Figure 23. Strands of Nigerian beryl in assorted colors.
mation. Photo by Kevin Schumacher.
In addition to the goods from Pakistan, the authors saw
a large koru, a spiral shape often used in Maori art, made
of nephrite from Afghanistan. Currently, Afghani materials
are scarce on the international market. The only informa-
tion the authors could find was that the rough comes from
the Kohi-Safi District of central Parwan Province. This lo-
cation is not far from the Pakistani deposit.
While buyers are still waiting to see more production
from these emerging sources, green nephrite from British
Columbia and Siberia still dominate the market. Siberian
goods in general demand higher prices than materials from
the rest of the world due to their brighter colors. British
Columbian production is still high and can supply finished
goods of all price ranges.
Tao Hsu and Jennifer Stone-Sundberg

166 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2020

 Figure 25. This 16.23 mm, near-round 28.38 ct natural
freshwater pearl was found along the banks of the
Cumberland River in 2019. Photo by Emily Lane;
Figure 24. This necklace makes use of the Arewa courtesy of Gina Latendresse.
knot, a symbol of unity. Photo by Kevin Schumacher.

Mina Stones aims to represent Nigerian gems and jew-

Plateau has a few larger operations, but small-scale min- elry. The country is an important supplier of many colored
ing is prevalent. gemstones that have remained largely unknown to the gen-
Nigerian sapphires are typically dark blue, although eral public.
goods found in the southeastern part of the country tend
to be lighter in color. Parti-colored, yellow, purplish, and Wim Vertriest
green sapphires are also common. GIA, Bangkok
Mina Stones’ jewelry is inspired by Nigerian traditions.
The country has long been known for its blacksmiths and Exceptional natural freshwater pearl. At the AGTA show,
goldsmiths, resulting in a strong jewelry culture. Mina we visited Gina Latendresse at the American Pearl Com-
Stones is trying to revive the skills, patterns, and heritage pany of Tennessee booth to find out about anything new
by making use of traditional alphabets and symbols. One in the pearl industry. We were excited to learn that an ex-
example is a necklace featuring the Arewa knot (figure 24), ceptional pearl, both in size and quality, was discovered
a symbol that represents the unity of the different people last year in the Cumberland River (figure 25). The pearl was
in northern Nigeria. found by a fisherman along the bank of the river in a “Pis-
Another collection draws inspiration from the city of tol Grip” (Tritogonia verrucosa) mussel (figure 26). This
Ife, in southwestern Nigeria. According to local beliefs, the 28.38 ct white pearl with natural color measured 16.23 mm
city was found by a supreme deity and served as the seat in diameter and was near round with good luster. This was
for a long line of dynasties. the first pearl of such size and quality found in the Ten-

Figure 26. The natural

pearl from figure 25
shown in its shell
(left), and the “Pistol
Grip” shell exterior
(right). Photos by
Emily Lane; courtesy
of Gina Latendresse.

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2020 167

 Figure 27. Bicolor sap-
phires from the Rock
Creek mine, Montana.
Cutters took advantage
of the color inhomo-
geneity of the rough to
create these interesting
color combinations.
These stones range
from 1.61 to 7.36 ct.
Photo by Emily Lane;
courtesy of Potentate
Mining.

nessee region since the 1970s. The pearl is so unusual that Montana. Stones from this deposit tend to be flat, showing
Ms. Latendresse described it as either a collector specimen pastel colors, and many stones show bicolor or have an or-
or suitable for museum display. ange or yellow core called “yolk” by the local miners. Gem
For context, the eastern portion of the United States cutters and jewelry designers have been finding innovative
was a source of many natural freshwater pearls to native ways to take advantage of these challenging attributes.
Americans as evidenced by John Smith’s 1608 writings not- Through careful design and cutting, some stones show at-
ing the “many chains of great pearls about his [Chief tractive color combinations such as yellow/green,
Powhatan of the Powhatan Confederacy near present day yellow/light blue, orange/blue, or simply different shades
Richmond, Virginia, also the father of Pocahontas] neck” of yellow (figure 27). This type of bicolor is very character-
(P.L. Barbour, Ed., The Complete Works of John Smith istic of Rock Creek sapphires.
(1580–1631), Vol. 1, University of North Carolina Press, As for the flat rough, Potentate offers several categories
Chapel Hill, 1983, p. 53). Two and a half centuries later, a based on color, pattern, and weight (figure 28). According
pearl rush started with the discovery of pearls of notable to the company’s marketing director, Warren Boyd, both
size and quality by fishermen in New Jersey, which re- cutters and designers have shown strong interest in these
sulted in people searching streams and rivers throughout slabs, especially those weighing half a gram and above. De-
the country. In 1857, David Howell found a near-round 100 signers seem to prefer slabs with patterns while cutters or
ct pearl in a fried mussel he fished from Notch Brook in carvers prefer slabs with solid colors to make special cuts.
New Jersey (which unfortunately had its luster destroyed
by the frying process). Shortly thereafter, Jacob Quacken-
bush discovered a 23.34 ct pink baroque pearl of fine luster,
later named the “Queen Pearl,” that he sold to Charles Figure 28. Flat rough sapphires from Rock Creek at-
Tiffany for $1,500. Good-quality pearls were found in the tract jewelry designers and cutters. The slabs are
Cumberland, Tennessee, and Clinch Rivers in the state of grouped and sold by their weight. Photo by Tao Hsu.
Tennessee for many years thereafter (G.F. Kunz and C.H.
Stevenson, The Book of the Pearl: The History, Art, Sci-
ence, and Industry of the Queen of Gems, The Century
Co., New York, 1908).
The recent discovery of such an extraordinary pearl
proves that it is prudent to check the inside of mollusks
known to produce pearls before consuming them!
Jennifer Stone-Sundberg and Tao Hsu

Interesting sapphires from Montana and Australia. While

sapphire has always been popular at trade shows, occasion-
ally the authors find some interesting stones with attrac-
tive color distribution or growth patterns that have a niche
in the market.
At this year’s AGTA show, Potentate Mining carried
some spectacular bicolor faceted sapphires and a variety of
slabs from its Rock Creek mine in Montana. The Rock
Creek operation is one of four active sapphire mines in

168 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2020

 2014, Mr. Hapeman has been focusing on Rock Creek ma-
terial in these nontraditional lighter blues to greens. They
have proven to be very popular, as several AGTA vendors
this year reported strong demand for these colors, not only
in Montana sapphire but from other sources as well. When
it comes to unheated material, only about 5% of the rough
from Rock Creek is cut without treatment, and most of
these are either fancy pinks, yellows, or lighter blues. In the
past year Mr. Hapeman has seen a shift in demand to parti-
colored and green sapphires, particularly deep bluish green.
Jennifer Stone-Sundberg, Tao Hsu, and Robert Weldon

New blue sapphire from Rakwana, Sri Lanka. The town of

Rakwana, in southern Sri Lanka, has been known for
decades as a premier source of royal blue sapphires. Often,
a simple mention that a gemstone is from Rakwana will
immediately explain the price and quality. In August 2019

Figure 30. The 17.65 ct “Lucky Sapphire” rough crys-

tal and 3.92 ct oval cut using the custom “Helena”
Figure 29. This pair of blue sapphire thin slabs dis- design, both from Rock Creek, Montana. Photo by
plays a very attractive color. They were sliced off a Robert Weldon/GIA; courtesy of Jeffrey Hapeman.
dark blue sapphire crystal from Inverell, Australia.
Mr. Coldham also offered thin slabs of blue and parti-
colored sapphires from Australia. Photo by Tao Hsu.

At the GJX show, Terry Coldham from Intogems

showed the authors a pair of thin blue sapphire slabs with
characteristic hexagonal growth patterns. The pair in figure
29 is from the Inverell sapphire field of New South Wales,
Australia. Blue sapphires from this area tend to show a very
saturated and dark inky blue color. Slicing them into thin
slabs makes the blue color much lighter and more desir-
able. However, the size and quality of this pair are ex-
tremely rare to find. Mr. Coldham also carries many
parti-colored and fancy-color Australian sapphires (see fig-
ure 4 on p. 157), which also sold well at the show.
To watch videos of the sapphire slabs from Montana
and Australia, go to https://www.gia.edu/gems-gemology/
spring-2020-gemnews-sapphires-montana-australia.
Tao Hsu and Jennifer Stone-Sundberg

Exceptional rough and cut blue sapphire from Rock Creek,

Montana. The Rock Creek sapphire deposit in Montana has
been active since the 1890s and is known as the state’s most
fruitful deposit (T. Hsu et al., “Rock Creek Montana sap-
phires: A new age of mining begins,” https://www.gia.edu/
gia-news-research/rock-creek-montana-sapphires-new-age-
mining-begins, August 29, 2016). This year in Tucson, Jef-
frey Hapeman of Earth’s Treasury showed the authors a very
fine untreated blue sapphire rough with a yellow spot in the
center, as well as a top-quality faceted light teal blue un-
treated stone with some yellow zoning (figure 30). Since

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2020 169

 Figure 31. A 49.99 ct
rough sapphire crystal
(5.8 cm in length) and
a 33.16 ct cut stone
with exceptional color
from the 2019 Rak-
wana find. Photo by
Robert Weldon/GIA.

a small pocket yielding numerous large sapphire crystals Trapiche gems. Jeffery Bergman is known in the trade as a
was discovered in the nearby Sinharaja Forest area of the collector of trapiche gems. These gems, characterized by
Ratnapura District, Sabaragamuwa Province. What was their six-rayed patterns, are elusive and seldom found in
unique about this find was the quality and size of the lus- high quality. Mr. Bergman believes people are attracted to
trous crystals. The 49.99 ct rough crystal in figure 31, the symmetrical pattern, which reflects a natural order in
measuring 5.8 cm in length and 1.6 cm at its widest point, the chaotic world we live in. According to Bergman, trapiche
was one of about 60 found in August 2019. It was also the gems are too rare to support a business, so he has also traded
largest complete crystal found. The beautiful 33.16 ct gem many other gemstones during his 50-year career.
beside it was cut from a 207 ct piece of rough, the end of a As with most gemologists, the first trapiche gems he
large portion of a crystal. This crystal and cut gem rival the saw were Colombian emeralds. But it wasn’t until 10 years
famous ones from Kataragama found four years ago. after this initial encounter at the Tucson gem show that
Jennifer Stone-Sundberg, Tao Hsu, and Robert Weldon he purchased his first trapiche gemstone: a thick Colom-
Dudley Blauwet bian emerald weighing about 20 ct. Mr. Bergman sliced it
Dudley Blauwet Gems in two, increasing the value after it became a matched pair
Louisville, Colorado of acceptable thickness.

170 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2020

 Trapiche emeralds from Colombia are well known, with
the Muzo mine producing the majority. The patterns can
exhibit many variations, from perfect six-ray symmetry
with high clarity to cloudy stones with broken or incom-
plete rays. Trapiche-like emeralds have recently been found
in Swat, Pakistan, but they show a very different appear-
ance and formation (Fall 2019 GNI, pp. 441–442). Beryls
such as aquamarine and bixbite (red beryl) are known to ex-
hibit trapiche or trapiche-like characteristics but they are
even rarer than emeralds.
Trapiche corundum gems are mostly known from
Myanmar. Trapiche ruby from Mong Hsu is often small.
In his 20 years of trapiche experience, Mr. Bergman has
only encountered two gem trapiche rubies from Mong Hsu
that were over 1 carat. The material is fairly dark, which
Figure 32. A pair of trapiche emerald slices from means it must be sliced thin to appreciate any color, thus
Colombia, 10.92 carats total. Photo by Kevin Schu- reducing the weight of the stone drastically. Heat treat-
macher; courtesy of Mayer & Watt. ment rarely improves the appearance; it alters the trapiche
pattern, making it chalkier and less pronounced.
The Tajik ruby mines have also produced trapiche ru-
bies, as well as the mines in Nepal. Batakundi (Pakistan)
He saw his first non-emerald trapiche gems at the Mae and Orissa (India) are other sources of ruby with six-rayed
Sot gem market on the Thai-Burmese border in the early patterns (Spring 2019 G&G Micro-World, p. 114).
2000s. Mr. Bergman was handed three Burmese trapiche sap- Trapiche sapphires from Mogok (figure 33) are often
phires from Mogok, something he never knew existed. Other dominated by gray tones or very saturated blues bordering
collectors were equally surprised and paid high amounts for black colors. Only the finest stones show a snow-white
these stones, encouraging Mr. Bergman to expand his collec- bodycolor with vivid blue arms. To date, these are the only
tion. During those days, many fine Burmese trapiche rubies true trapiche sapphires. Trapiche-like sapphires are much
came out of Mong Hsu while Mogok provided the trapiche more common from basalt-related sources such as southern
sapphires. Vietnam, Australia, and even lesser-known areas such as
Trapiche rough is extremely difficult to judge since the Scotland.
patterns often do not extend throughout the entire stone. Quartzes, including amethyst and smoky quartz, can
In most cases, the most desirable pattern can only be found exhibit trapiche patterns. A pocket of Zambian tourma-
in the middle of the crystal length. lines and some extremely rare garnets also show a six-
Trapiche and non-trapiche emerald evaluation are very rayed pattern.
similar, as the most important characteristics are always In Mr. Bergman’s opinion, trapiche only applies when
the color and transparency of the emerald itself (figure 32). the symmetry is sixfold. This excludes more common ma-
The symmetry and shape of the pattern are only of second- terials with similar growth patterns like chiastolite. As-
ary importance, followed by size and potential treatments. teriated diamonds, most often found in Zimbabwe,

Figure 33. Trapiche sap-

phire from Mogok,
Myanmar. This stone
shows a regular pattern
of snow white and blue
zones. Photos by Kevin
Schumacher.

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2020 171

Figure 34. These miniature opal carvings are favored by designers. They are hand-carved in Bali from different
varieties of Australian opal. Photos by Tao Hsu and Jennifer Stone-Sundberg.

resemble trapiche when viewed in certain angles, but in Another exhibitor that caught the author’s eye is Susan
reality they contain clouds extending in three dimensions Tereba. Ms. Tereba moved to Bali over 30 years ago and
at 90º angles. still lives there today. She brought to the show amazing
In the last decades, knowledge and appreciation of these hand-carved mammoth tusk and bovine bone. Ms. Tereba
unique stones has spread and they are now actively sought appreciates the remarkable talent of the Balinese carvers,
after by collectors, bespoke designers, and gem enthusiasts. who are fast learners and can easily grasp themes that are
This increased demand and availability has driven gemol- not indigenous. These exquisite carvings also show a wide
ogists to take a deeper look at these gems and their forma- variety of motifs, with animals and goddesses the most
tion, but there remains much to be discovered. popular and some pieces are gem studded (figure 35). The
Wim Vertriest bovine bone carvings are dyed, and according to Ms. Tereba

CUTS AND CUTTING Figure 35. Balinese carved mammoth tusk of a garden
goddess. Photo by Tao Hsu.
Carvings from Bali. Picturesque beaches and lush green
jungles are not the only attractions that make Bali the most
popular tourist destination in Indonesia. The local carving
tradition, derived mostly from woodcarving, adds a layer
of artistry and craftsmanship to this tropical paradise. Now
many carvers there are also involved in different types of
gem materials.
At the GJX show, the author found some authentic and
attractive Balinese carvings of Australian opal, mammoth
tusk, and bovine bone. Mr. Terry Coldham from Intogems
brought opal carvings of pendant sizes, which are highly de-
sired by jewelry designers (figure 34). Mr. Coldham utilizes
rough from different Australian mining areas, primarily
black and white opal, but some crystal opal as well. The
carvings feature natural themes such as plants, flowers, and
animals, as well as profiles of faces from different cultures.
The carving style includes both relief panel and three-di-
mensional sculpture. Some pieces show color blocks being
skillfully arranged, a technique commonly used by the Chi-
nese in jade carving. The quality of the rough dictates the
price, ranging from hundreds to thousands of dollars for
each finished piece. According to Mr. Coldham, opal carv-
ings have sold well over the past several years. He enjoys
working with the Balinese carvers and added that after a
couple of years of collaboration, the products have im-
proved to better fit the international market.

172 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2020

 Figure 36. Bovine bone
carvings are usually
dyed to show vivid col-
ors. The color is stable
over many years. Photo
by Tao Hsu.

the color is very durable and resistant to wear (figure 36). Nick Alexander (figure 37, left) of Alexanders Jewelers
Jewelry designers and hobbyists are her main clients. in Gilbert, Arizona, was exhibiting his work for the first
To watch videos of Terry Coldham and Susan Tereba time at the AGTA show. This year, Mr. Alexander won
displaying Balinese carvings, go to https://www.gia.edu/ second place in the carving category of the AGTA Cutting
gems-gemology/spring-2020-gemnews-carvings-bali. Edge Awards with his 42.05 ct Oregon sunstone piece at
Tao Hsu the impressive age of 17. A favorite material of his is elec-
tric blue gem silica from the Ray mine in Arizona (see fig-
Gem carvings, fantasy cuts, and master recutting. This ure 2, p. 157). Mr. Alexander is also skilled at fantasy
year at the AGTA show, the number of booths selling cutting with materials such as amethyst (figure 37, right)
skillfully carved, fantasy cut, or expertly recut items was and beryl.
noticeably higher than in past years. These forms of work- One of the carvers who has inspired Mr. Alexander,
ing rough or previously cut material are a way to add sig- Glenn Lehrer of Larkspur, California (figure 38, right), was
nificant value and desirability to material that might also at the AGTA show. Mr. Lehrer shared with us one of
otherwise be cut into standard shapes or with the intent his newest works, an exquisite lotus flower carved out of
to maximize weight at the expense of beauty. a top-quality rose quartz from Madagascar (figure 38, left).

Figure 37. Left: Nick

Alexander exhibits a
selection of his work.
Photo by Jennifer
Stone-Sundberg. Right:
A 64 ct fantasy cut
amethyst from Bahia,
Brazil. Photo by Kevin
Schumacher; courtesy
of Nick Alexander,
Alexanders Jewelers.

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2020 173

 Figure 38. A 322.72 ct
Madagascar rose
quartz lotus flower
carving with grayish
blue and padparadscha
Montana sapphires in
the center (5.56 and
0.65 ct, respectively),
shown by the artist,
Glenn Lehrer. Left
photo by Robert Wel-
don/GIA, courtesy of
Glenn Lehrer; right
photo by Jennifer
Stone-Sundberg.

This piece started as part of a museum collection project phenomenal gem, a natural color-change cat’s-eye alexan-
with designer Paula Crevoshay. It took two years to find drite (figure 39, right) that was polished to a double-sided
the 3.8 kg rose quartz rough meeting his stringent de- cabochon to display the cat’s-eye effect on both sides of the
mands: It had to be clean, clear, and with strong color. stone. For videos of David Nassi displaying his work, go to
The finished carving measures 79.5 × 69.8 mm and https://www.gia.edu/gems-gemology/spring-2020-
weighs 322.72 ct. To complement the piece, the center gemnews-carvings-fantasy-cuts-recutting.
features two sapphires from the Rock Creek area of Mon- Jennifer Stone-Sundberg, Tao Hsu, and Robert Weldon
tana: a 5.56 ct grayish blue “torus ring” carving and a 0.65
ct padparadscha.
Master gem cutter David Nassi from New York City JEWELRY DESIGN
(figure 39, left) showed us an impressive set of unheated nat-
ural spinels and several other gems he had recut to maxi- Fine-quality jadeite jewelry. At this year’s Pueblo Gem and
mize color and light return. Of particular interest was a Mineral Show, the authors met Frank Lau from Frank Lau

Figure 39. Left: David

Nassi of 100% Natural,
Ltd. showing a collec-
tion of unheated
spinels he has recut,
ranging in size from
7.10 to 18.53 ct. Photo
by Jennifer Stone-Sund-
berg. Right: A 14.49 ct
cat’s-eye alexandrite
from Sri Lanka cut by
Nassi, showing reddish
brown to green color
change. Photos by
Emily Lane; courtesy of
David Nassi.

174 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2020

 Figure 40. A selection
of jewelry offered at
Frank Lau’s booth.
Jadeite of fine quality
is not commonly seen
in Tucson. Photo by
Tao Hsu.

Jewelry. What caught the authors’ attention was the collec- air to oxidize the Fe-containing minerals to make the stone
tion of fine-quality jadeite jewelry, which is predominantly look red. This treatment is extremely hard to identify.
found in the mainland China market but not in Tucson. The colorless jadeite pieces, also called “ice jade,” are
Mr. Lau’s booth offered a wide variety of jadeite includ- of very fine quality. This variety is popular with younger
ing green, colorless, lavender, and red ranging in quality consumers. The exceptionally fine texture and relatively
from medium to fine (figure 40). The jadeites are polished high transparency make it an alternative to the top-quality
and carved in Guangzhou, China, while the jewelry is imperial green stones. When inclusions are present in the
manufactured either in China or Seattle, where the store colorless jadeite, sellers describe them as “snowflakes” to
is located. Clients of the jadeite jewelry include both Chi- attract buyers of different tastes (figure 42).
nese and Americans, with many repeat customers.
While all of the jadeites are natural, Mr. Lau informed
the authors that the reddish brown pendant is heated (figure Figure 42. This colorless jadeite pendant is of very
40). Reddish brown jadeite is extremely rare. This color is high transparency. Many sellers describe the whitish
produced from a very thin layer of jadeite boulder (figure 41) inclusions as “snowflakes.” Photo by Tao Hsu.
leaving little material to make pieces of reasonable sizes. A
good amount of reddish jadeite on the market is heated in

Figure 41. A rough jadeite boulder shows a very thin

layer of reddish brown skin. Red jadeite is even rarer
than the green variety. This boulder was offered at
the Myanmar Gem Emporium. Photo by Tao Hsu.

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2020 175

Figure 43. A cordierite Viking “sunstone” set in silver Figure 44. Examples of contemporary Scandinavian
with the “vegvísir” compass symbol. Note that Mr. designs in silver containing thulite. Photo by Jennifer
Berger’s ring also features a cordierite cabochon set in Stone-Sundberg; courtesy of Atle Berger.
silver. Photo by Jennifer Stone-Sundberg.

Mr. Lau sources his rough materials from Myanmar, Greenland (D. Száz and G. Horváth, “Success of sky-po-
though the skyrocketing price of rough makes this busi- larimetric Viking navigation: revealing the chance Viking
ness increasingly difficult. He looks forward to the avail- sailors could reach Greenland from Norway,” Royal Soci-
ability of more consumer-oriented jadeite education so ety Open Science, Vol. 5, 2018, No. 172187). An example
Western consumers can better appreciate this gem. of a silver pendant engraved with a vegvísir (a Nordic com-
Tao Hsu and Jennifer Stone-Sundberg pass “wayfinder” symbol) and a cordierite center stone was
shown to us by Atle Berger (figure 43).
Thulite, a manganese-containing pink variety of zoisite
Nordic gems and jewelry. At the JOGS show, Arctic Jew- with some white calcite mottling, is the national gemstone
elry (Axvalla, Sweden) featured a variety of Nordic gemmy of Norway, the country in which it was discovered by An-
materials set mainly in silver using ancient, traditional, ders Gustaf Ekeberg in 1820. The name comes from
and contemporary Scandinavian designs. The pieces spoke “Thule,” the ancient name for the mythical island (be-
to the mineral diversity, history, and artistry of the region. lieved to be modern-day Norway) that was considered the
The Vikings made navigation stones from locally northernmost part of the world. At the Arctic Jewelry
sourced transparent crystals of minerals such as cordierite, booth, contemporary Scandinavian designs in silver incor-
a biaxial magnesium iron aluminum silicate, to help cross porating this stone were featured in jewelry ranging from
the seas under cloudy weather. These Viking “sunstones” rings to bracelets and necklaces (figure 44).
are backed up by science, as cordierite and other materials A bright and attractive blue material found at the booth
such as calcite and tourmaline can be used to identify the is the slag byproduct of iron smelting in the Bergslagen re-
position of the sun through even thick clouds. These gion of central Sweden during the Middle Ages. This ma-
stones visibly split sunlight into two images that when ro- terial was reported on previously in G&G (Winter 2006
tated to make equally bright, show rings of polarized light GNI, p. 279) and characterized in the GIA lab. At that time,
around the sun’s position. Cordierite was found to be the EDXRF was the method used for chemical analysis, but
most accurate navigation stone in simulated journeys from today with LA-ICP-MS we were able to more exactly de-
Bergen, Norway, to the Viking settlement of Hyarf in termine the composition of this material from some rough

176 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2020

 and Anza Gems. They share a commitment to supply
chain ethics, mine-to-market traceability, environmental
and health and safety regulations, and supporting artisanal
miners and communities.
Monica Stephenson, founder and president of Seattle-
based Anza Gems, said her travels to East Africa to pur-
chase rough have shown her that miners need to be
educated about the value they bring to the supply chain.
Ten percent of Anza’s proceeds go to education at the min-
ing communities.
Anza Gems is a partner of Moyo Gemstones, a respon-
sibly sourced gemstone program that was sparked by GIA’s
artisanal mining education program. The pilot effort, a col-
laboration between GIA, Pact, and the Tanzania Women
Miners Association (TAWOMA), focused on women min-
ers in Tanga, Tanzania. After initial training, they were
more knowledgeable about the mined gemstones but
lacked a platform to bring them to the market. Stephenson
saw potential here, and together with Pact and TAWOMA
she founded Moyo Gemstones. The program provides a
marketplace for gemstone transactions, offers free occupa-
tional health and safety training, and helps ensure that all
miners are legally registered.
Stephenson said that before the program these miners
Figure 45. Rough “Swedish Blue” slag byproduct. would get a few dollars for gems from a broker, with no
Photo by Kevin Schumacher; courtesy of Atle Berger. visibility into their value in the supply chain. Moyo has
changed this through education. “For the miner to under-
stand what she has and the value of those gems, and also
to be taking home more like 95% of the export price, and
pieces given to us by Mr. Berger (figure 45). We identified pay her broker, is a complete reversal of the roles,”
it as a silica-rich glass with a composition of 56.35 wt.% Stephenson said. “This is truly a paradigm shift in this re-
SiO2, 24.70 wt.% CaO, 7.68 wt.% MgO, 5.05 wt.% Al2O3, gion. For her to have the income and financial sort of inde-
2.23 wt.% FeO, 2.21 wt.% K2O, 0.85 wt.% MnO, 0.27 pendence is enormous.”
wt.% Na2O, and 0.65 wt.% other elements (average of 10 Each gem can be traced back to the miner, Stephenson
ICP spot analyses). This gave a calculated chemical for- said, and it’s important to share that story with the con-
mula of (Na0.019, Mg0.418, Al0.217, K0.103, Mn0.026, Fe2+0.068, oth- sumer. “They need to know that what they’re buying not
ers0.012)0.864 Si2.058 Ca0.967O6, which is generally consistent only makes a difference, and not only does it not cause
with what was reported in 2006. The oxide components harm, but it actually is beneficial.”
are fairly normal for iron foundry slag, but the silica-rich Anza Gems’ offerings (figure 46) included sapphire,
nature is what gives this material its fine glassy aspect multiple colors of garnet, tourmaline, citrine, amethyst,
compared to those slags that are more CaO-rich. The zircon, and Mahenge spinel.
slightly greenish blue color comes from Fe2+ (FeO). Brian Cook, owner of Nature’s Geometry (Tucson),
Jennifer Stone-Sundberg and Ziyin Sun began selling crystals as a geology student almost 30 years
ago. This took him to a remote area of Bahia, Brazil, where
rutilated quartz is mined. He felt a strong connection to
RESPONSIBLE PRACTICES the place and made it his second home. Since then he has
advocated for the development of resources more sustain-
Ethical Gem Fair. Tucson’s first Ethical Gem Fair, a plat- able than mining so the community can rely on them
form for responsibly sourced gemstone suppliers to high- when mining ends.
light their products and projects, was held February 3–6 at Cook told us about their project in Bahia to benefit the
the Scottish Rite Cathedral. This was the first Ethical Gem artisanal mining community. The first initiative will raise
Fair in the United States after being held in London and funds for dust masks for the miners, at least half of whom
Edinburgh the last three years. have no dust protection for underground mining. Second, he
The Tucson event was a collaboration between suppli- envisions a model of investment in regenerative agriculture
ers Nineteen48, Capricorn Gems, Perpetuum Jewels, Na- to create food security, improve nutrition, and implement
ture’s Geometry, Agere Treasures, Columbia Gem House, carbon sequestration. The area’s soil is so rich that it allows

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2020 177

 Figure 46. A selection
of gemstones from
Anza Gems. All of
these stones were
sourced in Tanzania,
including the set of
rough sapphires that
was mined in Tanga
province. Each stone
can be traced back to
the individual miner
who unearthed the
gem. Photo by Kevin
Schumacher.

for farming of crops such as cacao, coffee, acai, and many ture’s Geometry has been cutting the material since 1989.
others. The third component will be teaching lapidary skills. Each piece is unique, which Cook said is something the
Eighty percent of colored gemstones are mined by small- younger generation likes.
scale and artisanal miners, according to Cook. “This project Cook said the response to the fair has been “phenome-
could serve as a model for our industry at large,” he said. nal.” He predicts it will grow in future years, with more
While he recognizes that every situation is different, food, vendors and a greater variety of stones.
water, and energy will always be the basic needs of every Ian Bone, manager of Capricorn Gems (Central Queens-
being, and guaranteeing them can uplift a community. land, Australia) said the Ethical Gem Fair was created to
Cook showed us several pieces of rutilated quartz from meet the needs of designers and customers who want a
Bahia, including a sizable piece of rough (figure 47). Na- gemstone’s mine-to-market story. “There’s undoubtedly a
generational shift happening,” he said. “Customers want
to know that the pieces are brought to market in the most
Figure 47. Rough rutilated quartz from Bahia, Brazil. responsible manner possible. We see this as the cutting
Photo by Kevin Schumacher. edge of jewelry design and in fact the future of the jewelry
industry.”
Bone said that because Australia’s mining industry al-
ready has strong regulations for safety, the environment,
energy, and land reclamation, it could serve as an example
for other mining communities around the world.
Bone, a Central Queensland native, has built his busi-
ness around local gemstones: boulder opal, chrysoprase, zir-
con, and sapphire. He is able to witness the entire process,
from mining to cutting to consumer sales. He showed us a
selection of chrysoprase from Marlborough, Central
Queensland (figure 48). “Marlborough has some of the best
chrysoprase in the world,” he said. “Chrysoprase is an un-
usual gem because you can get a whole set of qualities of
that stone, from dollars a pound up to dollars a carat.”
Bone said Capricorn Gems is breaking down bound-
aries between sourcing, production, and sales. They share
their images, videos, and stories of mining with designers
for use in their own marketing. One example is a booklet
showcasing the works of designers around the world who

178 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2020

 Figure 48. Capricorn
Gems’ chrysoprase from
Marlborough, Central
Queensland. Photo by
Kevin Schumacher.

use stones from Capricorn Gems. “The client knows that At the Denver Gem & Mineral Show, Zewdi’s first
their designer knows and trusts the gem supplier and trade show, she saw Ethiopian opal dealers undervaluing
knows where that material comes from,” he said. their material and advised them to increase their prices.
Hewan Zewdi, founder of Agere Treasures (Lynnwood, She also began buying stones from them to sell in her local
Washington), was born and raised in Ethiopia and moved market in Washington. Her commitment increased when
to Seattle 14 years ago. There she earned her G.G. through emeralds were discovered in southern Ethiopia. Agere
GIA Distance Education. When Ethiopian opal (figure 49) Treasures now offers opal, emerald, amazonite, yellow
was discovered in 2008–2009, people began calling from labradorite, garnet, zircon, tourmaline, and sapphire.
home to ask if she had a market for them. “That’s what I Part of Zewdi’s profits go to a new project to train
study—that’s what I want to do,” she told them. young women in basic gemology and jewelry design. She
also donates books on rocks and gems to the Ethiopian
Ministry of Mining, and she assists the organization in ob-
Figure 49. Ethiopian opal from Agere Treasures. Photo taining gemstone identification at GIA so the stones can
by Kevin Schumacher. be priced accordingly by exporters in Ethiopia.
“Here everybody has the same mission, the same vi-
sion,” Zewdi said of the Ethical Gem Fair. “To give back
and get the material traceably and responsibly, without
hurting the environment or the artisanal miner, by creating
a fair trade and a fair price.”
Buyers at the Ethical Gem Fair were younger people and
designers looking for a stone with a story. They want to
know about the supply chain for almost every product they
use, including jewelry. The show’s suppliers offer as much
transparency as possible to their customers, often providing
supporting information about the stones, such as images
from the mines and cutting factories. Everyone involved in
the show sees a growing demand for responsibly sourced
gemstones.
Articles on two other exhibitors, Perpetuum Jewels and
Columbia Gem House, can be found in the Spring 2019
G&G.
Wim Vertriest and Erin Hogarth

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2020 179

 and popular (figure 50). This material is grown by flame fu-
sion, and the chromophore component of the powder com-
position is altered during the growth process at the desired
length. Many different color combinations were available,
including some with the popular teal blue and purple hues
(figure 51, left). The inspiration for applying this method
of color zoning to these flame-fusion grown crystals came
from historical work on synthetic ruby laser rods where
colorless (or “undoped”) regions were desired for the ends
of the rods (figure 52). A process to grow a single crystal
having pure aluminum oxide for a certain length, then
adding the desired level of chromium to the growth for the
laser rod portion of the length, and then returning to pure
aluminum oxide was developed to achieve this.
We took the step-cut bicolor synthetic sapphire in figure
51 back to the GIA lab to determine the trace element
chemistry responsible for generating the teal and purple
hues. Using laser ablation–inductively coupled plasma–
mass spectrometry (LA-ICP-MS), we found that the teal end
contained no Be, Mg, Ti, V, Mn, Fe, or Ni but did contain
small amounts of Cr and Ga (approximately 4 ppma and 0.5
ppma, respectively) and a substantial amount of Co (approx-
imately 140 ppma). The purple end similarly contained no
Be, Mg, V, Mn, Fe, or Ni but did contain small amounts of
Figure 50. Split boules of bicolor synthetic sapphire
Ti (0.6 ppma) and Ga (0.4 ppma), with substantial amounts
grown by flame fusion. Photo by Jennifer Stone-Sund-
of Co (120 ppma), and Cr (470 ppma). Cobalt is giving this
berg; courtesy of RusGems.
crystal the greenish blue component of its color, and
chromium is modifying the hue to purple. We used UV-Vis
spectroscopy to detect absorption peaks that could be used
with the trace element chemistry to explain the two differ-
SYNTHETICS AND SIMULANTS ently colored regions. We identified chromium peaks at ap-
proximately 400 and 560 nm responsible for the red color
Bicolor synthetic sapphire. At the JOGS show, a wide se- component in the purple half. We also identified an addi-
lection of synthetic gem materials was offered by the tional peak in the purple half at around 640 nm, likely due
Bangkok office of RusGems, a synthetic crystal growth to Co3+. In the teal half, we did not find the Cr3+ peaks but
company out of Russia. Olga Tanskaia showed us their ex- did identify Co3+ peaks at 640 nm and at around 440 nm, and
tensive color range of synthetic sapphire, spinel, beryl, gar- a peak around 600 nm due to Co2+ (K. Schmetzer and A.
net, and other lab-grown crystals produced using a variety Peretti, “Characterization of a group of experimental Russ-
of crystal growth methods. She noted that blue and red syn- ian hydrothermal synthetic sapphires,” Journal of Gemmol-
thetic gems are always their best sellers. ogy, Vol. 27, No. 1, 2000, pp. 1–7).
As we found with natural sapphire this year, bicolor In inspecting the inclusions at the boundary between
synthetic sapphire, both rough and cut, was also available the two differently colored portions, we discovered bright

Figure 51. Left: Bicolor

14.83 ct synthetic sapphire
showing popular teal blue
and purple colors. Photo by
Kevin Schumacher; cour-
tesy of RusGems. Right:
Wisps of presumably
CoAl2O4 at the boundary of
the teal and purple colors
in the crystal. Photomicro-
graph by Nathan Renfro;
field of view 7.05 mm.

180 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2020

 Figure 52. Synthetic
ruby rod grown by the
flame-fusion technique
showing chromium-free
ends. Photo by Jennifer
Stone-Sundberg; cour-
tesy of RusGems.

blue wisps of what we believe to be CoAl2O4 (figure 51, who created them. It is in honor of Gianmaria Buccellati’s
right), a spinel-structure material referred to as “cobalt name that we will welcome, as our guest, Ms. Chiang to
blue” that is used to impart a blue color to glass, ceramics, Italy.”
and plastics.
The ability to generate just about any hue in sapphire,
including color zoning, was readily apparent when looking
through the offerings of RusGems. It was fascinating to see
evidence of the direct influence of the high-tech laser indus-
try on these synthetic gems.
Jennifer Stone-Sundberg, Ziyin Sun, Nichole Ahline,
and Nathan Renfro

ANNOUNCEMENTS

Third annual Gianmaria Buccellati Foundation Award

winner. Yi-Hsuan Chiang, a graduate of GIA’s Jewelry De-
sign program in Taiwan, received the third annual Gian-
maria Buccellati Foundation Award for Excellence in
Jewelry Design during the Tucson shows. More than 200
students competed this year, and Chiang was one of 18 fi-
nalists from seven GIA campuses. Her winning design, of
a necklace depicting a butterfly that has fallen into a spi-
der’s web, featured amethyst, diamond, moonstone, conch
pearls, coral, black opal, and baroque pearl (figure 53).
Chiang will travel to Italy to meet Rosa Maria Bres-
ciani Buccellati, president of the foundation, and visit its
collection.
Larry French of the foundation said, “The designs that
were selected as the 18 finalists and exhibited in Tucson
beautifully illustrate the passion and dedication of those

Figure 53. Yi-Hsuan Chiang’s amethyst, diamond,

moonstone, conch pearl, coral, black opal, and baroque
pearl necklace design won the third annual Gian-
maria Buccellati Foundation Award. Chiang is a grad-
uate of the Jewelry Design program at GIA in Taiwan.

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2020 181

 Figure 54. This exhibit
highlighted GIA’s re-
search into the origin of
several of the most fa-
mous diamonds. Photo
by McKenzie Santimer.

The 2020 Gianmaria Buccellati Foundation Award for of GIA’s Richard T. Liddicoat Gemological Library and In-
Excellence in Jewelry Design competition is underway and formation Center. In recent years he has also been closely
open to students in GIA’s Jewelry Design courses who involved in providing gemological training to independent
meet the eligibility requirements. For more information, African miners.
visit gia.edu/buccellati-foundation-award-jewelry-design.

GIA Museum’s Tucson exhibit receives two awards. The Figure 55. G&G contributor Robert Weldon received
GIA Museum received two awards for its exhibit at the 2020 the 2020 Bonanno Award for Excellence in Gemology.
Tucson Gem and Mineral Show. “The Science Behind World-
Class Diamonds” (figure 54) won Best Museum Exhibit from
the Tucson Gem & Mineral Society and Best Institutional
Educational Award from the Friends of Mineralogy. The ex-
hibit demonstrated how famous diamonds such as the Hope
and the Cullinan originated in extreme depths of the earth’s
lower mantle. The showcase featured extraordinary diamond
jewelry and diamond replicas.
More than 30 cases were displayed by renowned muse-
ums including the Smithsonian Institution, LA County
Museum of Natural History, and the Royal Ontario Mu-
seum. This is the second time GIA has won the award for
Best Museum Exhibit, having won previously in 2016.

Robert Weldon receives Bonanno Award. Longtime Gems

& Gemology author and photographer Robert Weldon (fig-
ure 55) has received the Accredited Gemologists Associa-
tion’s 2020 Antonio C. Bonanno Award for Excellence in
Gemology. The award was presented during AGA’s annual
gala in Tucson on February 5.
During his more than 30-year career, Weldon has con-
tributed scores of publications, serving as colored gemstone
editor at JCK and senior writer at Professional Jeweler. His
photography has appeared on more than 30 G&G covers,
and he has written numerous feature articles for the jour-
nal. He coauthored, with Dona Dirlam, Splendour & Sci-
ence of Pearls (2013). Since 2017, Weldon has been director

182 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2020

 nitrogen to obtain powder for analysis. Previously pub-
REGULAR FEATURES lished DNA extraction protocols for pearls were modified
to include newly designed primers for polymerase chain
reaction (PCR) amplification and NGS. Only high-quality
COLORED STONES AND ORGANIC MATERIALS reads assigned to correct Ion Xpress MID (molecular iden-
tifier) tags were used in NGS data analysis. Negative PCR
DNA barcoding and next-generation sequencing (NGS) of and negative extraction controls did not produce any valid
freshwater pearls. Research efforts on the deoxyribonucleic sequencing data. Blast algorithms were utilized to match
acid (DNA) species identification of biogenic materials, in resulting operational taxonomic units to a reference li-
particular pearls, have been the focus of some gemological brary database for mitochondrial cytochrome c oxidase
laboratories, including GIA’s, for several years (J.B. Meyer subunit I; results were imaged using MEGAN software.
et al., “DNA fingerprinting of pearls to determine their ori- DNA sequences recovered from the unknown pearl sam-
gins,” PLOS ONE, Vol. 8, No. 10, 2013, e75606; K. ples were compared against the species sequence reference
Saruwatari et al., “DNA techniques applied to the identi- library in the Barcode of Life Data System, accessible at
fication of Pinctada fucata pearls from Uwajima, Ehime http://www.boldsystems.org, and compared against the
Prefecture, Japan,” Spring 2018 G&G, pp. 40–50; K. Scar- National Center for Biotechnology Information database
ratt, CIBJO Special Pearl Report, 2019). Here we report on (https://www.ncbi.nlm.nih.gov).
the findings of a study on the DNA analysis of cultured DNA fragments from eight of the twenty-two samples
and natural freshwater pearls originating from both North were successfully recovered and identified. DNA-based
America and China using the next-generation sequencing species identification was highly consistent with the re-
(NGS) technique, in collaboration with the Canadian Cen- ported origin of the pearls (table 1). Results on the analyses
tre for DNA Barcoding at the University of Guelph. of the 14 remaining samples were not successful, and no
A total of 22 freshwater pearl samples were subjected valid yield was recovered. This can be attributed to many
to DNA barcode analysis (figure 56). Nine American fresh- factors, including insufficient DNA content due to limited
water natural pearls (sample numbers 1, 2, 8, 12, 13, 15, sample size, or the elimination of trace DNA during the
16, 17, and 18) and four American freshwater cultured various routine pearl treatments known to be applied.
pearls (sample numbers 3, 4, 9, and 10) from Gina Laten- While challenges remain on the extraction of DNA frag-
dresse (American Pearl Company, Inc., Nashville, Ten- ments from pearls, the results indicate that this technique
nessee), and nine Chinese freshwater cultured pearls can provide positive matches on individual pearls and
(sample numbers 5, 6, 7, 11, 14, 19, 20, 21, and 22) from prove which mollusk species they originated from, thus
the GIA research collection were analyzed. The samples aiding in the identification of some challenging pearls or
were either drilled with a hand drill using sterile tech- confirming mollusk origins in cases where greater detail
niques or crushed with a mortar and pestle using liquid may justify the time and expense of such analyses.

1 2 3 4 5 6 7

Figure 56. Twenty-two

pearls of various kinds
and species from vari-
8 9 10 11 12 13 14 ous freshwater sources
were used for the blind
DNA study. Photo by
Sood Oil (Judy) Chia.

15 16 17 18 19 20 21 22

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2020 183

TABLE 1. Successful DNA barcoding results of eight freshwater pearl samples, with their reported origins.

Sample no. Reported origin DNA barcoding result

3 American freshwater cultured pearl (Megalonaias nervosa—Washboard) Megalonaias nervosa
4 American freshwater cultured pearl (Megalonaias nervosa—Washboard) Megalonaias nervosa
8 American freshwater natural pearl (Potamilus alatus—Pink Heelsplitter) Potamilus alatus/purpuratus
9 American freshwater cultured pearl (Megalonaias nervosa—Washboard) Megalonaias nervosa
10 American freshwater cultured pearl (Megalonaias nervosa—Washboard) Megalonaias nervosa
15 American freshwater natural pearl (unknown) Megalonaias nervosa
19 Chinese freshwater cultured pearl (Hyriopsis species) Hyriopsis cumingii/schlegelii
22 Chinese freshwater cultured pearl (Hyriopsis species) Hyriopsis cumingii/schlegelii

This joint research was previously presented at the 7th Stephen Challener (Angry Turtle Jewelry), who acquired
International Barcode of Life Conference (Genome, Vol. 60, them from an Ethiopian gem dealer in Tucson in February
No. 11, pp. 1003–1004). GIA will continue to investigate 2019. Another four rough stones (figure 57, right) were pur-
this field of research, which is proving to be a highly valu- chased by author YK from Amde Zewdalem (Ethiopian Opal
able supplementary technique in the gemological exami- and Minerals) and Benyam Mengistu, who facilitates mining
nation of pearls. and exporting samples from Ethiopia, at the Tokyo Interna-
Chunhui Zhou tional Mineral Association show in June 2019. Prior to this
GIA, New York discovery, the only verified occurrences of Cu-bearing
feldspar were from Lake and Harney Counties in Oregon
Janet Topan and Evgeny V. Zakharov (e.g., the Dust Devil and Ponderosa mines). However, more
Canadian Centre for DNA Barcoding than a decade ago there was a controversy about Cu-bearing
University of Guelph, Ontario, Canada feldspar on the market purportedly from Asia or Africa with
an undetermined color origin, presumably Cu-diffused (G.R.
Sunstone plagioclase feldspar from Ethiopia. Ethiopia, tra- Rossman, “The Chinese red feldspar controversy: Chronol-
ditionally known for opal, has become an important source ogy of research through July 2009,” Spring 2011 G&G, pp.
for emerald and sapphire. After these significant discover- 16–30; A. Abduriyim et al., “Research on gem feldspar from
ies, a new type of Cu-bearing sunstone feldspar, first shown the Shigatse region of Tibet,” Summer 2011 G&G, pp.
in 2015 to Tewodros Sintayehu (Orbit Ethiopia Plc.), was 167–180). Gemological testing and advanced analytical
discovered in the Afar region (L. Kiefert et al., “Sunstone methods helped distinguish this new Ethiopian material
labradorite-bytownite from Ethiopia,” Journal of Gemmol- from the Oregon material and the controversial feldspar of
ogy, Vol. 36, No. 8, 2019, pp. 694–695). This material made questionable color origin mentioned above in order to en-
its way to the jewelry market last year in Tucson. sure GIA’s accurate reporting of the natural origin of Cu-
To fully characterize this new production, GIA obtained bearing feldspar.
48 Ethiopian sunstones for scientific examination. Among Two polished rough Ethiopian samples gave RI readings
them, 44 rough stones (figure 57, left) were borrowed from of nα = 1.562 and nγ = 1.571 and birefringence of 0.009. Optic

Figure 57. Left: Forty-four Ethiopian rough sunstones exhibiting different colors and clarities. The largest stone
weighs 54.41 ct. Right: Four Ethiopian rough sunstones ranging from 9.92 to 35.42 ct. All 48 rough stones were
identified as labradorite by LA-ICP-MS, except for one spot as bytownite. Photos by Diego Sanchez (left) and
Shunsuke Nagai (right).

184 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2020

Figure 58. Left: Clouds of copper platelets were observed in numerous Ethiopian sunstones. Field of view 3.13 mm.
Center: Dense networks of reddish wispy dislocations were seen in some Ethiopian sunstones, along with a
slightly greenish blue bodycolor. Field of view 2.72 mm. Right: Yellowish crystals of what are likely fayalite were
seen in one Ethiopian feldspar. These inclusions have also been documented in Oregon sunstone. Field of view
2.82 mm. Photomicrographs by Nathan Renfro.

signs were biaxial positive. Hydrostatic SG measurements = 1.559–1.566; birefringence—0.004–0.010; SG—2.68–2.69).

were 2.70 and 2.72. These RI and SG ranges overlapped with A total of 18 Ethiopian sunstones were tested under a desk-
Oregon sunstone (RI—nα = 1.560–1.570, nγ = 1.568–1.579; top UV lamp. Eleven showed very weak orange fluores-
birefringence—0.006–0.012; SG—2.70–2.72) but were signif- cence, while seven were inert under long-wave UV. All
icantly higher than other Cu-bearing feldspars with unde- showed very weak red fluorescence under short-wave UV.
termined natural or treated color origin (nα = 1.551–1.559, nγ Under magnification, many showed dense clouds of reflec-

TABLE 1. Generalized trace element profiles of sunstone feldspar in parts per million weight (ppmw) and mol.% end members.

Ponderosa mine: Natural Oregon sunstone

Key trace element chemistry Mol.% end membersa
Mg Cu Ga Sr Ab An Or
Range 879–1110 3.78–179 12.0–14.0 437–485 28.1–32.4 67.3–71.6 0.27–0.37
Average 1050 58.2 12.8 452 29.9 69.8 0.29
Median 1060 27.1 12.7 450 29.8 69.9 0.29
Dust Devil mine: Natural Oregon sunstone
Key trace element chemistry Mol.% end membersa
Mg Cu Ga Sr Ab An Or
Range 810–1090 0.74–104 12.0–20.1 449–671 27.7–35.7 63.5–72.0 0.28–0.82
Average 920 25.0 16.4 580 32.2 67.2 0.62
Median 903 17.5 17.1 593 32.5 66.8 0.68
Natural Ethiopian sunstone
Key trace element chemistry Mol.% end membersa
Mg Cu Ga Sr Ab An Or
Range 261–686 0.51–115 16.5–25.3 283–781 29.7–39.4 60.1–69.9 0.37–1.42
Average 421 15.5 19.4 367 34.2 65.3 0.59
Median 411 7.54 19.2 328 34.4 65.0 0.49
Cu-bearing feldspar with undetermined color origin
Key trace element chemistry Mol.% end membersa
Mg Cu Ga Sr Ab An Or
Range 315–492 405–653 16.3–18.8 1120–1210 46.3–50.8 46.3–51.3 2.36–3.05
Average 423 499 17.6 1160 48.7 48.7 2.60
Median 420 481 17.7 1160 48.7 48.8 2.57
a
Abbreviations: Ab = albite, An = anorthite, Or = orthoclase
Detection limits: 0.45–2.08 ppmw Na, 0.011–0.11 ppmw Mg, 0.26–0.59 ppmw K, 5.27–22.2 ppmw Ca, 0.014–0.050 ppmw Cu, 0.007–0.45 ppmw Ga,
0.002–0.008 ppmw Sr

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2020 185

 CHEMICAL COMPOSITION
1400 50000
A B
1200 45000

1000 40000

Na (ppmw)
Sr (ppmw)

800 35000

600 30000

400 25000

200 20000
200 400 600 800 1000 1200 60000 70000 80000 90000 100000 110000 120000
Mg (ppmw) Ca (ppmw)

30
C D
1000
25
Ga (ppmw)

Cu (ppmw)
100

20
10

15
1

10 0.1
200 400 600 800 1000 1200 60000 70000 80000 90000 100000 110000 120000
Mg (ppmw) Ca (ppmw)

30
E F
1000
25
Ga (ppmw)

Cu (ppmw)

100

20
10

15
1

10 0.1
200 400 600 800 1000 1200 1400 20000 25000 30000 35000 40000 45000 50000
Sr (ppmw) Na (ppmw)
Ponderosa Mine - Natural Oregon sunstone Natural Ethiopian sunstone
Dust Devil Mine - Natural Oregon sunstone Cu-bearing feldspar with undetermined color origin

Figure 59. Chemical comparison of Oregon and Ethiopian sunstone and Cu-bearing feldspar with undetermined
color origin. Mg (A and C) is the key trace element that separates Ethiopian from Oregon sunstone. The unknown
Cu-bearing feldspar has a much higher Cu and Sr concentration than other natural untreated sunstones (A, D, E,
and F). For major elements, Ethiopian sunstone overlaps with Oregon sunstone, with a slight lower Ca concentra-
tion (B and D). The unknown Cu-bearing feldspar contains more Na and less Ca than other natural untreated sun-
stones (B, D, and F).

tive copper platelets, much like those observed in material tals of what appeared to be fayalite, an inclusion also ob-
from Oregon (figure 58, left). Some stones also showed an served in Oregon sunstone (figure 58, right).
interesting wispy network of reddish dislocation stringers Laser ablation–inductively coupled plasma–mass spec-
with a greenish blue bodycolor in transmitted light (figure trometry (LA-ICP-MS) was used to measure the chemistry
58, center). Another example revealed several yellow crys- of all 48 Ethiopian sunstones, 26 Dust Devil sunstones, 19

186 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2020

 Orthoclase
and Microcline
KAlSi3O8
Natural Ethiopian sunstone
2.554 Dust Devil Mine - Natural Oregon sunstone
2.563 Ponderosa Mine - Natural Oregon sunstone
2.571
Cu-bearing feldspar with undetermined color origin
2.579
Refractive index
2.588
1.528 Specific gravity
2.596
2.605
2.613
2.621
2.630
ar
sp
rs

1.530 ld
2.638
pa

Fe
lds

8
3O

K-

2.647
Fe

lSi

2.655
1-x A
ali
Alk

2.663
x
Na

2.672
1.533 2.680
2.689
2.693
2.697
2.701
2.705
2.714
1.535
Anorthoclase
2.722
2.731
2.740
Andesine Labradorite An 2.747
ort
Albite Oligoclase Bytownite hit
e 2.756
Albite 1.543 1.548 1.554 1.559 1.561 1.566 1.569 1.572 1.577 1.582 1.587 Anorthite
NaAlSi3O8 1.541 1.546 1.551 1.556 1.563 1.567 1.570 1.574 1.580 1.585 CaAl2Si2O8
Plagioclase Feldspars

NaxCa1-xAl2-xSi2+xO8

Figure 60. Illustration of solid solution in the feldspars, modified after Deer et al. (Rock-Forming Minerals, Volume
4: Framework Silicates, 1963). The nomenclature of the plagioclase series and the alkali feldspars is also shown.
All 48 Ethiopian rough stones (yellow dots) were classified as labradorite except one bytownite. Forty-five Oregon
sunstones (blue and green dots) were classified as labradorite-bytownite. Twenty Cu-bearing feldspars with unde-
termined color origin (red dots) were classified as andesine-labradorite. Ethiopian sunstones are less calcic than
Oregon sunstones. Variation of refractive indices nγ and specific gravity with composition of feldspars were plotted
in the ternary plot as pink and blue straight lines, respectively. Note: nγ = gamma, the highest RI of a biaxial crys-
tal, light vibrating parallel to the Z optical direction.

Ponderosa sunstones, and 20 Cu-bearing feldspars with un- which is generally similar but less calcic than Oregon sun-
determined color origin. NIST 610 and USGS GSD-1G and stones (table 1). All analyses of Ethiopian material indicated
GSE-1G glasses were used as external standards. 29Si was classification as labradorite except one spot that gave by-
used as an internal standard. Ponderosa and Dust Devil sun- townite (figure 59B and figure 60, yellow dots). The Cu-bear-
stones yielded an end member composition of Ab28–32An67–72 ing feldspar with undetermined color origin yielded an end
Or0.3–0.4 and Ab28–36An64–72Or0.3–0.8, respectively (table 1). They member composition of Ab46–51An46–51Or2–3 (table 1). They
are classified as labradorite-bytownite using an albite- were classified as andesine-labradorite, distinct from the
anorthite-orthoclase (Ab-An-Or) ternary diagram (figures Oregon and Ethiopian material (figure 59B and figure 60, red
59B and 60, blue and green dots). Ethiopian sunstones dots). In addition to the differences with major elements Na,
yielded an end member composition of Ab30–39An60–70 Or0.4–1.4, Ca, and K, the analyses revealed that the trace elements Mg,

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2020 187

 R AMAN SPECTRA
60000
Cu, Ga, and Sr were the four best discriminators providing 463 Loose Chalcedony Tested
clear separations among Oregon and Ethiopian sunstone and Blue Chalcedony
50000 Chrysocolla
Cu-bearing feldspar with undetermined color origin. All
Ethiopian sunstone had a lower Mg concentration (261–686

INTENSITY (a.u.)
40000 3619
ppmw, table 1) than Oregon sunstone (>810 ppmw, table 1)
(figure 59A). The Cu-bearing feldspar with undetermined
30000
color origin had a higher Cu (>405 ppmw) and Sr (>1120
ppmw) concentration than Oregon and Ethiopian sunstone
20000
(figure 59, A, D, E, and F). Interestingly, Ponderosa samples 678
(figure 59, C and E) had a lower Ga concentration (<14.0
10000
ppmw) than the Ethiopian sunstone and Cu-bearing feldspar
with undetermined color origin. A group of Dust Devil 501
0
stones with higher Mg, Ga, and Sr concentrations were sep- 150 650 1150 1650 2150 2650 3150 3650
arated from all other sources in figures 59A and 59C, further
RAMAN SHIFT (cm–1)
differentiating them from Ponderosa stones.
Copper-bearing sunstones from different sources are vi-
sually indistinguishable from one another. Gemological Figure 61. The Raman spectrum of the loose chal-
properties are usually sufficient to separate Ethiopian and cedony compared to that of chrysocolla and blue
Oregon sunstones from these Cu-bearing feldspars with un- chalcedony reveals a mineral composition of quartz
determined color origin. However, accurate major and trace (463 cm–1) and moganite (501 cm–1) but a lack of
element chemical analysis obtained by methods such as LA- chrysocolla inclusions (3619 cm–1).
ICP-MS, XRF (Ga and Sr were first identified as reliable dis-
criminators for separating Ethiopian from Oregon sunstones
using XRF by author GRR before this work), or electron mi- chalcedony without the characteristic peaks of chrysocolla
croprobe is critical to separating Ethiopian, Oregon, and Cu- (figure 61).
bearing feldspar with undetermined color origin. With the owner’s consent, we cut one cabochon and pol-
ished the cross section displaying a blue mantle zoning from
Ziyin Sun, Nathan D. Renfro, Aaron C. Palke,
surface to center parallel to its profile (figure 62). The sample
Heidi Breitzmann, Jonathan Muyal, Dylan Hand,
was analyzed with EDXRF, and concentration mapping on
Maxwell Hain, and Shane F. McClure
the cross section confirmed that copper was concentrated
GIA Carlsbad
on the surface and decreased toward the interior, which is
Yusuke Katsurada and Makoto Miura strong evidence for dyeing with copper salts (figure 63).
GIA, Tokyo Twenty pieces of chalcedony dyed with copper salts
George R. Rossman were further analyzed with EDXRF and compared to results
California Institute of Technology, Pasadena from twenty pieces of natural blue chalcedony. EDXRF re-

SYNTHETICS AND SIMULANTS Figure 62. A blue mantle zoning on the cross section
of the loose chalcedony is due to the bath of copper
Dyed chalcedony imitation of chrysocolla-in-chalcedony. dye. Photo by Shu-Hong Lin.
Chrysocolla-in-chalcedony, also known as gem silica or blue
chalcedony in Taiwan, is the most valuable chalcedony va-
riety on the Taiwanese market. The beautiful greenish blue
color is derived from micro-inclusions of chrysocolla, which
can be identified by observation of a peak at 3619 cm–1 in
the Raman spectrum. This peak can be assigned to OH
groups in chrysocolla. Therefore, the color origin for this ma-
terial is fundamentally rooted in the presence of Cu2+ ions
in the structure of the chrysocolla inclusions. In the past few
years, a large number of dyed chalcedony imitations have
appeared in Taiwan’s market. The blue color of chalcedony
dyed by copper salts, and that of natural specimens contain-
ing chrysocolla, is caused by Cu2+ ions.
Recently, a parcel of loose chalcedonies was sent to the
Taiwan Union Lab of Gem Research (TULAB) for identifi-
cation. These stones were submitted as natural blue chal-
cedony, but Raman spectroscopy later confirmed them as

188 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2020

Figure 63. The concentration mapping for copper on
the cross section of dyed chalcedony, which shows
higher copper concentration in the periphery and
lower in the interior; the different colors on the right
represent the degree of relative concentration for cop-
per from high to low.
Figure 64. This “jadeite jade” carving submitted for
sults indicated that the Si/Cu ratio of chalcedony dyed with identification proved to be a jadeite jade and serpen-
copper salts was much higher than that of natural blue chal- tine doublet. Photo by Lai Tai-An Gem Lab.
cedony (400–600 and 4–50, respectively). The content of Cu
was relatively low in dyed chalcedonies tested.
There are many types of dye used for the color enhance- The client granted us permission to remove the coating
ment of chalcedony. Although a series of tests like those in order to analyze the item in more detail. After removal
used above provide a comprehensive comparison between of the coating, the four sides and base showed a lighter sat-
natural blue chalcedony and the dyed chalcedony analyzed uration of green color separated by a visible horizontal line
in this research, it requires further verification whether (figure 66, B and D). Before coating removal, the layering
these methods can be applied to other dyed chalcedonies. was not as obvious (figure 66, A and C). Magnification con-
Shu-Hong Lin firmed that the object was composed of two different ma-
Institute of Earth Sciences, National Taiwan Ocean terials. Standard gemological testing of the lighter colored
University, Keelung material yielded RIs of 1.56, consistent with those expected
Taiwan Union Lab of Gem Research, Taipei
Yu-Ho Li and Huei-Fen Chen
Institute of Earth Sciences, National Taiwan Ocean Figure 65. FTIR analysis revealed absorption valleys
University at 1167, 1077, 962, 853, 665, 587, 529, 476, and 433
cm–1 in the jadeite jade portion (red spectrum) and
Jiann-Neng Fang valleys at 1001, 646, 557, and 460 cm–1 in the serpen-
National Taiwan Museum, Taipei tine portion (green spectrum).

Jadeite jade and serpentine doublet. The Lai Tai-An Gem FTIR SPECTRA
Lab in Taipei recently received a carving presented as jadeite 2.1
jade. The rectangular, uneven green piece, carved on one
1.9
side but almost plain on the back and sides, weighed ap-
1.7
proximately 171.01 ct and measured approximately 50.4 ×
ABSORBANCE

39.9 × 7.4 mm (figure 64). Standard gemological testing re- 1.5

vealed a spot RI of 1.66 on an area of the carved side, but 1.3

surprisingly the smoother surfaces failed to yield clear read- 665
1.1
646 557
ings. Microscopic observation revealed a coating in the 1167 460
0.9
areas where the failed RI attempts were made. 1001 853
0.7
Subsequent infrared analysis proved the carved side was 1077
529
jadeite jade owing to the relevant absorption valleys ob- 0.5 Serpentine 962 433
Jadeite Jade
served at 1167, 1077, 962, 853, 665, 587, 529, 476, and 433 0.3 587 476
cm–1. However, the four sides and base revealed valleys at 1600 1500 1400 1300 1200 1100 1000 900 800 700 600 500 400

1001, 646, 557, and 460 cm–1, characteristic of serpentine de- WAVENUMBER (cm–1)
spite the surface coating observed (figure 65).

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2020 189

 A B

Figure 66. Removal of

the coating from the
four sides and base ex-
posed a clear boundary
between the two mate-
rials (B and D). Prior to
C D the coating’s removal
(A and C), there were
no indications the item
was a doublet. Photos
by Lai Tai-An Gem Lab.

for serpentine and supported by the FTIR and Raman tary series “Beautiful Gem Stories.” The program is broad-
analyses. Jadeite jade and serpentine can have a very simi- cast on BS-TBS channel, a subsidiary of Tokyo Broadcasting
lar appearance. The identification of this particular piece System, and hosted by longtime G&G contributor and ed-
proved relatively straightforward, but if it were mounted itorial board member Dr. Ahmadjan Abduriyim (Tokyo
in a closed-back setting with only the carved face visible, Gem Science, LLC and GSTV Gemological Laboratory).
a costly identification error could easily result. Hour-long episodes examine the growth of gemstones deep
Larry Tai-An Lai ([email protected]) under the earth’s surface and the mining methods to re-
Lai Tai-An Gem Laboratory cover them, as well as the characteristics, craftsmanship,
Taipei, Taiwan and cultural significance that make them special.
Since 2016, Dr. Abduriyim has filmed on location at im-
portant gem deposits in Madagascar, Mozambique (figure 67),
MISCELLANEOUS Tanzania, Namibia, Sri Lanka (figure 68), Myanmar, Thai-
land, Vietnam, Brazil, Colombia, and other countries. He has
Television documentary series: “Beautiful Gem Stories.” also gone to natural and cultured pearl localities in Bahrain,
The unique allure of jewelry and the excitement of the the United Arab Emirates, Vietnam, and Japan, as well as
global gem industry are captured in the Japanese documen- global manufacturing centers, museums, and trade shows.

Figure 67. Dr. Ahmad-

jan Abduriyim hosts an
episode of “Beautiful
Gem Stories,” filmed
on location at the
Montepuez ruby mine
in northern Mozam-
bique. Photo by
Tomoaki Miyake.

190 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2020

 Figure 68. In this
episode, Dr. Abduriyim
reports from the
Bogawatalawa sapphire
deposit in Sri Lanka.
Photo by Tomoaki
Miyake.

After monthly episodes from 2016 to 2018, four new been named in honor of Dr. Barbara Dutrow (figure 69,
episodes are broadcast per year now. With more than one left). Dutrowite (figure 69, right) was discovered in the
million regular viewers, “Beautiful Gem Stories” has been Apuan Alps of Tuscany, Italy, and has been recognized by
warmly received. the International Mineralogical Association. The Aus-
“The financial crisis that hit the Japanese economy in trian, Italian, and Swedish researchers who discovered the
2008 took a heavy toll on consumer interest in jewelry,” mineral named it in recognition of Dr. Dutrow’s teaching
said Dr. Abduriyim. “But I am very grateful to contribute and research contributions, particularly on tourmaline and
to restoring the jewelry industry in Japan and throughout its formation. She is the Gerald Cire and Lena Grand
the world with an enlightening look at gemstones.” Williams Alumni Professor in the Department of Geology
and Geophysics at Louisiana State University in Baton
ANNOUNCEMENTS Rouge. In 2007, she was a coauthor of the 23rd edition of
Manual of Mineral Science, a standard reference textbook
Dutrowite: New mineral species of tourmaline. A newly for the study of minerals. Since 2016, Dr. Dutrow has
discovered mineral species of the tourmaline group has served on the GIA Board of Governors.

Figure 69. Left: Profes-

sor Barbara Dutrow has
been honored with the
naming of the new min-
eral dutrowite. Photo
by Kevin Schumacher.
Right: Sample of
dutrowite (brown) and
dravite (blue) tourma-
lines in meta-rhyolite
from Italy. Courtesy of
Cristian Biagioni.

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2020 191

G&G launches Facebook group. On February 5, 2020,
Gems & Gemology launched a Facebook group dedicated
to gemology and research published in the journal. The
group has attracted jewelry professionals, researchers, stu-
dents, and those with a general interest in gemology who
want to share and expand their knowledge. With several
dozen posts (figure 70), the G&G Facebook group has be-
come a dedicated forum for the discussion of inclusions,
treatments, field gemology and gemstone mining, and
identification of laboratory-grown or cultured gemstones,
including identification challenge quizzes. Some of the
most popular posts to date have featured stunning pho-
tomicrographs of gersdorffite inclusions in quartz, dia-
monds with octahedral stellate cloud inclusions,
mysterious sapphires from the Andes Mountains in South
America, and the use of Google Earth for research in field
gemology. The enthusiastic base of more than 7,000 mem-
bers hails from all parts of the globe. To join, visit
www.facebook.com/groups/giagemsgemology.

ERRATUM

In the Winter 2019 article “Geographic Origin Determina-

tion of Emerald,” the figure 9 caption gave the source of
the faceted emerald as Zarajet, Afghanistan. The correct
location is Zarakhil, Afghanistan. We thank Ahmad Figure 70. The G&G Facebook group is a new com-
Khaled for this correction. munity for all gem and jewelry enthusiasts.

For online access to all issues of GEMS & GEMOLOGY from 1934 to the present, visit:

gia.edu/gems-gemology

192 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2020