APPROVED AS

AHERICAM STANDARD Alll.2·1961

BY AMERICAN STAMDARDS ASSOCIATION
UDC 666. 76 :543. 7

Standard Methods of
CHEMICAL ANALYSIS OF REFRACTORY MATERIALS 1

ASTM Designation: C 18 - 60
ADOPTED, 1939; REVISED, 1941, 194S, 19S1, 19S2, 1960.1
Thia Standard of the American Society for Testing Materials is issued under
the fixed designation C 18; the final number indicates the year of original
adoption as standard or, in the case of revision, the year of last revision.
Scope Lime.............................. 25
Magnesia. . . . . . . . . . . . . . . . . . . . . . . . . . 26
1. (a) These methods cover proce- Alkalies................... 27
dures for the chemical analysis of fire- Flame Photometer Method for Potas-
sium, Sodium, and Lithium Oxides.. 79
clay, silica, high-alumina, magnesite, and High-Alumina Refractories:
dolomite refractories, and of chrome ore Moisture........................... 29
and chrome refractories. The methods Loss on Ignition. .. . . . . . . . . . . . . . . . . . 30
apply to products as sold commercially Silica....... . . . . . . . . . . . . . . . . . . . . . . . 31
Separation of AlsOa and P101 from
but not necessarily after alteration in Fe,O., Ti<>t, CaO, and MgO...... . . 32
service. Iron O.r:ide......................... 33
(b) The analytical procedures appear Titania............................ 34
Phosphorus Oxide................... 3S
in the following order: Alumina .......................... 36
Fireclay Refractories: Section Lime.............................. 37
Moisture .......................... . 7 Magnesia.......................... 38
Loss on Ignition ................... . 8 Alkalies.. . . . . . . . . . .. . . .. .. . . .. . . . 39
Silica ............................. . 9 Flame Photometer Method for Potas-
Iron, Aluminum, and Titanium Ox- sium, Sodium, and Lithium Oxides.. 79
ides............................. 10 Magnesite and Dolomite Refractories:
Moisture........................... 42
¥l!!n?:.~~.· .·:::: ::::::::::::::::::: g Loss on Ignition.. .. . . . .. . . . . . . . . . . .
Silica............... . . . . . . . . . . . . . . .
43
46
Alumina ......................... , • 13
Lime.............................. 14 Iron and Aluminum Oxides.......... 47
Iron Oxides. . . . . . . . . . . . . . . . . . . . . . . . 48
~~:~·.-. ·. ·.:::::::::::::::::::::: ~~ Alumina ...........................
Lime..............................
49
SO
Flame Photometer Method for Potas-
sium, Sodium, and Lithium Oxides. . 79 Magnesia. . . . . . . . . . . . . . . . . . . . . . . . . . S1
Silica Refractories: Manganese as MnO. . .. . . . . . . . . .. . . . 52
Moisture........................... 18 Chrome Ore and Chrome Refractories:
Loss on Ignition.. . . . . . . . . . . . . . . . . . . 19 Moisture. . . . . . . . . . . . . . . . . . . . . . . . . . SS
Silica.............................. 20 Chromium Oxide. . . . . . .. . . .. . . . . .. . S6
Iron and Aluminum Oxides........... 21 Titania and Ferrous Oxide.... . . . . . . . S7
Iron Oxide. . . . . . . . . . . . . . . . . . . . . . . . . 22 Titania............................ 58
Titania.... . . . . . . . . . . . . . . . . . . . . . . . 23 Ferrous Oxide...................... S9
Alumina........................... 24 Silica.. . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
1 Under the atandardization p<ocedure of. the Society,
Alumina........................... 61
thae method• are under the )unadiction of the A.S.T.H. Lime.............................. 62
Committee C-8 on Refractories. Manganese as MnO. . . . . . . . . . . . . . . . . 63
• Prior to their present adoption u 1tandard, these Magnesia . . . . . . . . . . . . . . . . . . . . . . . • • 64
mcthoda ••re publiJbed u tentative from 1917 to 1920,
beinr rnised in 191&. They •ere adopted in 1920, pub- Carbon and Carbon-Ceramic Refractories:
liahed u atandard from 1920 to 1937, betq revbed In 19l5, Moisture. . . . . . . . . . . . . . . . . . . . . . . . . . 67
but were •itbdrawn and pub!abed u tentatln from 1937
to 1939. Loss on Ignition.. . . . . . . . . . . . . . . . . . . 68
453
A "tentative revision" of thls standard appears on p. 482.
U-9
454 CHEU:ICAL ANALYSIS OF REPR.ACTOllIES (C 18)

Silica.............................. 69 the specifications of the Committee on

Iron, Aluminum, and Titanium Oxides 70
Iron Oxide. . .. . . . . . . .. . . . . .. .. . . . . . 71 Analytical Reagents of the American
Titania............................ 72 Chemical Society, where such specifica-
Alumina........................... 73 tions are available.• Other grades may
Lime .............................. 74
Magnesia. . . . . . . . . . . . . . . . . . . . . . . . . . 7S be used, provided it is first ascertained
Alkalies. . . .. . .. .. .. .. .. . . . . .. . . . . . 76 that the reagent is of sufficiently high
Total Carbon . . . . . . . . . . . . . . . . . . . . . . 77 purity to permit its use without lessening
Alternative Procedure When Silicon
Carbide is Praent. . . . .. .. . .. .. . . 78 the accuracy of the determination. Un-
Non.-These methods havo been compiled a1 less otherwise indicated, references to
standard procedures for use in referee analyses. water shall be understood to mean dis-
These methods, however, when the determina- tilled water.
tion of iron oxide u Fe.O. is involved, are not Paragraphs (a) to {p) include those
intended to preclude the use of other procedure'
that give results within the permissible varia· reagents common to two or more of the
tions. For the sa.lte of uniformity the classical analytical procedures. Other reagents
Zimmerman-Reinhardt procedure is specified will be found listed with the particular
for the determination of iron oxide. It is method in which they are prescribed.
recognized that numerous other procedures are (a) Concentrated Acids and Ammo·
equally . accurate and often more convenient.
The other procedures commonly in use include nium Hydroxide.-Concentrated acids
reduction of an oxidized solution with zinc or and ammonium hydroxide of approxi·
other metal, and titration with standard K.MnOt mately the following specific gravities 01
or K1Cr10,, u well as titration with a standard concentrations will be required:
solution of titanoua chloride in an oxidized
solution. These proce4um 1ball he considered Hrdrochloric acid, HCl ............ 1.18 ap. p.
acceptable provided the analyst bas obtained Nitric acid, HNO. ................ 1.42 sp. gr.
results by his special procedure which check with Sulfuric acid, HsSO, ................ 1.841p. gr.
the Zimmerman-Reinhrdt procedure within Hydrofluoric add, HF .........•.... 40 per cent
the limits specified in Section 6. It is suggested Perchloric acid, HCI0,.. 60 to 70 per cent, c. p.•
Sulfurous a.cid ............. 6 per cent solution•
that the Nation&l Bureau of Standards' standard Ammonium hydroxide, NH.OH..... 0.90 sp. gr.
samples be used for checking the accuracy of
procedures. .
It will be underatoOcl that the making of a
complete analysis of a refractory materi&l is a
difficult procedure requbing a wide knowledge (b) Diluted Acids and Ammonium By-
of the chemistry involved in the opera.tions and droxide.-The diluted acids and ammo-
a thorough training in carrying out the work.
A skilled analyst of good training is therefore nium hydroxide referred to are of varying
required to do the work. The descriptions here percentages by volume. They shall be
given cover the vital points of procedure, but made up by mixing proportional volumes
frequent reference in regard to the details of of the concentrated reagent and water.
the various manipulations should be made to The diluted sulfuric acid mixtures shall
"Applied Inorganic Analysis" by Hillebrand
and Lundell• and to similar publications. Par- be made up by slowly stirring the acid
ticularly in the determination of a.lurnina, into the water. These diluted acids and
reference should he made to S'"ntijk Paper ammonium hydroxide are designated in
No. 286 oC the National Bureau of Standards.• the methods as (1:4), (1:9), etc., except
Reagents very diluted solutions which are referred
2. Unless otherwise indicated, it is in- to by the percentage of reagent added.
tended that all reagents shall conform to '° "Rea11:ent Chemicals, American Chemical
Society Specification!," Am. Chem. Soc., Wash-
3 W. F. Hillebrand and G. E. F. Lundell, "Ap-
in11:ton, D. C. For 11u11:11:estions on the testing of
plied Inorganic Analysis," Wiley and Son, New reagents not listed by the American Chemical
York City (1929). Society, eee "Rea11:ent Chemicals and Stand·
'\V. Blum, "Determination of Alumina 88 ards,'' by Joseph Rosin, D. Van Nostrand Co.,
Oxide," National Bureau of Standards Scienti./ic Inc., New York, N. Y., and the "United Statc3
Paper No. S86. Pharmacopoeia."
 CBElilICAL ANALYSIS OF ~EnACTOIUES (C 18) 455

The designation in parenth_eses indicates beakers, dilute, boil, and precipitate with
the ratio of the volume of the concen- NH,OH. Filter, and wash with hot
trated reagent to the volume of water; water. Place the papers in the original
for example, H,so. (1: 9) contains 10 beakers, add 15 ml. of HCI, stir to macer-
per cent by volume of H:SO. (sp. gr. ate the paper, dilute, and precipitate
1.84). The following will be required: again with NH.OH. Filter, and wash
with hot water until free of alkali
Percentace by Volume
Hydrochloric acid. . . . . . . . . . . . . . {~g salts. Ignite carefully, blast, and weigh.
From the weight determined, calculate
Sulfuric acid. . . . . . . . . . . . . . . . . . { i~ the strength of the solution.
(h) Standard Sodium Arsenite Solu-
Nitric acid.. .. . . . . . . . . . . . . . . . . { 3~ tion.-Dissolve 0.908 g. of arsenious
oxide, As/)1 , in a small amount of hot
Ammonium hydroxide.......... (50 Na1CO, solution, cool, filter, and dilute
I 5 to 1 liter. Standardize against a steel
(c) Ammonium Chloride (2 per cent). of known manganese content.
-Dissolve 2 g. of NH.CI in 100 ml. (i) Stannous Chloride Solution.-Dis-
of water. solve 50 g. of SnC11 in 100 ml. of HCI
(d) Ammonium Oxalate (Saturated s~ and dilute to 1000 ml. Keep a few
lution).-Dissolve 4 g. of (NH.)2C20c pieces of metallic tin in the bottle.
in 100 ml. of water. (j) Mercuric Chloride Solution.-Pre-
(e) Standard Potassium Permanganate pare a saturated solution of HgCl2.
(O.J N).- Dissolve 3.25 g. of K.MnO. (k) Manganese Sulfate Solution.-Dis-
in 1000 ml. of water. Allow to stand solve 70 g. of crystalline MnSO. in 500
for a week, filter through an asbestos ml. of water. Add 140 ml. of phosphoric
mat, porous glass, or porcelain filter, and acid, HaPO., (sp. gr. 1.7) and 130 ml. of
keep in a dark place. Standardize against H 2SO. (sp. gr. 1.84). Dilute to 1 liter.
the National Bureau of Standards' (/) Hydrogen Peroxide (30 per cent).
standard sample 40c of sodium oxalate. (m) Diammonium Phosphate Solution.
(fJ Standard Potassium Permanga- -Dissolve 10 g. of (NH.)tHPO. in 100
nate (0.04 N).-Dissolve 2.5 g. of ml. of water.
KMnO. in water and make up to 2 (n) Chloroplatinic Acid Solution (JO
liters. Allow to stand for a week, filter per cent).
through an asbestos mat, porous glass, (o) Ethyl Alcohol (80 per cent).-
or porcelain filter, and keep in a dark Prepare a solution containing 80 per cent
place. Standardize against the National by volume of ethyl alcohol in water.
Bureau of Standards' standard sample (p) Ethyl Alcohol (Absolftte).-Cer-
40c of sodium oxalate. tain commercial brands of denatured
(g) Standard Titania Solution.- absolute alcohol are satisfactory as well
Weigh out 0.05 g. of calcined TiOt. as being considerably Jess expensive than
Fuse with 10 g. of K2S207 in a clean the reagent grade absolute alcohol.
platinum crucible, keeping the tempera-
ture as low as possible to maintain fluid- Preparation of Samples for Analysis
ity. Cool, and dissolve in about 300 ml.
of diluted H2SO. (1:5). Cool, transfer 3. The sample, carefully obtained by
to a 500-ml. volumetric flask, dilute to one of the approved methods of sam ·
the mark with water, and mix thor- piing, shall be crushed in a small jaw or
oughly. To standardize the solution, roll-type crusher with hardened tool-
take two 50-ml. portions in 400-ml. steel faces to pass a No. 8 (2.38-mm)
MO
456 CBEVICAL ANALYSIS OJ' REFRACTORIES (C 18)

sieve.' The sample shall then be Blank Determination•_

crushed to pass a No. 20 (840-micron) 5. Blank determinations on the re·
sieve,' mixed, and quartered to about agents shall be made for each constit·
50 g. This 50-g. sample shall then be uent in the refractory and this blank
ground so that it will all pass a No. deducted in each case. For the de·
100 (149-micron) sieve,• unless other- termination of the silica blank, approx·
wise specified, mixed thoroughly, and imately 0.25 g. of AltOa should be added
placed in a container which will insure as aluminum chloride.
freedom from contamination. Pre--
cautions shall be taken to prevent con- Reproducibility of Result•
tamination of the sample by steel par- 6. In all cases, check determinations
ticles from the sampling equipment shall be made, and the results shall be
during crushing or grinding. Fine grind- redetermined if satisfactory checks are
ing shall be done in a suitable mor- not obtained. Results shall be con·
tar (agate, mullite, alumina, or boron sidered satisfactory if the differences
carbide) to prevent the introduction of between check determinations do not
impurities. exceed the following limits:
Statem'.ent of Analyaia Permi11lble Vuiat.ioaa Be·
tweea Checlt Dttcrmia-
4. Except in the case of dolomite re- ation1, mu., per cot
fractories (Note), moisture shall be deter- For silica or other constituent
amounting to 30 per cent- or
mined on the sample in its ordinary air- over........................ 0.3
dried condition. All other percentage For alumina or other ClOnatituent
compositions shall be determined on amounting to 10 to 30 per
moisture-free samples and reported, ac-
cent-....................... 0.2
For any other
constituent
cordingly, on a moisture-free basis. The amounting to under 10 per
drying temperature recommended for all cent-....................... 0.1
moisture determinations is 105 to 110 C., •'lb- li111n1 an 1ta'9d ID tennt of the whole -pie
u 100 per cent.
although it is recognized that in the
case of certain raw fire clays a tempera- FIRECLAY REPRACTORIES
ture as high as 140 C. may be required
for complete removal of all hygroscopic Moisture
moisture.• Whenever a sample is 7. Weigh 1.00 g. of the sample and
weighed out for any determination other heat to constant weight at a temperature
than moisture, it shall be moisture-free. not under 105 nor over 110 C. Record
If preferred, the sample may be dried in a the loss in weight as moisture.
weighing bottle from which the required
samples shall be weighed out. Loas on lpition
NoT.11:.-An exception is made in the case of 8. Weigh 1.000 g. of the moisture-free
dolomite refractories (see Sections 42 and 44), (105 to 110 C.) sample and heat to con·
on which the determination of free moisture is stant weight over a blast lamp, or in ao
omitted, and on which percentage compositions
are reported on the ignited basil. electric mutBe furnace, at 900 to 1000 C.
Record the loss in weight as the ignition
1 Detailed requirements for these sieves are loss.
given in the Specifications for Sieves for Te!tina:
Purposes (ASTM De!ignation: E 11), see p. Silica
1199.
1 See National Bureau or Standards' Certifi.
catll! of Analysis of Standard Samples or Fire 9. Weigh 0.5000 g. of the moisture·
Clay11, Nos. 97 and 98, Au&:U8t 31, 1931. free (105 to 110 C.) sample into a plati·
....
 CHEMICAL ANALYSIS 01' REl'llACTOJUES (C 18) 457

num crucible containing ~bout 5 g. of minutes at 1100 C. Cool the crucible in

powdered anhydrous NasCOa and mix a desiccator, weigh, and repeat blasting
well with a platinum wire. Cover the to constant weight. The loss in weight
mixture with a little more NasCOa. from the original silica residue represents
Heat gradually to the full heat of a good the SiO, content, except for that part of
burner (1000 to 1100 C.) maintained for the Si01 which is later recovered from
about 1 hr. until complete solution is ob- alumina, etc.
tained. Place the crucible cover on a
Non.-Another scheme to aid in subsequent
triangle, and when the melt has partially solution of the fused melt is to rotate the crucible
cooled pour it on the lid (Note). When as it cools, spreading the mass up the side walls.
cool, place the crucible and lid in a 150-
ml. beaker, placing the button on a Iron, Aluminum, and Titanium OD.des
watch glass above the beaker. Add 30 10. Fuse the residue with 1 g. of fused
ml. of HCl (1: 1). When solution is K1S,07 or NasS207, dissolve in a small
:omplete wash off the crucible and lid amount of water, and add to the filtrate
with HCl (1 :4), taking care to remove from the silica determination (Section
all SiO,. Place the button in the solu- 9). Add 5 g. of N~Cl and 3 drops of
tion. Transfer the contents of the methyl red solution (0.1 per cent).
beaker to an evaporating dish and evapo- Heat the solution almost to boiling, and
rate to dryness on a steam bath. Bake add slowly NH,OH (1: 1) until the indi-
for 1 hr. at 110 C. Add 20 to 30 ml. of cator has changed to a yellow color.
HCI (1: 1) and 50 ml. of hot water. Boil for several minutes to remove the
When all salts have been dissolved, allow excess ammonia. Allow to settle for 30
to settle for several minutes and then min. and decant through a No. 41 What-
filter through a No. 40 Whatman paper man filter paper or equivalent, trans-
or equivalent. Wash the SiO, three ferring the precipitate to the paper and
times by decantation using 20 to JO-ml. washing the beaker and paper several
portions of first hot water, then HCI times with a warm N~Cl solution (2
(1: 1), then hot water again. Transfer per cent). Reserve the filtrate, A, for
the precipitate to the filter paper, re- the determination of CaO and MgO
moving all SiO, from the dish with a (Section 14). Return the precipitate and
policeman. Wash the paper and pre- paper to the original beaker, add 50 ml.
cipitate with hot water until free from of hot water and 10 ml. of HCl (sp. gr.
salt. To recover the small amount of 1.19). Stir until the precipitate is dis-
Si(), remaining in the filtrate, evaporate solved and the paper is well macerated.
to dryness, using the same procedure for Dilute to about 200 ml. with bot water,
baking and filtering as before. Combine precipitate and filter as before. Com-
the two precipitates, place in a platinum bine this filtrate B with filtrate A. Wash
crucible, and bum off the paper carefully the paper and precipitate with a warm
to prevent any loss of SiO,. Ignite the NH,Cl solution (2 per cent). Place the
sample to constant weight at 1100 to precipitate in a weighed platinum cru-
1200 C. (15 to 20 min. is usually suffi- cible and ignite. Continue the ignition
cient), cool in a desiccator, and weigh. at 1200 C. to constant weight (15 to 20
Moisten the residue with several milli- min. is usually sufficient). Cool in a
liters of water, add 10 ml. of HF and 3 or desiccator, and weigh with the crucible
4 drops of HsS04. Evaporate the solu- covered with the lid. The R201 con-
tion to dryness, ignite carefully to pre- sists of the AbOa, 1;'i0i, and Fe20a pres-
vent decrepitation, and blast for several ent in the sample. In addition there
HD
458 CHEMICAL ANALYSIS OF REFRACTORIES (C 18)

may be small amounts of PsO,, ZrOi, VsO& crucible and then fused with KaS.Or or Na,S.O,.
This avoids loea of platinu& by the action of the
and CrsOa. pyroaulfate, and no platinum ia present in the
Iron Oxide filtrate to interfere with the iron determination.

11. (a) Method A: Fel), Determined (b) Method B: FesOa Determined on

on R.fJa Sampk.-Heat the RsOa pre- a SeparaJe Sample.-Weigh 1.00 g. of
cipitate (Note 1) obtained in the deter- the finely ground, moisture-free (105 to
mination of iron, aluminum, and tita- 110 C.} sample into a platinum crucible,
nium oxides (Section 10) with fused add 10 drops perchloric acid (HClO,)
KsSsOT or NasSs01 until solution is com- and 20 ml. of HF, and heat almost to
plete. Dissolve the fusion in 50 ml. of dryness on a hot plate. Add 5 to 10 ml.
HsSO, (1:9) and evaporate to fumes. of HClO, and heat until residue has dis·
Cool, dilute with water, and filter oti the solved (Note 2). Cool, place crucible ina
SiO,, washing with hot water. Reserve 400-ml. beaker, add 100 ml. of water, and
the filtrate for the determination of beat to boiling. Any residue present,
FesOa and TiOs. Ignite the SiOs in a other than SiOs, should be filtered off
platinum crucible and weigh. Treat the and fused with KsSs01 or NasSsOT in a
precipitate with 5 ml. of HF and 2 or porcelain crucible and added to the main
3 drops of HsSO.. Evaporate to dry- solution.
ness, ignite, and weigh. The loss in Non 2.-Decomposition of raw lire clays
weight represents extra SiOs which and some tired refractories can be effected
should be added to that determined equally u well by 1ubstituting an equal volume
previously and also deducted from the ef ~SO, (1: 1) for the HCIO,. In this case,
beat to fumes once, cool, dilute, filter, and fuse
weight of the R,O, precipitate. Evapo- any residue remaining undissolved.
rate the filtrate obtained in correcting
the R,O, precipitate for SiOs to about (c) Determine iron, using one of the
75 ml. Cool, and dilute to 100 ml. in a approved methods referred to in the
volumetric tlask. Reserve 25 ml. for note under Section 1.
the determination of TiOs (Section 12). Titania
To the remainder, add 25 ml. of HCl 12. Determine Ti02 colorimetrically
(1: 1) and heat to boiling. Reduce the by use of a photometer, as follows: Trans·
iron by adding SnCh solution drop by fer the 25-ml. portion reserved for the
drop from a pipette with constant swirl- determination of Ti02 (Section 11) to a
ing of the beaker until the solution is 100-ml. volumetric flask. Add 5 ml. of
colorless. Then add one drop in excess. H2so. (1: 1). Cool to room temperature
Cool quickly in running water, then add and dilute to the mark. Transfer exactly
at one stroke 15 ml. of saturated HgCl1 half of the solution to another 100-ml.
solution. Allow to stand for 3 min., volumetric flask. Dilute one of the flasks
then transfer with washing to a 1000-ml. to the mark with H,so. (1:19). To the
beaker containing 300 ml. of cold dis- other add 5 ml. of 3 per cent H20 2, then
tilled water and 25 ml. of MnSO, solu- dilute nearly to the mark with H 2SO,
tion. Titrate with standard 0.04 N (1: 19), adjust temperature to 25 ::!: 1 C.,
KMnO,, added very slowly while stir- then adjust volume exactly and let stand
ring constantly, until a permanent pink at least 5 min. Transfer a portion from
end point is obtained. the first flask to a cuvette and set the
Non !.-Instead of fusing directly in the potentiometer scale reading at zero. Then
platinum crucible in which the R101 was Ignited, measure the absorption of the solution in
the precipitate may be brushed into a porcelain the second flask. Read the percentage of
MO
 CllE)(ICAL ANALYSIS 01' REl'JlACTORIES (C 18) 459

titania present from a cali.bration curve. Return the paper to the beaker contain-
Construct this curve by adding varying ing the precipitate, add 100 ml. of a
amounts of the standard titania solution H.SO. solution (5 per cent), warm, and
to H2S04(1:19), develop color with 5 ml. titrate to a faint pink end point with
of 3 per cent H20 2, let stand 5 min. and standard 0.04 N KMnO, solution. A
read the absorption, using H2SO, (1: 19) blank should be previously determined
for the zero setting of the potentiometer for the effect of the paper.
scale. -
NOT1t.-For greater accuracy a double
NOTE 1.-If a spectrophotometer is used,
precipitation should be made, in which cue,
measure the absorption at mean transmission of
after precipitating the calcium on.late u
420 mµ. If a filter photometer is used, use a glass
deacribed above, decant the liquid and wash the
with a maximum transmission in the region of beaker and paper several times with wa.nn water.
420 mµ.
Dissolve the precipitate on the paper with warm
NOTE 2.-As an alternate method, Ti0 2 can
HCl (1 :4), allowing it to run into the beaker
be determined in the 25-ml. portion reserved for containing the major portion of the calcium
this purpose (Section 11) by oxidizing both the oxalate. Wash the paper with hot water. To
sample and the standard Ti02 solution with the solution (about 75 to 100 ml. in volume)
several drops of a H,o, solution (30 per cent). add several milliliters of saturated (NH.),C,o.
Compare the colors either in Nessler tubes or in solution and NH.OH in sli&ht excess. Heat for 2
a suitable colorimeter. Use a H2SO, solution (5 hr., filter, wash, and titrate as described above.
per cent) for diluting purposes in matching the
colors. Mapesia
Alumina
15. Evaporate the filtrate from the
13. Subtract the calculated weight of CaO determination (Section 14) to about
Fes01 (Section 11 (a) or (b)), TiOt (Sec- 150 to 200 ml. and add 2 to 3 g. of di-
tion 12), and SiOs (Section 9) from the ammonium phosphate((NH,)JIPO,), stir
weight of R,O, (Section 10). The re- until dissolved, and then add NH,OH
mainder is the weight of AlsOa plus small until alkaline and then 20 ml. in excess.
amounts of the oxides which may in- Allow the solution to stand overnight.
clude those previously mentioned in Filter and wash withNH40H(Spercent).
Section 10. These are generally con- Dissolve the precipitate on the paper
sidered as Al10i in reporting the analysis with hot HCl (1:4), allowing it to run
of fireclay refractories. into the beak.er containing the precipi-
Lime tate. Wash the paper with hot water.
14. Evaporate the combined filtrates To the solution, which should be not
reserved (Section 10) for the determina- more than 100 ml. in volume, add 0.1 to
tion of CaO and MgO to about 200 ml., 0.2 g. of (NH,)JIPO,. Make ammo-
add 10 to 15 ml. of the saturated am- niacal, and then add a slight excess while
monium oxalate solution and 2 to 3 ml. of stirring constantly until the precipitate
NH,OH. Heat for 1 to 2 hr., by which is well formed. Then add 10 ml. more
time the volume should be about 75 to of NH,OH and allow to stand overnight
100 ml. Allow the precipitated calcium or at least for 4 hr. Filter through a
oxalate to settle. Decant through a No. 40 Whatman paper or equivalent.
dense filter paper (Whatman No. 40 or Transfer the precipitate to the paper and
equivalent), taking care to retain the wash well with NH,QH {5 per cent).
precipitate in the beaker, wash several Place the paper in a weighed platinum
times with warm water by decantation, or porcelain crucible, burn off the paper
and then wash the paper until free from at a low temperature (below 900 C.),
;oluble 3alts. Reserve the filtrate for and ignite to constant weight at 1050 to
the MgO determination (Section 15). 1100 C. {15 to 30 min. is sufficient).
460 CBE14ICAL ANALYSIS OP REFRACTORIES (C 18)

Alkalies7 2 ml. of NH.OH to precipitate the last

16. (a) Weigh 1.00 g. of the moisture- trace of calcium. Heat for 30 to 45
free (105 to 110 C.) sample (ground to an min., filter, and wash with water con-
impalpable powder) and 1.0 g. of NH.Cl taining (NH,)1CsO, (0.1 per cent)
into an agate mortar and mix well. Add Catch the filtrate in a weighed platinum
7 to 8 g. of CaC01 (Note) and again mix dish. Add several drops of HCl and
intimately. Place a i-in. layer of CaCOa evaporate to dryness. Ignite gently as
in the bottom of a platinum crucible, and before and weigh as NaCl and KCl.
then add the above mixture, tapping the Non:.-Calcium carbonate of the A.C.S.
grade ''Jow in alkalies and heavy metals" shall
crucible occasionally to obtain a dense be used.
mass. Place a l-in. layer of CaC01 on
the top. Heat the crucible over a low (b) The separation of the potassium
flame until ammonia fumes are no longer and sodium must be carried out in an
given off, then increase the heat so that atmosphere free from ammonia fumes.
the bottom half of the crucible is a dull Add to the solution of the combined
red and maintain this temperature for chlorides in a small porcelain dish suffi·
about 1,hr. Cool, fill the crucible three- dent chloroplatinic acid solution' to re·
fourths full of water, and heat until the act with all of the sodium and potassium.
contents can be taken out and crushed in The necessary amount to use is readily
an agate mortar. Transfer to a plati- calculated from the known strength of
num or porcelain dish by means of a jet the platinum solution and the weight of
of water. Evaporate to a low volume, mixed chlorides counted as NaCl. The
decant through a No. 40 Whatman filter dilution of the resulting solution should
paper, or equivalent, and wash the ma- be such that when heated on the steam
terial in the dish several times by de- bath any precipitate that may have
cantation with warm water. Transfer formed entirely dissolves. This is to
to the paper and wash several times with prevent incluition of mother liquor in a
hot water. Acidify with several milli- mass of crystals suddenly formed.
liters of HCl and evaporate to a volume Evaporate until the solution is just
of 150 to 200 ml. Add several milliliters syrupy enough to solidify on cooling.
of NH,QH and sufficient (NH,),C01 to Do not evaporate to dryness, as this will
precipitate the lime, keeping the dish dehydrate the sodium salt and render it
covered with a watch glass. Warm until less soluble in alcohol. Drench the
the precipitate settles out. Filter and residue with alcohol (80 per cent) filter
wash with warm water. Evaporate the by decantation through a small ~aper,
solution to a low volume, then add a wash by decantation with more of the
small lump of (NH,)iCOa to determine alcohol, crushing the crystals with a
whether practically all calcium has been small pestle or a widened and rounded
precipitated. If no precipitate forms, end of a short glass rod. Reserve the
evaporate to dryness, otherwise precipi- filtrate and washings if sodium is to
tate and filter as before. Drive off the be determined directly. The residue
ammonium salts by heating just short should be golden yellow. An orange-red
of a dull red. Dissolve the residue in color indicates incomplete removal of
water and add a few milliliters of a satu- th~ aodium salt. It is unnecessary to
rated solution of (NH,)2C 20, and 1 to bnng the mass of the precipitate upon
the filter. Dry the dish and paper for a
' This procedure for the determination of alkali is few moments to remove adhering alco-
t~e J. Lawrenc;e Smi!h method. An alternate method
will be found in Section 79. hol. Dissolve the precipitate on the
HI
 CHEMICAL ANALYSIS OF REFRACTORIES (C 18) 461

filter with hot water, catc&ing the solu- The final temperature should be regu-
tion in a weighed crucible or small dish lated to prevent boiling or spattering.
of platinum. Evaporate to dryness and Heating in this manner for 1 hr. is suffi·
add the salt that is still in the porcelain cient to decompose the sample. Rinse
dish. If the salt is at all lumpy, re- the underside of the crucible lid with a
dissolve it in water and again evaporate few drops of water, allowing it to run
to dryness. Heat for 1 hr. at 130 C. in into the crucible. Evaporate the cru-
an air oven (100 C. suffices for very small cible contents to dryness. Cool, add 10
amounts of fine-grained precipitate). It ml. of HNOa (1: 1), and again evapo-
is necessary to cover the receptacle at rate to dryness. Introduce the acid
first because the precipitate is prone to slowly by pouring it down the sides of
decrepitate. When dry, cool and weigh the crucible. Evaporate three times
as KsPtCla. Calculate the oxides, as with the HNO. to insure complete vola-
follows: tilization of the fluorides. Ignite gently
KCl • wt. of KJ>tCla X 0. 3068 at first and then finally to constant
KtO - wt. of KJ>tCla X 0.1941 weight at about 1200 C. Cool in a desic-
NaCl - total chlorides - KCI cator, and weigh. The loss in weight
Na,O .. NaCl X 0.5303
minus the loss on ignition (Section 19)
SILICA REFRACTORIES represents all but a possible trace of the
Special Solutions and Reagents Re- Si01 in the sample.
quired Non.-The alblie1 present will be volatilo
ized also by this treatment and thus will be in-
17. (a) Std/uric Acid (25 per cent by cluded in the percentage of Si01 reported. Be-
volume). cause of the very low alkali content of commer·
(b) Nitric Acid (50 per cent by cial silica refractories, the error introduced by the
volume). use of thia procedure is not ordinarily lignificant
with respect to the total percentage of lilica
Moisture present.
18. Weigh 1.00 g. of the sample and Iron and Aluminum Ozides
heat to constant weight at a temperature
not under 105 nor over 110 C. Record 21. Fuse the residue in the crucible
the loss in weight as moisture. from the silica determination (Section
20) by heating gently with about 2 g.
Lou on Ignition of potassium pyrosulfate, KtS 10r, or
19. Weigh 1.000 g. of the moisture-free sodium pyrosulfate, Na1S20r; cool, dis-
(105 to 110 C.) sample and heat to con- solve in water, and acidify with 3 to 4 ml.
stant weight over a blast lamp, or in an of HCI. Dilute to 150 to 200 ml., add
electric muffie furnace, at 900 to 1000 C. 5 g. of NH,CI and 3 drops of methyl red
Record the loss in weight as the ignition (0.1 per cent). Ht<it to boiling, add
lose. ~OH (1: 1) slowly, finally drop by
drop until the solution is just alkaline
Silica to the indicator. Boil for 1 to 2 min.
20. Transfer 1.0000 g. of the moisture- and filter through a 9-cm. rapid filter
free (105 to 110 C.) sample to a 30-ml. paper. Do not wash. Reserve the fil-
tared platinum crucible. Add 10 ml. of trate. Transfer the paper and precipi-
HN01 (1:1) and 5 ml. of HF. Cover the tate to the beaker in which the precipita-
crucible, place on a hot plate on an asbes- tion was made, add 50 ml. of HCI (1:4),
tos pad, and heat. Heat gently at first, and digest covered on the hot plate,
gradually increasing the temperature. shaking frequently until the precipitate
9-45
462 CHEMICAL ANALYSIS OF REFRACTORIES (C 18)

is dissolved and the paper well macer- tion. Titrate with standard 0.04 N
ated. Dilute, add methyl red, reprecipi- KMnO., added very slowly while stirring
tate with NH.OH (1: 1), boil, and filter, constantly, until a permanent pink end
exactly as before. Add this second fil- point is obtained.
trate to the first filtrate for the subse- Titania ·
quent determination of lime (Section
25). Swab the beaker and rod, and 23. Fuse the ignited R20a residue after
wash· the paper and precipitate with recovery of Si02 (Section 21) with a small
NILCl (2 per cent). Dry, then ignite piece of K2S201 or Na2S201. Cool, and
the paper and precipitate to constant dissolve the melt in 30 ml. of H 2S0 4
weight in a tared platinum crucible of (1: 5). Transfer to a 100-ml. volumetric
30-ml. capacity at 1000 to 1100 C. with flask, and dilute almost to the mark.
access to air. Cool in a desiccator, and Cool to 25 ± 1 C. and adjust volume
weigh as R 20 3 which is mainly the iron exactly. Transfer a portion of the solu-
and aluminum oxides. To recover the tion to a photometer cuvette and set the
trace of Si02 which may have remained potentiometer scale reading at zero. To
with the RiOs group, add 2 drops of the balance of the solution in the flask
H2so. ·(1 :4) and 1 ml. of HF to the add 5 drops of 30 per cent H 20 2, mix,
crucible and evaporate to dryness. let stand at least 5 min., and measure the
Cover the crucible and ignite to constant absorption of the colored solution. Read
weight at about 1100 C. Cool and the percentage of titania present from a
weigh. The loss in weight represents calibration curve. Construct this curve
additional Si02 which should be added by adding varying amounts of the stand-
to that determined in Section 20. ard titania solution to H2SO. (1: 19), de·
velop color with 30 per cent H 20 2, let
Iron Oxide s~nd 5 min., and read the absorption,
22. Transfer 1.00 g. of the moisture- usmg H2SO, (1: 19) for the zero setting
free (105 to 110 C.) sample to a 30-ml. of the potentiometer scale.
platinum crucible and moisten with a NOTE 1.-If a spectrophotometer is used,
few drops of water. Add 3 ml. of HNOs, measure the absorption at mean transmission
and about 15 ml. of HF. Evaporate at a of 420 ?1"· If a filter photometer is used, use a
moderate heat to dryness, taking care glass with maximum transmission in the regioo
not to allow any loss by spattering. De- of 420 m,..
NOTE 2.-As an alternate method fuse the
compose the nitrates and fluorides by ignited RtOs residue after recovery of SiO,
ignition at about 1000 C. for 15 min. (Section 21) with a small piece of K,S,0 7 or
Now add to the residue in the crucible a Na2S:iOr. Cool, and dissolve the fusion by the
small piece of K1S:01 or Na2S201, and addition of about 25 ml. of warm water and t
to 2 ml. of H2SO. to the crucible. It is important
fuse. Dissolve the fusion in 25 ml. of at this point to keep the solution to a small
HCl (1: 1) and heat to boiling. Reduce volume. Transfer to a 50-ml. volumetric flask,
the iron by adding SnCl2 solution drop add a few drops of H202, and dilute to the mark.
by drop from a pipette with constant Place the SO ml. of solution in a small Nes&er
swirling of the beaker until the solution tube. Compare the color of this solution with the
color of a known standard solution. A satis-
is colorless, then add 1 drop in excess. factory solution is one of such strength that 1
Cool quickly in running water, then add ml. equals 0.0001 g. of Ti02. To make the com-
at one stroke 15 ml. of saturated HgCl1 parison use any standard colorimeter, or place
solution. Allow to stand for 3 min., an appropriate amount of standard solution in
transfer with washing to a 1000-ml. a second Nessler tube and dilute with water
from a buret until the color is matched. From
beaker containing 300 ml. of cold dis- the amount of water added, calculate the per·
tilled water and 25 ml. of MnS04 solu- centage of Ti0 2 in the sample.
 CHEMICAL ANALYSIS OF REFRACTORIES (C 18) 463

Alumina to about 300 ml., adcl 2 to 3 g. of

24. Subtract the calculated weight of (NHJ,HPO,. Stir until dissolved and
Fe20i (Section 22), the weight of Si01 then add NH,OH slowly until alkaline,
recovered (Section 21), and the calcu- and then 30 ml. of NH.OH in excess.
lated weight of Ti02 (Section 23) from Allow the solution to stand for at least
the weight of RtO, (Section 21). The 4 hr., and preferably overnight if the
remainder is the weight of the Al 20i, total amount of solution exceeds 400 ml.
plus small amounts of oxides of any other in volume. Filter, wash with diluted
elements that were precipitated. NH.OH (S per cent). Dissolve from
the paper with hot HCl (1: 1) and
NoTE.-These may include P 20, and Zr0 2
which are usually present in such smali reprecipitate by adding 0.1 g. of
amounts as to be disregarded, but silica from (NH,)tHPO,, diluting, rendering ammo-
some geographical sources may contain suffici- niacal, and letting stand as before.
ent phosphorus to warrant a separate P 20 6 de- After filtering and washing as before,
termination. char the paper and burn off the carbon
Lime below 900 C., and then ignite in an
25. Acidify the combined filtrate electric mufile furnace at about 1100 C.
from the R,O, precipitation with several to constant weight. The ignited mate-
milliliters of HCl (1: 1) (Section 21) and rial consists of magnesium pyrophos-
evaporate to a volume of about 250 ml. phate, Mg2P207. Calculate the weight
Heat to incipient boiling, add 25 ml. of of MgO by multiplying the Mg 2P 207 by
a saturated (NHi.)sC10,, solution and 0.3621.
NH.OH until slightly alkaline. Stir Alkalies
thoroughly and allow the precipitate to 27. Determine the alkalies by the
settle out, keeping the solution warm for procedure described in Section 16 (a)
3 to 4 hr. but not boiling. Filter and and (b) for the analysis of fireclay
wash five times with cold water. Save refractories, or by the alternate procedure
the filtrate. Dissolved the washed oxa- described in Section 79.
late precipitate with 50 ml. of diluted
HCl (1: 1), dilute to about 150 ml., and HIGH-ALUMINA REFRACTORIES
precipitate as before. Wash the beaker, (Includes Bonded Refractories made from
precipitate, and paper with cold water Diaspore, Fused Alumina, and the
until (NRthC20, is removed. Avoid Sillimanite Minerals, as well as
excessive washing. Combine the fil- EJectrocasl.)
trates for the magnesia determination Special Solutions Required
(Section 26). To the beaker used for
the precipitation, add 150 ml. of a 28. (a) Potassium Nitrate (1 per cent).
-Dissolve 1 g. in 100 ml. of water.
H2SO, solution (5 per cent), then intro-
(b) M olybdate Reagent.-M ix 100 g.
duce the paper containing the precipi-
tate. Heat to almost boiling and titrate of molybdic anhydride or 118 g. of
with 0.1 N KM nO,. A blank should molybdic acid (85 per cent) with 400
be previously determined for the effect ml. of water, add 80 ml. of NRtOH,
of the paper. and filter when solution is complete.
Prepare a second solution containing
Magnesia 400 ml. of HNOa and 600 ml. of water.
26. To the combined filtrates from Agitate the HNOa solution by means
the lime determination (Section 25), of a current of air and add the molybdate
which have been reduced by evaporation solution very slowly. Continue agita-
f-45
464 Cm;)(lCAL ANALYSIS or REFRACTORIES (C 18)

tion for 1 hr. Add a small crystal of rated with dry HCl gasp. Evapora.te, add
(NH.)JIPO., let stand over night, filter, methyl alcohol solution, and evaporate
and keep in a glass-stoppered bottle. again. Bake to dehydrate the SiOi.
(c) Standard Sodium Hydroxide Solu- Cool, add 15 ml. of HCl and 50 ml. of
tion.-Dissolve 60 g. of NaOH in 100 ml. water, b<>il 1 to 2 min., and filter on a
of distilled water in a 150-ml. test tube No. 40 Whatman or similar filter paper.
of resistant glass. Stopper tightly with Wash four to five times with hot diluted
a cork covered with tin foil and let stand HCl (1 :4), then seven to eight times
until the supernatant liquid is clear. with hot water. Reserve the paper con-
By means of a pipette take out 5.0 ml. taining the Si01 precipitate. Transfer
of the clear solution and dilute to 1 liter the filtrate to the original dish and evapo-
with freshly boiled water (Note). rate. to dryness and bake. Add 15 ml.
{<l) Standard Nitru Acid Solution.- of HCI and 50 ml. of water, bring just
Dilute 5.0 ml. of HNOa (sp. gr. 1.42) to to a boil, tilter on a No. 42 Whatman or
1 liter with freshly boiled water (Note). similar filter paper, and wash as before.
Non:.-Adjust the NaOH and HNO, solu- Reserve the filtrate.
tions until they exactly agree, and standardize Nou 1.-Diluted a.so, is used to prevent
against National Bureau of Standards' standard TiO. from hydrolyzing and precipitating with
aample No. 73 of known phosphorus content. the SiO..
Moisture (b) Place the two filter papers con-
29. Weigh 1.00 g. of the sample and taining the Si01 precipitate in a platinum
heat to constant weight at a temperature crucible, ignite carefully until the paper
not under 105 nor over 110 C. Record is burned off, then heat at 1100 to
the loss in weight as moisture. 1200 C. for 10 to 15 min. Cool, weigh,
and repeat heating at 1100 to 1200 C.
Loss on Ignition until constant weight is obtained.
30. Weigh 1.000 g. of the moisture- Moisten the ignited precipitate with
free (105 to 110 C.) sample and heat to water, a<Jd HF slowly until the crucible
constant weight over a blast lamp, or in is hall-full, then add 5 drops of H,S04
an electric muffle furnace, at 900 to and evaporate to fumes. Cool, add 2 to
1000 C. Record the loss in weight as 3 ml. of HF and 2 drops of H1SO., and
the ignition loss. evaporate to dryness. Heat the cruct'ble
to bright redness and blast at 1100 to
Silica
1200 C. for 5 min. Cool, weigh, and
31. (a) Weigh 1.0000 g. of the mois- repeat blastin~ until constant weight is
ture free (105 to 110 C.) sample into a obtained. The loss in weight represents
platinum crucible which is free from any the SiOs content except for the small
iron stain. Add 0.40 g. of anhydrous amount which is later recovered from
borax {Na,B.01) and 2.00 g. of Na,C01• the alumina precipitate.
Mix well and fuse first at a low red heat,
gradually increasing to the full tempera- Separation of Als01 and P10, from
ture of a blast burner. Blast for 15 min. Fe10a, Ti01, CaO, and MgO
Dissolve the melt in diluted HCl (1: 4) 32. Fuse the residue from the HF
in an evaporating dish, add 2 to 3 ml. evaporation (Section 31 (b)) with a
of diluted HsSO. (1: 1) (Note 1), and little K1S10, or NatSi01, dissolve the
evaporate to dryness. To volatilize bo- melt in diluted HCl (1:4) and add to the
ron, wash down the sides of the dish with
25 ml. of absolute methyl alcohol satu-
..
filtrate from the SiOt precipitate which
was reserved. Reduce the volume to
 CHE~UCAL ANALYSIS OF REFRACTORIES (C 18) 465

about 75 ml. and nearly neutralize with lowed) and reserve for the determination
NaOH (10 per cent). Slowly and while of Ti0 2 (Section 34). Pour the remaining
swirling constantly pour into a 250-ml. solution into a 250-ml. beaker and evap-
volumetric fl.ask containing 100 ml. of orate to dryness. Add 25 ml. of HCl
NaOH (10 per cent} to which has been (1: 1), and heat to boiling. From this
added 1 g. each of Na2COa and Na,Ot point proceed in accordance with the
(Notes 1 and 2). Warm on a water bath procedure described in Section 22 for the
for 1 hr. and cool to room temperature. determination of Fe 20 1 in the analysis of
Dilute with water to the mark, mix well, silica refractories.
and let stand a few minutes. Filter off
Titania
the precipitate on a filter paper which
has been previously washed with NaOH 34. Add to the 10.0-ml. portion previ-
(10 per cent) and dried, catching the ously reserved (Section 33) 10 ml. of
filtrate in a dry beaker. Do not wash H 2SO, (1: 1) and evaporate to fumes.
the precipitate. From the filtrate take Dilute with 10 to 20 ml. of water, and
a 100-ml. aliquot for the determination cool. Transfer to a 100-ml. volumetric
of A\ 20 3 (Section 36) and another 100- fl.ask, and dilute to the mark. Transfer
ml. aliquot for the determination of exactly half of the solution to another
PsOb {Section 35). 100-rnl. volumetric fl.ask. Dilute one of
NOTE 1.-The Na.COa is used to precipitate
the flasks to the mark with H 2S0 4 (1: 19).
calcium and magnesium along with iron and To the other add 5 ml. of 3 per cent
titanium. H 20 2 , then dilute nearly to the mark
NOTE 2.-The NaiOi is used to oxidize any with H 2SO, (1: 19), adjust temperature
chromium present so that it may later be sepa- to 25 ± 1 C., then adjust volume exactly
rated from the AlaOa.
and let stand at least 5 min. Transfer a
Iron Oxide portion from the first flask to a photom-
33. Dissolve the precipitate on the eter cuvette and set the potentiometer
filter paper and that adhering to the fl.ask scale reading at zero. Then measure the
with hot diluted HCl (1 :4). Adjust the absorption of the solution in the second
volume to about 100 ml., add 5 g. of fl.ask. Read the percentage of titania pres-
NII,Cl, and heat to boiling. Add ent from a calibration curve. Construct
NH.OH while stirring slowly until there this curve by adding varying amounts of
is a slight excess. Again beat to boiling, the standard titania solution to HiSO,
filter hot, and wash five times with hot (1: 19), develop color with 5 ml. of 3 per
NH,Cl (2 per cent). Place the paper cent H202, let stand 5 min. and read the
containing the precipitate in the original absorption, using HiSO, (1: 19) for the
beaker, add 15 ml. of HCl, stir to mace- zero setting of the potentiometer scale.
rate the paper, dilute with hot water to NOTE 1.-If spectrophotometer is used, meas-
100-ml. volume, and precipitate with ure the absorption at mean transmission of 420
NH,OH. Filter, and wash as before. mµ. If a filter photometer is used, use a glass
with maximum transmission in the region of
Combine the two filtrates and reserve 420mµ.
for the determination of CaO and MgO. NOTE 2.-As an alternate method add to the
Dissolve the precipitate in hot diluted 5.0-ml. portion previously reserved (Section 33)
HCl (1:4). and cool. Pour into a 10 ml. of HaSO, (1: 1) and evaporate to fumes.
!~(}.ml. volumetric flask, dilute with Dilute with 10 to 20 ml. of water, and cool.
Pour into a 100-ml. volumetric flask, add a few
water to the mark, and mix. Draw drops of Ha02 (30 per cent), dilute to the mark,
off exactly 10.0 ml. of the solution (or and mix. To a 100-ml. volumetric flask, add ex-
5.0 ml. if the alternate method is fol- actly 20.0 ml. of the standard Ti02 solution and
!H3
466 Cm:MICAL .ANALYSIS O• RERACTORIES (C 18)

10 ml. of H1SO, (1:1). Add a few drops of H102 to 2 min., filter hot, and wash five to six
(30 per cent), dilute to the mark, and mix. times with hot NH.Cl (2 per cent).
Compare the colors either in Nessler tubes or in
a suitable colorimeter. Place the paper containing the precipi-
tate in the original beaker, add 15 ml.
Phosphorus Oxide of HCl, and, with the aid of a glass
35. Acidify with HN01 the 100-ml. stirring rod, macerate the paper. Add
aliquot previously reserved (Section 32). boiling water to make up to a volume
Dilute to 200-ml. volume and pre- of about 250 ml. Repeat the precipita-
cipitate the Als01 and P,O, by adding tion with NH.OH and filter as before.
a slight excess of NILOH solution. Wash thoroughly with hot NH.Cl (2 per
Filter, wash a few times with hot water, cent), using at least 250 ml. of wash
and dissolve the precipitate in 100 ml. of solution.
diluted HNOa (1 :9). Transfer to a (b) Place the paper containing the
300-ml. Erlenmeyer flask, heat to boiling precipitate in a platinum crucible which
and add a saturated solution of KMnO, has been previously weighed with its
until a pink color persists. Decolorize cover. Carefully burn off the paper,
with a, few drops of sulfurous acid, add blast at not less than 1200 C. for 10 min.,
15 ml. of HNO,, cool, adjust the volume cool, and weigh. Repeat blasting and
to about 125 ml., and add 20 ml. of weighing until constant weight is ob-
NH.OH and 40 ml. of molybdate re- tained. The increase in weight is the
agent. Shake for 10 min., let stand for Al,O, and P,O, plus a little Si02 which
30 to 40 min., and filter. Wash the has escaped precipitation. To obtain
flask, paper, and precipitate at least the percentage of Al,O,, subtract the
ten times with KN01 solution (1 per P,O, (Section 35) and SiO, (Section
cent). Return the paper and precipi- 36 (c)).
tate to the flask. Add an excess of (c) To determine the Si01, fuse the
standard NaOH solution and 25 ml. of precipitate with about S g. of K1SJ>,
water (both free from CO,). Shake or Na,S,O,. Dissolve the melt in warm
until the precipitate is dissolved and water, add about 30 ml. of diluted H1S01
dilute to a 100-ml. volume. Add 2 to 3 (1: 1) solution. Evaporate to copious
drops of phenolphthalein (0.2 per cent), fumes of H1SO, and cool. Take up with
then titrate with standard HN01 to the water, filter, wash, ignite in a platinum
disappearance of the pink color. Sub- crucible, and weigh. Treat the residue
tract from the original NaOH added the with 5 ml. of HF and 2 to 3 drops of
volume equivalent to the HNOa required H2SO,, evaporate to dryness, ignite, and
to discharge the color. The difference weigh. The loss in weight is Si01• Add
gives the milliliters of NaOH equivalent to the original silica determination.
to the P,O, present. Cakulate the per- Lime
centage of P20..
37. Determine CaO in the filtrate
Alumina from the Fe,Oa precipitate according to
36. (a) Dilute the 100-ml. aliquot, the method described for the determina-
previously reserved (Section 32), to a tion of lime under analysis of fueclay
250-ml. volume, make acid with HCl refractories (Section 14).
(1: 1), and add 5 g. of NH,Cl. Heat to
boiling, add NH,OH very slowly, while Magnesia
stirring constantly, until there is a slight 38. Determine MgO by the method
excess as shown by methyl red. Boil 1 described for the determination of mag-
~•
 Cm:KICAL ANALYSIS OP hl'RACTOJUES (C 18) 467

nesia under analysis of fireclay refrac- as a tightly rubber-stoppered bottle) un-

tories (Section 15). til analyzed, since finely ground, dead-
Alkalies burned magnesite gradually absorbs and
fixes moisture and carbon dioxide.
39. Proceed as in the determination (b) For dolomite refractories, two sam-
of alkalies under fireclay refractories in ples are required. These shall consist of
Section 16, or alternatively in Section 79. the following:
If the procedure in Section 16 is followed, (1) A sample for the determination of
use these precautions: the loss on ignition. It shall consist of
(a) Fuse the sample at 1200 C. to 10 lb. carefully obtained and then re-
obtain complete decomposition of the duced as rapidly as possible to pass com-
material. pletely a No. 16 (1.19-mm) sieve' by
(b) Cool the top of the crucible to one pass through a suitable mill. This
prevent any loss of alkalies by volatiliza- crushing must be done in 3 min. or less.
tion. Cooling can be accomplished more The crushed material shall be split re-
euily if a platinum crucible of the J. peatedly without delay (preferably in a
Lawrence Smith, long type is used. mechanical splitter) until a 4-oz. sample
MAGNESITE AND DOLOMITE results. This shall be placed immedi-
REFRACTORIES ately in an airtight container (such as a
tightly rubber-stoppered bottle).
Special Solutions Required (2) A sample for the determination of
40. The following solutions will be the Si02, Fe20a, Al201, CaO, and MgO
required: shall be prepared as described in Para-
(a) Hydrochloric Acid (2.5 per cent graph (a).
by TJolurm). Moisture
(b) Potassium Permanganate (Ap-
42. For magnesite refractories, weigh
proximately O.S N).-Dissolve 16 g. of
KMnO. in 1000 ml. of water.
1.00 g. of the sample and heat to constant
(c) Ammonium Chloride - Hydro- weight at a temperature not under 105
nor over 110 C. Record the loss in
chloric Acid Solution.-Dissolve 50 g.
weight as moisture.
of NH.Cl in water, add 50 ml. of HCl
(sp. gr. 1.18), and dilute to 1000 ml. NOTE.-The determination of moisture in
dolomite refractories is omitted, since no free
Preparation of Samples for Analysis moisture exists in these products. Further, they
are sometimes coated with sealing oil or tar.
41. (a) For magnesite refractories, a
10-lb. sample carefully obtained after the Losa on Ignition
manner of one of the approved methods 43. (a) For magnesite refractories,
of sampling shall be crushed to pass a weigh 1.000 g. of the moisture-free (105
No. 8 (2.38-mm) sieve,' mixed thor- to 110 C.) sample and heat to constant
oughly, and reduced by quartering to weight over a blast lamp, or in an electric
about 0.5 lb. This shall be crushed to furnace, at 900 to 1000 C. Record the
pass a No. 20 (841-micron) sieve, 6 mixed, loss in weight as the ignition loss.
and quartered to about 50 g. This 50-g. (b) For dolomite refractories, weigh
sample shall then be ground in a suitable approximately 10 g. of the specially pre-
mortar (agate, mullite, alumina, or boron pared sample (Section 41 (b)) quickly
:arbide) so that it will all pass a No. 100 into a tared porcelain crucible and heat
(149-micron) sieve, 6 mixed thoroughly, to constant weight over a blast lamp, or
and placed in an airtight container (such in an electric muffle furnace, at 900 to
9-«
468 C1rnMJCAL ANALYSIS OF REFRACTORIES (C 18)

1000 C. The loss in weight is calculated rapid filter. Wash the paper and pre-
to percentage loss on ignition. cipitate ten times with a hot HCl so-
Statement of Analysis for Dolomite Re- lution (2.5 per cent), then three times
with hot water. Return the filtrate to
fractories
the casserole in which the first evapo-
44. Prepared samples of dolomite re- ration was made, add 2 ml. of HCI01,
fractories are known to pick up moisture :md evaporate to fumes on the hot plate,
and C0 2 on storage in the laboratory. boiling as before. Cool, take up in acid
For this reason the determinations which and water, filter, and wash as before.
follow are reported on the ignited basis. Reserve the filtrate. Place both papers
The loss on ignition value to be used and precipitates in a platinum crucible.
shall be determined on the material pass- Dry, then burn off the papers and ig-
ing the No. 100 (149-micron) sieve (Sec- nite to constant weight in the muffie or
tion 41 (b) (2)). This determination over a moderate blast. Cool in a desic-
shall be made on the same day the por- cator and determine the combined weight
tions for chemical analysis are weighed. of the crucible plus impure Si01. Add
Char1tcter of the Sample to be W eigbed two drops of diluted H1S04 (1: 1) and 1
ml. of HF to the impure Si01 in the
45. In the following sections, the sam-
crucible. Digest for 5 to 10 min.,
ple weighed shall be as follows:
covered, on the hot plate, then remove
(a) Magnesite.-Use a portion of the
the cover and evaporate to dryness.
material passing the No. 100 (149-
Fume off the HiSO, over a moderate
micron) sieve (Section 41 (a)) which has
heat, and ignite or blast for a few min-
been dried to constant weight at 105 to
utes. Cool in a desiccator, and weigh
110 c.
again. Subtract the weight of the cruci-
(b) Dolomite.-Use a portion of the
ble and residue from the weight of the
material passing the No. 100 (149-
crucible and impure Si02; the difference
micron) sieve (Section 41 (b) (2)).
is the corrected Si01. Fuse the residue
Silica in the crucible by heating gently with a
46. Transfer 1.0000 g. of the sample to small piece of K1St01 or N a1S201 about
a 210-ml. casserole. Moisten the solids the size of a pea, cool, dissolve in water,
with a few milliliters of water, slowly add and add to the reserved filtrate from the
20 ml. of HCl, and cover the casserole. silica determination.
While covered, digest on the hot plate NOTE: Caution.-Boiling perchloric acid is
until no further dissolution occurs and all violently explosive when contaminated by easily
gritty particles have disappeared, break- oxidizable material or organic matter. When
ing up any solid cake from time to time analyzing dolomite refractories containing scal-
with a stirring rod. Cool, uncover (Cau- ing oil or tar, 5 to 10 ml. of HNOa (sp. gr. 1.42)
are added just be/ore the HCIO, to oxidize the
tion, see Note), add 8 ml. of perchloric organic materials, and eliminate the danger of
acid, HCIO. (60 to 70 per cent), and a perchloric acid explosion.
then evaporate to heavy white fumes of
HClO.. Cover and boil so that the acid Iron and Aluminum Oxides
refluxes for 10 to 15 min. Remove from 47. Add to the filtrate from the silica
the hot plate, cool, wet with a few milli- determination (Section 46), which should
liters of water and 5 ml. of HCI, warm be in a 400-ml. beaker, 5 g. of NH.Cl.
for a few minutes on the hot plate, add Dilute from 150 to 200 ml., add three
SO ml. of hot water, heat to approxi- drops of methyl red (0.1 per cent), heat
mately boiling, and filter through a 9-cm. to boiling. Add NH.OH (1: 1) slowly.
H4
 CHEMICAL ANALYSIS OP REFRACTORIES (C 18) 469

finally drop by drop until the solution is Allow to stand for 3 min., transfer with
just alkaline to the indicator. Boil for washing to a 600-ml. beaker containing
1 or 2 min. and filter through a 9-cm. 300 ml. of cold distilled water and 25 ml.
rapid filter. Do not wash. Reserve of MnSO, solution. Titrate with stand-
the filtrate. Transfer the paper and ard 0.1 N KMnO, solution, added very
precipitate to. the beaker in which the slowly while stirring constantly until a
precipitation was made, add 50 ml. of permanent pink end point is obtained.
HCl (1 :4), and digest covered on the
hot plate, shaking frequently until the Alumina
precipitate is dissolved and the paper 49. Subtract the calculated weight of
well macerated. Dilute, add methyl iron oxide (Section 48) from the weight
red, reprecipitate with NH.cOH (1: 1), of R203 (Section 47). The remainder
boil, and filter, exactly as before. Add is the weight of the Alt<)3 plus small
this second filtrate to the first filtrate amounts of oxides of any other elements
for the subsequent determination of lime that were precipitated (Note).
(Section 50 (a)). Swab the beaker and NOTE.-In the complete analysis of magne-
rod and wash the paper and precipitate site refractories, these may include P.0,, Ti01
with NH,Cl solution (2 per cent). Dry, and small amounts of SiOt, present in such
then ignite the paper and precipitate in a small amounts as to be disregarded. Dolomite
tared platinum crucible at 1000 to refractories sometimes contain appreciable
amounts of Ti02, in which case it should be
1100 C. for 1 hr. with access to air. Cool determined, and a correction made for the FCtOr
in a desiccator, and weigh as Rt03 which and Al.Or determined according to Sections 48
is mainly the Fe20a and Al20a. and 49.

Iron Oxides Lime

48. Weigh 2.00 g. of the sample into a 50. (a) Method A: WhenMgOistobe
150-ml. beaker, and cover. Dissolve by Determined Directly.-Acidify with sev-
slowly adding HCl, keeping the beaker on eral milliliters of HCI (50 per cent) the
the hot plate. Finally, add about 5 ml. combined filtrate from the RsOa precipi-
of HCl in excess. Heat to gentle boiling tation (Section 47) and evaporate to a
until the siliceous residue undissolved ap- volume of about 250 ml. Heat to in-
pears iron-free. Filter, and fuse the cipient boiling, add 25 ml. of a saturated
residue with a very little K1Ss01 or (NH,)tC204 solution (about 4 per cent)
Nas~01 in a porcelain or vitreous silica and NH.cOH until slightly alkaline.
crucible to avoid introducing platinum. Stir thoroughly and allow the precipitate
Dissolve the fused material and add to to settle out, keeping the solution warm
the filtrate. Add a few drops of a strong for 3 to 4 hr. but not boiling. Filter and
solution of KMnO, (about 0.5 N) to oxi- wash five times with cold water. Save
dize any organic matter that may be the filtrate. Dissolve the washed oxa-
present. Boil. When the excess of late precipitate with SO ml. of diluted
KMnO, is decomposed and chlorine HCl (1: 1), dilute to 250 ml. and precipi-
fumes have disappeared, reduce the iron tate as before. Wash the beaker, pre-
by adding SnC12 solution drop by drop cipitate, and paper with cold water until
from a pipette while swirling the beaker (1'11I.)2C20t is removed. Avoid exces-
constantly until the solution is colorless, sive washing. Combine the filtrates for
then add 1 drop in excess. Cool quickly the determination of MgO (Section 51).
in running water, then add at one stroke To the beaker used for the precipitation,
15 ml. of saturated HgCl2 solution. add 150 ml. of HiSO, (5 per cent), then
t-4'
470 Cm:lllCAL ANALYSIS OF RERACTORIES (C 18)

introduce the paper containing the using abcut 35 ml. -If the Si01 is high
precipitate. Heat to almost boiling and (over 6 per cent) it is advisable to filter
titrate with 0.1 N KMn04 • A blank it off at this point before precipitating
should be previously determined for the the lime. Dilute to about 150 ml. with
effect of the paper. hot water, make alkaline with NH40H,
A small part of the CaO occasionally beat to boiling, and precipitate the cal-
escapes the oxalate precipitation and cium by slowly adding 15 ml. of a satu-
comes down subsequently with the MgO. rated solution of (NH.hC 20 4 • Digest
To the CaO precipitated by the oxalate on the hot plate just below boiling for
should accordingly be added that found 30 min. Filter on a 9-cm. paper, wash
in the MgO. with cold water until free from oxalate,
(b) Method B: When MgO is to be testing final washing with a drop of
Determined by Difference. 8-Weigh 2.50 g. KMnO, after acidification. Transfer
of the sample into a 250-ml. beaker, add the paper and precipitate to the original
10 ml. of water, cover, slowly add HCI beaker, dissolve with 150 ml. of H,S01
until no further dissolution occurs, stir- (5 per cent), and titrate as in Method A
ring frequently and avoiding excessive (Paragraph (a)). Correct for a blank
boiling. Then add about 5 ml. of HCI on the filter paper.
in excess. Evaporate on a hot plate un-
til very thick and syrupy. At this stage Magnesia
crusts will be forming on the surface and 51. (a) Acidify with several milli-
a drop of material placed on the wall of liters of HCl (SO per cent) the combined
the beaker will immediately solidify. filtrates reserved during the determina·
Remove the beaker from the hot plate tion of CaO (Section SO (a)). Dilute to
and allow to partially cool. Add 5 ml. of 500 ml. in a calibrated flask and take a
H2SO. (10 per cent). Cover with a 50 ml. (T~) aliquot portion. Since a 1-g.
watch glass, heat on a hot plate until sample was originally used, this will yield
all soluble matter is dissolved. Cool aliquot portions containing between
slightly, but not until solids separate. 0.08 to 0.09 g. of MgO. Dilute to
Add 50 ml. of alcohol, stir thoroughly, 250 ml. and add 2 to 3 g. of diammonium
and heat just to boiling on a hot plate.
phosphate ((NH.hHP04), stir until dis·
Filter with suction through a Gooch cru- solved. Add NH.OH slowly until alka·
cible prepared with a moderately thin line and then 30 ml. of NH,OH in excess.
asbestos mat. If the Si02 is less than Allow the solution to stand for at least
3.5 per cent, filter through paper. Rinse
4 hr., preferably overnight if the total
the beaker twice with alcohol, pouring amount of solution exceeds 400 ml. in
the washings through the crucible. volume. Filter, and wash with diluted
Drain under suction. Remove the cruci- NIL.OH (5 per cent). Dissolve from
ble, scrape out the mat with a glass rod the paper with hot HCI ( 1: 1) and
into the beaker in which the precipitation reprecipitate by adding 0.1 g. of
of sulfate was made, wash the crucible (NH.)2HP04, diluting, rendering ammo-
clean with a hot NH 4Cl - HCl solution,
niacal, and letting stand as before.
s This procedure i8 applicable to magnesite After filtering and washing as before,
refractories only. In this specific procedure very
much less lime escapes precipitation and thus char the paper and burn off the carbon
the magnesia "by difference" ia not subject to below 900 C., then ignite to constant
this plus error. The procedure further has the weight in an electric muffi.e furnace at
advantage that an individual lime determination
may be made without separating the other usual about 1100 C. The ignited material
constituents. consists of magnesium pyrophosphate,
9-41
 CHEMICAL ANALYSIS OP REPRACTORIES (C 18) 471

Mg2P:Ch, as well as some of tht: CaO and per cent) until the residue is free from
MnO. Correction may be made for permanganic acid, as determined by the
these as follows: absence of the pink color in the wash-
(b) Dissolve the ignited pyrophos- ings. Titrate, at once, the perrnanganic
phate in a little diluted H2S04 and add acid formed, with standard sodium
enough absolute alcohol to make up 90 arsenite (1 ml. - 0.0002 g. of Mn).
to 95 per cent of the final volume. Any The end point is the disappearance of
calcium present will settle out as CaS04• the pink color. The factor for convert-
Filter, wash with alcohol, and dissolve ing Mn to MnO is 1.291.
the precipitate in hot diluted HCl.
NoTE.-A good grade of asbestos, acid
Precipitate with oxalate and titrate with washed, and specially prepared for Gooch
permanganate. Add this CaO to that crucibles, should be used.
previously found. Calculate the CaO
CHROME ORE, CHROME REFRACTORIES,
to tricalcium phosphate, Caa(PO.)t, by
multiplying by the factor 1.845 and de- CHROME-MAGNESITE REFRACTORIES,
duct this amount from the weight of the AND MAGNESITE-CHROME
REFACTORIES
Mg2P201. The correction to the first
calcium precipitate is ten times the NOTE 1.-This method is not intended to
weight of the CaO precipitated as sulfate. cover electrically fused &luminous refractories
(c) The manganese which contami- containing chrome ore.
NOTE 2.-No procedure is given for loss on ig-
nates the Mg2P201 is determined in the nition, since the exact ignition conditions to yield
alcoholic filtrate from the CaS04• precision results for chrome ore and chrome
Evaporate this filtrate until strong refractories have not yet been established.
fumes are evolved and all organic matter
Reagents
destroyed. Cool, dissolve the residue
in 25 ml. of HN01 (30 per cent) and 53. (a) Ammonium Hydroxide (2.5 per
determine the manganese by the bis- ant by volume).
muthate method, as described in Section (b) Ammonium Hydroxide (JO per
52, omitting the Na 2COa fusion. Cal- cent by volume).
culate any manganese found to MniP10 7 (c) Ammonium Hydroxi<k (25 per
by multiplying by the factor 2.585 and cent by volume).
deduct this weight from the MgiP10 7 • (d) Hydrochloric Acid (5 per cent by
volume).
Manganese as MnO (e) Hydrochloric Acid (JO per cent by
52. Mix intimately 1.0 to 5.0 g. of the volume).
sample (depending upon the amount of (j) Phosphoric Acid, HaPO, (sp. gr.
manganese expected) with 0.5 to 2.5 g. of 1.71).
anhydrous Na2C01 and ignite in a plati- (g) Ammonium Chloride Solution (I
num crucible for 15 min. at about 1000 C. per cent).-Dissolve 1 g. of NH,Cl in
Dissolve the sintered melt in 50 ml. of 100 ml. of water.
HNOa (30 per cent), and if there is any (h) Ammonium Persulfate Solution.-
evidence of brown hydrated Mn01 add Dissolve 25 g. of (NH.)2S20s in 80 ml.
sulfurous acid drop by drop. Boil the of water.
solution, filter if necessary, cool to room (i) Standard Ferrous Ammonium Sul-
temperature or lower, and add 2 to 3 g. fate Solution (0.2 N).-Dissolve 78.4 g.
of sodium bismuthate. Agitate the solu- of Fe(NH,)2(S0,)1·6H20 in 1 liter of
tion for 1 min. and filter through as- water (Note). Standardize against 0.2 N
bestos (Note). Wash with HNOa (3 KMnO,. This standardization must
H9
472 CHEMICAL ANALYSIS 0!' RuR.AcroRIES (C 18)

be made each day the solution is used, walls, and cool. Place crucible and lid
as it gradually weakens due to oxidation. in a covered 600-ml. beaker, and add
Non.-Ferrous ammonium sulfate crystals
250 ml. of cold water. After reaction
may be used instead of the standard solution. has stopped and the melt is completely
dissolved, remove the crucible and lid,
(j) Standard Potassium Permanganate
and wash them thoroughly (Note 3).
(0.2 N).-Dissolve 6.5 g. of K.MnO, in
1 liter of water. Standardize against Non 1.--0rdinary iron crucibles, which
the National Bureau of Standards' may contain significant &mounts of chromium,
should be avoided. Crucibles made of ingot
standard sample 40c of sodium oxalate.
iron are satisfactory.
(k) Potassium Permanganate Solution Non 2.-If about 0.2 g. of very finely pul·
(2.5 per cent).-Dissolve 2.5 g. of KMnO, verized sugar charcoal is mixed with the Nas(),,
in 100 ml. of water. only about S g. of the peroxide is required. Not
(l) Silver Nitrate Solution (0.5 per more than 0.2 g. of sugar charcoal is recom·
mended in order to avoid the poSl!ibility of explo-
cenl).-Dissolve 0.5 g. of AgN01 in sion. After heating for about 30 sec., ignition
100 ml. of water. of the sugar charcoal takes place, and the cruciblt
(m) Sodium Chloride Solution (JO per will suddenly become a dull red on the outside.
cent).7 Dissolve 10 g. of NaCl in 100 ml. The total time required to complete the fusion
of water. is about 2 min. By thus shortening the time
for fusion and contact with the flux, the life of
(n) Ortho-Phenanthroline Indicator the crucible is extended.
(Ferroin). Non 3.-If sugar charcoal is used, it ii
advisable to add 1to2 g. of N&t<>s at thia point.
Preparation of Samples for Analysis
54. The 50-g. sample referred to in (b) Boil for 10 min., or digest at 80
Section 3 shall be ground so that it to 90 C. for 30 to 40 min., stirring
will all pass a No. 200 (74-micron) occasionally to decompose the excess
sieve.1 NasO.. Cool, and add 60 ml. of HsSO,
(1: 1) and 5 ml. of HNO,. Boil until
Moisture the iron scale from the crucible is dis-
55. Weigh 1.00 g. of the sample and solved. Add 20 to 25 ml. of AgN01
heat to constant weight at a tempera- (0.5 per cent), 1 to 2 drops of K.MnO,
ture not under 105 nor over 110 C. (2.5 per cent), and 10 to 15 ml. of
Record the loss in weight as moisture (NH,)sSsOa solution. Boil for 5 min.
Add 20 ml. of NaCl (10 per cent) and
Chromium Oxide boil for 5 to 10 min. after the per·
56. (a) Transfer 0.5000 g. of the mois- manganate or MnO. has completely
ture-free (105 to 110 C.) sample to a disappeared. Cool to room tempera·
30-ml. pure iron crucible (Note 1). Mix ture, and add 3 ml. of HaPO,.
with about 8 g. of dry Na101, cover (c) Now add a measured excess of
with an iron lid, and heat over a very low ferrous ammonium sulfate, either as
Bame (Note 2). After the charge is crystals or as a standard solution. If
melted, increase the temperature slowly the crystals are used, add exactly 5 g.
to a medium red (650 to 750 C.), and and stir until completely dissolved. If
continue heating for 3 to 5 min. Do a solution is used, add from a burette
not continue heating longer than this, the standard 0.2 N solution with 5 ml.
because attack of the fl~ is quite rapid in excess. The chromium will thus be
at red heat or above. After removing reduced to the trivalent state and will
the crucible from the heat, rotate to change the color of the solution from
spread the contents over the crucible amber to a bright green. Add two drops
~
 CIIEMICAL ANALYSIS OF REFRACTORIES (C 18) 473

of ortho-phenanthroline (Feuoin) as an (Note) of dry Na10 1 , cover with lid,

indicator. (The solution will become and fuse as described in Section 56.
turbid and pink in color.) Titrate the Non.-The u&e of only 2 to 3 g. of N320t
ferrous ammonium sulfate, in excess of reduces the amount of nickel dissolved which
that required to reduce the chromium, might interfere with the subsequent colorimetric
with 0.2 N KMnO, (Note 4). The end determination of Ti01. The sm&!l amount of
nickel actually dissolved is entirely removed by
point is reached when one drop of the the double NH.OH precipitation.
standard KMnO, solution causes the
turbidity to disappear and the solution to (b) Take up the fusion in about
become clear green in color. Designate 250 ml. of cold water. After reaction
the first titration in either case as T.. If has stopped and the soluble constituents
ferrous ammonium sulfate crystals were of the melt have completely dissolved,
used for reduction, titrate 5 g. of the crys- remove the crucible and cover and wash
tals dissolved in H 1SO, (5 per cent) with thoroughly. Boil for 5 min. Filter
the 0.2 N KMnO, solution. Designate the iron and titanium oxides, and wash
the second titration as T 1 • If the 10 times with hot water. Dissolve the
standard ferrous ammonium sulfate so- precipitate from the paper with diluted
lution was used, Ts is the milliliters of HCl (1: 1), and precipitate with NH,OH
0.2 N KMnO, equivalent to the total in. excess. Filter, and wash thoroughly
volume of 0.2 N ferrous ammonium with hot water. Again dissolve with
sulfate originally added, including the HCl (1: 1) and reprecipitate with excess
5 ml. excess. In either case, T1 - T1 NH40H, filter, and wash. Finally dis-
then equals the milliliters of 0.2 N solve the precipitate with hot H 2S0 4 (1:
KMnO, equivalent to the weight of 19) using a total of 125 ml. for dissolving
ferrous ammonium sulfate required to and washing, and catching the solution
reduce the chromium in the sample. and washings in a 250-ml. volumetric
Calculate the percentage of chromium flask. Cool, and dilute to the mark. In
oxide, as follows: case the alternate method is followed
dissolve the precipitate with hot diluted
CriC>i, per cent -
H,SO, (1: 9), cool, add several drops of
(T1 - T1) X N of KMnO, X 0.02534 H201, and dilute to 250 ml.
w x 100
Titania
where:
0.02534 = Cr20a equivalent of the NOTE.-This determination is optional since
it is made infrequently. '
KMnO,, and
W = weight of the sample in 58. Transfer two 50-ml. aliquots to
grams. 100-ml. volumetric flasks. Dilute one
flask to the mark with H 2SO, (1: 19).
NOTE 4.--A 0.2 N K1Cr10r solution may be
To the other flask add 5 ml. of 3 per cent
satisfactorily substituted for the 0.2 N KMnO,
tolution, to titrate the excess ferrous ammonium H202, and dilute nearly to the mark.
sulfate. If the K1Cr10r solution is used, the in- Adjust the temperature to 25 ± 1 C.
dicator should be sodium cliphenylaminesulfon- then make final volume adjustment. Let
ate and the titration should be made to a stand. at least 5 min. before reading ab-
permanent blue color. sorption. Transfer a portion from the first
flask to a photometer cuvette and set the
Titania and Ferrous Oir:ide
potentiometer scale reading at zero. Then
57. (a) Transfer 0.5000 g. of the mois· measure the absorption of the solution in
ture-free (105 to 110 C.) sample to a 30- t?e ~cond flask. Read the percentage of
ml. nickel crucible. Mix with 2 to 3 g. titama present from a calibration curve.
t-42
474 CHEMICAL ANALYSIS OF R.EFRACTORmS (C 18)

Construct this curve by adding varying or 1000-ml. beaker coontaining 300 ml.
amounts of the standard titania solution of cold distilled water and 25 ml. of
to H:aSO, (1: 19), develop color with 5 ml. the MnSO, solution. Titrate with
of 3 per cent H 101, let stand 5 min. and standard 0.04 N KMnO, solution, added
read the absorption, using H2SO, (1: 19) very slowly while stirring constantly,
for the zero setting of the potentiometer until a permanent pink end point is
scale. obtained. A blank should also be
NOTE 1.-If a spectrophotometer is used,
determined. Calculate the iron to FeO.
measure the absorption at mean transmission Silica
of 420 m11• If a filter photometer is used, use a
glass with maximum transmission in the region 60. Fuse 0.5000 g. of the dried (105
of 420 IDJ.I. to 110 C.) sample in approximately
Nou 2.-As an alternate method take 100-
10 g. of fused K:aS10J or NazSz01 in a vi-
ml. aliquot portion from the 250-ml. acid solution
(Section 59) and determine the Ti01 colori- treous silica or porcelain crucible (Note).
metrically by placing the solution in a 100-ml. The usual precautions of a pyrosulfate
Nessler tube and comparing it with the color of fusion should be observed; that is, the
an appropriate amount of the standard Ti01 flame should be turned very low at the
solution in another Nessler tube. Both the solu-
tion of the sample and the reference standard
start so that the pyrosulfate just melts,
solution shall be cooled to room temperature then the flame increased so that the
before making the comparison. Dilute the refer- pyrosulfate remains in a liquid condition,
ence standard with a measured amount of water and finally the temperature gradually
from a buret until the color is matched. From raised until the crucible reaches a duU
the amount of water added, calculate the per-
centage of Ti01 in the sample. A standard red heat. The object of the slow heating
colorimeter may be used instead of the Nessler is to retain the H1SO,, which is the
tubes. effective part of the fiux. Excessive
fuming is to be avoided because it may
Ferrous Ozi.de
result in "crawling" of the melt up the
59. (a) Use the remaining 150-ml. por- sides of the crucible, producing the diffi-
tion of the acid solution (Section 57), cultly soluble anhydrous Cr1(S0,)1. The
but boil off the H,O,; or, if TiOz is not most refractory materials should be
to be determined, use the entire 250-ml. thoroughly fused in 1to1l hr. Remove
of solution (without adding H101). the crucible from the burner, and allow
Nor1t.-It is equally satisfactory to add the to cool in the air. The pyrosulfate
100-ml. aliquot portion used for the TiO. deter- melt will shrink from the crucible, and
mination in Section 58 to the remaining 150 ml. the main mass may be removed easily
in order to obtain the advantage in accuracy
in one cake. Dissolve the cake and the
of using the entire 0.5-g. original sample.
small portion retained on the crucible
(b) Evaporate to 100 ml. or less, add walls in 200 ml. of hot diluted HCl (1: 1),
25 ml. of diluted HCl (1:1), and heat to boiling until the solution acquires a
boiling. Reduce the iron by adding clear, bright green color. Evaporate to
SnCia solution drop by drop from a dryness, and bake at not over 120 C.
pipette, while swirling the beaker con- until free from fumes of HCl, which
stantly, until the solution is colorless. will require about 1 hr. Moisten with
Then add one drop in excess. Cool hot diluted HCl (1: 1), then add about
quickly in running water, then add at 25 ml. more acid, warm until all salts
one stroke 15 ml. of saturated HgClt are in solution, and filter off the SiO..
solution. Allow to stand for 3 min.; Evaporate the filtrate to dryness, bake,
then transfer with washing to a 600-ml. and filter again. Wash the SiOz thor·
Ml
 CHEMICAL ANALYSIS OE' REFRACTORIES (C 18) 475

oughly with diluted HCl 15 per cent). boil, and filter, exactly as before. Add
Ignite in a platinum crucible to con- this second filtrate B to the first filtrate
stant weight in an electric muffie for the subsequent determination of
furnace or over a moderate blast, and CaO and MgO (Section 62). Swab
weigh. Add two drops of diluted H1S04 the beaker and rod, and wash the paper
(1: 1) and about 10 ml. of HF to the and precipitate with hot NH,Cl (2
crucible. Evaporate, ignite, and weigh per cent). Place the paper and pre-
as before. The difference represents the cipitate in a tared platinum crucible and
amount of Si01 present in the sample. dry. Carefully bum off the paper,
Fuse the residue left in the crucible with and then blast at 1200 C. to constant
a small amount of pyrosulfate and add to weight (30 min. is generally sufficient).
the filtrate from the Si01. Cool in a desiccator, and weigh with the
Non:.-Decomposition by means of per- crucible covered by the lid. The ig-
chloric acid is permissible where it has been nited precipitate consists of Fe.O,,
found that the particular chrome ore or burned A1101, Cr,O,, and TiO,. To obtain the
refractory is amenable to such treatment. percentage of Al201, subtract from the
High-alumina ores and hard burned refractories
are generally difficult to get completely into weight of ignited precipitate the FeO
solution with perchloric acid.' (Section 59) calculated to Fe201, plus
the Ti01 (Section 58) and the Cr101
Alumina
(Section 56).
61. (a) Method A: R,O, Precipitated (b) Method B (Optional): RsOa Pre-
with NH,OH.-Combine the filtrates cipitated 'Wi.th Urea.-Dilute the entire
from the silica determination (Section combined filtrates from the silica
60) and dilute to exactly 500 ml. determination (Section 60) to about
in a graduated flask. Take a 300- 400 ml., preferably in a 600-ml.
ml. aliquot portion and add 15 g. of beaker. Add 15 g. of NHcCl, then add
NH,Cl and 3 drops of methyl red solu- NH,OH slowly until a slight permanent
tion (0.1 per cent). Heat the solution cloudiness appears. Clear the solution
almost to boiling, and add slowly NH,OH by adding diluted HCI (1:1) drop by
(1:1) until the indicator has changed drop and then 2 to 3 drops in excess.
to a yellow color. Boil for 1 or 2 min. Now add 6 g. of urea (Note). Bring
to remove excess ammonia, and allow to a boil and continue boiling gently
the precipitate to settle. Filter through on the back of a hot plate for 2 hr.
an 11-cm. rapid filter paper. Do not after the first opalescence appears. The
wash. Reserve the filtrate A for the decomposition of the urea causes a
determination of CaO and MgO (Sec- gradual increase in the pH of the solu-
tion 62). Transfer the paper and pre- tion, which is uniform throughout, and
cipitate to the beaker in which the pre- precipitates the R10 1 group as the dense
cipitation was made, add 50 ml. of HCl basic sulfates. The final pH value of
(20 per cent), and digest covered on the the solution is about the same as when
hot plate, shaking frequently until the using NH,OH and the proper indicator.
precipitate is dissolved and the paper Remove from the hot plate, and allow
well macerated. Dilute, add methyl the precipitate to settle, keeping the
red, reprecipitate with NH,OH (1: 1), solution warm. Filter through two
' For a description of a procedure using dense filter papers folded together, and
pP.rchloric acid, see T. R. Cunningham and T. R. transfer the bulk of the precipitate to
~fcNeil, "Analysis of Chrome Ores,'' Industrial
and Enur. Chemistry, Anal. Ed., Vol. I, p. 70
the papers. The basic sulfate precipi-
(1929). tate occupies only about one third
t-39
4i6 CHEMICAL ANALYSIS OF REFRACTORIES (C 18)

of the volume of the equivalent hy- allow to stand overajght at room tern·
droxide and is, accordingly, easily trans- perature. Filter the solution on a
ferred and washed. For transferring 9-cm. filter paper, and wash the precipi-
the . precipitate, use NH,Cl solution tate three or four times by decantation
(1 per cent) made just alkaline to with cold NH,OH (2.5 per cent). Pour
methyl red indicator with NaOH. 25 ml. of HCl (1: 1) through the filter
The small amount of precipitate adher- paper, collecting the filtrate in the
ing to the sides of the beaker may be original beaker containing the bulk of
removed by thorough scrubbing with a the precipitate. Wash the paper thor·
rubber policeman, or better, by dis- oughly with HCl (5 per cent). Dilute
solving with diluted HCl (5 per cent) the solution to a volume of 150 ml.
directed against the side of the beaker with cold water, add 3 ml. of
from a wash bottle. Heat gently, add (NH,)1HPO, (10 percent), and repeat the
one drop of methyl red indicator, and precipitation. Allow to stand several
then diluted NH.OH (10 per cent) hours. When cool, filter, wash 10 to 12
until just alkaline. Transfer the small times with cold NH,OH (2.5 per cent),
amount of precipitated hydroxide to the and ignite in a weighed platinum or por-
filter· paper holding the basic sulfate celain crucible, first at a dull red heat
precipitate, using the NH,Cl solution until the carbon has been removed, and
(1 per cent) referred to above. Scrub finally to constant weight at about 1100
the beaker, and then wash the combined C. (20 to 30 min. is generally sufficient).
precipitates 10 times with the above The pyrophosphates thus obtained con·
solution. Reserve the filtrate for the tain all the calcium, magnesium, and any
determination of CaO and MgO (Sec- manganese present.
tion 62), and complete the procedure as (b) Dissolve the ignited pyropbos-
described in Paragraph (a). phates in 10 ml. of H 2SO, (1: 1), and
Non:.-The precipitate obtained with urea evaporate to fumes of S03• Cool, and
is fine grained and much smaller in volume than add 5 ml. of water and enough absolute
the gelatinous hydroxides obtained with NH.OH. ethyl or methyl alcohol to constitute 90
Thi! permits rapid filtering and more thorough to 95 per cent of the total volume of
washing." liquids present. Stir vigorously for
Lime several minutes. Allow the CaSO, pre-
cipitate to settle for 2 or 3 hr., or over·
62. (a) Make the combined filtrates night if preferred, filter on a 9-cm.
A and B from Section 61 (a), or the paper, and wash free of H,PO, with
filtrate from Section 61 (b), slightly acid ethyl alcohol (80 per cent). Dissolve
with HCL, evaporate to a volume of the precipitate from the paper with HCI
approximately 150 ml., and cool. Add (10 per cent), and wash with about 25 ml.
20 ml. of diammonium phosphate of the acid. Dilute to about 150 ml.
((NHt)1HPO.) solution (10 per cent), with hot water, make alkaline with
and slowly introduce NH,OH drop by NH,OH, heat to boiling, and precipitate
drop, while stirring vigorously, until the calcium by slowly adding 10 to 15
the solution is ammoniacal and a ml. of a saturated solution of ammonium
crystalline precipitate appears. Add oxalate. Digest on the hot plate just
15 ml. of NH,OH, stir thoroughly, and below boiling for 30 min. Filter on a
'°For a description of an investigation of this pro·
cedure, '"" Hobart H. Willard and Ning Kang Tang.
9-cm. paper, and wash with cold wata
"Quantitative Determination of Aluminum by Precipita· until free from oxalate. Transfer the
tion with Urea/' Industrial and Engr. Chemislry, Anal.
Ed., Vol. 9, pp. 357 to 363 (1937). paper and precipitate to the original
~39
 CHEMICAL ANALYSIS 01!' REFRACTORIES (C 18) 477

beaker, and dissolve with..150 ml. of determined in Section 63 to Mn2Pi01

H,so. (5 per cent). Heat almost to by multiplying by the factor 2.001.
boiling, and titrate with 0.04 N Deduct the sum of these two phosphates
KMnO •. from the weight of the combined phos-
phates of calcium, manganese, and mag-
Manganese as MnO nesium in Section 62. The difference
63. (a) Evaporate the alcoholic fil- is the weight of Mg2P201. The factor
trate from the Caso. (Section 62) until for converting M~P~1 to MgO is
strong fumes of SOa are evolved and all 0.3621.
organic matter is destroyed. After the
solution has cooled, dissolve the residue CARBON AND CARBON - CERAMIC
in 25 ml. of HNOa (30 per cent). The REFRACTORIES
oxidation of the manganese from this
Application
point may be effected by either of the
following two methods: 65. (a) Sections 67 to 76 are intended
Pusulfale Method.-Add 75 ml. of to apply only to baked or fired carbon -
water, 6 ml. of AgN0 3 solution (0.5 ceramic refractories containing no silicon
per cent), and 10 ml. of (1'.ti,)zStOs carbide. On these refractories the mois-
solution. Boil for 1 min. and cool to ture determination shall be made on,
room temperature. and reported as the percentage of, the
Bismuthate M ethod.-If there is any as-received or natural-basis sample;
evidence of brown hydrated Mn().z the loss on ignition determination shall
after solution in the HNOa, add H2SOa be made on a moisture-free (105 to
drop by drop. Boil the solution, filter 110 C) sample; and all other determina-
if necessary, cool to room temperature tions shall be made on moisture-free
or lower, and add 2 to 3 g. of sodium and carbon-free samples. The carbon-
bismuthate. Agitate the solution for 1 free samples shall be obtained either
min., and filter through asbestos (Note).by heating 5 to 20 g of the sample as
Wash with HNOa (3 per cent) until described under loss on ignition (Sec-
the residue is free from permanganic tion 68) and taking the prescribed weight
acid, as determined by the absence of from the residual ash, or by placing the
the pink color in the washings. prescribed weight in a platinum crucible,
Non.-A good grade of asbestos, acid heating as described under loss on igni-
washed and specially prepared for Gooch tion, and then continuing with the pro-
crucibles, should be used. cedure. All results except moisture shall
(b) Titrate at once with standard be calculated to, and reported as, per-
sodium arsenite (1 ml. is equivalent to centages of the moisture-free sample.
0.0002 g. of Mn), the permanganic acid (b) Section 77, covering the determi-
loaned from either the persulfate or the nation of total carbon, is intended to
bismuthate method (Paragraph (a)). apply to baked or fired carbon - ceramic
The end point is the disappearance of refractories, including those that contain
the pink color. The factor for con- silicon carbide, and to natural graphite.
verting Mn to MnO is 1.291. The carbon in the silicon carbide will
be included in the result obtained.
Magnesia (c) Section 78 covers precautions and
64. Calculate the CaO obtained in exceptions to be made in analyzing baked
Section 62 to Caa(P0,)2 by multiplying or fired carbon - ceramic refractories
by the factor 1.845, and the MnO containing silicon carbide.
478 CHEMICAL ANALYSIS OF °REFRACTORlES (C 18)

Preparation of Samples for Analysis 0.500-g of moisture-free sample taken

66. Grind 50 g of the sample to pass from a portion of the carbon-free mate-
a No. 100 (149-micron) sieve5 as de- rial described in Section 6S(a).
scribed in Section 3 except for the analy- Iron, Aluminum, and Titanium Oxides
sis of graphite, in which case the sample 70. Determine the iron, aluminum,
may be ground to pass a No. 60 (250- and titanium oxides by the procedure
micron) sieve. The exception is made described in Section 10.
because the lubricating characteristics
and the flaking tendencies of graphite Iron Oxide
make grinding to pass a No. 100 sieve 71. (a) Method A: Fe,P 3 Determined
very difficult. on R,P3 Sample.-Determine the iron
A

A-Gages with needle valve control.

8-Bubbling tower containing KOH solution (50 per cent) to remove C02 from oxygen.
C-U tube filled with anhydrous magn011ium perchlorate {Mg{CJO,)ll to dry the oxygen.
D-Gairtight combustion tube of porcelain, sillimanite, fused quartz, or platinum.
E-Bubbling tower contsining concentrated sulfuric acid - chromic acid solution to catch SO,.
F-Fleming jar containing anhydrous magnesium perchlorate {Mg{ClO,)i) to catch water.
G-Johnson absorption bulb half filled with KOH solution {50 per cent). Caution.-It is best to
use a fresh solution for each determination.
H-Drying tube filled about one-half with soda-asbestos {Ascarite) and one-half with anhydrous
magnesium perchlorate (Mg(Cl04)s) to prevent atmospheric moisture and C0 2 from enterin11 the
absorption bulb, G.
I-Electrically heated combustion furnace capable of heati1tg the sample to 1000 to 1100 C.
NOTE.-Apparatus containing dry materials {C, F, Hin Fig. 1) shall be plugged on both ends
with glass wool to catch dust and organic matter.
Fm. 1.-Apparatus for Determining Total Carbon.

Moisture oxide by the procedure described in

67. Determine the moisture by the Section ll(a).
procedure described in Section 7. (b) Method B: Fe.Pa Determined 011
a Separate Sample.-Determine the iron
Loss on Ignition oxide by the procedure described in
68. Determine the loss on ignition by Section 11 (b) on 1.00 g of moisture-free
the procedure described in Section 8. sample taken from a portion of the
carbon-free material described in Section
NoTE.-The above procedure can also be used 65(a).
for the determination of the approximate per-
centage of carbon, provided there is no apprecia- Titania
ble loss from other volatile materials, such as
sulfur, clay, organic matter, or combined mois- 72. Determine the titania. (Ti02) by
ture. the procedure described in Section 12.
Silica Alumina
69. Determine the silica (Si02) by 73. Determine the alumina (Al~)
the procedure described in Section 9 on by the procedure described in Section 13.
31-7
 CHEMICAL ANALYSIS OF REFltACTORIES (C 18) 479

Lime 1000 C with oxygen flowing as described

in Paragraph (b). Remove the absorp-
74. Determine the lime (CaO) by
tion bulb from the train, bring to room
the procedure described in Section 14.
temperature, open momentarily to the
Magnesia air, and reweigh. The increase in weight
is C02.
75. Determine the magnesia (MgO)
(d) Calculation.-Calculate the per-
hy the procedure described in Section
centage of carbon in the sample as
15.
follows (Note):
Alkalies AX 0.2729
Carbon, per cent = X 100
i6. Determine the alkalies by the B
procedure described in Section 16, or where:
alternatively in Section 79, on 1.00 g of A = grams of C02, and
moisture-free sample taken from a por- B = grams of sample used.
tion of the carbon-free material described
Nom.-Some natural graphites contain ap-
in Section 65 (a). preciable amounts of carbonate rock. In such
cases the C01 from the carbonate rock will be
Total Carbonll added to the C01 obtained from the combustion
7i. (a) Apparatus.-The apparatus of the carbon. Correction may be made by de-
shall conform to that shown diagram- termining carbonate C0 2 on a separate sample
decomposed in Hcr.12
matically in Fig. 1.
(b) Purging the Apparatus.- Pass Alternative Procedure When Silicon Car-
oxygen through the apparatus with bide is Present
the furnace at operating temperature
until it has been established that there 78. If the presence of silicon carbide is
are no leaks and until the COrabsorp- suspected or known, the loss on ignition
tion bulb has attained constant weight. (Section 68) may be determined at 800
The pressure and rate of flow of the to 850 C, that is, below the decomposi-
oxygen should be such as to provide tion temperature of SiC. The residue may
about 125 to 150 bubbles per minute be digested in hydrofluoric acid (HF),
through the liquids in the train. After then fused with potassium bisulfate
removal from the train, bring the ab- (KHS04), the melt taken up in distilled
5orption bulb to room temperature, and water containing a few drops of H 2S04,
open momentarily to the air before the insoluble SiC removed by filtration,
weighing. and its approximate percentage deter-
(c) Procedure.-Put a weighed COr mined. Analysis of the filtrate would
ahsorption bulb in place. Weigh a por- then follow the procedures outlined for
tion of the moisture-free sample weigh- samples containing no silicon carbide.
ing from 0.2000 g to 1.0000 g, depending Analysis of the silicon carbide itself, if
on the amount of carbon present, into desired, may be done in accordance with
a previously ignited boat. Place the the procedures used for Bureau of Stand-
hoat in the combustion tube, immedi- ards standard sample No. 112. 1a
ately reseal the tube, and heat the 12 See Scott's "Standard Methods or Chemical
sample until the COrabsorption bulb Analysis," Fifth Edition, D. Van Nostrand Co .•
has attained constant weight at 900 to Inc., New York, N. Y.. p, 1616 (1939).
13 See the National Bureau of Standards "Cer-
tificate or Analysis" for standard sample No.
11 See Section 68 for an alternate method for 112. See p. 202 of the Manual of ASTM Stand-
lctermining carbon. ards on Refractory Materials (Nov., 1957).
Ml
480 CHEMICAL ANALYSIS OF REFRACTORIES (C 18)

FIRECLAY, SIUCA, HIGH-ALUMINA (9) Standard Solt1tions for Direct In-

REFRACTORIES tensity M ethod.-The standard solu-
Flame Photometer Method for Potas- tions for the direct intensity method
may be either of the following:
sium, Sodium, and Lithium Oxides
(A) A solution of a standard sam-
(Alternative Procedure)
ple whose composition is similar to
79. (a) Apparatus.-A flame pho- the unknown.
tometer. (B) A synthetic solution that con-
(b) Reagents: tains alkalies, interfering ions, and
(I) Sulfuric Acid (JO per cent by acids in concentrations approxi-
volume). mating those obtained in the solu-
(2) Nitric Acid (50 per cent by vol- tion of the unknown sample.
ume). (c) Treatment of Sample.-Transfer
(J) Hydrofluoric Acid (48 per cent by 0.5000 g of the finely ground sample,
weight). passing a No. 100 (149-micron) sieve and
(4) Lithium Chloride Stock Solution largely retained on a No. 200 (74-micron)
A (24.4 g per liter). sieve, and moisture-free (dried at 105 C),
(5) Lithium Chloride Stock Solution to a 30-ml platinum crucible. If the
B. (1.4 g per liter). material contains crude clay, ignite the
(6) Sodium Chloride Stock Solution sample and crucible for 1 hr at 800 C.
(0.9429 g per liter). Add 5 ml of H 2SO, (10 per cent) (Note 4)
(7) Potassium Chloride Stock Solu- and 15 to 20 ml of HF to the crucible and
tion (1.5830 g per liter). stir with a platinum rod. Partially cover
(8) Standard Solutions for Internal and heat just below boiling for about
Standard Method.-Prepare the fol- 2 hr. Rinse the lid with water and evapo-
lowing standard solutions for the in- rate to dryness at a low heat to avoid
ternal standard method (Notes 1, 2, spattering. Add 20 to 25 ml of water and
and 3): warm until all soluble salts have dis-
Solu-
solved (Note 5). Proceed in accordance
tion Reagent Equivalent to with the following Item (J) or (2):
No.
(1) Internal Standard Method.-
NaCl 0, 5, 10, 20, 30, 40, 50 ppm 118
Filter the contents of the crucible into
Na20 a 100-ml flask using a 9-cm ashless,
2 KCl 0, 10, 20, 40, 60, 80, 100 ppm rapid paper. Rinse the crucible and
118 K,O
3 LiCl 0, 6, 10, 15. 20, 30, 40 ppm 118 wash the paper well with warm water.
Li20 Cool, dilute to the mark, and mix well.
Pipet a SO-ml aliquot for the deter-
Non 1.-The NaCl and KCI standards may
be prepared as one solution, and must be remade
mination of LbO. To the solution re-
whenever a new LiCl stock solution A is pre- maining in the flask, add 5.0 ml of the
pared. LiCl stock solution A. Dilute to the
NoTE 2.-The amount of LiCl to be used as mark with water and mix well. Use
the internal standard will depend upon the sensi- this solution for the determination of
ivity desired. The use of 50 ml of LiCl stock
solution A per liter will provide approximately Na20 and K20.
500 ppm LisO. (2) Direct Intensity Method.-Trans-
NOTE 3.-The LiCl standard solutions are fer the contents of the crucible to a
prepared from the LiCl stock solution B, the 100-ml flask, rinse the crucible with
actual concentration of which is determined by
evaporating a 25-ml aliquot to dryness with a
water, cool, and dilute to the mark.
!l!ight excess of H2SO, and heating to constant Mix well, and allow to settle. Filter
weight as Li.SO,. through a dry, 9-cm, ashless, rapid
HS
 CHEMICAL ANALYSIS OF REFRACTORIES (C 18) 481

paper until sufficient filtrate for the r2 = instrument reading of unknown

determination of sodium, potassium, solution,
and lithium is obtained. c1 = concentration of element in
NOTE 4.-For silica brick or rock use 2.0000 g known solution, and
of the sample and substitute HN0 1 (1: 1) for c2 = concentration of element m un-
theH1SO, (10 per cent). known solution.
NOTE 5.-Although complete solution of
mullite-type refractories may not always be Variation between unknown and
attained, experience has indicated that essen- known samples should not be greater
tially all the alkalies do dissolve. Refractories than 10 per cent if the highest accu-
very incompletely attacked by HF - H.SO,, as
for example the higher alumina types, must be racy is to be obtained. An exploratory
brought into solution by the J. Lawrence Smith run on the flame photometer may be
procedure. I< necessary to establish the order of
(d) Determination of Na{), K./J, and magnitude of alkali in unfamiliar
Lh().-Allow the instrument to warm, samples.
and make the necessary adjustments and (e) Determination of Correction, when
settings for the type of instrument used. Required.-If the curves of concentra-
(1) Internal Standard Method.- tion versus instrument reading deviate
Atomize each of the test solutions, and appreciably from a straight line, then a
then the standard solutions necessary correction must be made. This shall be
to cover the same range of concentra- done by drawing a number of tangents
tions. As the series of standard solu- to a concentration versus instrument
tions are run, rerun the test solutions reading curve for each element and ex-
according to their proximity in concen- tending them to the "instrument read-
tration to the standard solution. Re- ing" axis (ordinate or Y-axis). A series of
peat this procedure for each of the instrument readings and corresponding
elements sought. intercepts are thereby obtained. These
(2) Direct lntensity Method.---Alter- are plotted to give an instrument reading
nately compare the standard and the (Y-axis) i•ersus intercept (X-axis) curve.
unknown in the flame photometer The correction for nonlinearity is then
until successive readings are essen- made by taking a point midway between
tially duplicates. If previously pre- the two instrument reading values (of
pared concentration versus instrument known and unknown) and determining
reading curves are essentially straight the intercept Yalue (I). Calculate the
lines, then a simple comparison of in- corrected concentration of the unknown
strument readings and concentration as follows 15 (Note 6):
may be made, as follows:
.., - I C1
'1 c, --=-
-· = - r1 - I cs

where: where the terms are defined as in Para·

r1 = instrument reading of known graph (d).
solution,
14
See Scott's "Standard Methods of Chemical 15
For details and explanations, consult J. P.
Analysis,'' Fifth Edition, D. Van Nostrand Co.. Williams and P. B. Adams, Journal, Am.
Inc., New York, N. Y., p. 882 (1939). Ceramic Soc., Vol. 37, p. 306 (1954).

18-19
482 CHEMICAL ANALYSIS OF REFRACTORIES (C 18)

EDITORIAL NOTE

The following "tentative revision" of this standard has been accepted by the Society
for the purpose of eliciti~g criticisi:1s o~ which due cogniza~~e. will be taken before the
revision is approved for mcorporat10n m the Standard. Cnt1c1sms should be addressed
to the Society, 1916 Race St., Philadelphia 3, Pa. ·

Add the following new Sections 80 to drops of concentrated nitric acid (HNOa, sp gr
1.42) and boil to oxidize the iron. Add a few
84: drops of methyl red indicator and carefully add
METHOD OF ANALYSIS OF SILICON CARBIDE ammonium hydroxide (NH,OH) until the color
REFRACTORIES WITH Low of the solution changes to a distinct yellow. Boil
BOND CONTENT the solution for 1 or 2 min to precipitate the
hydroxides of iron, aluminum, and titanium.
80. Preparation of the Sample.-Crush the Filter and wash the precipitate thoroughly with
sample in a hard tool steel mortar to pass a No. hot ammonium chloride solution (NH.Cl, 20 g
100 (149-µ) standard sieve or equivalent mesh, per liter) (be sure the NH,Cl solution is on the
and mix thoroughly. alkaline side). Dissolve the precipitate in dilute
NoTE.-The sample must be crushed by im- HCl and reprecipitate, using the same procedure
pact and not by grinding. A pneumatic hammer as ahove.
and tool steel mortar are preferred. (b) Combine the filtrates from the R,01 pre·
cipitation and evaporate to about 200 ml if vol-
81. Silicon Carbide Determination.-(a) Weigh ume is large. ::\fake the solution slightly am-
a 1-g sample into a 30-ml platinum crucible; add moniacal and add a few ml of a saturated
15 to 20 ml of concentrated hydrofluoric acid solution of ammonium oxalate ((NH,),C.0,·
(HF, 48 per cent) and 3 to 4 drops of concen- H 20), boil for several minutes, let stand for sev-
trated sulfuric acid (H,SO,, sp gr 1.84) and eral hours (preferably overnight), filter and wash
evaporate slowly on a sand bath to volatilize the with hot water until the washings give no test
free silica and silica combined as silicates. Evap- for oxalate. Dissolve in hot H,SO, (1:4) and
orate excess sulfur trioxide (SOa) fumes, and fuse titrate with 0.05 N potassium permanganate
with approximately 10 g of potassium bisulfate (KMn04) at a temperature 21 to 27 C (70 to
(KHSO,) at red heat, cool, extract the melt with 80 F). Report as per cent calcium oxide (CaO).
75 ml of hot water and 20 ml of concentrated 83. M agnesiz&m Oxide Determinalion.-Acid-
hydrochloric acid (HCI, sp gr 1.19). Filter and ify the filtrate from CaO determination with
wash the silicon carbide (SiC) residue with hot about 40 ml of concentrated HNOa (sp gr 1.42),
wateracidulated with HCL. Burn off the filter pa- boil until dry, and heat on hot plate to volatilize
per and ignite the SiC residue at about 800 C the excess ammonium salts. Take up with a few
(1470 F) in a weighed platinum crucible, cool, ml of dilute HCI, make ammoniacal, filter, and
and weigh. wash. To the filtrate add sodium ammonium
(b) A small amount of SiC is decomposed by phosphate solution (Na!\~HP0,·4H20, 1:10)
the bisulfate fusion and is determined by adding and then add with constant stirring one tenth of
5 ml of concentrated H,SO, (sp gr 1.84) to the the volume of NH,OH. J,et stand several hours
filtrate obtained after filtering off the silicon (preferably overnight), filter, and wash with cold
carbide. Evaporate on a sand bath until fumes NH,OH solution (1: 20). The filter paper and its
of S01 appear, in order to dehydrate the silica contents are placed in a weighed platinum cruci-
due to the decomposition of SiC. Cool, dilute ble and after drying and charring, it is ignited to
with hot water, filter, and wash with hot water constant weight. The residue is magnesium pyro-
acidulated with HCl. Burn off the filter paper phosphate (Mg 2P 20,); calculate and report as
and ignite to constant weight. The silica is per cent magnesium oxide (MgO).
volatilized according to the standard procedure, 84. Alkalies.-Determine alkalies by the J.
and is calculated to the equivalent amount of Lawrence Smith method for alkalies in clay,
SiC. This is added to the SiC previously weighed. separating potassium as perchlorate (Section
Report as per cent SiC. 16).
82. Calcium Oxide. Detcrmination.-(a) To the
filtrate from the SiC determination, add a few Tentative revision issmd J uru, 1960.