Journal of Analytical Toxicology 2012;36:293 –302

doi:10.1093/jat/bks025 Article

Use of SPME-HS-GC –MS for the Analysis of Herbal Products Containing

Synthetic Cannabinoids
Anderson O. Cox1, Richard C. Daw1, Michele D. Mason1, Megan Grabenauer1, Poonam G. Pande1, Kenneth H. Davis1,
Jenny L. Wiley1, Peter R. Stout1, Brian F. Thomas1* and John W. Huffman2
1
RTI International, Research Triangle Park, North Carolina 27709-2194, and 2Department of Chemistry, Clemson University,
Clemson, South Carolina 29634-0973

*Author to whom correspondence should be addressed. Email: [email protected]

The increasing prevalence and use of herbal mixtures containing the piperazines and cathinone derivatives in the 2000s (2).
synthetic cannabinoids presents a growing public health concern Smokable herbal mixtures containing non-controlled synthetic
and legal challenge for society. In contrast to the plant-derived can- cannabinoids have been sold on the Internet and in various
nabinoids in medical marijuana and other cannabinoid-based thera- retail shops as “Spice” and other misleadingly labeled products
peutics, the commonly encountered synthetic cannabinoids in these since at least 2006. The majority of these products are mix-
mendaciously labeled products constitute a structurally diverse set tures of plant/herbal ingredients of varied name and constitu-
of compounds of relatively unknown pharmacology and toxicology. tion, and little is known about the toxicology of the plant
Indeed, the use of these substances has been associated with an materials purportedly contained in these products (3).
alarming number of hospitalizations and emergency room visits. Synthetic cannabinoids that have most commonly been
Moreover, there are already several hundred known cannabinoid reported to be present in herbal Spice and Spice-like samples
agonist compounds that could potentially be used for illicit include JWH-018, JWH-073, JWH-398, JWH-250, HU-210, and
purposes, posing an additional challenge for public health profes- CP-47,497 and its homologues (4 –8). These and other known
sionals and law enforcement efforts, which often require the detec- and plausible synthetic cannabinoids also have relatively
tion and identification of the active ingredients for effective unknown toxicology. Hence, accurate assessment of the poten-
treatment or prosecution. A solid-phase microextraction headspace tial health risks of these herbal mixtures has become increas-
gas chromatography –mass spectrometry method is shown here to ingly difficult. Furthermore, although little is known about the
allow for rapid and reliable detection and structural identification prevalence of their use or the extent to which smoking mix-
of many of the synthetic cannabinoid compounds that are currently tures containing non-controlled synthetic cannabinoids have
or could potentially be used in herbal smoking mixtures. This replaced cannabis in the United States or globally, their preva-
approach provides accelerated analysis and results that distinguish lence and abuse appears to have significantly increased. In
between structural analogs within several classes of cannabinoid 2009, a survey conducted among 1,463 students aged between
compounds, including positional isomers. The analytical results 15 and 18 years at schools providing general and vocational
confirm the continued manufacture and distribution of herbal training in Frankfurt found that around 6% of respondents
materials with synthetic cannabinoids and provide insight into the reported having used synthetic cannabinoid-containing Spice
manipulation of these products to avoid legal constraints and products at least once (9). In the US, a survey conducted by
prosecution. the Idaho Hospital Association revealed that of the 20 hospitals
polled across the state, 11 had knowledge of the drug and
more than 80 cases of suspected synthetic cannabinoid (Spice)
overdoses occurred between February and August of 2010
(10). According to the United States Drug Enforcement
Introduction Administration (DEA), Office of Diversion Control’s special
According to the 2010 World Drug Report issued by the report on synthetic cannabinoid and cathinone data within the
United Nations Office on Drugs and Crime (UNODC), between National Forensic Laboratory Information System, the incidence
155 and 250 million people, or 3.5 to 5.7% of the population of synthetic cannabinoids that were submitted to state and
aged 15 –64, is estimated to have used illicit substances at least local forensic laboratories rose from 13 to 2,977 cases between
once in the previous year. Cannabis users comprise by far the the same reporting periods in 2009 to 2010 (11).
largest number of illicit drug users (129–190 million people), Several forces may exert some control on the continued
followed by amphetamine-type stimulants, opiates and cocaine. design, production or abuse of synthetic designer drugs.
Although these numbers have remained relatively stable over Demand reduction is obviously one of the most important
the last decade, the incidence of abuse of “designer drugs,” factors in controlling abuse, which to the clandestine supplier
clandestinely produced drugs that are structurally and pharma- equates to maintenance of user desire and acceptance of the
cologically very similar to a controlled substance but are not marketed drug. Another important market controlling factor is
themselves controlled substances, has increased (1). Indeed, law enforcement control of their manufacture and distribution.
the appearance of synthetic cannabinoids marks the latest With respect to law enforcement, the status of control mea-
stage in the development and abuse of controlled substance sures and regulation against synthetic cannabinoids also varies
analogs, the history of which includes the design of fentanyl in across the globe. None of the synthetic cannabinoids are under
the 1980s; the development of the ring-substituted phenethyla- international control by virtue of the UN drug control conven-
mines in the late 1980s and the tryptamines in the 1990s; and tions, but JWH-018, JWH-073, HU-210 and CP 47,497 (together

# The Author [2012]. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected]
with its C6, C8 and C9 homologues) are scheduled drugs in some products appeared to have more uncommon or previously
some Member States. The following countries control Spice unreported active compounds present. The results described
and/or other synthetic cannabinoids: Austria, Canada, Chile, here provide evidence to support the SPME-GC –MS approach
Denmark, Germany, Estonia, Finland, France, Ireland, Italy, as an effective means for the rapid analysis, detection and mass
Japan, Latvia, Lithuania, Luxembourg, New Zealand, Poland, spectral confirmation of synthetic cannabinoids in both bulk
Romania, Russia, South Korea, Sweden, Switzerland and the drug substance and herbal formulations. This method was used
United Kingdom (12). In the United States, D9-THC and similar to analyze products obtained before and after the DEA’s emer-
forms of synthetic cannabinoids of the “classical cannabinoid” gency scheduling control of five commonly encountered syn-
structural motif (HU-210) have been scheduled under federal thetic cannabinoids. The results show that the legislation has
law for many years. However, in response to the perceived risk not prevented manufacturers or distributers from dispensing
associated with these synthetic cannabinoids and herbal these newly controlled chemicals; however, the use of new can-
products, on November 24, 2010, the United States DEA nabinoids in these herbal formulations was also observed. In
announced that it intended to use its emergency scheduling addition to the analysis of herbal products, over 20 synthetic
authority to temporarily control five synthetic cannabinoids cannabinoid standards were analyzed to confirm the structural
(JWH-018, JWH-073, JWH-200, CP-47,497 and cannabicyclo- identification of the synthetic cannabinoid compounds detected
hexanol). This scheduling took effect on March 1, 2011. The in the herbal products based on both retention time and mass
DEA clearly considers these, and other synthetic cannabinoids, spectrum. Finally, the gas chromatograms and mass spectra for
to be a drug class of concern. Before the action by the DEA, both the herbal products and the synthetic cannabinoid stan-
many states had passed acts making it illegal under state law to dards have been submitted to a publically available spectral data-
sell Spice and/or synthetic cannabinoids. Since the DEA edict, base (www.forensicdb.org) for future use by the scientific
other states have also passed or are considering similar legisla- community. The further adoption and continued application of
tion (13). Additional bans in the US are being enacted on city the described method for rapid analysis and identification of
and county levels. There is little doubt that the actions of the synthetic cannabinoid-containing formulations, and the expan-
UN member states, the DEA, and other regulatory agencies will sion of the spectral information that this method has allowed to
reduce access to these synthetic cannabinoids in herbal formu- be made publically available to date, should prove useful for fo-
lations, and may dramatically impact their abuse. However, rensic and special testing laboratories in identifying previously
there are hundreds of compounds with cannabinoid receptor described and novel synthetic cannabinoids in bulk substances
affinity and activity, in several chemical classes (14 –17), and it and herbal formulations. In turn, these efforts will enable the
is likely that new substances will continue to appear in various scientific community to better address the intent of our
formulations for illicit use. This large array of possible com- society’s legislation and the effectiveness of the concomitant
pounds presents a constant challenge for the forensic and toxi- law enforcement efforts.
cological identification of new substances that is needed for
the prompt assessment of risk and, where necessary, imple-
mentation of control measures by both public health and law Experimental
enforcement agencies.
To analyze commercially available herbal formulations sold as Materials
incense, but suspected to contain synthetic cannabinoids, Methanol (99.9 þ %, capillary GC grade), and 2,6-dichlorotoluene
various researchers have developed and applied chromato- (99 þ %) were purchased from Sigma-Aldrich (St. Louis, MO).
graphic methods and mass spectrometry. Dresen et al. and Ethanol (200 proof, USP grade) was purchased from Pharmco-
Uchiyama et al. have used gas chromatographs coupled to mass AAPER (Brookfield, CT). HPLC standard mixture I (item
spectrometers (GC –MS) in electron ionization mode to iden- number 13830) containing (+)-CP 47,497, (+)-CP 47,497-
tify synthetic cannabinoids in extracts of herbal products, C8-homolog, (+)-CP 55,940, HU-308, HU-331, JWH 015, JWH
while Hudson et al. and Uchiyama et al. also performed ana- 018, JWH 019, JWH 073, JWH 200, JWH 250 and WIN 55212-2,
lysis of herbal extracts with liquid chromatography –mass spec- was purchased from Cayman Chemical (Ann Arbor, MI). AM-
trometry (LC –MS) (5, 7, 8, 18). These methods have allowed 694 was obtained from LGC (Teddington, UK). All JWH analogs
for the qualitative identification of numerous synthetic cannabi- used in these studies as reference standards were synthesized
noids in herbal product matrices. Each of these methods used by Dr. John W. Huffman and provided by Dr. Jenny L. Wiley.
an extraction step to isolate the analytes of interest followed by Herbal products were purchased in convenience stores and
concentration before analysis. various commercial venues in the Raleigh, Durham and Chapel
In an effort to develop rapid methods to characterize herbal Hill area.
formulations, an automated headspace solid-phase microextrac-
tion and gas chromatography –mass spectrometry (HS-SPME-
GC –MS) method was developed. The analytical approach allows Standard and sample preparations
for the detection of JWH-018, JWH-073 and other synthetic can- Herbal samples were prepared by accurately weighing approxi-
nabinoid analogs from small quantities of materials (50 mg) mately 50 mg of each sample into individual 10 mL headspace
without requiring extraction, concentration or derivatization. vials and then spiking each with a 20 uL aliquot of a
Analysis of over 20 products, obtained from retail stores in the 2,6-dichlorotoluene internal standard solution (0.1 mg/mL solu-
Raleigh, Durham and Chapel Hill areas of North Carolina, tion in methanol). The vials were capped with 20-mm magnet-
revealed the presence of at least one synthetic cannabinoid in ic crimp seal caps with PTFE/silicone septa (Restek; Bellefonte,
each product. Several products had two or more analogs, and PA) before analysis.

294 Cox et al.

 Standard synthetic cannabinoid samples for SPME-HS-GC– MS cannabinoids in the herbal products were volatilized, absorbed
analysis were prepared by diluting the synthetic cannabinoid onto the SPME fiber, desorbed in the GC injection port, chro-
standards to a final concentration of 1 mg/mL in ethanol. matographically resolved and detected reliably, with very little
Sonication was used to aid in the dissolution. A 20-mL aliquot indication of degradation (Table I). The direct SPME from the
of each standard solution was then removed and added directly headspace produced by heating as little as 20 mg of a synthetic
to individual 10-mL headspace vials and dried in a TurboVap cannabinoid standard, including JWH-018, JWH-073, CP-55,940,
LV evaporation system (Caliper LifeSciences; Hopkinton, MA) CP-47,497 (and its octyl analog), HU-308, HU-331, and several
until dry. The vials were capped as described previously and other JWH analogs using the SPME method produced intense
subsequently analyzed. chromatographic peaks. The high resolution of the capillary
chromatography resulted in baseline separation and unique
retention times for all of the isobaric synthetic cannabinoid
HS-SPME-GC–MS standards analyzed. This is an important consideration in the
An Agilent 6890 Plus gas chromatograph coupled to a 5973 present work, because most of the synthetic cannabinoids
MSD mass selective detector (Agilent Technologies; Santa encountered in these herbal products have positional isomers
Clara, CA) was used for all analysis. A CTC analytics combipal that are relatively indistinguishable from each other based on
autosampler fitted with a SPME sampling system (Leap their mass spectrum alone, but are able to be differentiated
Technologies; Carrboro, NC) was used to extract and inject all based on capillary gas chromatographic retention time. In add-
samples. The SPME system consisted of an SPME fiber holder, a ition, all of the analytes were baseline separated and well
fiber conditioning station and a sample incubation block resolved from each other, with the exception of CP-55,940 and
capable of heating and agitation of the sample. A stableflex JWH-302, which only differed by 0.08 min in retention time
23-gauge 85-mm carboxen/polydimethylsiloxane (CAR/PDMS) (Table I). Nevertheless, these two compounds are clearly differ-
SPME fiber was used for the analysis (Supelco; Bellefonte, PA). entiated by their different molecular ions and fragmentation
The autosampler was controlled by Pal (firmware version 2.5.2) patterns. Hence, suspected synthetic cannabinoid compounds
software (Leap Technologies). present in the herbal products were identified based on con-
Samples were incubated at 2008C with pulse-agitation at sistencies between both retention time and mass spectrum
250 rpm (cycles of 5 s on, 2 s off ). After one minute of sample when compared to those obtained from the analysis of stan-
equilibration, the SPME fiber was inserted and exposed to the dards, as well as agreement between compound mass spectrum
sample headspace for five minutes of incubation. The fiber was and published structure. This method of analysis, using reten-
then injected into the GC inlet for 15 min to desorb the ana- tion time with mass spectral data, can help to identify both
lytes, and then allowed to condition for 5 min before the fol- compounds that have been explicitly banned by the DEA along
lowing injection. with their structural analogs that are not, which is important
An Agilent DB-5MS capillary column (30 m  0.25 mm  given what appears to be the recent migration by herbal
0.25 mm, Agilent Technologies) was used for separation with a product manufacturers away from the use of compounds
helium carrier gas flowing at 1.0 mL/min. A split injection ratio specifically banned by the DEA to other synthetic analogs that
of 1:50 was used at 3008C. The oven temperature program con- may delay or confound detection.
sisted of a one minute hold at 408C, followed by a 108C/min During the analysis of herbal products, the most intense
ramp to 3008C, a 9 min hold at 3008C, a 158C/min ramp to chromatographic peaks were most often found to correspond
3258C and then a one minute hold. The MSD transfer line to synthetic cannabinoid compounds. Although this has impli-
heater was set to 3008C. The temperatures of the quadrupole cations regarding the relative abundance and volatility of the
and the ion source were 150 and 2308C, respectively. The dur- synthetic cannabinoid analogs in and between the various
ation of each run was 38.67 min. The analytes were detected herbal samples, it also indicates that the choice of SPME fiber
after electron impact ionization (70 eV) in the SCAN-mode material was appropriate for the type of analyses conducted.
from m/z 50 to 550. SPME has previously been shown to discriminate during ad-
sorption between different molecules based on size and affinity
to the fiber material (19). This discrimination is often also the
means used to concentrate the analytes relative to the matrix.
Data analysis
Thus, these data and other data obtained in our laboratory by
All GC –MS data acquisition and analysis was performed using LC–MS and other analytical approaches (data not shown)
the enhanced MS Chemstation G1701BA versions B.01.00 soft- suggest that the synthetic cannabinoid compounds are present
ware (Agilent Technologies, 1998) and E.02.01.1177 (Agilent in considerable quantity and are readily volatilized or extracted
Technologies, 2010). Samples were analyzed to obtain reten- from the herbal material. Previously reported methods of
tion time and mass spectra for analytes of interest to appropri- sample preparation such as that of Dresen (5), or the LC–MS
ately identify each analyte. Active compounds found in the method provided by Hudson (18), involve a solvent extraction
samples were identified based on the spectra and retention step entailing a vortexing or sonicating period. Although these
times of the standards. extractions are relatively simple, when compared to the auto-
mated SPME fiber, they require solvents for extraction and add-
itional efforts to purify and concentrate the sample before
Results and Discussion delivery to the GC-inlet. In contrast, SPME provides the ability
The SPME-HS-GC –MS analysis of synthetic cannabinoid stan- to extract volatile analytes from the headspace vapors pro-
dards revealed that most of the commonly encountered duced simply by heating a complex solid sample, thereby

Use of SPME-HS-GC–MS for the Analysis of Herbal Products Containing Synthetic Cannabinoids 295
296 Cox et al.
Table I
Structures and Retention Times of Synthetic Cannabinoid Standards

Compound Structure Mass* Time† Compound Structure Mass* Time†

AM-694 435 29.88 JWH-408 371 38.42

JWH-348 435 31.45 JWH-197 371 31.94

JWH-363 435 30.88 JWH-019 355 32.58

JWH-372 435 30.57 JWH-018 341 31.30

JWH-200 384 38.21 JWH-201 335 29.36

Pravadoline (WIN 48098) 378 34.20 JWH-250 335 28.51

CP 55,940 376 28.78 JWH-302 335 28.86

JWH-081 371 36.06 CP 47,497 C8-homolog 332 26.47

30.22

29.61

25.71
concentrating them and carrying them to the inlet, all in one
single sample preparation step. Because there are few sample
manipulations, the potential for confounding variables such as
solvent artifacts or reactivity, or variations in recovery and
detection due to sample solubility, are potentially reduced or
eliminated. Additionally, while 2,5-dichlorotoluene was used
327

327

318
here as an internal standard to verify suitable method perform-
ance, deuterated cannabinoid ligands are commercially avail-
able that could prove useful for enabling more quantitative
analyses.
In all of the 31 herbal products tested, active synthetic
cannabinoids were readily detected (for example, see
Figure 1). In this particular herbal product (K2 Summit), the
most intense peak, at a retention time of 31.3 min, was identi-
fied as JWH-018 based on its mass spectrum (Figure 2) and re-
tention time match to a synthetic JWH-018 standard. Fifteen
products were found to contain a single synthetic cannabinoid
(Table II). JWH-018, a compound previously reported to be
present in numerous herbal products (5, 6, 8, 18), was the
analyte most often detected during this analysis; being present
in 23 of 31 products sampled. The prevalence of JWH-018 in
herbal products is corroborated by previous analyses, including
those reported by Hudson et al., as is the occurrence of mul-
tiple compounds in certain products (18). Other previously
reported synthetic cannabinoids detected in this analysis
include JWH-073, JWH-019 and JWH-081. Interestingly,
CP 47,497
JWH-073

JWH-015

JWH-073 is often found in combination with JWH-018

(Figures 3 and 4), and is seldom present alone in herbal pro-
ducts, suggesting that these analogs are most commonly avail-
able for use, or that the combination of these active
cannabimimetic analogs affords some advantage to the formula-
tion, such as a pharmacological property of preference to the
33.56

35.84

32.27

consumer.
JWH-250 is another synthetic indole cannabinoid that has
previously been reported as a constituent in Spice products
(5), and was also determined to be present in many of our
samples. JWH-250 was found in seven out of 31 herbal pro-
371

371

371

ducts. This compound was the second most abundant psycho-

active constituent found in our analyses, which is in contrast to
previous reports of the synthetic cannabinoids found in herbal
products, where JWH-250 was not as prevalent (17). This may
reflect the recent DEA scheduling of JWH-018 and JWH-073,
which did not specifically mention JWH-250, leading to a
Retention time of compound.
*Nominal mass of compound.
JWH-164

JWH-166

JWH-267

Figure 1. Total ion chromatogram recorded for the herbal product K2 Summit
revealing a peak eluting at 31.3 min that was determined to be due to the presence
of JWH-018.
†

Use of SPME-HS-GC–MS for the Analysis of Herbal Products Containing Synthetic Cannabinoids 297
Figure 2. Mass spectrum recorded for the peak eluting at 31.3 min in the total ion chromatogram obtained during the analysis of the herbal product K2 Summit. This
constituent was determined to be JWH-018 based on comparison to the retention time and mass spectrum recorded for a synthetic standard. The structure of JWH-018 is
provided in the figure for reference.

switch to this cannabinoid compound as an effort to avoid de- Figures 2, 4 and 5). Another indole cannabimimetic compound
tection and law enforcement. Indeed, the mass spectrum of was detected in an herbal formulation, JWH-019, but failed to
JWH-250 (Figure 5) is substantially different from JWH-073 and follow the trend of producing an intense chromatographic
JWH-018 due to the added methylene group between the two signal. The spectrum was determined to be a match to a syn-
ring systems (compare the structures and spectra presented in thetic standard with respect to retention time and mass

Table II
Synthetic Cannabinoids Detected in Herbal Products by SPME-HS-GC –MS

Type Description JWH-018 JWH-073 JWH-250 JWH-081 JWH-019 Pravadoline AM-694 AM-2201
K2 Summit (Batch 1) X
K2 Summit (Batch 2) X X
Happy Shaman Herbs Humboldt Gold X X
Happy Shaman Herbs Pink Tiger X X
Happy Shaman Herbs Brain Freeze X
Meditation Strawberry (1.5g Package) X X X
Meditation Strawberry (3.0g Package) X X
Meditation Mango X X
Meditation Blueberry X X
Spice99 Bad 2 the Bone (cinnamon) X X
Mr. Nice Guy Herbal smoke blend X X
Solowholesale Cloud 49 (ultra strong, Irish Crème) X X
Unknown Ultra Cloud 10 X X
Unknown Generic (ultimate) Spice! Double strength X
Unknown Generic (Premium) Spice! X
Hot Hawaiian Exotic Herbal Blend X
Tranquility Herbal Incens Kush X
Deviltracks Newprot (mint) X X
Flawless Strawberry X
Black Magic Smoke Herbal incense X
Purple Flake N/A X
Southern Spice Herbal incense (mango) X
Berry Twist N/A X
Destiny Exotic Herbal Blend X
Southern Spice Blue Dream (melon) X
Southern Spice Blue Dream (melon) X
Happy Shaman Herbs Humboldt Gold X X
Received After DEA Ban
Spice99 Stinger X X
Spice99 GI Jane X
Spice99 GI Joe X
Purple Puff Kush X
Yeah Right Head Funk X

298 Cox et al.

spectrum, but the peak was barely distinguishable in the gas product line except one contained JWH-073 and JWH-018 (see
chromatogram of the herbal product. This suggests that Table I). Similarly, all products from the Meditation brand con-
JWH-019 was present at a relatively low concentration in this tained JWH-081 and JWH-250. Spice99 was also noted as the
herbal product or sample aliquot, or that it was less efficiently only brand that we sampled that contained the AM (i.e.,
extracted from this material using the SPME technique. AM-694) class of compounds. While varying other aspects of
However, the presence of this compound was confirmed with their products, this consistency suggests that brands or manu-
data from a second study using high-resolution LC–MS (data facturers may attempt to establish a means to recognize or dif-
not shown), where the compound appeared to be present in ferentiate their products by the mixture of cannabinoids
relatively high concentrations. A tentative conclusion, based on present within them. AM-694 was detected in our analysis of
our ability to detect JWH-019 easily in a SPME analysis of the both Spice99 Bad 2 the Bone (cinnamon) and Spice99 Stinger,
synthetic standard, is that heterogeneity in the product manu- samples from before and after the DEA legislation took effect,
facture may result in varied concentration across the herbal respectively. The identity of this compound eluded identifica-
material, which is consistent with variation noted across tion for some time, as we analyzed JWH compounds with the
packages of the same material (20). Finally, in cases in which same nominal mass, such as JWH-363, JWH-348 and JWH-372,
more than one type or flavor of herbal product was analyzed and found that their mass spectra did not match the peaks in
from the same brand, it was observed that the active constitu- our herbal product samples. Further analysis of additional syn-
ents and their relative proportions often remain quite similar. thetic cannabinoids allowed us to arrive at the identification of
For example, all products tested from the Happy Shaman Herbs AM-694 (Figure 6). This identification has been confirmed
through analysis of a synthetic standard, and is consistent with
the observation that AM-694 has been previously reported in
European samples (4).
The herbal products obtained after the DEA ban of five syn-
thetic cannabinoids show both an unexpected continuation in
the use of DEA Schedule I controlled substances, along with
new active substances, indicating that some manufacturers are
modifying their products to attempt to continue to supply
illicit user demand in light of new legal constraints and law en-
forcement activities. For example, Kush, a product from Purple
Puff, was shown to contain JWH-018. This may indicate that
some producers of herbal smoke products are reluctant to
move away from the use of existing supplies, or an available
supply of a now illegal compound. However, three other
samples analyzed after the ban (Spice99’s Stinger, GI Joe and
Figure 3. Total ion chromatogram recorded for the herbal product Happy Shaman
GI Jane) also contained what appear to be new or rarely
Herbs Humboldt Gold revealing peaks eluting at 30.2 and 31.3 min that were
determined to be due to the presence of JWH-073 and JWH-018, respectively. encountered components in herbal smoking blends, such as

Figure 4. Mass spectrum recorded for the peak eluting at 30.2 min in the total ion chromatogram obtained during the analysis of the herbal product Happy Shaman Herbs
Humboldt Gold. This constituent was determined to be JWH-073 based on comparison to the retention time and mass spectrum recorded for a synthetic standard. The
structure of JWH-073 is provided in the figure for reference.

Use of SPME-HS-GC–MS for the Analysis of Herbal Products Containing Synthetic Cannabinoids 299
Figure 5. Mass spectrum recorded for the peak eluting at 28.5 min in the total ion chromatogram obtained during the analysis of the herbal product Destiny Exotic Herbal
Blend. This constituent was determined to be JWH-250 based on comparison to the retention time and mass spectrum recorded for a synthetic standard. The structure of
JWH-250 is provided in the figure for reference.

Figure 6. Mass spectrum recorded for the peak eluting at 29.9 min in the total ion chromatogram obtained during the analysis of the herbal product Spice99 Stinger. This
constituent was determined to be AM-694 based on comparison to the retention time and mass spectrum recorded for a synthetic standard. The structure of AM-694 is
provided in the figure for reference.

WIN-48,098 ( pravadoline, see Figure 7). Pravadoline was devel- The absence of CP-47,497 and its side chain analogs is
oped originally as an analgesic, but was later found to have can- notable. Hudson et al. were able to detect these compounds
nabimimetic properties and is considered to be a predecessor more frequently in their herbal product samples than any other
of the well known WIN-55,212-2. In addition, the herbal synthetic cannabinoid (18). The high frequency with which
product Head Funk was found to contain what appears to these compounds have been reported may have contributed to
be AM-2201 based on GC –MS (Figure 8) and high resolution the reasoning used by the DEA when choosing which com-
LC–quadrupole time-of-flight analysis (data not shown). The pounds to place on schedule in March of 2011, because
putative identification of AM-2201 in this herbal product CP-47,497 and its homologs were specifically scheduled.
is also consistent with previous published mass spectral Because all herbal samples found to contain CP-47,497 were
information identifying AM-2201 in samples obtained from the purchased before the DEA ban, and subsequently no herbal
internet (21). product has been found to contain this psychoactive

300 Cox et al.

Figure 7. Mass spectrum recorded for the peak eluting at 34.3 min in the total ion chromatogram obtained during the analysis of the herbal product Spice99 Stinger. This
constituent was determined to be pravadoline based on comparison to the retention time and mass spectrum recorded for a synthetic standard. The structure of pravadoline is
provided in the figure for reference.

Figure 8. Mass spectrum recorded for the peak eluting at 32.4 min in the total ion chromatogram obtained during the analysis of the herbal product Head Funk. This
constituent is suspected to be AM-2201 based on the observed mass spectrum and the chemical structure and proposed fragmentation, as provided in the figure.

ingredient, it again suggests that these products are being spe- and other compounds of interest in herbal products and other
cifically designed with the legality and policing of synthetic items of interest to the health sciences, drug enforcement and
cannabinoids in mind. Indeed, we have also noted that once regulatory agencies, and law enforcement efforts.
the US states and federal agencies began to identify and ban or
schedule specific substances, the prevalence of the prohibited
substances decreased and new analogs began to appear and
become more prevalent. Finally, it is important to emphasize Conclusions
that all of our spectral data, including ongoing analytical A novel SPME-HS-GC– MS method for the qualitative analysis of
efforts, are available online at www.forensicdb.org. This data- cannabinoid compounds in herbal products was developed and
base provides a searchable, publicly available cheminformatic shown to be effective for the detection and identification of a
library that allows online users to compare and search for wide variety of synthetic cannabinoids in several structural
spectra to aid in the identification of synthetic cannabinoids classes, in both a neat standard form as well as in herbal

Use of SPME-HS-GC–MS for the Analysis of Herbal Products Containing Synthetic Cannabinoids 301
product matrices. The analysis of more than 25 herbal products receptor agonists in ‘Spice.’ International Journal of Drug Policy,
with this method identified several active compounds, some 22, 274–277.
previously reported such as JWH-073, 019, 018 and 081, and 5. Dresen, S., Ferreiros, N., Putz, M., Westphal, F., Zimmermann, R.,
Auwarter, V. (2010) Monitoring of herbal mixtures potentially con-
newer constituents such as AM-694 and WIN-48,098 ( pravado- taining synthetic cannabinoids as psychoactive compounds.
line). It is clear that although the use of JWH-018 and illegal Journal of Mass Spectrometry, 45, 1186– 1194.
compounds in herbal products has continued after the DEA 6. Emanuel, C.E.J., Ellison, B., Banks, C.E. (2010) Spice up your life:
ban, manufacturers of these products are pursuing modifica- screening the illegal components of ‘Spice’ herbal products.
tions and innovations to the chemical constituents in an Analytical Methods, 2, 614– 616.
attempt to elude detection, identification and law enforcement. 7. Hudson, S., Ramsey, J. (2011) The emergence and analysis of syn-
thetic cannabinoids. Drug Testing and Analysis, 3, 466–478.
The SPME-HS-GC–MS method provides sufficient resolution to
8. Uchiyama, N., Kikura-Hanajiri, R., Ogata, J., Goda, Y. (2010)
discriminate between many of the positional isomers in the Chemical analysis of synthetic cannabinoids as designer drugs in
various series of synthetic cannabinoids, which could also be herbal products. Forensic Science International, 198, 31 –38.
employed by manufacturers to evade detection and legal pros- 9. European Monitoring Centre for Drugs and Drug Addiction. (2010)
ecution. The minimization of sample preparation inherent in Drug profile—Synthetic cannabinoids and ‘Spice,’ European
SPME analysis allows for ease and rapid access to the analysis of Monitoring Centre for Drugs and Drug Addiction, Lisbon.
10. Hurst, D. (2010) Idaho Reporter.Com—Hospital survey shows more
the herbal products and synthetic cannabinoids, of particular
than 80 suspected cases of Spice overdose in past six months.
interest to the forensic and toxicological communities. http://www.idahoreporter.com/2010/hospital-survey-shows-more-
The increasing popularity and availability of synthetic canna- than-80-suspected-cases-of-spice-overdose-in-past-six-months/
binoids and herbal formulations is cause for considerable (accessed 5 November 2010).
public concern. For this reason, the implementation of control 11. U.S. Drug Enforcement Administration, Office of Diversion Control.
actions and enforcement by regulatory authorities, and the (2011) National Forensic Laboratory Information System: Special
demand for scientific investigation of these synthetic cannabi- Report: Synthetic cannabinoids and synthetic cathinones reported
in NFLIS, 2009–2010., U.S. Drug Enforcement Administration,
noids and herbal products, has intensified. These pharmaco- Springfield, VA.
logical and toxicological studies and regulatory control actions 12. Cooper, D.A. Future synthetic drugs of abuse. http://www.erowid.
require the development of sophisticated analytical methods org/library/books_online/future_synthetic/future_synthetic.shtml
that are able to readily sample a variety of complex products, (accessed 5 November 2010).
bulk chemical samples or biological matrices, while providing 13. National Conference of State Legislatures. Synthetic cannabinoids
reliable and robust separation and discrimination of their illicit legislation. http://www.ncsl.org/?TabId=22431 (accessed 5
November 2010).
chemical constituents. Finally, the application of these analytic-
14. Pertwee, R.G. (2000) Cannabinoid receptor ligands: Clinical and
al approaches and the interpretation of their results are further neuropharmacological considerations, relevant to future drug dis-
supported by the inclusion of the data into publically available covery and development. Expert Opinion on Investigational
analytical databases, such as those described here. Drugs, 9, 1553–1571.
15. Pertwee, R.G. (2009) Emerging strategies for exploiting cannabin-
oid receptor agonists as medicines. British Journal of
Pharmacology, 156, 397–411.
Acknowledgments
16. Pertwee, R.G., Howlett, A.C., Abood, M.E., Alexander, S.P.H., Di
This work was supported by the National Institute on Drug Marzo, V., Elphick, M.R. et al. (2010) International Union of Basic
Abuse (N01DA-0-7301, HHSN271201000003C). The ForensicDB and Clinical Pharmacology. LXXIX. Cannabinoid receptors and
website is funded in part through a Cooperative Agreement their ligands: Beyond CB1 and CB2. Pharmacological Reviews, 62,
from the National Institute of Justice, Office of Justice Programs, 588– 631.
17. Thakur, G.A., Tichkule, R., Bajaj, S., Makriyannis, A. (2009) Latest
U.S. Department of Justice (# 2008-DN-BX-K180). Neither the advances in cannabinoid receptor agonists. Expert Opinion on
U.S. Department of Justice nor any of its components operate, Therapeutic Patents, 19, 1647–1673.
control, are responsible for, or necessarily endorse, this Web site 18. Hudson, S., Ramsey, J., King, L., Timbers, S., Maynard, S., Dorgan, P.I.
(including, without limitation, its content, technical infrastruc- et al. (2010) Use of high-resolution accurate mass spectrometry to
ture, and policies and any services or tools provided). detect reported and previously unreported cannabinomimetics in
“herbal high” products. Journal of Analytical Toxicology, 34,
252– 260.
19. Zhang, Z., Pawliszyn, J. (1993) Headspace solid-phase microextrac-
References tion. Analytical Chemistry, 65, 1843–1852.
1. United Nations Office on Drugs and Crime. (2010) World Drug 20. Hudson, S., Ramsey, J., King, L., Timbers, S., Maynard, S., Dargan, P.I.
Report. http://www.unodc.org/unodc/en/data-and-analysis/WDR- et al. (2010) Use of high-resolution accurate mass spectrometry to
2010.html (accessed 5 November 2010). detect reported and previously unreported cannabinomimetics in
2. Griffiths, P., Sedefov, R., Gallegos, A.N.A., Lopez, D. (2010) How herbal high products. Journal of Analytical Toxicology, 34, 252–260.
globalization and market innovation challenge how we think about 21. Nakajima, J.I., Takahashi, M., Nonaka, R., Seto, T., Suzuki, J., Yoshida,
and respond to drug use: ‘Spice’ a case study. Addiction, 105, M. et al. (2011) Identification and quantitation of a benzoylindole
951–953. (2-methoxyphenyl)(1-pentyl-1H-indol-3-yl)methanone and a
3. European Monitoring Centre for Drugs and Drug Addiction. (2009) naphthoylindole 1-(5-fluoropentyl-1H-indol-3-yl)-(naphthalene-
Addiction. Understanding the ‘Spice’ phenomenon. European 1-yl)methanone (AM-2201) found in illegal products obtained via
Monitoring Centre for Drugs and Drug Addiction, Lisbon. the Internet and their cannabimimetic effects evaluated by in vitro
4. Dargan, P.I., Hudson, S., Ramsey, J., Wood, D.M. (2011) The impact [35S]GTPgS binding assays. Forensic Toxicology, 29, 132– 141.
of changes in UK classification of the synthetic cannabinoid

302 Cox et al.