Discrimination and identification of the six aromatic positional isomers of trimethoxyamphetamine (TMA) by gas chromatography-mass spectrometry (GC-MS)

Article abstract of DOI:10.1002/jms.1347

A reliable and accurate GC-MS method was developed that allows both mass spectrometric and chromatographic discrimination of the six aromatic positional isomers of trimethoxyamphetamine (TMA). Regardless of the trifluoroacetyl (TFA) derivatization, chromatographic separation of all the investigated isomers was achieved by using DB-5ms capillary columns (30 m x 0.32 mm i.d.), with run times less than 15 min. However, the mass spectra of the nonderivatized TMAs, except 2,4,6-trimethoxyamphetmine (TMA-6), showed insufficient difference for unambiguous discrimination. On the other hand, the mass spectra of the TFA derivatives of the six isomers exhibited fragments with significant intensity differences, which allowed the unequivocal identification of all the aromatic positional isomers investigated in the present study. This GC-MS technique in combination with TFA derivatization, therefore, is a powerful method to discriminate these isomers, especially useful to distinguish the currently controlled 3,4,5-trimethoxyamphetmine (TMA-1) and 2,4,5-trimethoxyamphetmine (TMA-2) from other uncontrolled TMAs. Copyright

Full text of DOI:10.1002/jms.1347

JOURNAL OF MASS SPECTROMETRY
J. Mass Spectrom. 2008; 43: 528–534
Published online 26 November 2007 in Wiley InterScience
(www.interscience.wiley.com) DOI: 10.1002/jms.1347
Discrimination and identification of the six aromatic
positional isomers of trimethoxyamphetamine (TMA)
by gas chromatography-mass spectrometry (GC-MS)
Kei Zaitsu,1∗ Munehiro Katagi,1 Hiroe Kamata,1 Tooru Kamata,1 Noriaki Shima,1
Akihiro Miki,1 Tatsunori Iwamura2 and Hitoshi Tsuchihashi1
1
Forensic Science Laboratory, Osaka Prefectural Police Headquarters, 1-3-18, Hommachi, Chuo-ku, Osaka 541-0053, Japan
Faculty of Pharmaceutical Sciences, Matsuyama University, 4-2, Bunkyo-Cho, Matsuyama, Ehime 790-8578, Japan
2
Received 28 July 2007; Accepted 9 October 2007
A reliable and accurate GC-MS method was developed that allows both mass spectrometric and
chromatographic discrimination of the six aromatic positional isomers of trimethoxyamphetamine (TMA).
Regardless of the trifluoroacetyl (TFA) derivatization, chromatographic separation of all the investigated
isomers was achieved by using DB-5ms capillary columns (30 m × 0.32 mm i.d.), with run times less
than 15 min. However, the mass spectra of the nonderivatized TMAs, except 2,4,6-trimethoxyamphetmine
(TMA-6), showed insufficient difference for unambiguous discrimination. On the other hand, the mass
spectra of the TFA derivatives of the six isomers exhibited fragments with significant intensity differences,
which allowed the unequivocal identification of all the aromatic positional isomers investigated in
the present study. This GC-MS technique in combination with TFA derivatization, therefore, is a
powerful method to discriminate these isomers, especially useful to distinguish the currently controlled
3,4,5-trimethoxyamphetmine (TMA-1) and 2,4,5-trimethoxyamphetmine (TMA-2) from other uncontrolled
TMAs. Copyright  2007 John Wiley & Sons, Ltd.
KEYWORDS: trimethoxyamphetamine; trifluoroacetyl derivatization; gas chromatography-mass spectrometry; designer
drug; psychotomimetic property; forensic science
TMA-1 was the first psychotomimetic drug synthesized
in 1947 by modifying the structure of mescaline on the basis
of the principles discovered in studying the relationships
between chemical structures and biological activity, and has
about twice the potency of mescaline.12 TMA-2, the geometric
isomer of TMA-1, was first synthesized in 1933, and its
psychotomimetic properties were first reported by Shulgin
in 1964. This was thus called TMA-2 as the second of the
six possible positional isomers found to be psychotomimetic.
TMA-2 was reported to possess some 20 times the potency
of mescaline and is the most active of the six isomers of
TMA.13 The remaining isomers were numbered by Shulgin
in the order of their progressive position of substitution, as
abbreviated above. The values for relative human potencies
of TMA-4, -5, and -6 compared to that of mescaline were
evaluated to be 4, 10, and 10, respectively. TMA-3 was
found to exhibit no psychotomimetic activity.14 The isomers,
except TMA-3, have both stimulant and psychedelic effects,
and their qualitative effects on mood alteration and sensory
enhancement are similar to those of mescaline. Although
their physical health risks and toxic doses in humans have
never been clarified, mental health risks are assumed to be
similar to those of other hallucinogens.14
INTRODUCTION
Drug abuse, which affects human behavior and causes
numerous crimes, has become a serious problem through-
out the world. Amphetamine and its analogs have been
most extensively and increasingly abused as central ner-
vous system stimulants, and are well documented in the
literature.1–6 There are hundreds of possible amphetamine
analogs that modify the basic amphetamine structure and
retain or vary the stimulant effects of the parent compound.
The amphetamine analogs are known usually to be unex-
pectedly introduced onto the clandestine market as ‘designer
drugs’, and their large number makes identification a difficult
task in the forensic science field.7–9
More recently, trimethoxyamphetamines (TMAs) have
emerged on the illicit drug markets in Europe, USA and
even in Japan.10,11 As shown in Fig. 1, there exist six
possible aromatic positional isomers of TMA; i.e. 3,4,5-
trimethoxyamphetmine (TMA-1), 2,4,5-trimethoxyamphet-
mine (TMA-2), 2,3,4-trimethoxyamphetmine (TMA-3), 2,3,5-
trimethoxyamphetmine (TMA-4), 2,3,6-trimethoxyamphet-
mine (TMA-5) and 2,4,6-trimethoxyamphetmine (TMA-6).
Recently, TMA-2 and TMA-6 have been encountered in
the Japanese drug market as well as in Europe and USA and
have been increasingly abused as new designer drugs. In
ŁCorrespondence to: Kei Zaitsu, Forensic Science Laboratory, Osaka
Prefectural Police Headquarters, 1-3-18, Hommachi, Chuo-ku,
Osaka 541-0053, Japan. E-mail: k zaitsu fsl opp@ybb.ne.jp
Copyright  2007 John Wiley & Sons, Ltd.

Discrimination of the six TMA-isomers by GC–MS 529
for TMA-4 and -5 were 3 and 6%, respectively. Every synthe-
sized compound was ensured to be of >98% purity, based on
LC-MS by the flow-injection method. Stock standard solu-
tions of these six compounds were prepared in methanol
(1 mg/ml each), and these solutions were then diluted to the
appropriate concentrations with distilled water, immediately
prior to use.
The trifluoroacetyl (TFA) derivatives of the analytes
were prepared by adding 200 µl of trifluoroacetic anhydride
(Wako, Osaka) and 200 µl of ethyl acetate, followed by
°
reacting at 60 C for 20 min. The reaction mixture was
carefully evaporated to dryness under a gentle nitrogen
stream, and then reconstituted in 100 µl of ethyl acetate.
Then, 1 µl aliquots were automatically injected into the GC-
MS systems.
Figure 1. Chemical structures of trimethoxyamphetamines.
Abbreviations: TMA-1, 3,4,5-trimethoxyamphetamine; TMA-2,
2,4,5-trimethoxyamphetamine; TMA-3,
2,3,4-trimethoxyamphetamine; TMA-4,
2,3,5-trimethoxyamphetamine; TMA-5,
2,3,6-trimethoxyamphetamine; TMA-6,
Chemical synthesis
TMA-4
To concentrated nitric acid (80 ml) cooled in an ice bath,
2-hydroxy-5-methoxybenzaldehyde (7.60 g, 50 mmol) was
added dropwise with stirring, and then the mixture was
2,4,6-trimethoxyamphetamine.
USA all the aromatic positional isomers of TMA have been
controlled under U.S. Federal Law as Schedule I substances,15
while TMA-1 has been controlled in Europe. More recently,
TMA-2 has been controlled under most international laws in
Europe,16 and has also been placed in Schedule I by the U.N.
Narcotics Commission.
On the other hand, in Japan, only TMA-1 had been
controlled until Japanese authorities banned TMA-2 under
the Narcotics and Psychotropics Control Law of 2006 because
of a growing number of its encounter. The other aromatic
positional isomers are not currently classified as illegal drugs,
and nothing prevents their sale and abuse; thus there is much
apprehension of its popularity in Japan.
As mentioned above, some positional isomers of TMA are
currently not controlled in Japan and Europe. Also, a large
variation in the psychoactive properties and toxicity of the
isomers is expected. Discrimination of the six isomers is then
indispensable for drug enforcement and forensic toxicology.
There are no full analytical data available on the aromatic
positional isomers of TMA. Therefore, development of
convenient and reliable analytical methods differentiating
and identifying the isomers is strongly demanded, and has
become an urgent subject in the forensic science field.
In the present study, we have synthesized authentic stan-
dards of the six aromatic positional isomers of TMA. Using
these authentic standards, reliable and accurate methods for
their unequivocal discrimination were developed by GC-MS.
°
stirred for 6 h at 10–15 C. The reaction mixture was
poured into crushed ice and extracted with chloroform.
The organic layer was successively washed with saturated
sodium bicarbonate solution and brine. It was dried over
sodium sulfate and evaporated to give a crude residue, which
was chromatographed on silica gel (hexane–benzene–ethyl
acetate, 5 : 1 : 0.1–5 : 1 : 0.5) to give the pure 2-hydroxy-5-
methoxy-3-nitrobenzaldehyde (5.32 g, 54%).
To a mixture of 2-hydroxy-5-methoxy-3-nitrobenzalde-
hyde (9.85 g, 50 mmol) and potassium carbonate (6.90 g,
50 mmol) in dimethylformamide (100 ml), dimethyl sulfate
(9.5 ml) was added dropwise. The mixture was stirred under
an argon atmosphere at room temperature for 4 days. The
reaction mixture was filtered, evaporated in vacuo, diluted
with water and extracted with chloroform. The organic layer
was washed with brine. Evaporation of the organic solvent
gave a crude residue, which was chromatographed on
silica gel (hexane–benzene–ethyl acetate, 5 : 1 : 0.1–5 : 1 : 0.5)
to give the pure 2,5-dimethoxy-3-nitrobenzaldehyde (9.71 g,
92%).
A
mixture of 2,5-dimethoxy-3-nitrobenzaldehyde
(10.55 g, 50 mmol), powdered iron (2.80 g) and iron (II) sul-
fate heptahydrate (209 g) in an aqueous calcium carbonate
(75 g) suspension (H2O, 1 l) was stirred for 6 h at 75 C. The
solid in the reaction mixture was removed by filtration and
the filtrate was acidified to pH 1 with concentrated HCl in
an ice bath.
°
To a solution of the crude 3-amino-2,5-dimethoxy-
benzaldehyde in aqueous hydrochloric acid, an aqueous
sodium nitrite (3.45 g, 50 mmol) solution (10 ml) was added
EXPERIMENTAL
Materials
°
dropwise with stirring at 5 C. After continuous stirring
TMA-1, -2, -3, and -6 were synthesized according to the
methods of Shulgin.1 These were obtained by lithium
aluminum hydride (Wako, Osaka) reduction of 2-nitro-1-
(trimethoxyphenyl)propenes synthesized from the appro-
priate trimethoxybenzaldehydes (Aldrich, Milwaukee), and
finally purified as their hydrochloride. TMA-4 and -5 were
both synthesized as described in detail in the next section. The
overall yields for TMA-1, -2, -3 and -6 were 14–59%, and those
°
for 0.5 h, the aqueous mixture was increased to 80 C
and stirred for 1 h. The reaction mixture was filtered
and then extracted with dichloromethane. The organic
layer was successively washed with a saturated aqueous
sodium bicarbonate solution and brine and dried over
sodium sulfate. Evaporation of the organic solvent gave
a crude residue, which was chromatographed on silica
Copyright  2007 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2008; 43: 528–534
DOI: 10.1002/jms

530
K. Zaitsu et al.
gel (hexane–benzene–ethyl acetate, 5 : 1 : 0.1–5 : 1 : 0.5) to
provide the pure 3-hydroxy-2,5-dimethoxybenzaldehyde
(1.73 g, 19%).
hydroxide solution. The inorganic material was removed by
filtration and the filter cake was washed with tetrahydrofu-
ran. Evaporation of the combined filtrate and washings gave
the crude residue, which was chromatographed on silica gel
(hexane–benzene–ethyl acetate, 5 : 1 : 0.1–5 : 1 : 0.5) to give
the pure 2,3,6-trimethoxybenzyl alcohol (1.09 g, 55%).
The mixture of 2,3,6-trimethoxybenzyl alcohol (1.98 g,
10 mmol) and manganese dioxide (6.0 g) in benzene was
refluxed for 8 h with vigorous stirring. The inorganic
salts in the reaction mixture were removed by filtration,
and the filter cake was washed with acetone. Evapora-
tion of the combined filtrate and washings gave a crude
residue, which was chromatographed on silica gel (hex-
ane–benzene–ethyl acetate, 5 : 1 : 0.1–5 : 1 : 0.5) to provide
pure 2,3,6-trimethoxybenzaldehyde (1.45 g, 74%).
Under an argon atmosphere, a mixture of 3-hydroxy-2,5-
dimethoxybenzaldehyde (1.82 g, 10 mmol), methyl iodide
(5.0 ml) and sodium carbonate (1.06 g, 10 mmol) in ace-
tone (50 ml) was stirred under a gentle reflux for 6 h.
The mixture was filtered and evaporated to give a crude
residue, which was chromatographed on silica gel (hex-
ane–benzene–ethyl acetate, 5 : 1 : 0.1–5 : 1 : 0.5) to give pure
2, 3, 5-trimethoxybenzaldehyde (1.61 g, 82%).
A solution of 2,3,5-trimethoxybenzaldehyde (1.96 g,
10 mmol)andanhydrousammoniumacetate(0.39 g, 5 mmol)
in nitroethane (7.2 ml) was refluxed for 2 h under an argon
atmosphere. The excess nitroethane was removed under
vacuum, and the residue was dissolved in ethyl acetate. The
organic layer was washed with brine and dried over sodium
sulfate. Evaporation of the organic solvent gave a crude
residue, which was chromatographed on silica gel (hex-
ane–benzene–ethyl acetate, 5 : 1 : 0.1–5 : 1 : 0.5) to give pure
2-nitro-1-(2,3,5-trimethoxyphenyl)propene (1.77 g, 70%).
To a suspension of lithium aluminum hydride (1.52 g,
40 mmol) in dry tetrahydrofuran (30 ml) under an argon
atmosphere, a solution of 2-nitro-1-(2,3,5-trimethoxyphenyl)
propene (2.53 g, 10 mmol) in dry tetrahydrofuran (10 ml)
was added dropwise at room temperature, and the mixture
According to the above-mentioned procedure for TMA-
4, TMA-5 hydrochloride (0.70 g, 32%) was prepared, with
a slight modification: instead of 2,3,5-trimethoxybenzalde-
hyde, 2,3,6-trimethoxybenzaldehyde was used.
NMR data for the synthesized compounds
The 1H NMR spectra were obtained in CDCl3 solutions
using a Varian Gemini spectrometer operating at 300 MHz.
The detailed spectral data are given below.
TMA-1: 1H NMR (300 MHz, CDCl3) υ 6.41 (2H, s), 3.85
(6H, s), 3.83 (3H,s), 3.17 (1H, ddq, J D 4.8, 6.0, 8.3 Hz), 2.68
(1H, dd, J D 4.8, 13.4 Hz), 2.42 (1H, dd, J D 8.3, 13.4 Hz),
1.14 (3H, d, J D 6.0 Hz).
°
was kept at 40 C with stirring for 12 h, and then the reaction
mixture was cooled in an ice bath. The excess hydride was
decomposed by adding water and a sodium hydroxide
solution. The inorganic salts were removed by filtration,
and the filter cake was washed with tetrahydrofuran.
Evaporation of the combined filtrate and washings gave the
crude base. This was dissolved in isopropyl alcohol, and then
neutralized with methanolic hydrogenchloride. The solution
was saturated with diethyl ether and allowed to stand for
a few days to form a residue, which was recrystallized
from isopropanol–ether to afford fine needles of TMA-4
hydrochloride (1.18 g, 54%).
TMA-2: 1H NMR (300 MHz, CDCl3) υ 6.69 (1H, s), 6.52
(1H, s), 3.88 (3H, s), 3.83 (3H, s), 3.79 (3H, s), 3.15 (1H, ddq,
J D 5.4, 6.3, 8.0 Hz), 2.67 (1H, dd, J D 5.4, 13.1 Hz), 2.47
(1H, dd, J D 8.0, 13.1 Hz), 1.10 (3H, d, J D 6.3 Hz).
TMA-3: 1H NMR (300 MHz, CDCl3) υ 6.82 (1H, d,
J D 8.4 Hz), 6.61 (1H, d, J D 8.4 Hz), 3.874 (3H, s), 3.867
(3H, s), 3.84 (3H, s), 3.13 (1H, ddq, J D 5.4, 6.3, 7.8 Hz), 2.66
(1H, dd, J D 5.4, 13.2 Hz), 2.46 (1H, dd, J D 7.8, 13.2 Hz),
1.11 (3H, d, J D 6.3 Hz).
TMA-4: 1H NMR (300 MHz, CDCl3) υ 6.39 (1H, d,
J D 3.0 Hz), 6.28 (1H, d, J D 3.0 Hz), 3.83 (3H, s), 3.77
(3H, s), 3.75 (3H, s), 3.19 (1H, ddq, J D 5.7, 6.3, 8.1 Hz), 2.70
(1H, dd, J D 5.7, 12.9 Hz), 2.54 (1H, dd, J D 8.1, 12.9 Hz),
1.13 (3H, d, J D 6.3 Hz).
TMA-5
To a stirred solution of 1,2,4-trimethoxybenzene (1.68 g,
10 mmol) in tetrahydrofuran (15 ml) was added dropwise
a 1.0 M n-butyl lithium solution in tetrahydrofuran (11 ml)
TMA-5: 1H NMR (300 MHz, CDCl3) υ 6.73 (1H, d, J D 9.0),
6.55 (1H, d, J D 9.0), 3.82 (3H, s), 3.81 (3H, s), 3.76 (3H, s),
3.15 (1H, ddq, J D 5.9, 6.0, 7.8 Hz), 2.73 (1H, dd, J D 5.9,
12.5 Hz), 2.64 (1H, dd, J D 7.8, 12.5 Hz), 1.11 (3H, d,
J D 6.0 Hz).
°
at ꢀ78 C under an argon atmosphere. The reaction mixture
was held at that temperature while stirring for 1 h, and then
quenched by the addition of solid carbon dioxide. Aqueous
ammonium chloride was added to the mixture, followed by
extraction with ethyl acetate. The organic layer was washed
with brine and dried over sodium sulfate. Evaporation
of the organic solvent gave a crude residue, which was
chromatographed on silica gel (hexane–ethyl acetate, 5 : 1) to
give pure 2,3,6-trimethoxybenzoic acid (1.04 g, 49%).
TMA-6: 1H NMR (300 MHz, CDCl3) υ 6.13 (2H, s), 3.81
(3H, s), 3.78 (6H,s), 3.08 (1H, ddq, J D 5.6, 6.3, 7.8 Hz), 2.63
(1H, dd, J D 5.6, 12.9 Hz), 2.57 (1H, dd, J D 7.8, 12.9 Hz),
1.07 (3H, d, J D 6.3 Hz).
To a suspension of lithium aluminum hydride (0.38 g,
10 mmol) in dry tetrahydrofuran (10 ml) under an argon
atmosphere, a solution of 2,3,6-trimethoxybenzoic acid
(2.12 g, 10 mmol) in dry tetrahydrofuran (30 ml) was added
dropwise at room temperature. The mixture was kept at
Instrumentation
GC-MS was carried out using a GCMS QP-2010 instru-
ment (Shimadzu, Kyoto, Japan). Three different fused-
silica capillary columns, DB-1MS, DB-5MS and DB-17MS
(30 m ð 0.32 mm i.d.; 0.25 ꢀ µm thickness; J&W Scientific,
Rancho Cordova, CA, USA), were investigated for the
chromatographic separation. Samples were automatically
°
40 C with stirring for 12 h. Cooled in an ice bath, the excess
hydride decomposed by the addition of water and a sodium
Copyright  2007 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2008; 43: 528–534
DOI: 10.1002/jms

Discrimination of the six TMA-isomers by GC–MS 531
°
injected in the splitless mode at 250 C. The column oven
-5, -4, -1, -2 and -6. However, they were all eluted with
somewhat tailing peaks (Fig. 2).
°
temperature was maintained at 80 C for 1 min, and then
°
°
raised at 10 C/min to 320 C. The transfer line temperature
For improvement of the peak shapes, TFA derivatization
of the TMAs was then examined. As depicted in Fig. 3,
all TFA derivatives of the six TMAs eluted into the
baseline-separated peaks in the same order as those of
the nonderivatized TMAs with DB-1ms or DB-5ms within
15 min. Also, the tailing of the analytes observed for the free
bases could be overcome. However, with DB-17ms, TMA-
1, -2 and -6 were eluted with insufficient resolution. The
calculated Kovats indices17 are summarized in Table 1.
On the basis of the above comparison, we finally chose
DB-5ms as the analytical column, which provided a better
resolution among the free bases of TMA-4, -1 and -2 than
from DB-1ms.
°
was set at 250 C. High-purity helium, at the flow rate of
3 ml/min, was used as the carrier gas. The electron ion-
ization (EI) operating parameters were as follows: source
temperature, 200 C; electron energy, 70 eV; ion multiplier
gain, 1.2kV. Mass spectra were collected from m/z 40 to 500
°
at the scan rate of 0.5 s/scan.
RESULTS AND DISCUSSION
Chromatographic separation
In order to optimize the separation of the six aromatic
positional isomers of TMA, three different column phases
were explored, i.e. DB-1ms, DB-5ms and DB-17ms, under
the same GC-MS operating conditions as described in the
experimental section. As samples, artificial mixtures of the
six TMAs as their free base in ethyl acetate (1 µg/ml) were
first used.
Mass spectrometry
As shown in Fig. 4, each EI mass spectrum of the six
nonderivatized TMAs is characterized by the predominant
immonium ions at m/z 44 produced by an ˛-cleavage
reaction, and the significant ions at m/z 182 and 167. These
fragment ions at m/z 182 and 167 may be generated by
With all columns, all six TMAs could be baseline-
separated from each other within 15 min in the order TMA-3,
Figure 2. Total-ion and extracted-mass chromatograms obtained from a mixture of the nonderivatized TMAs using the different
columns; top: DB-1ms, middle: DB-5ms, bottom: DB-17ms. Peak identification: 1, TMA-1; 2, TMA-2; 3, TMA-3; 4, TMA-4; 5, TMA-5;
6, TMA-6.
Table 1. Retention indices of the TMAs and the TFA-derivatized compounds
DB-1ms
DB-5ms
DB-17ms
column type
Compound
form
Free
base
TFA
Free
base
TFA
Free
base
TFA
derivative
derivative
derivative
TMA-1
TMA-2
TMA-3
TMA-4
TMA-5
TMA-6
1678
1691
1593
1669
1634
1711
1788
1802
1712
1767
1732
1820
1718
1730
1631
1709
1670
1761
1817
1830
1737
1793
1755
1846
1876
1888
1758
1852
1806
1920
2118
2124
1996
2069
2014
2128
Copyright  2007 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2008; 43: 528–534
DOI: 10.1002/jms

532
K. Zaitsu et al.
Figure 3. Total-ion and extracted-mass chromatograms obtained from a mixture of the TFA-derivatized TMAs using the different
columns; top: DB-1ms, middle: DB-5ms, bottom: DB-17ms. Peak identification: 1, TMA-1; 2, TMA-2; 3, TMA-3; 4, TMA-4; 5, TMA-5;
6, TMA-6.
Figure 4. EI mass spectra of the nonderivatized TMAs.
an H-rearrangement from the amino group to the aromatic
part of the molecule with a subsequent ˛-cleavage of the
benzyl bond by the radical electron at the nitrogen, and the
subsequent elimination of a methyl radical.18,19: Especially,
in the mass spectra of TMA-1, -2, -3, -4 and -5, the ions at
m/z 182 exhibited an abundance of about 15–45%, though
TMA-6 showed the ion with an abundance of about 95%. It
could be due to the stabilizing effect caused from the ortho-
disubstitution of two methoxy groups on positive charge. At
least, this higher relative abundance of the fragment ions at
m/z 182 seems to be a discriminating indicator for TMA-6.
However, there appears no significant fragment ion
to allow the unequivocal discrimination between the five
isomeric TMAs, and therefore the mass spectral information
is not sufficient to unambiguously identify the five TMAs
except TMA-6, without derivatization.
In our previous study, TFA derivatization was success-
fully applied to the discrimination of the ortho-, meta-, and
para-methoxyamphetamines, which was able to be mass-
spectral-differentiate with derivatization.4,5 In the present
study, TFA derivatization of the isomeric TMAs was
explored.
As shown in Fig. 5, the mass spectra of the trifluoroacety-
lated TMA-1, -2, -3 and -6 were characterized by base peak
ions at m/z 181 owing to their trimethoxy-substituted benzyl
or tropylium cations and their relatively intense molecular
ions at m/z 321 (5–15%).
Also, the spectra of the four isomeric TFA derivatives
showed a significant intensity difference for the fragments
at m/z 208, 193, 166, 151, 136 and 121: TMA-1-TFA exhibited
relatively intense ions at m/z 208 (18%), and 193 (8%); TMA-
2-TFA at m/z 151 (26%), 136 (8%) and 208 (7%); TMA-3-TFA
Copyright  2007 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2008; 43: 528–534
DOI: 10.1002/jms

Discrimination of the six TMA-isomers by GC–MS 533
Figure 5. EI mass spectra of the TFA-derivatized TMAs.
Table 2. Relative intensities (%) of the ions in the EI mass spectra obtained from the TFA derivatives of the six TMA isomers
m/z
TMA-1-TFA
TMA-2-TFA
TMA-3-TFA
TMA-4-TFA
TMA-5-TFA
TMA-6-TFA
321
208
193
181
166
151
136
121
15
18
8
100
–
13
7
–
100
–
26
8
11
28
–
100
55
5
65
43
54
18
100
6
40
52
5
73
100
11
18
18
5
2
–
100
–
2
12
20
2
4
17
5
12
18
–
–
at m/z 166 (55%), 208 (28%), 136 (17%) and 121 (5%); TMA-
6-TFA at m/z 121 (20%), 136 (12%) and 208 (2%). These mass
spectrometric data for the TFA derivatives are summarized
in Table 2.
CONCLUSIONS
A reliable and accurate GC-MS method was developed
that allows both mass spectrometric and chromatographic
discrimination of the six aromatic positional isomers of
TMA. Regardless of the derivatization, chromatographic
separation of all the investigated isomers was achieved
by using the DB-5ms capillary column, with run times
less than 15 min. The mass spectra of the nonderivatized
TMAs, except TMA-6, showed insufficient difference for
their unambiguous determination. The fragments of the mass
spectra of their TFA derivatives, on the other hand, exhibited
a significant intensity difference, which allowed the mass
spectral differentiation of all the isomers investigated in the
present study. GC-MS with TFA derivatization, therefore, is
a powerful method for the discrimination of these isomers,
especially useful to distinguish the currently controlled
TMA-1 and TMA-2 from other uncontrolled TMAs.
The mass spectra of the trifluoroacetylated TMA-4 and
-5, on the other hand, were characterized by base peak ions
at m/z 166 and intense molecular ions at m/z 321 (40–65%).
Also, the spectra of the two isomeric TFA derivatives showed
a significant intensity difference for the fragments at m/z 208,
193, 181, 151, 136 and 121: TMA-4-TFA exhibited relatively
intense ions at m/z 193 (54%), 208 (43%), 181 (18%), 121 (18%),
136 (12%) and 151 (6%), and TMA-5-TFA at m/z 181 (73%),
208 (52%), 136 (18%), 121 (18%), 151 (11%) and 193 (5%).
The fragment ions at m/z 208 would be generated by
the inductive route of the McLafferty rearrangement, and
the fragment ions at m/z 151 and 121 would be produced
from their trimethoxy-substituted benzyl cations at m/z 181
by loss of a formaldehyde molecule and two formaldehyde
molecules, respectively.18,19
These variations of the relative intensities between the
significant fragment ions, as mentioned above, would allow
discrimination of the six trifluoroacetylated TMAs, leading
to the unequivocal identification of the positional isomeric
TMAs. TFA derivatization, therefore, has been found to be
very effective for the mass spectral differentiation of the
positional isomers of TMA.
REFERENCES
1. Shulgin AT. In Pihkal, A Chemical Love Story, Dan Joy (ed).
Transform Press: Berkley, 1991.
2. Ellenhorn MJ (ed). Amphetamines and designer drugs.
Ellenhorn’s Medical Toxicology, 2nd ed. Williams & Wilkins:
Baltimore, 1997; 340.
3. Logan BK. Amphetamines: an update on forensic issues. Journal
of Analytical Toxicology 2001; 25: 400.
Copyright  2007 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2008; 43: 528–534
DOI: 10.1002/jms

534
K. Zaitsu et al.
4. Munehiro K, Hiroe T, Akihiro M, Kunio N, Hitoshi T. Analyses
of clandestine tablets of amphetamines and their related designer
drugs encountered in recent Japan. Japanese Journal of Forensic
Toxicology 2002; 20: 303.
5. Munehiro K, Hitoshi T. Update on clandestine amphetamines
and their analogues recently seen in Japan. Journal of Health
Science 2002; 48: 14.
11. Tsujikawa K, Kanamori T, Kuwayama K, Miyaguchi H, Iwata Y,
Inoue H. Analytical profiles for 3, 4, 5-, 2, 4, 5-, and 2, 4, 6-
trimethoxyamphetamine. Microgram Journal 2006; 4: 12.
12. Snyder SH, Richelson E. Psychedelic drugs: steric factors that
predict psychotropic activity. Biochemistry 1968; 60: 206.
13. Shulgin AT. Profiles of psychedelic drugs: TMA-2. Journal of
Psychedelic Drugs 1976; 8: 169.
6. de Boer D, Bosman I. A new trend in drug-of-abuse; the 2C-
series of phenethylamine designer drugs. Pharmacy World and
Science 2004; 26: 110.
7. Marson C, Schneide S, Meys F, Wenning R. Structural
elucidation of an uncommon phenethylamine analogue in urine
responsible for discordant amphetamine immunoassay results.
Journal of Analytical Toxicology 2000; 24: 17.
14. Shulgin AT. Psychotomimetic drugs: structure-activity relation-
ships. In The Handbook of Psychopharmacology, vol. 11, Stimulants,
Iverson LL, Iverson SD, Snyder SH (eds). Plenum Publishing
Company: New York, 1978; 243.
15. Drug and Chemical Evaluation Section, Office of Diversion
Control, Drug Enforcement Administration, U.S. Department
of Justice, Scheduling Actions Controlled Substances Regulated
Chemicals, April 2007.
16. Council of the European Union, Council decision 2003/847/JHA
of 27 November 2003 concerning control measures and criminal
sanctions in respect of the new synthetic drugs 2C-1, 2C-T-2,
2C-T-7 and TMA-2, official journal of the European union, 2003; 6:
64.
8. Koper C, Ali-Tolppa E, Bozenko JS, Dufey V, Puetz M, Weyer-
mann C, Zrcek F. Identification of a new amphetamine type stim-
ulant:
3,4-methylenedioxy-N-(2-hydroxyethyl)amphetamine
(MDHOET). Microgram Journal 2005; 3: 166.
9. Ro¨sner P, Quednow B, Girreser U, Junge T. Isomeric fluoro-
methoxy-phenylalkylamines:
a new series of controlled-
substance analogues (designer drugs). Forensic Science
International 2005; 148: 143.
17. Kovats ES. Gas chromatographic characterization of organic
substances in the retention index system. Advances in
Chromatography 1965; 1: 229.
18. McLafferty FW, Turecek F. Interpretation of Mass Spectra, 4th ed.
University Science Books: Mill Valley, 1993.
10. Ewald AH, Fritschi G, Maurer HH. Designer drug 2, 4, 5-
trimethoxyamphetamine (TMA-2): studies on its metabolism
and toxicological detection in rat urine using gas
chromatographic/massspectrometric techniques. Journal of Mass
Spectrometry 2006; 41: 1140.
19. Smith RM, Busch KL. Understanding Mass Spectra – A Basic
Approach. Wiley: New York, 1999.
Copyright  2007 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2008; 43: 528–534
DOI: 10.1002/jms