Gas chromatographic/mass spectrometric assay for profiling the enantiomers of 3,4-methylenedioxymethamphetamine and its chiral metabolites using positive chemical ionization ion trap mass spectrometry

Article abstract of DOI:10.1002/(SICI)1096-9888(199711)32:11<1236::AID-JMS584>3.0.CO;2-K

A qualitative GC/MS profile was obtained and its mass spectrometric features characterized for the analysis of the enantiomers of (RS)-3,4-methylenedioxmethamphetamine (MDMA) and its metabolites (RS)-3,4- methylenedioxyamphetamine (MDA), (RS)-4-hydroxy-3-methoxymethamphetamine (HMMA) and (RS)-4- hydroxy-3-methoxyamphetamine (HMA). A chiral derivatization method was selected to obtain the diastereomers required for the separation of the respective enantiomers with a non-chiral GC stationary phase. The selected derivatization consisted of a reaction with N-heptafluorobutyryl-(S)-prolyl chloride combined with a consecutive reaction with N-methyl-N-trimethylsilyltrifluoracetamid, resulting in N-[heptafluorobutyryl-(S)-prolyl]-O-trimethylsilyl derivatives. Detection was carried out with electron ionization and positive chemical ionization (PCI) ion trap mass spectrometry. Mass spectra of the derivatives of reference standards of the compounds of interest obtained with PCI demonstrated that this method simultaneously induces proton and charge-transfer reactions in the ion trap. The advantage is that high mass information is provided while some fragmentation remains to elucidate structural details. Subsequently, in three urine samples obtained from different and unrelated MDMA intoxications the enantiomers of MDMA and MDA were identified. In some urine samples also HMMA and/or HMA were found. In addition to these compounds, an unexpected compound and/or additional chiral metabolite, N-hydroxy-(RS)-3,4-methylenedioxyamphetamine, was identified in two out of three urine samples. Preliminary results also indicated an enantioselective metabolism in the N-demethylation pathway for MDMA in humans.

Full text of DOI:10.1002/(SICI)1096-9888(199711)32:11<1236::AID-JMS584>3.0.CO;2-K

JOURNAL OF MASS SPECTROMETRY, VOL. 32, 1236È1246 (1997)
Gas Chromatographic/Mass Spectrometric Assay
for ProÐling the Enantiomers of
3,4-Methylenedioxymethamphetamine and its
Chiral Metabolites using Positive Chemical
Ionization Ion Trap Mass Spectrometry
D. de Boer,1* L. P. Tan,1 P. Gorter,1 R. M. A. van de Wal,1 J. J. Kettenes-van den Bosch,2
E. A. de Bruijn1,3 and R. A. A. Maes2
1 Department of Human Toxicology, Utrecht Institute for Pharmaceutical Sciences, University of Utrecht, Sorbonnelaan 16,
3583 TC Utrecht, The Netherlands
2 Department of Pharmaceutical Analysis, Utrecht Institute for Pharmaceutical Sciences, University of Utrecht,
The Netherlands
3 Laboratory of Radiotherapy and Experimental Oncology, Catholic University Leuven, Leuven, Belgium
A qualitative GC/MS proÐle was obtained and its mass spectrometric features characterized for the analysis of the
enantiomers of (RS)-3,4-methylenedioxymethamphetamine (MDMA) and its metabolites (RS)-3,4-
methylenedioxyamphetamine (MDA), (RS)-4-hydroxy-3-methoxymethamphetamine (HMMA) and (RS)-4-
hydroxy-3-methoxyamphetamine (HMA). A chiral derivatization method was selected to obtain the diastereomers
required for the separation of the respective enantiomers with a non-chiral GC stationary phase. The selected
derivatization consisted of a reaction with N-heptaÑuorobutyryl-(S)-prolyl chloride combined with a consecutive
reaction with N-methyl-N-trimethylsilyltriÑuoroacetamide, resulting in N-[heptaÑuorobutyryl-(S)-prolyl]-O-
trimethylsilyl derivatives. Detection was carried out with electron ionization and positive chemical ionization (PCI)
ion trap mass spectrometry. Mass spectra of the derivatives of reference standards of the compounds of interest
obtained with PCI demonstrated that this method simultaneously induces proton and charge-transfer reactions in
the ion trap. The advantage is that high mass information is provided while some fragmentation remains to eluci-
date structural details. Subsequently, in three urine samples obtained from di†erent and unrelated MDMA intoxi-
cations, the enantiomers of MDMA and MDA were identiÐed. In some urine samples also HMMA and/or HMA
were found. In addition to these compounds, an unexpected compound and/or additional chiral metabolite, N-
hydroxy-(RS)-3,4-methylenedioxyamphetamine, was identiÐed in two out of three urine samples. Preliminary
results also indicated an enantioselective metabolism in the N-demethylation pathway for MDMA in humans.
( 1997 John Wiley & Sons, Ltd.
J. Mass Spectrom. 32, 1236È1246 (1997)
No. of Figures: 7 No. of Tables: 3 No. of Refs: 31
KEYWORDS: 3,4-methylenedioxymethamphetamine; metabolism; enantioselectivity; gas chromatography/mass spectrom-
etry; ion trap
INTRODUCTION
The importance of enantiometric proÐling in phar-
macodynamics and toxicology in general1 and in 3,4-
methylenedioxymethamphetamine (MDMA) studies in
particular2,3 is well established. MDMA, more com-
monly known as “EcstasyÏ or “XTC,Ï is an amphetamine
analogue (Fig. 1), which is a common recreational drug
of abuse in Europe. It has the reputation of being a mild
intoxicant, although serious neurotoxic e†ects in rats
Figure 1. Molecular structures of the enantiomers of MDMA,
some of its analogues and/or its chiral metabolites. Asterisks
indicate
a
chiral
center.
MDMA ¼3,4-methylenedi-
* Correspondence to: D. de Boer, Department of Human Toxi-
cology, Utrecht Institute for Pharmaceutical Sciences, University of
Utrecht, Sorbonnelaan 16, 3583 TC Utrecht, The Netherlands.
E-mail: D. de Boer=far.ruu.nl.
oxymethamphetamine; MDA ¼3,4-methylenedioxyamphetamine;
MDAOH ¼ N - hydroxy - 3 , 4 - methylenedioxyamphetamine;
HMMA ¼4-hydroxy-3-methoxymethamphetamine;
4-hydroxy-3-methoxyamphetamine.
HMA ¼
CCC 1076È5174/97/111236È11 $17.50
( 1997 John Wiley & Sons, Ltd.
Received 22 May 1997
Accepted 5 August 1997

METHYLENEDIOXYMETHAMPHETAMINE AND CHIRAL METABOLITES
1237
have been described.4,5 In humans several com-
plications after MDMA use have been reported.6,7 Also
sold under the names of “EcstasyÏ or “XTCÏ are tablets
containing the 3,4-methylenedioxy analogues 3,4-
methylenedioxyamphetamine (MDA) and N-hydroxy-3,
4-methylenedioxyamphetamine (MDAOH) (Fig 1).8
Illicitly synthesized MDMA usually is a racemic
mixture. The two enantiomers of MDMA, however,
show di†erent pharmacodynamic properties. In animal
models, (S)-MDMA is a more potent psychomimetic
agent than the (R)-enantiomer and reportedly only the
(S)-enantiomer is neurotoxic.4 The major chiral metabo-
lites, which in addition to the parent compound have
been found in urine of animals and humans after decon-
jugation, are MDA, 4-hydroxy-3-methoxymeth-
Amphetamine, (RS)-methamphetamine and (RS)-
propylhexedrine were kindly provided by Professor Dr
J. M. van Rossum (Catholic University of Nijmegen,
The Netherlands). A mixture of b-glucuronidase and
arylsulfatase from Helix pomatia was obtained from
Serva (Heidelberg, Germany). Anhydrous ammonium
acetate ([98%), concentrated hydrochloric acid (37%,
w/v), deuterotrichloromethane (CDCl ) ([99%),
3
dichloromethane ([99%), diethyl ether ([99%), ethyl
acetate ([99.5%), iron(III) chloride (99%), formic acid
([98%), hexadeuterodimethyl sulfoxide (DMSO-d )
6
([99.8%), hydrogen peroxide (30% in water), propan-2-
ol ([99%), lithium aluminum hydride (97%), anhy-
drous magnesium sulfate ([70% water free), methanol
(99.8%), picric acid ([99.8%), potassium hydroxide
([85%), sodium chloride ([99.5%), sodium hydroxide
([99%), sulfuric acid (96%) and thionyl chloride were
purchased from Merck Nederland (Amsterdam, The
Netherlands). Iron powder (325 mesh, 97%), nitroethane
(96%), safrole (97%), tetrahydrofuran ([99%), tetra-
methylsilane and vanillin (99%) were purchased from
Acros Chimica (Geel, Belgium). HeptaÑuorobutyryl
anhydride, (S)-proline ([99%) and sodium cyano-
borohydride (95%) were obtained from Aldrich Chemie
amphetamine
(HMMA)
and
4-hydroxy-3-
methoxyamphetamine (HMA)9h12 (Fig. 1). The metabo-
lism in animal models is enantioselective,2 resulting in
di†erent R/S ratios of MDA and HMA (R/S ratio \1)
compared with MDMA and HMMA (R/S ratio [1).
For humans, such data are not yet available.
The goal of the present study was to develop a gas
chromatographic/mass spectrometric (GC/MS) assay
for the enantiomeric proÐling of MDMA and its chiral
metabolites and to characterize its mass spectrometric
features. Because suitable and stable chiral stationary
phases for GC are not readily commercially available,
chiral derivatization methods are being recommended
for the assay of enantiomers in MDMA studies.2 In this
way, enantiomers are converted into diastereomers
which can be separated with a non-chiral GC stationary
phase. Recently reported chiral derivatization methods
for MDMA and/or MDA involved reaction with N-
(Axel, The Netherlands). Tetrachloromethane (CCl )
4
([99%) was purchased from J. T. Baker (Deventer, The
Netherlands). Methylamine hydrochloride ([98%) and
deuterotriÑuoroacetic acid (TFA-d) were purchased
from
Fluka
(Buchs,
Switzerland).
N-
Methylbis(triÑuoroacetamide) (MBTFA) and N-methyl-
N-trimethylsilyltriÑuoroacetamide (MSTFA) were
obtained from MachereyÈNagel (Duren, Germany). Tri-
Ñuoroacetic anhydride (TFAA) was obtained from
Pierce Europe (Oud-Beijerland, The Netherlands).
Hydroxylamine hydrochloride (99%) was obtained from
Sigma (St Louis, MO, USA). Silica (dry-column grade,
activity III, 30 mm) for chromatography was obtained
from Woelm (Eschwege, Germany). If not speciÐed, the
reagents and chemicals were of analytical grade.
triÑuoroacetyl-(S)-prolyl
Ñuoropropionyl-(S)-prolyl chloride13 or N-hepta-
resulting
chloride,3,13,14
N-penta-
Ñuorobutyryl-(S)-prolyl
chloride,2,13,14
in
N-[triÑuoroacetyl-(S)-prolyl]
(\N-(S)-TFAP),
N-[pentaÑuoropropionyl-(S)-prolyl]
(\N-(S)-PFPP)
and N-[heptaÑuorobutyryl-(S)-prolyl] (\N-(S)-HFBP)
derivatives, respectively. For the metabolites HMMA
and HMA with additional hydroxy groups to be
derivatized, a reaction with N-heptaÑuorobutyryl-(S)-
Synthesis of reference standards of metabolites
prolyl chloride followed by
methylbis(triÑuoroacetamide) was reported to be
a reaction with N-
The reference standards of metabolites and their start-
ing materials were synthesized according to published
synthetic routes and were characterized by proton
nuclear magnetic resonance (1H NMR) spectrometry
and electron ionization (EI) GC/MS and, if speciÐed,
also by 13C NMR spectrometry. 1H NMR (90 MHz)
was carried out with a Varian EM 390 instrument
(Varian, Palo Alto, CA, USA) with tetramethylsilane as
internal standard. 1H NMR (300.07 MHz) and 13C
NMR (75.5 MHz) spectra of end products were record-
ed with a Gemini-300 BB instrument (Varian). In 1H
useful.2 This reaction results in N-[heptaÑuorobutyryl-
(S)-prolyl]-O-triÑuoroacetyl
(N-(S)-HFBP-O-TFA)
derivatives. In our study, the second step in this chiral
derivatization method for the metabolites HMMA and
HMA was carried out with N-methyl-N-trimethylsilyl-
triÑuoroacetamide, resulting in O-trimethylsilyl (O-
TMS) derivatives, instead of O-TFA derivatives.
In addition to the chiral derivatization, the usefulness
of ion trap mass spectrometry and the preliminary
results of the enantiomeric proÐles of MDMA and its
chiral metabolites in human urine are described.
spectra the CHCl line was set at 7.26 ppm and in 13C
3
spectra the central CDCl signal was set at 76.9 ppm.
3
GC/MS was performed with a Hewlett-Packard GC
5890 Series II instrument equipped with a CP-Sil 8 CB
low-bleed MS fused-silica column (30 m ] 0.25 mm i.d.,
Ðlm thickness 0.25 lm) (Chrompack, Middelburg, The
Netherlands) coupled to a Hewlett-Packard Model
5972 mass-selective detector (Hewlett-Packard, Palo
Alto, CA, USA). Mass spectral data were recorded from
m/z 50 to 650 and are reported as normalized to the
EXPERIMENTAL
Chemicals and reagents
(RS)-MDMA was a generous gift from Dr H. Huizer
(Forensic Laboratory, Rijswijk, The Netherlands). (S)-
( 1997 John Wiley & Sons, Ltd.
JOURNAL OF MASS SPECTROMETRY VOL. 32, 1236È1246 (1997)

1238
D. DE BOER ET AL .
base peak. Prior to GC/MS and if speciÐed, the synthe-
sized compounds were converted into triÑuoroacetyl
(TEA) derivatives. Ethyl acetate (50 ll) and 50 ll of
TFAA were added to dry residues and the mixture was
heated for 20 min at 60 ¡C. Excess of organic solvent
and TFAA were removed by evaporation and the
residue was dissolved in 100 ll of ethyl acetate. A 1 ll
aliquot was injected into the GC/MS instrument.
Melting points were determined in capillary tubes and
are uncorrected.
was extracted with two consecutive portions of 50 ml of
dichoromethane, brought to pH 10 with solid sodium
hydroxide, saturated with sodium chloride and
extracted three times with 50 ml portions of dichorome-
thane. The combined extracts were dried (magnesium
sulfate) and the solvent was removed in vacuo. The
resulting oil was dissolved in propan-2-ol and acidiÐed
with concentrated hydrochloric acid. A white solid sub-
stance crystallized spontaneously, was Ðltered o† and
washed Ðrst with propan-2-olÈdiethyl ether (1 : 1) and
then with diethyl ether. The product was isolated as
(RS)-MDA hydrochloride (m.p. 187È188 ¡C, lit.21 187È
Starting materials (IÈIII) for the synthesis of the refer-
ence standards were produced in several batches.
188 ¡C; yield 31%). 1H NMR (300.07 Hz; CDCl ÈTFA-
3
d (6 : 1; v/v)): d 1.45 (d, 3 H, J \ 6.6 Hz, CCH ), 2.86
3
Piperonyl methyl ketone (I). Safrole was isomerized to
isosafrole15 by heating with 2.5 molar equivalents of
potassium hydroxide for 6.5 h at 120 ¡C. After cooling
to room temperature, the clear golden yellow solution
was decanted (average yield [90%). The crude product
was used in the next step as soon as possible, because
isosafrole polymerizes slowly in the presence of mois-
ture. Piperonyl methyl ketone was obtained by hydro-
gen peroxide oxidation of crude isosafrole followed by
an acid-catalyzed rearrangement.16 Distillation under
(dd, 1 H, J \ 14.5 Hz, J \ 8.2 Hz, H-1), 2.95 (dd, 1
AB
BX
H, J \ 14.5 Hz, J \ 6.3 Hz, H-1), 3.68 (m, 1 H,
AB
AX
CHN), 5.96 (s, 2 H, OCH O), 6.67 (dd, 1 H, J \ 7.8 Hz,
2
o
J \ 1.8 Hz, H-6@), 6.69 (d, 1 H, J \ 1.8 Hz, H-2@), 6.79
m
m
(d, 1 H, J \ 7.8 Hz, H-5@), 6.9 (bs, 3 H, NH `). 13C
o
3
NMR (75.5 Hz; CDCl ÈTFA-d (6 : 1; v/v)): d 18.0
3
(CH ), 40.3 (CH ), 51.2 (CH), 101.2 (CH O), 109.0 (C-2@),
3
2
2
109.2 (C-5@), 122.5 (C-6@), 127.8 (C-1@), 147.3 (C-4@), 148.3
(C-3@). GC/MS (TFA derivative): m/z, M`~ 275(15),
162(38), 135(100), 77(16), 69(11).
reduced pressure yielded
a clear yellow product
(average yield 40%). 1H NMR (90 Hz; CCl ): d 1.99 (s,
4
3 H, CH ), 3.47 (s, 2 H, CH CO), 5.88 (s, 2 H OCH O),
(RS)-HMMA. (RS)-HMMA (1-(3-methoxy-4-hydroxy-
phenyl)-2-methylaminopropane) was prepared from III
by reductive amination with sodium cyano-
borohydride.21,22 To a solution of 3.0 g of methylamine
hydrochloride in 20 ml of methanol were added 2.3 g of
III followed by 1.0 g of sodium cyanoborohydride. The
solution was treated as in the reductive amination of
(RS)-MDA, except that the extraction was done with
ethyl acetate instead of dichloromethane. Light brown
crystals were obtained, which were Ðltered o† and
washed Ðrst with propan-2-olÈdiethyl ether (1 : 1) and
then with diethyl ether. The product was isolated as
(RS)-HMMA hydrochloride (m.p. 181È189 ¡C; yield
3
2
2
6.49È6.72 (m, 3 H, aromatic). GC/MS: m/z, M`~
178(26), 135(100), 77(23).
b-Nitroisoeugenol (II). Condensation of vanillin with
nitroethane was carried out by the Knoevenagel con-
densation.17,18 b-Nitroisoeugenol was obtained as
yellow crystals, Ðltered o† and air dried (average
yield[55%). 1H NMR (90 Hz; CDCl ): d 2.48 (s, 3 H,
3
CCH ), 3.95 (s, 3 H, OCH ), 5.91 (bs, 1 H, OH), 6.85È
3
3
7.17 (m, 3 H, aromatic), 8.08 (s, 1 H, CH). GC/MS (TFA
derivative): m/z, M`~ 305(100), 258(52), 192(22), 178(16),
162(32), 161(49), 151(19), 147(35), 139(12), 131(36),
119(14), 115(66), 105(17), 103(34), 102(13), 91(27), 89(18),
79(18), 77(32), 69(68), 65(22), 63(15).
12%). 1H NMR (300.07 Hz; CDCl ÈTFA-d (6 : 1; v/v)):
3
d 1.41 (d, 3 H, J \ 6.6 Hz, CCH ), 2.80 (m, 3 H, NCH ),
3
3
2.88 (dd, 1 H, J \ 14.2 Hz, J \ 7.2 Hz, H-1), 3.00
AB
(dd, 1 H, J \ 14.2 Hz, J \ 7.2 Hz, H-1), 3.53 (m, 1
AB AX
H, CH), 3.88 (s, 3 H, OCH ), 6.70 (dd, 1 H, J \ 8.1 Hz,
BX
Vanillyl methyl ketone (III). b-Nitroisoeugenol was
reduced with iron(III) chloride, iron powder and con-
centrated hydrochloric acid.19 The resulting dark
brown oil was puriÐed by “dry ÑashÏ column chromatog-
raphy.20 The product was eluted from a silica gel
column with tolueneÈdiethyl ether (1 : 2) (yield 68%). 1H
NMR (90 Hz; CDCl ): d 2.12 (s, 3 H, COCH ), 3.64 (s,
3
o
J \ 1.8 Hz, H-6@), 6.73 (d, 1 H, J \ 1.8 Hz, H-2@), 6.90
m
m
(d, 1 H, J \ 8.1 Hz, H-5@), 7.10 (bs, 2 H, NH `), 11.38
o
2
(s, 1 H, OH). 13C NMR (75.5 Hz; CDCl ÈTFA-d (6 : 1;
3
v/v)): d 15.3 (CH ), 30.8 (NCH ), 39.4 (CH ), 55.9
3
3
2
(OCH ), 58.2 (CH), 112.0 (C-5@), 115.2 (C-2@), 122.0 (C-6@)
3
3
3
2 H, CH ), 3.90 (s, 3 H, OCH ), 5.95 (bs, 1 H, OH),
, 126.9 (C-1@), 144.1 (C-4@), 147.1 (C-3@). GC/MS (TFA
2
3
6.66È7.02 (m, 3 H, aromatic). GC/MS (TFA derivative):
m/z, M`~ 276(100), 233(75), 205(16), 137(16), 107(15),
106(10), 105(16), 77(16), 69(32), 65(13).
derivative): m/z, M`~ 387(\1), 260(24), 154(100),
110(24), 69(14).
(RS)-HMA. (RS)-HMA (1-(3-methoxy-4-hydroxyphenyl)
-2-aminopropane) was prepared from II by reduction
with lithium aluminium hydride essentially according to
Ramirez and Burger.17 Instead of the Soxhlet extraction
technique, 7.0 g of II were dissolved in 50 ml of dry
tetrahydrofuran and the solution obtained was added
dropwise to a well stirred mixture of 2.5 g of lithium
aluminium hydride and 200 ml of dry diethyl ether. The
dark red picrate obtained was dissolved in propan-2-ol
and acidiÐed with concentrated hydrochloric acid. A
brown solid crystallized spontaneously and was Ðltered
o† and washed Ðrst with propan-2-olÈdiethyl ether
(RS)-MDA. (RS)-MDA (1-(3,4-methylenedioxyphenyl)-2-
aminopropane) was prepared from I by reductive amin-
ation with sodium cyanoborohydride.21,22 To a solu-
tion of 17.7 g of anhydrous ammonium acetate in 60 ml
of methanol were added 4.2 g of I followed by 1.7 g of
sodium cyanoborohydride. The solution was stirred at
ambient temperature for 72 h and kept at pH 6È7 by
gradual addition of concentrated hydrochloric acid.
After adding 100 ml of demineralized water, the pH was
brought to 2 with concentrated hydrochloric acid.
Methanol was removed in-vacuo. The aqueous solution
( 1997 John Wiley & Sons, Ltd.
JOURNAL OF MASS SPECTROMETRY, VOL. 32, 1236È1246 (1997)

METHYLENEDIOXYMETHAMPHETAMINE AND CHIRAL METABOLITES
1239
(1 : 1) and then with diethyl ether. The product was iso-
MDMA and non- and deconjugated metabolites from
lated as (RS)-HMA hydrochloride (m.p. 244È246 ¡C;
urine samples were carried out according to Lim et al.2
yield 9%). 1H NMR (300.07 Hz; CDCl ÈTFA-d (3 : 1;
Instead of b-glucuronidase,
a
mixture of b-
3
v/v)): d 1.46 (d, 3 H, J \ 6.6 Hz, CCH ), 2.89 (dd, 1 H,
glucuronidase and arylsulfatase from Helix pomatia was
used for enzymatic hydrolysis. (RS)-Propylhexedrine
was used as an internal standard. The extraction was
carried out with a mixture of dichloromethane and
propanol (3 : 1, v/v). The organic solvent was removed
in a gentle stream of nitrogen at 40 ¡C.
3
J
\ 14.5 Hz, J \ 8.4 Hz, H-1), 3.00 (dd, 1 H, J
\
AB
BX
AB
14.5 Hz, J \ 6.2 Hz, H-1), 3.73 (m, 1 H, CH), 3.90 (s, 3
AX
H, OCH ), 6.73 (dd, 1 H, J \ 7.9 Hz, J \ 1.8 Hz,
3
o
m
H-6@), 6.78 (d, 1 H, J \ 1.8 Hz, H-2@), 6.92 (d, 1 H, J \
m
o
7.9 Hz, H-5@), 11.23 (s, 3 H, NH `), 11.23 (s, 1 H, OH).
3
13C NMR (75.5 Hz; CDCl ÈTFA-d (3 : 1, v/v)): d 18.0
3
(CH ), 40.3 (CH ), 51.2 (CH), 56.0 (OCH ), 112.2 (C-5@),
3
2
3
N-(S)-HFBP-O-TFA and -O-TMS derivatization
115.4 (C-2@), 122.2 (C-6@), 127.5 (C-1@), 143.8 (C-4@), 147.3
(C-3@). GC/MS (TFA derivative): m/z, M`~ 373(3),
260(55), 233(12), 140(100), 69(21).
The N-(S)-HFBP-O-TFA derivatization was carried out
according to Lim et al.2 with the following modiÐ-
cations. To the dry residue after extraction, 0.5 ml of
borate bu†er (pH 9.0) was added. The solution was
vortex mixed for 20 s and cooled in an ice-bath for 10
min. A 40 ll volume of (S)-HFBPCl reagent was added
and the solution was again vortex mixed for 20 s and
placed in an ice-bath for 20 min. A 200 ll volume of
toluene was added, the solution was vortex mixed for 20
s and the phases were separated by centrifugation. The
samples were stored at [20 ¡C for 30 min and the
organic layer was decanted from the frozen aqueous
phase into a clean tube. The solvent was removed in a
gentle stream of nitrogen at 60 ¡C and the residues were
dried under reduced pressure over phosphorous pentox-
ide and potassium hydroxide for at least 1 h. To the dry
residue 50 ll of MSTFA were added. The solution was
vortex mixed for 20 s and kept at 80 ¡C for 40 min.
Excess reagent was removed in a gentle stream of nitro-
gen at 60 ¡C and the residue was dissolved in 200 ll of
toluene. A 1 ll aliquot was analyzed by GC/MS.
(RS)-MDAOH. (RS)-MDAOH (1-(3,4-methylenedioxy-
phenyl)-2-hydroxylaminopropane)
was
prepared
from
I
by reductive amination with sodium
cyanoborohydride.21,22 To a solution of 14.8 g of
hydroxylamine hydrochloride in 120 ml of methanol
were added 3.6 g of I followed by 1.0 g of sodium
cyanoborohydride. The solution was treated as in the
reductive amination of (RS)-MDA. A white solid crys-
tallized spontaneously and was Ðltered o† and washed
Ðrst with propan-2-olÈdiethyl ether (1 : 1) and then with
diethyl ether. The product was isolated as (RS)-
MDAOH hydrochloride (m.p. 148È149 ¡C, lit.21 149È
150 ¡C; yield 17%). 1H NMR (300.07 Hz;
CDCl ÈDMSO-d (500 : 1, v/v)): d 0.89 (d, 3 H, J \ 6.0
3
6
Hz, CCH ), 2.25 (dd, 1 H, J \ 13.0 Hz, J \ 11 Hz,
AB BX
3
H-1), 2.96 (dd, 1 H, J \ 13.0 Hz, J \ 1.6 Hz, H-1),
AB
AX
3.15 (m, 1 H, CHN), 5.58 (s, 2 H, OCH O), 6.08 (dd, 1
2
H, J \ 7.8 Hz, J \ 1.8 Hz, H-6@), 6.35 (d, 1 H, J \
o
m
m
1.8 Hz, H-2@), 6.34 (d, 1 H, J \ 7.8 Hz, H-5@), 11.12 (s, 2
o
H, NH `), 10.2 (bs, 1 H, NOH). 13C NMR (75.5 Hz;
2
CDCl ): d 12.8 (CH ), 35.5 (CH ), 57.9 (CH), 100.0
3
3
2
GC/MS analysis of enantiomers
(CH O), 107.4 (C-2@), 108.6 (C-5@), 121.5 (C-6@), 128.9 (C-
2
1@), 145.5 (C-4@), 146.8 (C-3@). GC/MS (TFA derivative):
m/z, M`~ 291(6), 162(31), 135(100), 105(5), 77(15), 69(11).
The GC/MS analyses were carried out with an Saturn
II ion trap (Varian). The GC was provided with a
septum-equipped programmable injector and
a
Synthesis of (S)-HFBPCl
DB-5MS fused-silica capillary column (15 m ] 0.25 mm
i.d., Ðlm thickness 0.25 lm) (J&W ScientiÐc, Folsom,
CA, USA). The injector temperature program was as
follows: the initial temperature (110 ¡C) was maintained
for 1.5 min, then the temperature ramped to 280 ¡C at
100 ¡C min~1 and maintained there for 1 min. The oven
temperature program was as follows: the initial tem-
perature was maintained at 105 ¡C for 7.5 min, raised to
200 ¡C at 10 ¡C min~1, maintained there for 10 min and
then raised to 280 ¡C at 20 ¡C min~1. The GC/MS
analysis was performed either in the EI mode or in the
positive chemical ionization (PCI) mode with isobutane
N-HeptaÑuorobutyryl-(S)-prolyl chloride ((S)-HFBPCl)
was synthesized according to Lim et al.23 The optical
purity of (S)-HFBPCl was checked by GC/MS analysis
of the derivatives of (S)-amphetamine, (RS)-meth-
amphetamine and (RS)-MDMA. The reagent (S)-
HFBPCl was stored under nitrogen at [20 ¡C as a 0.02
M solution in toluene and was stable for at least 3
months.
Urine sample collection
(UCAR
Union
Carbide,
Nieuw-Vennep,
The
Netherlands) as the reactant gas. All parameters were
set according to the manufacturerÏs manual. Ratios of
enantiomers were calculated as the peak height ratios of
the protonated molecules in the PCI mode.
Urine samples were collected from intoxicated persons
and delivered to the Department of Human Toxicology
for analysis.
Isolation of MDMA and metabolites from urine
Statistical analysis
Apart from three modiÐcations, the enzymatic hydro-
lysis of the conjugated metabolites and the extraction of
Statistical signiÐcance was assessed according to a
StudentÏs t-test.
( 1997 John Wiley & Sons, Ltd.
JOURNAL OF MASS SPECTROMETRY VOL. 32, 1236È1246 (1997)

1240
D. DE BOER ET AL .
RESULTS AND DISCUSSION
Selection of chiral derivatization
Because N-(S)-HFBP derivatives of compounds in
general14 and of MDMA in particular13 are known to
result in better chromatographic resolution and shorter
retention times than N-(S)-TFAP and N-(S)-PFPP
derivatives, it was decided to choose the N-(S)-HFBP
derivatization for enantiomeric separation. An addi-
tional advantage is that the derivatization with (S)-
HFBPCl can be carried out in aqueous solution. In our
study, MSTFA was used in the second step of the chiral
derivatization method necessary for the analysis of the
metabolites HMMA and HMA, resulting in O-
trimethylsilyl (O-TMS) derivatives (Fig. 2). Lim et al.2
used MBTFA for the second step in order to produce
O-TFA derivatives. However, in our hands the N-(S)-
HFBP-O-TMS derivatization method proved to be
more reproducible and to yield higher signals than the
N-(S)-HFBP-O-TFA method (data not shown). The
N-(S)-HFBP-O-TMS derivatization method was there-
fore chosen for the chiral GC/MS assay in this study.
Probably N-(S)-HFBP-O-TFA derivatives are more
useful for the analysis in the negative ionization mode,
as reported by Lim et al.2 In our study, only positive
ions were analyzed.
Assessment of the optical purity of the (S)-HFBPCl
reagent by GC/MS analysis of the (S)-amphetamine and
(RS)-methamphetamine derivatives showed that
racemization may occur after storing the reagent for
more than 3 months. Racemization results in undesired
(N-(RS)-HFBP derivatives. Similar problems were
reported previously.23 Storage of small amounts of (S)-
HFBPCl reagent at a relatively low stock concentration
under nitrogen, and therefore minimizing exposure to
air during handling, extended the storage life.
Scheme 1.
fragmentation mechanisms of some characteristic EI
fragments are given.
The EI spectra are dominated by the fragment at m/z
266, which is characteristic for N-HFBP derivatives, but
not for the compound concerned (Scheme 2). Also
typical for the N-HFBP derivative is the fragment at
m/z 294 (Scheme 2). In the EI spectra most of the m/z
values of the other fragment ions are smaller and their
abundances lower than those at m/z 266 and although
some of them are characteristic (fragments a , a , b and
1
2
R NHxCHCH `), others are not (fragments TMS`
1
3
and CF `). The characteristic fragments a , a and b
3
1
2
result from cleavage of the C ÈN bond (Scheme 1) and
a
the C ÈC bond (Scheme 2), respectively. Except for
b
a
MDA and HMA, the abundances of M`~ or com-
binations of M`~ with the loss an F~ radical are not
signiÐcant. SigniÐcant loss of a CH ~ radical was
3
Selection of ionization mode
In Table 1 the mass spectrometric data obtained with
the EI mode are shown and in Schemes 1 and 2 the
Figure 2. Formation of the N-heptafluorobutyryl-(S)-prolyl-O-
trimethylsilyl derivatives of the (R)-enantiomers of the compounds
studied. Asterisks indicate
a chiral center. (S)-HFBPCl ¼N-
heptafluorobutyryl-(S)-prolyl chloride; MSTFA ¼N-methyl-N-tri-
methylsilyltrifluoroacetamide.
Scheme 2.
( 1997 John Wiley & Sons, Ltd.
JOURNAL OF MASS SPECTROMETRY, VOL. 32, 1236È1246 (1997)

Table 1. Partial mass spectrometric dataa for N-HFBP-O-TMS derivatives of reference standards of MDMA and its chiral metabolites using the electron ionization mode
m/z values of characteristic fragments (normalized on the base peak)d
Compoundb
Enantiomer No.c
M
M½~
ÍM ÉCH ˽
3
ÍM ÉF˽
C F CONC H CO½
7
C F CONC H ½
4
a
a
b
TMS½
CF ½
3
R NHxCHCH ½
1 3
r
3
7
4
3
7
7
1
2
MDMA
1
2
1
2
1
2
1
2
1
2
486
472
560
546
560
486 (Ä1)
n.o.
n.o.
n.o.
467 (1)
467 (Ä1)
453 (1)
453 (Ä1)
n.o.
294 (Ä1)
294 (Ä1)
294 (8)
294 (8)
294 (3)
294 (3)
294 (5)
294 (6)
294 (1)
294 (Ä1)
266 (100)
266 (100)
266 (100)
266 (100)
266 (100)
266 (93)
266 (81)
266 (87)
266 (100)
266 (100)
163 (12)
163 (15)
163 (12)
163 (12)
237 (23)
237 (24)
237 (22)
237 (23)
163 (17)
163 (19)
162 (56)
162 (71)
162 (75)
162 (71)
236 (90)
236 (100)
236 (100)
236 (100)
162 (37)
162 (67)
135 (14)
135 (14)
135 (15)
135 (15)
209 (14)
209 (14)
209 (7)
n.o.
n.o.
69 (14)
69 (14)
69 (12)
69 (11)
69 (10)
69 (10)
69 (8)
58 (36)
58 (31)
n.m.
MDA
472 (5)
472 (2)
560 (1)
n.o.
n.o.
n.o.
n.o.
n.o.
n.m.
HMMA
HMA
n.o.
73 (15)
73 (14)
73 (17)
73 (18)
73 (8)
73 (2)
58 (37)
58 (31)
n.m.
545 (Ä1)
531 (Ä1)
531 (1)
545 (8)
545 (6)
n.o.
546 (14)
546 (13)
n.o.
n.o.
n.o.
209 (7)
69 (9)
n.m.
MDAOH
n.o.
135 (22)
135 (26)
69 (8)
n.o.
n.o.
n.o.
69 (9)
n.o.
a Peaks were considered to be of significant importance if the signal-to-noise ratio was ¿3.
bN-HFBP-O-TMS ¼N-heptafluorobutyryl-(S)-prolyl-O-trimethylsilyl; MDMA ¼3,4-methylenedioxymethamphetamine; MDA ¼3,4-methylenedioxyamphetamine; HMMA ¼4-hydroxy-3-
methoxymethamphetamine; HMA ¼4-hydroxy-3-methoxyamphetamine.
c Sequence based on consecutive retention times.
d n.o. ¼Not observed; n.m. ¼not measured, because outside the selected mass range.

1242
D. DE BOER ET AL .
observed especially for the N-O-TMS derivative of
MDAOH. Also for the O-TMS derivatives of one enan-
tiomer of HMMA and both enantiomers of HMA, the
the helium bu†er gas has an adverse e†ect and may
compete with the overall increased fragmentation. As
discussed by Boswell et al.,26 di†erent ionization mecha-
nisms operate under PCI conditions in the ion trap,
including proton transfer giving rise to protonated mol-
ecules and charge transfer resulting in EI-like molecular
ions. With respect to the latter process, transfer of elec-
trons from the sample molecule to reactant gas ions
takes place, resulting in odd-electron molecular ions,
which may fragment further by radical-induced frag-
mentation. The PCI mode has the advantages that it
provides high-mass information and that some fragmen-
tation remains, allowing elucidation of structural
details, making it the ionization mode of choice for the
purposes of the present study.
[M [ CH ]` fragment was observed, although at very
3
low abundance. The CH ~ radical probably originates
3
from the TMS moiety; this fragmentation is very
general, occurring during EI of TMS ethers and
esters.24
In Table 2, mass spectrometric data are presented for
the PCI mode. Although in all spectra the MH` ions
were found to be the base peak, the characteristic EI
fragments a and a were also prominent. The main dif-
1
2
ference between the spectra of the two ionization modes
is that, compared with fragment a , the presence of
2
fragment a is more pronounced in the PCI mode than
1
it is in the EI mode. Such subtle di†erences in the rela-
tive abundances of ions which di†er by only one atomic
mass unit and are formed by di†erent mechanisms have,
for example, also been observed between the ions at m/z
70 and 71 in the EI and PCI mass spectra of N,N@-
dimethylpiperazine.25
Enantiomeric ratios of MDMA and metabolites in urine
The selected ion chromatograms of the GC/MS analysis
of the urine sample obtained from an MDMA intoxi-
cated person are shown in Fig. 3. Peak identiÐcation of
MDMA, MDA, MDAOH, HMMA and HMA was
based on both the mass spectrum and the relative reten-
tion time. In all three samples MDMA and MDA were
identiÐed. In one sample, both HMMA and HMA were
found, in the second one only a trace of HMMA and in
the third no additional metabolites at all. The presence
of these metabolites in human urine is in agreement
with earlier reports.10h12 In two of the samples studied,
also the presence of the enantiomers of N-hydroxy-(RS)-
3,4-methylenedioxyamphetamine (MDAOH; Fig. 1) was
observed at retention times of 22.43 and 24.03 min,
respectively (mass spectra, Fig. 4(A) and (B)).
Characteristic for the PCI mode is the formation of
fragment c and of the fragments d Èd (Table 2). Espe-
1
8
cially the d fragments are of interest because they form
characteristic ion clusters, easily recognized in the
spectra. The formation of the series of d fragments prob-
ably includes a cleavage of the prolyl ring followed by a
rearrangement. Selective deuterium and oxygen labeling
combined with tandem mass spectrometry and/or high-
resolution mass spectrometry should give a more
detailed understanding of the fragmentation and mecha-
nism.
Increased fragmentation in the ion trap when using
the PCI mode has been reported by Boswell et al.26 and
may be caused by the relatively long residence time of
the ions, by which the chances of collisions are
increased. However, collisional stabilization of ions by
Based on the peak heights of the protonated mol-
ecule, the absolute or “analyticalÏ enantiomeric ratio of
the reference standards was found to be [1 for MDMA
Figure 3. Selected ion chromatograms of the protonated molecules (see Table 2) of the N-(S)-HFBP-O-TMS derivatives of MDMA, MDA,
MDAOH, HMMA and HMA, as observed with positive chemical ionization GC/MS after the isolation of the compounds from urine of an
intoxicated person, and derivatization with N-heptafluorobutyryl-(S)-prolyl chloride and N-methyl-N-trimethylsilyltrifluoroacetamide.
MDMA ¼3,4-methylenedioxymethamphetamine; MDA ¼3,4-methylenedioxyamphetamine; MDAOH ¼N-hydroxy-3,4-methylenedioxy-
amphetamine; HMMA ¼4-hydroxy-3-methoxymethamphetamine; HMA ¼4-hydroxy-3-methoxyamphetamine.
( 1997 John Wiley & Sons, Ltd.
JOURNAL OF MASS SPECTROMETRY, VOL. 32, 1236È1246 (1997)

Table 2. Partial mass spectrometric dataa for N-HFBP-O-TMS derivatives of MDMA and its chiral metabolites using the positive chemical ionization
mode
m/z values of characteristic fragments (normalized on the base peak)d
Compoundb
Enantiomer No.c
M
MH½
ÍM ÉCH ˽
3
c
C F CONC H ½
7
d
/d
d
/d
d
d
/d
d
a
a
r
3
7
4
1
2
3
4
5
6
7
8
1
2
MDMA
1
2
1
2
1
2
1
2
1
2
486
472
560
546
560
487 (100)
487 (100)
473 (100)
473 (100)
561 (100)
561 (100)
547 (100)
547 (100)
561 (100)
561 (100)
n.o.
n.o.
365 (2)
365 (2)
351 (2)
351 (2)
365 (7)
365 (9)
351 (1)
351 (1)
439 (4)
439 (3)
266 (2)
266 (2)
266 (8)
266 (4)
234 (Ä1)
234 (Ä1)
220 (4)
220 (3)
n.o.
232 (3)
232 (3)
218 (4)
218 (3)
306 (3)
306 (3)
292 (1)
292 (1)
n.o.
205 (8)
205 (8)
205 (11)
205 (8)
279 (7)
279 (5)
279 (2)
279 (2)
205 (8)
205 (9)
203 (8)
203 (9)
203 (7)
203 (6)
277 (22)
277 (26)
277 (4)
277 (4)
203 (11)
203 (16)
201 (15)
201 (7)
163 (64)
163 (78)
163 (24)
163 (20)
237 (56)
237 (69)
237 (5)
162 (24)
162 (28)
162 (26)
162 (24)
236 (38)
236 (43)
236 (8)
MDA
n.o.
201 (15)
201 (11)
275 (29)
275 (31)
275 (8)
n.o.
HMMA
HMA
n.o.
266 (Ä1)
266 (Ä1)
266 (1)
n.o.
n.o.
n.o.
294 (4)
294 (4)
n.o.
n.o.
266 (1)
275 (8)
237 (6)
236 (9)
MDAOH
545 (18)
545 (12)
266 (15)
266 (7)
201 (11)
201 (13)
163 (20)
163 (26)
162 (24)
162 (27)
n.o.
n.o.
ahd See Table 1.

1244
D. DE BOER ET AL .
Figure 4. Positive chemical ionization mass spectra of the derivatives of two enantiomers of N-hydroxy-3,4-methylenedioxyamphetamine
at the retention times of (A) 22.43 and (B) 24.03 min (see also Fig. 3).
and MDAOH and \1 for the other reference stan-
dards. The observation that an “analyticalÏ enantiomeric
ratio is not equal to 1 is not uncommon; it can be
caused by a di†erence in instrumental response and/or
derivatization efficiency of the individual enantiomers.23
The “metabolicÏ enantiomeric ratios, normalized to the
“analyticalÏ ratio, of all N-demethylated metabolites
were signiÐcantly di†erent from the “analyticalÏ ratios
(Table 3). The “metabolicÏ enantiomeric ratios of
MDMA and HMMA were [0.5 and those of the N-
demethylated metabolites MDA, MDAOH and HMA
were \0.5. These preliminary results of the analysis of
urine samples suggest an enantioselective metabolism
for MDMA in humans. This would be in agreement
with studies in mice and in rats,2 where the enantio-
meric proÐles of the N-demethylated metabolites MDA
and HMA (R/S ratio \1) are the reverse of those of
MDMA and HMMA (R/S ratio [1) in 24 h pooled
urine. In our study, we used the racemic mixtures as
reference material. Therefore, we could not assign the
(R)- or (S)-conformation to the individual enantiomers
of the compounds.
N-Hydroxylated chiral metabolites
N-Hydroxylated metabolites of amphetamine-like com-
pounds in human urine have been reported for phenter-
mine and chlorphentermine27 and were suggested to be
intermediates in the metabolic route leading to hippuric
acid conjugates (Fig. 5). However, the presence of
MDAOH in the urine samples studied was a surprise
as, even if formed, N-hydroxylated compounds in
general28 and MDAOH itself29 are known to be rela-
tively unstable. In our study MDAOH was only
observed after enzymatic hydrolysis, which means that
it was present in urine as a conjugate with either gluc-
uronic acid or sulfate. MDAOH may therefore be more
stable in a conjugated form than when unconjugated.
Because MDAOH itself is sometimes the main phar-
macologically active compound in illicitly produced
“EcstasyÏ tablets, it cannot be excluded that the presence
of MDAOH is caused by the combined use of MDMA
and MDAOH tablets, especially since up to now meta-
bolic studies of 3,4-methylenedioxyamphetamine-like
Table 3. Absolute and normalized ratios of the enantiomers of MDMA and its chiral metabo-
litesa
Mean of reference
standards (n ¼5)
Absolute
Sample 1
Sample 2c
Sample 3c
Compoundb
Absolute
Normalized
Absolute
Normalized
Absolute
Normalized
MDMA
MDA
1.12 À0.09
0.61 À0.03
0.85 À0.10
0.67 À0.07
1.47 À0.13
1.21
1.08
0.38
0.87
0.43
0.27
1.19
0.33d
n.q.
1.06
0.54
—
1.27f
0.28d
n.q.
1.14
0.46
—
0.23d
0.74e
0.29d
0.40d
HMMA
HMA
n.o.
—
n.o.
—
MDAOH
n.o.
—
0.45d
0.31
a First peak/second peak height ratios were normalized on the absolute ratio of the respective refer-
ence standards, which were considered to contain equal amounts of both enantiomers.
bN-HFBP-O-TMS ¼N-heptafluorobutyryl-(S)-prolyl-O-trimethylsilyl; MDMA ¼3,4-methylene-
dioxymethamphetamine; MDA ¼3,4-methylenedioxyamphetamine; HMMA ¼4-hydroxy-3-methoxy-
methamphetamine; HMA ¼4-hydroxy-3-methoxyamphetamine; MDAOH ¼N-hydroxy-3,4-methy-
lenedioxyamphetamine.
c n.o. ¼Not observed; n.q. ¼present, but impossible to quantify correctly.
dp Ä0.005.
ep Ä0.1.
fp Ä0.05.
( 1997 John Wiley & Sons, Ltd.
JOURNAL OF MASS SPECTROMETRY, VOL. 32, 1236È1246 (1997)

METHYLENEDIOXYMETHAMPHETAMINE AND CHIRAL METABOLITES
1245
Figure 5. Assumed metabolism of 3,4-methylenedioxymethamphetamine in humans.12
compounds did not report the presence of MDAOH in
urine.10h12,30 However, the low frequency of the pres-
ence of MDAOH in “EcstasyÏ tablets8 and the fact that
MDAOH was found in two out of three urine samples
from totally di†erent and unrelated intoxications
suggest that the observed MDAOH is a metabolite
rather than a co-administered compound. The fact that
the“metabolicÏ enantiomeric ratios of MDAOH were in
the same range as those of MDA (Table 3) also suggests
that MDA is the precursor (Fig. 5) of MDAOH in these
two cases. The signiÐcance of this observation is that
MDAOH is converted in the body into 3,4-methy-
lenedioxyphenylacetone (Fig. 5), probably via the inter-
the developed methodology for quantitative purposes as
for typical pharmacokinetic experiments and more
extensive toxicological studies.
The presence of the minor chiral metabolites 3,4-
dihydroxymethamphetamine
and
3,4-dihydroxy-
amphetamine11,12 was not observed. However,
because the compounds were not available as reference
compounds in our laboratory, the isolation procedure
used in this study could not be validated for these com-
pounds. In principle, our GC/MS assay and the deriva-
tization method should also be applicable to detect the
dihydroxy metabolites.
mediate
3,4-methylenedioxy-2-nitroso-1-phenyl-
propane.31 Since nitroso compounds are known to be
relatively toxic, a sudden accumulation of 3,4-methy-
lenedioxy-2-nitroso-1-phenylpropane, for several pos-
sible metabolic reasons, could contribute to acute
MDMA intoxications.
CONCLUSIONS
A chiral assay for MDMA and its major metabolites
has been selected and developed. The procedure was
based on the N-(S)-HFBP-O-TMS derivatization
method and GC/MS analysis in the PCI mode. Prelimi-
nary results of the analysis of urine samples showed the
usefulness of the assay and indicated an enantioselective
metabolism in the N-demethylation pathway for
MDMA in humans.
Unambiguous identiÐcation or exclusion of MDAOH
as a metabolite of MDMA in biological samples from
MDMA excretion experiments with human subjects
and from intoxicated persons is therefore of special
interest. It would not only be important for the toxico-
logical aspects of MDMA and other 3,4-
methylenedioxyamphetamine-like compounds, but also
for their pharmacodynamics, since MDAOH is known
to be pharmacologically active.21 This would mean that
the overall pharmacological e†ect of MDMA is deter-
mined by three compounds, since the parent compound
and the metabolite MDA are also pharmacologically
active. Further research will be performed in order to Ðt
Acknowledgements
We thank Dr J. Boersma and Mr J. Schuring of the Chemistry
Department for their assistance in the synthesis of the reference com-
pounds.
REFERENCES
1. E. J. Ariens, Trends Pharmacol. Sci. 14, 68 (1993).
2. H. K. Lim, Z. Su and R. L. Foltz, Biol. Mass Spectrom. 22, 403
(1993).
3. K. M. Hegadoren, G. B. Baker and R. T. Couts, Res. Commun.
Substances Abuse 14, 67 (1993).
4. M. Johnson, A. A. Letter, K. Merchant, G. R. Hanson and J.
W. Gibb, J. Pharmacol. Exp. Ther. 244, 977 (1988).
5. A. R. Green and G. M. Goodwin, Br. Med. J. 312, 1493
(1996).
6. J. A. Henry, K. J. Jeffreys and S. Dawling, Lancet 340, 384
(1992).
9. H. K. Lim and R. L. Foltz, Chem. Res. Toxicol. 1, 370 (1988).
10. H. K. Lim and R. L. Foltz, Chem. Res. Toxicol. 2, 142 (1989).
11. H.-J. Helmlin, K. Bracher, D. Bourquin, D. Vonlanthen and R.
Brenneisen, J. Anal. Toxicol. 20, 432 (1996).
12. H. H. Maurer, Ther. Drug Monit. 18, 465 (1996).
13. D. de Boer, R. M. A. van de Wal, K. L. Bos, R. A. A. Maes, in
Proceedings of the 12th Cologne Workshop on Dope
Analysis, 10th to 15th April 1994, Recent Advances in Doping
Analysis (2), edited by M. Donike et al., pp. 329–338. Sport
und Buch Strauss, Cologne (1995).
14. S. Caccia, C. Chiabrando, P. de Ponte and R. Fanelli, J. Chro-
matogr. Sci. 16, 543 (1978).
15. H. I. Waterman and R. Priester, Recl. Trav. Chim. Pays-Bas 47,
1027 (1928).
16. T. Fujisawa and Y. Deguchi, J. Pharm. Soc. Jpn. 74, 975
(1954).
17. F. Ramirez and A. Burger, J. Am. Chem. Soc. 72, 2781
(1950).
18. E. Knoevenagel and L. Walter, Chem. Ber. 37, 4502 (1904).
7. R. A. de Man, J. H. P. Wilson and S. L. M. Tjen, Ned. Tijdschr.
Geneeskd. 137, 727 (1993).
8. E. Fromberg, F. Janssen, A. de Loor and H. Matser, in De
XTC-Markt Juli 1993–Juli 1994. Netherlands Institute for
Alcohol and Drugs, Utrecht (1994); M. W. Van Laar, M. van
Ooyen-Houben and I. P. Spruit, in Fact Sheet. Hard Drug
Policy XTC. Netherlands Institute for Alcohol and Drugs,
Utrecht (1996).
( 1997 John Wiley & Sons, Ltd.
JOURNAL OF MASS SPECTROMETRY VOL. 32, 1236È1246 (1997)

1246
D. DE BOER ET AL .
19. I. A. Pearl and D. L. Beyer, J. Org. Chem. 16, 221 (1951).
20. L. M. Harwood and C. J. Moody, in Experimental Organic
Chemistry: Principles and Practice, p. 185. Blackwell, Oxford
(1994).
26. S. M. Boswell, R. E. Mather and J. F. J. Todd, Int. J. Mass
Spectrom. Ion Processes 99, 139 (1990).
27. A. H. Beckett and P. M. Belanger, J. Pharm. Pharmacol. 26,
205 (1974).
21. U. Braun, A. T. Shulgin and G. Braun, J. Pharm. Sci. 16, 192
(1951).
28. A. H. Beckett and S. Al-Sarray, J. Pharm. Pharmacol. 24, 174
(1972).
22. R. F. Borch, M. D. Bernstein and H. D. Durst, J. Am. Chem.
Soc. 93, 2897 (1971).
29. A. K. Valaer, W. R. Ravis, C. R. Clark, J. Chromatogr. Sci. 28,
482 (1990).
23. H. K. Lim, J. W. Hubbard and K. K. Midha, J. Chromatogr.
378, 109 (1986).
24. J. Diekman and C. Djerassi, J. Org. Chem. 32, 1005 (1967).
25. T. A. Whitney, L. P. Kleman and F. H. Field, Anal. Chem. 43,
1048 (1971).
30. H. K. Ensslin, H. H. Maurer, E. Gouzoulis, L. Hermie, K. A.
Kovar, Drug Metab. Dispos. 24, 813 (1996).
31. B. Lindeke and E. Anderson, Acta Pharm. Suec. 12. 183
(1975).
( 1997 John Wiley & Sons, Ltd.
JOURNAL OF MASS SPECTROMETRY, VOL. 32, 1236È1246 (1997)