Determination of the synthetic origin of methamphetamine samples by 2H NMR spectroscopy

Article abstract of DOI:10.1021/ac052105w

Samples of methamphetamine, prepared according to the most common synthetic pathways, were submitted to natural-abundance ²H NMR spectroscopy. The deuterium content at the various sites of the molecule was found to depend on its synthetic history. The technique provides a chemical fingerprint of methamphetamine samples and can give hints to trace back the starting materials and the synthetic procedures.

Full text of DOI:10.1021/ac052105w

Anal. Chem. 2006, 78, 3113-3117
Determination of the Synthetic Origin of
2
Methamphetamine Samples by H NMR
Spectroscopy
Silvia Armellin,† Elisabetta Brenna,*,† Samuele Frigoli,† Giovanni Fronza,*,‡ Claudio Fuganti,† and
Daniele Mussida†
Dipartimento di Chimica, Materiali, Ingegneria Chimica, Politecnico di Milano, and
Istituto CNR per la Chimica del Riconoscimento Molecolare, Via Mancinelli 7, I-20131 Milano, Italy
Chart 1
Samples of methamphetamine, prepared according to the
most common synthetic pathways, were submitted to
natural-abundance 2H NMR spectroscopy. The deuterium
content at the various sites of the molecule was found to
depend on its synthetic history. The technique provides
a chemical fingerprint of methamphetamine samples and
can give hints to trace back the starting materials and the
synthetic procedures.
Methamphetamine (1) is a synthetic stimulant drug (Chart 1)
that induces a strong feeling of euphoria and generates addiction.
It was developed in the 1930s from its parent drug amphetamine
(2) and was originally used in nasal decongestants, bronchial in-
halers, and the treatment of narcolepsy and obesity. In the 1970s,
it became a Schedule II drugsa drug with a high abuse risk, but
also accepted medical uses in the United States. Nowadays, it is
illegally produced in clandestine laboratories and sold in pill form,
capsules, powder, and chunks.
Methamphetamine was first obtained in the work1 of structural
elucidation of ephedrine (3), the main active component isolated
from Ephedra vulgaris. Structure 1 was tentatively assigned to
the product obtained by reaction of ephedrine with red phospho-
rus and iodine. Compound 1 was synthesized by reduction of the
condensation product of phenyl acetone (4) and methylamine, to
verify the structural assignment. Methamphetamine is a chiral
compound: (S)-1 is a stimulant, while the (R)-enantiomer is a
decongestant with no stimulant activity.
The synthetic methods, which are most widely employed in
clandestine laboratories, are still making use of ephedrine or of
phenyl acetone as starting materials. Also pseudoephedrine (5),
which differs from ephedrine in the configuration of the benzylic
stereocenter, can be used to prepare illegal stimulant metham-
phetamine.
Commercial ephedrine is produced in substantial amounts for
medical purposes by extraction from Ephedra plants, by chemical
synthesis starting from propiophenone (6) or by fermentation
from benzaldehyde.
Phenyl acetone (4) is prepared by reaction of phenylacetic acid
(7) with acetic anhydride. Pseudoephedrine is the active component
of Sudafed, NyQuil, and several other over-the-counter deconges-
tants. It is obtained together with ephedrine in the extraction from
Ephedra or by treatment of ephedrine with boiling hydrochloric acid.
Ephedrine, pseudoephedrine, phenylacetic acid, acetic anhy-
dride, and phenyl acetone are all substances included in the Red
List,2 i.e., the list of precursors and chemicals frequently used in
the illicit manufacture of narcotic drugs and psychotropic sub-
stances under international control. This list has been prepared
by the International Narcotics Control Board (INCB) as a tool to
be used for the identification of substances scheduled in Tables
1 and 2 of the United Nations Convention against Illicit Traffic in
Narcotic Drugs and Psychotropic Substances (1988 Convention).
The export-import of the substances belonging to the Red List
is strictly controlled and is completely traceable (name and
address of the exporter and importer and, when available, the
consignee; quantity of the substance to be exported; expected
point of entry and expected date of dispatch).
In the fight against mom-and-pop meth labs, it thus becomes
extremely important to trace the synthetic path of seized meth-
amphetamine samples, to establish whether products from dif-
ferent seizures are from a common origin, and to identify the
reagents involved in the synthetic procedure. These are important
clues for use in the criminal investigation.
The analysis of isotope ratios has been found to be a powerful
tool to trace the origin of organic compounds. The distribution of
the stable isotopes within the atomic species of an organic mole-
cule is not random, but depends on the way the molecule was
obtained.
* To whom the correspondence shoul be addressed. E-mail: elisabetta.brenna@
polimi.it. Phone ++39 02 23993077. Fax ++39 02 23993080. E-mail:
giovanni.fronza@polimi.it. Phone ++39 02 23993027. Fax ++39 02 23993080.
† Politecnico di Milano.
‡ Istituto CNR per la Chimica del Riconoscimento Molecolare.
(1) Ogata, A. Yakugaku Zasshi 1919, 451, 751-764.
(2) http://www.incb.org/incb/red_list.html.
10.1021/ac052105w CCC: $33.50 © 2006 American Chemical Society
Published on Web 03/24/2006
Analytical Chemistry, Vol. 78, No. 9, May 1, 2006 3113

As for synthetic products, such as drugs, their isotopic compos-
ition is dependent on both the isotopic composition of the reactants
and on the synthetic isotopic fractionation of the manufacturing
process employed. A change in either of these variables produces
a drug product having a different isotopic profile. Recent work
has shown that when both the source of the starting materials
and the manufacturing process are presumably held relatively
constant during manufacture of the bulk drug substance, similar
product isotope ratios are observed.3 Two different techniques can
be employed to determine isotope ratios: (i) isotope ratio mass
spectrometry (IRMS) and (ii) Site-specific natural isotope frac-
tionation measured by NMR spectroscopy (SNIF NMR). The latter
is particularly advantageous because it allows the simultaneous
determination of the deuterium content at each site of a molecule,
expressed as D/H (ppm). This technique allows the D/H of each
hydrogen atom to be directly correlated to the reagents employed
to insert it. The information is very punctual: by combining the
knowledge of the mechanisms of organic reactions with the
determination of the deuterium enrichment or depletion of a
certain position, it is possible to gain clues to trace the origin.
Isotope ratio analysis has been useful in the recognition of the
geographical origin of natural extractive drugs such as heroin and
cocain (SNIF NMR)4 and in the linkage of different seizures of
ecstasy tablets to a possible common origin (IRMS).5-8 In the
past,9 we submitted nine samples of N-acetyl-3,4-methylene-
dioxyamphetamine, prepared according to the most common
Scheme 1. Methamphetamine Samples from
Phenyl Acetone
Scheme 2. Methamphetamine from
Benzaldehyde
2
synthetic procedures, to H NMR spectroscopy. We could show
that the relative deuterium content at the various sites of the
molecule was dependent on its synthetic history.
In a recent work published in this journal10 carbon and nitrogen
stable isotope analysis obtained by IRMS was used to determine
the origin of ephedrine. The values of δ13C and δ15N were
employed to discriminate natural, semisynthetic, or synthetic
ephedrine samples.
Scheme 3. Methamphetamine Samples from
Ephedrine
We, now, report on the use of 2H NMR spectroscopy to estab-
lish the synthetic origin of methamphetamine samples. Twelve
samples of methamphetamine, prepared according to the most
common routes, were submitted to 2H NMR spectroscopy in order
to establish whether the deuterium pattern of the molecule could
retain a memory of the synthetic steps.
(3) (i) Jasper, J. P.; Westenberger, B. J.; Spencer, J. A.; Buhse, L. F.; Nasr, M.
J. Pharm. Biomed. Anal. 2004, 35, 21. (ii) Wokovich, A. M.; Spencer, J. A.;
Westenberger, B. J.; Buhse, L. F.; Jasper, J. P. J. Pharm. Biomed. Anal.
2005, 38, 781. (iii) Jasper, J. P.; Fourel, F.; Eaton, A.; Morrison, J.; Phillips,
A. Pharm. Technol. 2004, 28, 60. (iv) Jasper, J. P.; Weaner, L. E.; Duffy, B.
J. J. Pharm. Biomed. Anal. 2005, 39, 66.
(4) Hays, P. A.; Remaud, G.; Jasmin, E.; Martin, Y.-L. J. Forensic Sci. 2000, 45,
552.
(5) Mas, F.; Beemsterboer, B.; Veltkamp, A. C.; Verweij, A. M. A. Forensic Sci.
Int. 1995, 71, 225-231.
(6) Palhol, F.; Lamoreux, C.; Naulet, N. Anal. Bioanal. Chem. 2003, 376, 486-
490.
(7) Carter, J. F.; Titterton, E. L.; Murray, M.; Steeman, R. Analyst 2002, 127,
830-833.
(8) Carter, J. F.; Titterton, E. L.; Graant, H.; Steeman, R. Chem. Commun. 2002,
2590.
EXPERIMENTAL SECTION
Materials and Methods. Twelve samples of methamphet-
amine were prepared in our laboratory according to the most
common synthetic procedures, as shown in Schemes 1-3. The
origins of the chemicals were as follows: phenyl acetone com-
mercial from Solmag (4b), phenyl acetone prepared by synthesis
from phenylacetic acid and acetic anhydride (4a); ephedrine
unknown (3b) available from the collection of chemicals of the
department; ephedrine commercial from Fluka (3a), Schuchardt
(3c), and Solmag (3d); benzaldehyde commercial from Fluka;
LiAlH4 and NaBH4 from Aldrich.
The 2H experiments were performed on a Bruker Avance 500
spectrometer equipped with a 10-mm probe head and a 19F lock
channel (C6F6), under CPD (Waltz 16 sequence) proton decou-
pling conditions at the temperature of 300 K. The spectra were
(9) Armellin, S.; Brenna, E.; Fronza, G.; Fuganti, C.; Pinciroli, M.; Serra, S.
Analyst 2004, 129, 130-133.
(10) Kurashima, N.; Makino, Y.; Semita, S.; Urano, Y.; Nagano, T. Anal. Chem.
2004, 76, 4233-4236.
3114 Analytical Chemistry, Vol. 78, No. 9, May 1, 2006

recorded dissolving 0.7-0.8 g of material in ∼3.0 mL of CH2Cl2
adding ∼50 mL of C6F6 for the lock and 130-150 mg of hexameth-
yldisiloxane as internal standard. Hexamethyldisiloxane was chosen
because the signal of the six equivalent methyl groups occurs in
a well-isolated region of the spectrum allowing precise peak inte-
gration. It shows a D/H ratio of 130.7 ppm determined by calibra-
tion against the official standard TMU (Community Bureau of Ref-
erences, BCR) with a certified D/H ratio. Hexamethyldisiloxane
is rather volatile and thus extreme care had to be taken for precise
weighing. The solutions were prepared introducing and weighing
directly in the NMR tube the methamphetamine followed by the
standard and the solvent, maintaining the tube carefully closed.
The signal assignment of the spectrum of methamphetamine,
obtained from the analysis of the proton spectrum taken in the
same conditions of the deuterium spectrum, is the following (δ,
ppm): 7.34-7.25 (5H, m, Ph), 2.81 (1H, m, J ) 6.5 Hz, CH), 2.75
(1H, dd, J ) 6.5 and 13.2 Hz, Ha), 2.62 (1H, dd, J ) 6.4 and 13.2
Hz, Hb), 2.40 (3H, s, NCH3), 1.06 (3H, d, J ) 6.2 Hz. CH3). Four
spectra were run for each sample, collecting 3000-4000 scans
using the following parameters: 5.9-s acquisition time, 1380-Hz
spectral width, 16 K time domain, and 2-3-s delay. In these
conditions, the signal/noise ratio calculated for the smallest peaks
using a FID exponential multiplication of 0.5 Hz is always superior
to 15. An estimation of the relaxation times (T1) of the deuterium
nuclei showed that the longest T1 is that of the reference material
(∼1.0 s). Thus, the total time of 7-8 s between successive pulses
is certainly sufficient to ensure a complete nuclear relaxation. Each
FID was Fourier transformed with a line broadening of 0.5-1.0
Hz, manually phased, and integrated after an accurate correction
of the spectral baseline. For partially overlapped signals, the peak
areas were determined through the deconvolution routine of the
Bruker TopSpin NMR software using a Lorentzian line shape.
The absolute values of the site-specific D/H ratios were
calculated according to the formula
(D/H)i ) nWSgWS(MW)LSi(D/H)WS/(nigL(MWWS)SWS
)
where WS stands for the working standard with a known isotope
ratio (D/H)WS and L for the product under examination; nWS and
ni are the number of equivalent deuterium atoms of the standard
and of the ith peak; gWS and gL are the weights of the standard
and the sample; MWL and MWWS are the corresponding molecular
weights; Si and SWS are the areas of the ith peak and of the
standard, respectively.
RESULTS AND DISCUSSION
We have examined samples of methamphetamine prepared
from phenyl acetone (Scheme 1), benzaldehyde (Scheme 2), and
ephedrine (Scheme 3). Two different batches of phenyl acetone
had been used, a commercial one (4b) received from Solmag
(Italy) and one prepared in our laboratory (4a) from phenylacetic
acid according to the same industrial synthesis. The conversion
of phenyl acetone into methamphetamine requires the transforma-
tion of the carbonyl group into the CHNCH3 moiety. This can be
achieved according to the four different paths reported in Scheme
1: the hydrogen atom at C(2) in final compound 1 comes from
the reducing agent (NaBH4, H2, HCOOH) and the methyl group
is essentially that of methylamine.
The reaction starting from benzaldehyde needs the addition
of two carbon atoms, and this can be obtained by condensation
with nitroethane (Scheme 2). The two hydrogen atoms at C(1)
and C(2) of 1 are inserted by reduction with LiAlH4-H2SO4 and
the methyl group is generated by LiAlH4 reduction of the urethane
intermediate.
In the transformation of ephedrine into methamphetamine
(Scheme 3), the only modification in the pattern of the hydrogen
atoms is the insertion of a hydrogen atom in the benzylic position
by means of molecular hydrogen.
The values of D/H (ppm) for the different positions of the 12
samples of methamphetamines are reported in Table 1. The most
Table 1. (D/H)i Isotope Ratios (ppm)a
origin
samples
D/H(CHa)
138.7 (3.1)
D/H(CHb)
139.4 (2.1)
D/H(CH)
111.0 (3.2)
D/H(CH3)
115.9 (2.8)
D/H(NCH3)
129.4 (1.5)
D/H(Ph)
from phenyl acetone
(see Scheme 1)
A
152.8 (1.5)
B
C
D
E
F
140.7 (3.2)
168.4 (3.7)
134.8 (3.2)
128.4 (3.5)
137.1 (2.9)
141.9 (3.2)
137.0 (2.9)
158.6 (3.0)
143.8 (3.2)
136.7 (3.5)
139.4 (3.5)
140.8 (3.7)
107.7 (3.8)
150.4 (3.4)
36.5 (4.2)
26.3 (2.9)
135.6 (5.7)
114.6 (3.4)
116.7 (0.8)
137.7 (1.3)
125.6 (3.2)
117.2 (1.9)
121.1 (2.3)
122.9 (2.1)
130.4 (0.8)
130.6 (0.7)
135.6 (0.6)
131.6 (1.2)
134.8 (1.4)
137.8 (1.7)
151.0 (1.5)
152.9 (2.2)
153.1 (1.6)
152.3 (1.0)
150.9 (1.9)
151.6 (2.5)
G
from benzaldehyde
(see Scheme 2)
H
102.0 (3.1)
102.4 (1.2)
38.2 (3.4)
139.9 (1.4)
44.6 (0.6)
153.1 (2.0)
from ephedrine
(see Scheme 3)
I
59.3 (3.5)
88.3 (2.4)
35.3 (3.8)
102.1 (2.9)
130.5 (1.2)
145.1 (1.8)
J
K
L
61.9 (1.9)
161.9 (4.1)
160.6 (3.1)
83.2 (1.8)
131.1 (2.4)
158.3 (3.6)
28.9 (1.0)
62.3 (5.0)
27.9 (3.1)
101.9 (1.1)
125.3 (1.4)
98.8 (1.6)
127.5 (1.8)
102.1 (1.7)
118.9 (1.6)
147.9 (1.5)
129.6 (1.6)
154.9 (2.1)
a Values in parentheses are (D/H)i standard deviations.
Analytical Chemistry, Vol. 78, No. 9, May 1, 2006 3115

Scheme 4. Commercial Syntheses of Ephedrine
interesting isotope ratio values are those of the CH, CH2 (CHa
and CHb),1 and NCH3 positions, because these hydrogen atoms
were manipulated during the synthetic procedure.
stereoselective reductive amination by reaction with methylamine
and molecular hydrogen.
Optically active ephedrine can be obtained by resolution of
the racemate prepared by bromination of propiophenone (6),
reaction with methylamine, and reduction of the carbonyl group.
Natural ephedrine is obtained by extraction from Ephedra plants.
According to the synthetic procedure, the hydrogen atom at C(2)
of the final ephedrine is that of the starting propiophenone or is
generated by reductive amination. The low deuterium content at
the CH position of samples I-L could be tentatively attributed to
the use of H2 as a reducing agent.
On the basis of the values of (D/H)CH, the samples can be
divided into two groups: one (A, B, C, F, G) with (D/H)CH
>
100 ppm (from 107.7 to 150.4 ppm, underscored values in Table
1) and the other (D, E, H-L) with (D/H)CH < 100 ppm (from
26.3 to 62.3 ppm, highlighted values in Table 1). The analysis of
the (D/H)NCH3 shows that all the samples have comparable values
except sample H (D/H ) 44.6 ppm). The values of (D/H)CHa and
of (D/H)CHb appear to be randomly dispersed.
Isotope Ratios (D/H)CH. Table 1 shows that samples A, B,
C, F, and G, with (D/H)CH > 100 ppm can be distinguished from
samples D, E, H, I, J, K, and L, characterized by (D/H)CH < 100
ppm. Samples A-G are prepared starting from phenylacetone,
sample H is prepared from benzaldehyde, samples I-L are
obtained from ephedrine.
Isotope Ratios (D/H)NCH . The D/H values for the methyl
3
group linked to the nitrogen atom are very similar. The only
exception is sample H: the D/H is 44.6 ppm. This low value is in
accord with the fact that the methyl group is generated by LiAlH4
reduction of the urethane intermediate. This low value is diag-
nostic to distinguish this method of insertion of the methyl group
at the nitrogen atom from those involving methylamine or
compounds prepared thereof (PhCH2NHCH3 from PhCH2Cl and
CH3NH2 HCONHCH3 from HCOOH and CH3NH2 in Scheme 1).
Isotope Ratios (D/H)CHa and D/HCHb. Figure 1A shows the
D/H values of samples A-C, F, and G, with (D/H)CH > 100 ppm.
The six samples are almost superimposed: no relevant differences
are noted between samples A and F, obtained from phenyl acetone
4a, and samples B, C, and G, obtained from phenyl acetone 4b.
In all these samples, the benzylic CH2 is that of the parent phenyl
acetone and the corresponding values of (D/H)CHa and (D/H)CHB
are rather similar. Sample C, which is the one prepared from
phenylacetone by reduction followed by amination, shows the
highest D/H values for Ha and Hb. This might be tentatively
attributed to a scrambling of the hydrogen atoms in the benzylic
position due to a keto-enolic equilibrium promoted by the basic
conditions of NaBH4 carbonyl reduction.
Figure 1B shows the D/H values of samples D, E, and I-L,
with (D/H)CH < 100 ppm (sample H excluded). Samples I and J
are almost identical: they were obtained from two different
batches of ephedrine. Samples D, E, K, and L appear to be
similar: D and E were prepared from phenyl acetone by reaction
with benzyl methylamine and molecular hydrogen; K and L were
obtained from two different ephedrines.
The experimental results obtained for samples I-L are rather
intriguing. All the samples were prepared by converting ephedrine
into the chloro intermediate, which was then submitted to H2
reduction: hydride substitution at a sp3 carbon atom was involved
For the samples obtained from phenyl acetone, the deuterium
content of the hydrogen atom of this position was found to depend
on the reduction reagent. The use of molecular hydrogen
produced a deuterium depletion (samples D and E), compared
to the reduction performed with sodium boron hydride and formic
acid (samples A-C, F, G). The evidence in our hands9 is as
follows: (i) the reduction of CdN by H2 or LiAlH4, and the
reduction of double bonds by H2 or LiAlH4-H2SO4 (addition to
sp2 carbon atoms) leads to the introduction of deuterium-depleted
hydrogen atoms; (ii) the use of NaBH4 in carbonyl reductions
gives the insertion of hydrogen atoms with high deuterium
content. In the meth series, these observations are also confirmed
by the low D/H of the CH position of H. This latter was obtained
by LiAlH4-H2SO4 reduction of the condensation product between
benzaldehyde and nitroethane.
For the samples prepared starting from ephedrine, the (D/
H)CH values were found to be rather low. This hydrogen belongs
to the starting ephedrine molecule; thus, the synthetic methods
employed to prepare ephedrine should be considered. The
samples of methamphetamine examined had been prepared from
four different batches of ephedrine, but we have no idea of their
synthetic history.
The largest manufacturer of (1R,2S)-3 is Knoll Pharmaceuti-
cals, a division of BASF. The key intermediate is (R)-phenylace-
tylcarbinol (8), which is prepared by biotransformation of ben-
zaldehyde by pyruvate decarboxylase in whole cells of fermenting
baker’s yeast (Scheme 4). Compound 8 is then submitted to
3116 Analytical Chemistry, Vol. 78, No. 9, May 1, 2006

Figure 1. (A) Graphical representation of the D/H values of samples A-C, F, and G, with (D/H)CH > 100 ppm; (B) Graphical representation
of the D/H values of samples (D, E, I-L, except H) with (D/H)CH < 100 ppm.
similarity of samples A-C and F-G and of samples I-J and D-E.
It shows also the clear different origin of the ephedrine batches
employed as starting materials for samples I-L. Sample H has
its own synthetic history from benzaldehyde.
CONCLUSIONS
Interesting information can be obtained from the 2H NMR
spectra of methamphetamine samples. A high value of (D/H)CH
(>100 ppm) allows the identification of samples prepared from
phenylacetone using NaBH4 or HCOOH-HCONHCH3. When low
values of (D/H)CH are found (<100 ppm), the deuterium content
of the benzylic position becomes discriminant and should be
considered. Deuterium depletion of Ha and Hb is correlated to
the use of ephedrine as a starting material. If high values of (D/
H)CHa and (D/H)CHb are found, the choice of the synthetic
procedure is between phenyl acetone/H2 reduction or ephedrine
dehalogenation. The use of benzaldehyde and nitroethane gives
a very typical and easily recognizable deuterium pattern. The (D/
H)NMe is diagnostic to discriminate the use of methylamine as a
source of the NCH3 moiety.
Such knowledge of the reagents and starting materials and
the definition of the synthetic procedure are valuable tools in the
fight against synthetic illicit drugs. All the reagents are traceable,
and the probable suppliers of specific compounds can be detected.
The deuterium pattern of the methamphetamine molecule can
evidently provide a useful “fingerprint” to be considered in criminal
investigations.
Figure 2. Ternary plot (D/H)CH vs (D/H)CH vs (D/H)NCH
.
2
3
(Scheme 3). Low D/H were determined for samples I and J: (D/
H)CHA ) 59.3-61.9 ppm and (D/H)CHB ) 88.3. -83.2 ppm,
respectively. Samples K and L showed, on the contrary, a relevant
deuterium enrichment at the CH2 position, (D/H)CHA ) 161.9-
160.6 ppm and (D/H)CHB ) 131.1-158.3 ppm. These values are
comparable to those of the samples obtained from phenyl acetone.
No conclusion can be drawn from these data. First, we do not
know the synthetic origin and the deuterium pattern of the starting
ephedrines. Second we expect that, the mechanism of the
substitution reaction of an hydride at a sp3 carbon atom being
different from that of the addition to a sp2 carbon atom, a different
kinetic isotope fractionation could occur during this reaction step.
A comprehensive pictorial view of the diagnostic D/H values
is given Figure 2, which was obtained by plotting (D/H)CH versus
(D/H)NCH versus (D/H)CH (these latter data were the mean
ACKNOWLEDGMENT
Financial support of COFIN-MIUR is acknowledged.
Received for review November 30, 2005. Accepted March
3, 2006.
3
2
values of (D/H)CHa and (D/H)CHb). The plot highlights the
AC052105W
Analytical Chemistry, Vol. 78, No. 9, May 1, 2006 3117