Archived at the Flinders Academic Commons:
http://dspace.flinders.edu.au/dspace/
‘This is the peer reviewed version of the following article:
Barnes, C., Madaras, S., Pigou, P. E., Johnston, M. R., &
Paul Kirkbride, K. (2019). Origins of N-
formylmethamphetamine and N-acetylmethamphetamine in methamphetamine produced by the hydriodic acid and red phosphorus reduction of pseudoephedrine. Forensic Chemistry, 13, 100158. https://doi.org/10.1016/
j.forc.2019.100158
which has been published in final form at https://doi.org/10.1016/j.forc.2019.100158
© 2019 Elsevier B.V.. This manuscript version is made available under the CC-BY-NC-ND 4.0 license:
http://creativecommons.org/licenses/by-nc-nd/4.0/
Accepted Manuscript
Origins of N-formylmethamphetamine and N-acetylmethamphetamine in meth- amphetamine produced by the hydriodic acid and red phosphorus reduction of pseudoephedrine
Christopher Barnes, Simone Madaras, Paul E Pigou, Martin R Johnston, K Paul Kirkbride
PII: S2468-1709(18)30078-X
DOI: https://doi.org/10.1016/j.forc.2019.100158 Article Number: 100158
Reference: FORC 100158
To appear in: Forensic Chemistry Received Date: 9 August 2018 Revised Date: 14 March 2019 Accepted Date: 19 March 2019
Please cite this article as: C. Barnes, S. Madaras, P.E. Pigou, M.R. Johnston, K. Paul Kirkbride, Origins of N- formylmethamphetamine and N-acetylmethamphetamine in methamphetamine produced by the hydriodic acid and red phosphorus reduction of pseudoephedrine, Forensic Chemistry (2019), doi: https://doi.org/10.1016/j.forc.
2019.100158
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Origins of N-formylmethamphetamine and N-acetylmethamphetamine in methamphetamine produced by the hydriodic acid and red phosphorus reduction of pseudoephedrine
Christopher Barnes1, * Simone Madaras1 Paul E Pigou1,2 Martin R Johnston1 K Paul Kirkbride1
1 College of Science and Engineering, Flinders University, GPO Box 2100, Adelaide, South Australia, Australia.
2 Forensic Science SA, GPO Box 2790, Adelaide, South Australia, Australia Corresponding author:
KP Kirkbride, [email protected]
College of Science and Engineering, Flinders University, GPO Box 2100, Adelaide 5001, South Australia, Australia.
KEYWORDS Nagai reaction Leuckart reaction
N-Acetylmethamphetamine N-Formylmethamphetamine
ABSTRACT
N-Formylmethamphetamine (FMA) and N-acetylmethamphetamine (AMA) are suspected to be trace by-products in methamphetamine (MA) produced from pseudoephedrine using the Nagai method. However, these amides are not rational by-products of the Nagai method.
FMA is an intermediate in the synthesis of MA using the Leuckart method. However, as there is the possibility that FMA is a by-product of the Nagai method, the significance of FMA as an indicator of the Leuckart method has been debated. It is therefore important to establish whether AMA and especially FMA are by-products of the Nagai reaction and thus establish their significance as synthetic route markers.
From the work presented here, FMA is a by-product of the Nagai reaction but the mechanism by which FMA arises could be not determined.
AMA was also shown to be a by-product of the Nagai reaction, most likely due to reaction between MA and phenyl-2-propanone (P-2-P), itself a by-product of the Nagai reaction.
Furthermore, during GC analysis of Nagai reaction products, MA has been shown to react with P-2-P or ethyl acetate in the injector to form AMA.
Caution is recommended if the relative abundance of AMA and/or FMA are used as a basis for determining whether MA samples have a common source or not. Furthermore, it is clear that FMA cannot be considered to be a route-specific by-product for the Leuckart reaction - it is the abundance of FMA in a reaction mixture or profile, not simply its presence, that points to the involvement of the Leuckart reaction.
Caution is recommended if the relative abundance of AMA and/or FMA are used as a basis for determining whether MA samples have a common source or not. Furthermore, it is clear that FMA cannot be considered to be a route-specific by-product for the Leuckart reaction - it is the abundance of FMA in a reaction mixture or profile, not simply its presence, that points to the involvement of the Leuckart reaction.
1. INTRODUCTION
The use of gas chromatography-mass spectrometry (GC-MS) for the generation of profiles of synthesis by-products, adulterants, impurities and unchanged precursors in illicit drugs is a well- established means of gathering tactical and strategic intelligence relevant to illicit drug distribution and manufacture [1, 2, 3, 4, 5, 6, 7, 8, 9].
Methamphetamine (MA) is an illicit drug of major importance in many countries around the world, and as a consequence a substantial effort has gone into the development of methods to profileMA seizures. As of the date of this publication, two methods for the preparation of MA samples prior to GC-MS have become dominant. First, the method recommended by Tanaka et al. [3], where ethyl acetate is used as solvent to extract MA free base from a basified, aqueous phosphate buffer at pH 9. Second, the method developed by Lock [4], and later adopted as a harmonized method [5, 6, 7], which involves the use of toluene as solvent to extract MA free base from aqueous tris(hydroxymethyl)aminomethane (TRIS) buffer at pH 8.1.
Both sample treatment methods, when combined with GC-MS, provide valuable insight as to the synthetic pathway that may have been used in the manufacture of a particular sample of MA, and each is being used as means of indicating whether samples of powder may have originated from a single batch or not. However, in order for profiling to provide reliable results, it is critical that any analytical methodology does not discriminate against certain compounds nor cause the appearance of additional compounds, as this could result in type 2 errors in interpretation (i.e., falsely indicating a common source) or type 1 errors (falsely rejecting a common source), respectively. In regards to the use of GC-MS as an analytical technique, it is well-established that the high temperatures of the injection port and column can cause unimolecular reactions to take place—such as cyclization of chloroephedrine to yield 2-phenyl-2,3-dimethylaziridine [10]—or bimolecular reactions such as transesterification [9].
Two compounds are noteworthy in regards to their detection during GC-MS profiling of illicit MA samples; N-acetylmethamphetamine (AMA, compound 1, Figure 1) and N- formylmethamphetamine (FMA, compound 2, Figure 1).
N
(1) O
N
O H
(2)
Figure 1 structures of AMA and FMA.
AMA is noteworthy because it does not seem to be a rational by-product in those synthetic sequences in which it has been reported to form. These reactions include the Leuckart reaction [10], aluminium amalgam reductive alkylation of phenyl-2-propanone (P-2-P) [12], reductions of ephedrine or pseudoephedrine based upon the Nagai1 method [13, 14], sequences employing hydrogenolysis of chloroephedrine/chloropseudoephedrine [13] and dissolving metal reduction of ephedrine/pseudoephedrine [15]. A common element to these reports is that the GC-MS sample preparation method of Tanaka is used, where ethyl acetate is the extraction solvent. Lee et al. [16] attributed the detection of AMA samples examined by them to pyrolysis of MA when it was injected into their gas chromatograph, citing as supporting evidence the work of Sekine and Nakahara [17], who observed FMA, AMA and N-propionylmethamphetamine (PrMA) in their GC-MS chromatograms when methamphetamine/tobacco mixtures are smoked. Conn et al. [18] detected AMA in street seizures and surmised, after consideration of police intelligence reports, that its origin was a transesterification reaction that takes place when clandestine drug laboratory operators used propyl acetate for azeotropic removal of water from MA hydrochloride salt.
The presence of FMA in any MA drug profile is noteworthy because FMA is a known intermediate in the synthesis of MA using the Leuckart reaction. As a consequence, the presence of FMA in a MA profile has been described as a route-specific marker for the involvement of the Leuckart reaction in the preparation of the drug (see, for example, [19] and [20]). In their work however, Qi et al. [21] issued a caution that traces of FMA can be detected in profiles obtained from MA that was most likely not produced by the Leuckart reaction. Qi et al. also presented the observation that FMA, when it was present as a trace in MA, was always accompanied by traces of AMA.
As there is some evidence that either AMA or FMA or both might be analytical artefacts, and because these substances are members of a suite of profiling substances used for tactical and strategic intelligence purposes, we felt that a detailed examination of the origin of these amides in drug profiles was warranted. This article describes experiments and results relating to such an investigation.
2. MATERIALS AND METHODS
1 For convenience, in this article the term “Nagai reaction” will be used to refer to the method where hydriodic acid, red phosphorus and ephedrine or pseudoephedrine are boiled together in an unsealed container.
Phenyl-2-propanone (Merck brand), toluene, sodium bisulfite, cyclohexylamine, phenyl-2- butanone, phosphorus, propanoyl chloride, acetic anhydride, sodium sulfate (anhydrous) and benzyl acetate were sourced from Sigma Aldrich (Sydney, Australia). 1-Phenyl-1,2-propanedione (TCI brand) and nonadecane (TCI brand, for GC analysis) were sourced from ChemSupply (Adelaide, Australia).
MA as its hydrochloride salt (secondary GC quantification material calibrated against a primary standard provided by the National Measurement Institute, Sydney, Australia) was provided by Forensic Science SA (Adelaide Australia). The method of Windahl et al. [22] was used for the preparation of other samples of MA used in this research, where indicated. Briefly, pseudoephedrine hydrochloride (1.0 g; 5.0 mmol) and red phosphorus (20 mg; 0.65 mmol) were added to hydriodic acid (2.1 g; 3 mL of 47% solution; 17 mmol) and the mixture was heated under reflux for 5 hours. The resulting suspension was diluted with water (3 mL) and left to stand overnight. The solution was filtered to remove any remaining phosphorus and basified to pH 10.5 with NaOH solution (20%). The organic products were extracted into a solution of nonadecane (as internal standard, 50 mg/100mL) in toluene (3 x 10 mL) and dried with MgSO4. When MA salt was required HCl gas was passed through the solution, which yielded a white solid that was collected by filtration.
For experiments where the Nagai reaction was carried out in the presence of phenyl-2-butanone or cyclohexylamine, the above conditions were reproduced but approximately 400 µL of the ketone or amine were added to the reaction mixture.
All NMR spectra were collected on a Bruker Avance III NMR Spectrometer. 1H NMR and 13C NMR spectra were collected at 600 MHz and 150 MHz, respectively. (CD3)2SO was used as a solvent and internal lock. Chemical shift was reported in ppm and coupling constants in Hz.
Abbreviations: s=singlet, d=doublet, t=triplet, m=multiplet, b=broad, o=overlapping peaks.
GC-MS operating conditions were adapted from Qi et al. [21]. Analysis was performed using an Agilent 7890 gas chromatograph coupled to an Agilent 5975C mass spectrometer (Agilent Technologies, USA) and an Agilent 7693 autosampler. In most analyses splitless injection of 1 µL of solution was used with a split/splitless injector equipped with a deactivated quartz wool- packed liner (SGE brand, p/n 092218). The state of contamination of the GC liner may affect the abundance of artefacts produced; for this study a new liner was used prior to each set of experiments.
Helium was the carrier gas and an Agilent DB-5MS column (30 m x 0.25 mm i.d. x 0.25 µm film thickness) was used under the following conditions: flow rate 1 mL/min; injector
temperature 280˚C; oven program 50˚C for 1 min; 10˚C/min to 300˚C; 300˚C for 10 min. The GC-MS interface temperature was maintained at 250˚C, the MS quadrupole temperature was
150˚C, the ion source temperature was 230˚C and the mass scan range was m/z 43-350.
SPME was carried out using a Supelco manual fibre holder equipped with a fused silica fibre coated with a 100 µm-thick polydimethylsiloxane (PDMS) film. To detect FMA and AMA in MA HCl, the salt (20.6 mg) was placed into a GC vial (2 mL capacity). The fibre was inserted into the headspace in the vial for 1.5 hours and inserted into a GC inlet, desorbed at 280˚C under splitless conditions and chromatographed using the following temperature program: 90C (5 min) – 280C @20C/min.
For the following GC-MS analyses MA hydrochloride samples (25 mg, a secondary quantitative reference material provided by Forensic Science SA) were dissolved in an aqueous solution of sodium hydroxide (NaOH (20%), 1 mL, pH 10.5, diluted with water) and extracted with a solution of nonadecane (as internal standard) in toluene (1 mL, 500 mg/L) using a vortex mixer (1 min). The organic layer was transferred to an autosampler vial for analysis.
For experiments where reactions between MA and P-2-P or phenyl-2-butanone in a GC injection port were monitored, or for derivatization of MA, solutions of MA in toluene
(approximately 1 mg/mL) were treated with approximately 2 mL of reagent before injection into the chromatograph. For reactions between ketones and cyclohexylamine in the GC injection port, solutions of P-2-P or phenyl-2-butanone in toluene (2.5 µL in 2 mL) were treated with cyclohexylamine (2.5 µL) and injected.
The bisulfite addition complex of P-2-P was prepared by adding commercial P-2-P ( 3.3 g, 25 mmol) dropwise to a rapidly stirring solution of sodium bisulfite ( 5.5 g, 53 mmol) in a mixture of water (9 mL) and ethanol (6 mL). The solution was stirred for 10 min then allowed to stand.
The precipitate that formed was filtered under reduced pressure, washed with ethanol (7 x 50mL) and dried in vacuo. For reactions between P-2-P and MA in an injection port the bisulfite addition complex of P-2-P (10 mg) and MA hydrochloride (10 mg) were placed into a GC vial, treated with sodium hydroxide solution (2 N, 1 drop) and then extracted with a solution of nonadecane in toluene (1-2 mL, ca. 0.1mg/mL). Anhydrous sodium sulfate (ca. 10 mg) was added to the vial, which was then shaken for ca. 1 min before the supernatant was transferred to a fresh vial for analysis.
For reactions between P-2-P and MA on a hotplate, mixtures of the bisulfite addition complex of P-2-P (10 mg) and MA hydrochloride (10 mg) were treated as above for GC analysis but then the toluene solutions were evaporated to dryness under a stream of nitrogen, capped and placed on a hotplate set at various temperatures between 130 and 180C. After the reaction time the vial was cooled and toluene (1 mL) was added prior to GC-MS analysis.
For reactions between MA and benzyl acetate or 1-phenyl-1,2-propanedione, MA as its hydrochloride salt (10.2 mg) was treated with sodium hydroxide solution (2 N, 200 µL) and a
solution of nonadecane in toluene (1.5 mL, 3 mg in 50 mL) and vortexed for 1 min. The aqueous layer was drawn off, the organic phase was dried with anhydrous sodium sulfate and then split between 2 GC vials (2 mL capacity). One vial was treated with benzyl acetate (100 µL) while the other was treated with 1-phenyl-1,2-propanedione (100 µL) and the solutions were immediately analyzed by GC-MS.
AMA and PrMA were prepared for GC-MS by adding either acetic anhydride (100 µL; 1.1 mmol) or propanoyl chloride (100 µL; 1.1 mmol), respectively to MA (5 mg; 27 µmol) freebase in a solution of nonadecane in toluene (5 mL, 500 mg/L) and mixing using a vortex mixer for 1 min.
FMA was prepared by following a procedure outlined by Jung et al. [25], which involves heating a mixture of MA (5 mg ; 27 µmol), formic acid (0.1 mL; 2.6 mmol), and toluene (2 mL) under reflux for 5 hours. It was also prepared using the Leuckart reaction by treating P-2-P (2.7 mL; 20 mmol) with N-methylformamide (6.7 mL; 110 mmol) at 160 – 170˚C for 24 h.
Synthesis of 3,4-dimethyl-5-phenyl-1,3-oxazolidine hydrochloride was adapted from Lewis et al. [23].
Briefly, pseudoephedrine hydrochloride (2.0 g; 10 mmol) and formaldehyde (1.2 g; 3 mL of 37%
solution; 40 mmol) were added to methanol (50 mL). The solution was stirred following the conditions outlined in Table 7. The volume of the solution was reduced to approximately 3 - 4 mL under a stream of nitrogen at 40˚C. Milli-Q water (2 mL) was added and the organic products extracted with chloroform (3 x 10 mL). The combined extracts were dried over MgSO4 and evaporated to give a white solid. The solid was recrystallised from ethyl acetate/methanol. The identity of the product was confirmed by NMR.
1H NMR (600 MHz, (CD3)2SO): δ (ppm) 7.37 – 7.35 (4H, m, ArH), 7.32 – 7.31 (1H, m, ArH), 4.66 (1H, d, J=3.2 Hz, OCHaHbN), 4.44 (1H, d, J=8.3 Hz, ArCHO), 4.27 (1H, d, J=2.8 Hz, OCHaHbN), 2.45 (1H, b, CH3CHN), 2.33 (3H, s, NCH3), 1.11 (3H, d, J=6.2 Hz, CHCH3). 13C NMR (150 MHz, (CD3)2SO): δ (ppm) 141.1, 128.8, 128.2, 126.8, 88.4, 85.5, 68.2, 45.8, 36.9, 14.2, 8.9.
LC-MS analysis was carried out using a Waters Acquity Ultra-Performance Liquid Chromatograph (Waters, Milford, USA) coupled to a Waters Quattro Micro API ESCI triple quadrupole mass spectrometer (Waters, Manchester, UK). Samples (10 µL) were injected onto a Waters Acquity Ethylene Bridged Hybrid (BEH) C18, 1.7 µm 2.1 mm x 50 mm column. Mobile phase A was aqueous formic acid (0.1%), mobile phase B was acetonitrile and flow rate was maintained at 0.2 mL/min. The elution solvent program was 0.00 -1 min 2% B, 1.1-15.0 min 2- 100% B, 15.1-20.00 min 98% A. Positive ESI mode was used with nitrogen as the desolvation gas at a flow rate of 510 L/h and temperature of 350˚C. The capillary was maintained at 3.0 kV, the MS source temperature was 80˚C and the mass range was m/z 50-300. For LC, MA hydrochloride (1 – 5 mg, prepared using the Nagai reaction) was dissolved in 1 mL deionised water and filtered (0.45 µm nylon membrane) before analysis. For chromatography of crude reaction mixtures, the solvent was removed using rotary evaporation or under a steady stream of nitrogen,
depending on the volume. This produced the crude salt, which was dissolved in aqueous acetonitrile solution (1 mL, 20% acetonitrile) and filtered through a 0.45 µm nylon filter prior to analysis.
3. RESULTS AND DISCUSSION 3.1 Formation of AMA
A significant body of research dealing with the profiling of MA makes use of ethyl acetate as the extraction solvent. In a number of these articles where the Nagai reaction has been used to produce the drug, or where there was a suggestion of the involvement of the Nagai reaction, it has been indicated that AMA is a by-product [13, 14, 15, 21]. However, it has also been reported that MA and ethyl acetate can react to form AMA under conditions used for GC-MS analysis [22]. The question therefore arises as to whether AMA actually is a by-product of the Nagai reaction or whether the AMA detected in MA samples that have been extracted with ethyl acetate arises as an analytical artefact. One aim of the research presented here is to attempt to resolve that question.
The key observation of Sasaki [24] - that when a solution of MA in ethyl acetate was injected into a GC the abundance of AMA was positively correlated to injection port temperature - was reproduced in the present research. This supports the notion that a chemical reaction is occurring that is significantly influenced by temperature. In the present study this was demonstrated by measuring the peak area for AMA relative to the peak area for one of the arylalkylnaphthalene by-products for a solution of MA free base in ethyl acetate injected over a temperature range of 250-350C (see Figure 2).
250 280 300 330 350
0 2 4 6
Relat ive Peak Are as
Injection Port Temperature (C)
Figure 2 GC-MS response of signal due to AMA (red line, AMA peak area relative to the peak area due to 1,3-
dimethyl -2-phenylnaphthalene) and to FMA (green line, FMA peak area relative to the peak area due to 1,3- dimethyl -2-phenylnaphthalene) arising from a solution of MA in ethyl acetate as a function of injection port temperature. Data were calculated from a series of single injections.
It therefore is evident that AMA can be produced by a reaction between MA and ethyl acetate in the GC inlet. However, a correlation between FMA abundance and injection port temperature was not observed (see Figure 2). The abundance of FMA was monitored in order to determine whether a formylating agent, perhaps ethyl formate, was present as an impurity in the ethyl acetate used in experiments. The presence of a trace of FMA that did not increase in abundance as the injection temperature is raised suggests that it is present in the MA used in analysis (i.e., as a by-product in the Nagai reaction) and not likely to be an injection artefact in these experiments.
It is important to establish whether transesterification in the injection port is the sole origin of AMA in MA extracts or whether some AMA is already present as a by-product of the Nagai reaction prior to any involvement of ethyl acetate and GC analysis. To this end, several samples of solid MA hydrochloride prepared using the Nagai reaction (without any involvement with any ethyl acetate whatsoever) were analyzed using headspace SPME-GC-MS (1.5 hr exposure of fibre to headspace at room temperature). Traces of both AMA and FMA were detected using this approach (see Figure 3) indicating their formation as by-products under the Nagai reaction conditions.
Figure 3 Typical total ion chromatogram acquired using SPME collection (1.5 hr, room temperature) of headspace above a solid sample of MA. The peak at retention time 7.168 min is MA, 10.665 is FMA, 10.869 is AMA, 13.413 is 1,3-dimethyl-2-phenylnaphthalene and the peak at 13.536 was due to 1-benzyl-3-methylnaphthalene. P-2-P was not detected. Data were treated to remove background siloxane interference at m/z 73, 147, 207, 221 and 281.
The proposition that AMA and FMA are by-products in the Nagai reaction was confirmed by subjecting the same MA salt produced by the Nagai reaction to LC-MS; both AMA and FMA were detected in the sample. Furthermore, when toluene was used to extract a basified solution (aqueous sodium hydroxide) of the same MA and then analyzed using GC-MS, both AMA and FMA were detected. Taken together, the SPME, GC and LC findings suggest that ethyl acetate does induce formation of AMA via transesterification of MA during its analysis by GC, but they also suggest that AMA and FMA are also present as in MA as by-products of the Nagai reaction.
Conceivably, AMA present in the Nagai reaction mixture could be a residue carried through from the starting material. In order to test that possibility, the pseudoephedrine used to manufacture MA in the experiments reported here was examined as concentrated solutions using GC-MS and LC-MS and the solid was analysed using SPME GC-MS. Neither AMA nor FMA was detected in any of these experiments (results not shown) again indicating their formation under the Nagai reaction conditions.
Analysis of the reaction conditions indicated that during the Nagai reaction the only substances present other than pseudoephedrine, hydriodic acid, red phosphorus and atmospheric carbon dioxide that could lead to the formation of AMA are the abundant by-products, i.e., P-2-P, 1,3- dimethyl-2-phenyl aziridine and the arylalkylnaphthalenes. Of these, only the ketone possesses a reactive moiety that potentially could be involved in adventitious acetylation of MA. Although
7.00 8.00 9.00 10.00 11.00 12.00 13.00
500000 1000000 1500000 2000000 2500000 3000000 3500000 4000000 4500000 5000000 5500000 6000000 6500000 7000000 7500000 8000000 8500000 9000000 9500000 1.0500000 1.1500000
Time-->
Abundance 7.168
8.571 10.413
10.665
10.869
11.214
11.426 12.383
13.413 13.536
there is no chemical precedent for P-2-P acting in this manner (i.e., acting as an acetylating agent), experiments relating to this possibility were carried out. Co-injection of MA free base (known to be free of AMA) and P-2-P in toluene into a GC-MS produced AMA (see Figure 4).
Figure 4 Total ion chromatogram produced when MA and P-2-P are co-injected into a GC-MS. The peak at retention time 15.135 min is due to AMA, internal standard (nonadecane) elutes at 18.46, MA and P-2-P are the peaks that are above scale at approximately 9.8 and 8.9 min, respectively, while the peaks at 9.296 and 9.401 min are due to benzyl acetate and 1-phenyl-1,2-propanedione, respectively.
However, in an experiment where P-2-P was co-injected with MA protected using trifluoroacetic anhydride, AMA was not formed. This is presumably because of the weaker nucleophilicity of the nitrogen atom when present as an amide.
During review of this manuscript it was suggested by a reviewer that it may be possible for a by-product in the P-2-P used in experiments, rather than P-2-P itself, might have been responsible for the formation of AMA. Examination of the data presented in Figure 4 indicated that benzyl acetate (compound 3 in Figure 5) and 1-phenyl-1,2-propanedione (compound 4) were present in the mixture at low levels (retention times 9.296 and 9.401 min, respectively in figure 4). In order to test whether these compounds can acetylate P-2-P, experiments using P-2-P free of compounds 3 and 4 were carried out. This was achieved by mixing P-2-P bisulfite addition complex, MA HCl (highly pure, secondary quantitative analysis reference material) and a few drops of sodium hydroxide solution together, which produced P-2-P of extremely high purity in situ. Extraction of the products of reaction using toluene with internal standard (nonadecane) followed by GC-MS indicated the presence of AMA. When mixtures of MA and P-2-P generated in this fashion were evaporated to dryness and then placed on a hotplate (at temperatures ranging from 130 and 180C) approximately an 8-10-fold increase in AMA was observed (see
9.00 10.00 11.00 12.00 13.00 14.00 15.00 16.00 17.00 18.00
1000000 2000000 3000000 4000000 5000000 6000000 7000000 8000000 9000000 1.0000000 1.1000000 1.2000000 1.3000000 1.4000000 1.5000000 1.6000000 1.7000000 1.8000000
Time-->
Abundance
12.410 12.610
15.135
16.495
16.785 17.084
17.161
17.366 17.664
17.971 18.006 9.296
9.401
Supplementary Information, Figure 1). Interestingly, when 1-phenyl-1,2-propanedione was co- injected with MA, AMA was also detected, but AMA was not detected when benzyl acetate and MA were co-injected. The same outcomes were achieved when phenyl-1,2-propanedione and benzyl acetate were heated with MA on a hotplate at high temperatures. When P-2-P generated from its bisulphite complex was heated on a hotplate without MA, neither phenyl-1,2- propanedione nor benzyl acetate were produced; GC-MS indicated that P-2-P remained pure.
In order to examine the generality of some of these findings, more experiments were carried out. First, MA free base was co-injected with phenyl-2-butanone into a GC-MS as were mixtures of cyclohexylamine and P-2-P or phenyl-2-butanone. From these experiments PrMA (compound 5, Figure 5), N-acetylcyclohexylamine (compound 6) and N-propylcyclohexylamide (compound 7), respectively were all detected.
N
O (3)
O (4)
O
(5) O N
(6)
N
(7)
HN HN
O
H O
O O
O
(9) (8)
Figure 5, Structures for benzyl acetate (3), 1-phenyl-1,2-propanedione (4), N-propionylmethylamphetamine (PrMA, 5), N-acetylcyclohexylamine (6) and N-propylcyclohexylamide (7) 3,4-dimethyl-5-phenyl-1,3-oxazolidine (8), cyclohexylamine formamide (9).
The structures of the amide products were confirmed by comparison of their retention time and mass spectral fragmentation pattern with those of authentic samples produced by the reaction between MA and propionyl chloride, cyclohexylamine and acetyl chloride, and cyclohexylamine
and propionyl chloride. Only traces of these amides, approximately the same abundance as AMA shown in Figure 4, are produced when these amines and ketones are co-injected into a GC-MS;
therefore the acetylations and propionylations discussed obviously are not favourable reactions in the GC injection port.
These experiments indicate that P-2-P can directly acetylate MA, and given that it is present in the Nagai reaction at a much higher level than either benzyl acetate or 1-phenyl-1,2- propanedione, it is the most likely compound of the three to be responsible for the formation of AMA during the Nagai reaction. The temperature at which the Nagai reaction is carried out (approx. 127C, the boiling point of hydriodic acid) is quite high, but nevertheless the possibility that the reaction conditions were sufficient to allow acetylation of MA were explored. This was carried out by spiking a Nagai reaction with phenyl-2-butanone; PrMA was detected in the reaction mixture.
In summary, there are three potential origins for AMA detected during MA profiling. As demonstrated in Figure 6, AMA can arise during the Nagai reaction when P-2-P is present; it can form during GC analysis if MA and P-2-P are co-injected; and it can form during GC analysis if MA and ethyl acetate are co-injected. Furthermore, the relative abundance of AMA will depend upon the temperature of the injection port of the chromatograph and/or the amount of P-2-P present in the analytical extract.
NH OH
N NH O
EtOAc GC-MS injection port AMA
NH GC-MS injection port P-2-P
P-2-Pin situ HI / P
Figure 6. Three potential pathways by which AMA may be present in a MA profile.
3.2 Formation of FMA
Having successfully found various routes for the formation of AMA in drug profiles, we turned our attention to the origins of FMA. Exactly how FMA might be produced during the Nagai
reaction, or during GC of its extracts, is not obvious since there is a lack of an obvious formylating agent. As clandestine drug laboratory operators usually do not attempt to exclude carbon dioxide from the Nagai reaction (and neither did the initial experiments reported here) the possibility that adventitious carbon dioxide is responsible for formylation was explored.
Rather than relying on natural amounts of carbon dioxide, the Nagai reaction was carried out with carbon dioxide bubbling through the reaction mixture. This resulted in an approximately four-fold increase in the amount of FMA detected compared to a Nagai reaction carried out under nitrogen. Given that the amount of carbon dioxide present during this experiment would greatly exceed the amount present when the reaction is carried out in air, and given that FMA was detected even when the reaction was carried out under nitrogen, it appears unlikely that atmospheric carbon dioxide produces the FMA observed during MA profiling.
Traces of 3,4-dimethyl-5-phenyl-1,3-oxazolidine (compound 8, Figure 5), which is known to form from the reaction between pseudoephedrine and formaldehyde [21], were detected in Nagai reaction extracts. To determine whether an oxidative ring opening of the oxazolidine takes place during the Nagai reaction to form FMA (i.e., the oxazolidine acts as ‘masked’ FMA, with unmasking taking place as the Nagai reaction proceeds or upon injection of extracts into the GC), a separately synthesized sample of oxazolidine was treated with boiling hydriodic acid and red phosphorus. A trace of FMA was observed using GC-MS analysis, but given that the oxazolidine is itself a trace by-product in the Nagai reaction it can be ruled out as a major source of FMA.
Attention then turned to the possibility another formylating agent is present in the Nagai reaction.
When cyclohexylamine was added to a Nagai reaction cyclohexylamine formamide (compound 9, Figure 5) was detected in the reaction mixture. This result, together with the detection of 3,4- dimethyl-5-phenyl-1,3-oxazolidine in the Nagai reaction mixture, suggests that a formylating agent, perhaps formaldehyde itself, is present in the reaction mixture. However, at this stage we have not been able to identify the agent.
4. Summary and conclusions
Both AMA and FMA are produced as by-products in the Nagai reaction. AMA has three potential origins: it is a GC artefact arising as a result of transacetylation when ethyl acetate is used to extract MA samples; it is produced as a result of reaction between P-2-P and MA during the Nagai reaction; and it is produced as a result of reaction between P-2-P and MA (or MA and 1-phenyl-1,2-propanedione) when the two are co-injected into a GC-MS. The abundance of AMA detected for a particular sample will depend upon the amounts produced by each of these
sources. Therefore caution must be exercised if the relative abundance of AMA is used as a basis for determining whether MA samples have a common source or not.
In comparison, the origin of FMA is still obscure. It is evident that FMA can be produced by the Nagai reaction. Therefore other information pertaining to the route of manufacture must be taken into account before the presence of FMA in a drug profile is used as evidence to implicate the involvement of the Leuckart reaction in the production of MA. The points of view of Sanger et al. and Qi et al., published many years ago, still hold true - that a high relative abundance of FMA in a reaction mixture or profile provides some evidence pointing to the involvement of the Leuckart reaction. As it is not yet clear whether reactions that lead to the formation of FMA can take place upon injection of Nagai reaction mixtures or extracts into the GC, it is recommended that caution is exercised when using the relative abundance of FMA as means of determining whether MA samples have a common source.
5. ACKNOWLEDGEMENT
The authors wish to acknowledge financial support from the Ross Vining Research Fund, which is provided by Forensic Science SA.
6. REFERENCES
[1] http://www.swgdrug.org/Monographs/METHAMPHETAMINE.pdf, accessed February 2, 2019.
[2] http://www.swgdrug.org/Monographs/AMPHETAMINE.pdf, accessed February 2, 2019.
[3] Tanaka, K., Ohmori, T., Inoue, T., Seta, S., Impurity profiling analysis of illicit methamphetamine by capillary gas chromatography, Journal of Forensic Sciences, (1994), 39(2): 500-511.
[4] Lock, E., Development of a harmonized method for the profiling of amphetamine, (2005), University of Lausanne, PhD Thesis.
[5]Andersson, K., Jalava, K., Lock, E., Finnon, Y., Huizer, H.,Finnon, Y., Huizer, H., Kaa, E., Lopes, Poortman-van der Meer, A., Cole, M.D., Dahlen, J., Development of a harmonised method for the profiling of amphetamines:III. Development of the gas chromatographic method, Forensic Science International, (2007), 169(1): 50-63.
[6]Aalberg, L., Andersson, K., Bertler, C., Borén, H., Cole, M.D., Finnon, Y., Huizer, H., Jalava, K., Kaa, E., Lock, E.,
Lopes, A., Poortman-van der Meer, A., Sippola, E., Dahlén, J., Development of a harmonised method for the profiling of amphetamines: I. Synthesis of standards and compilation of analytical data, Forensic Science International, (2005), 149(2–3): 219-229.
[7]Aalberg, L., Andersson, K., Bertler, C., Borén, H., Cole, M.D.,Finnon, Y., Huizer, H., Jalava, K., Kaa, E., Lock, E.,
Lopes, A., Poortman-van der Meer, A., Sippola, E., Dahlén, J., Development of a harmonised method for the profiling of amphetamines: II. Stability of impurities in organic solvents, Forensic Science International, (2005), 149(2–3): 231-241.
[8] Doughty, D., Painter, B., Pigou, P., Johnston, MR., The synthesis and investigation of impurities found in
clandestine laboratories: Baeyer-Villiger route part II; synthesis of Phenyl-2-propanone (P2P) analogues from substituted benzaldehydes, (2018), Forensic Chemistry, Volume 9, 1-11.
[9] Toske, S.G., Brown, J.L., Miller, E.E., Phillips, M.Z., Hays, P.M., Recent Methamphetamine Profiling Trends: Tracking the Nitrostyrene Method Used for P2P Production, Forensic Chemistry,
https://doi.org/10.1016/j.forc.2018.12.003.
[10] Allen, A., Kiser, W., Methamphetamine from ephedrine: I. chloroephedrines and aziridines, Journal of Forensic Sciences, (1987), 32(4): 953-962.
[11] Dybowski, R., Gough, T.A., A Study of Transacetylation between 3,6-Diacetylmorphine and Morphine, Journal of Chromatographic Science, (1984), 22: 465-469.
[12] Kunalan, V., Kerr, W.J., NicDaéid, N., Characterization of Route Specific Impurities Found in Methamphetamine Synthesized by the Leuckart and Reductive Amination Methods, Anal Chem, (2009), 81: 7342- 7348.
[13] Ko, B.J..Suh, S. Suh, Y.-J., In, M.-K., Kim, S.-H., The impurity characteristics of methamphetamine synthesized by Emde and Nagai method, Forensic Science International, (2007), 170, 142–147.
[14] Kunalan, V., Kerr, W.J., Nic Daéid, N., Investigation of the Reaction Impurities Associated with Methylamphetamine Synthesized Using the Nagai Method, Anal Chem, (2012), 84: 5744-5752.
[15] Kunalan, V., Kerr, W.J. NicDaéid, N., Clarification of route specific impurities found in methylamphetamine synthesised using the Birch method, Forensic Science International, (2012), 223(1–3): 321-329.
[16] Lee, J. S., Han, E. Y., Lee, S. Y., Kim, E. M., Park, Y. H., Lim, M. A., Chung, H. S., Park, J. H., Analysis of the impurities in the methamphetamine synthesized by three different methods from ephedrine and pseudoephedrine, Forensic Science International, (2006), 161(2-3): 209-215.
[17] Sekine, H., Nakahara, Y., Abuse of methamphetamine mixed with tobacco: 1. inhalation efficiency and pyrolysis products of methamphetamine, Journal of Forensic Sciences, (1987), 32(5): 1271-1280.
[18] Conn, C., Dawson, M., Baker, A., Keegan, J., Fryirs, B., Identification of N-acetylmethamphetamine in a sample of illicitly synthesized methamphetamine. Journal of Forensic Sciences, (1996), 41(4): 645-647.
[19] Kram, T.C., Kruegel, A.V., The identification of Impurities in Illicit Methamphetamine exhibits by gas chromatography/mass spectrometry and Nuclear magnetic resonance spectroscopy, Journal of Forensic Sciences, (1977), 22(1): 40-52.
[20] Sanger, D.G., Humphreys, I.J, Joyce, J.R., A review of analytical techniques for the comparison and characterization of illicit drugs, J. Forensic Sci. Soc, (1979), 19: 65–71.
[21] Qi Y., Evans, I., McCluskey, A., New impurity profiles of recent Australian imported ‘ice’: Methamphetamine impurity profiling and the identification of (pseudo)- ephedrine and Leuckart specific marker compounds, Forensic Science International, (2007), 169, 173–180.
[22] Windahl, K.L., McTigue, M.J., Pearson, J.R., Pratt, S.J., Rowe, J.E., Sear, E.M., Investigation of the impurities found in methylamphetamine synthesised from pseudoephedrine by reduction with hydriodic acid and red phosphorus, Forensic Science International, (1995), 76: 97–114.
[23] Lewis, R.J., Huffine, E.F, Mattson, J, Formation of an interfering substance, 3,4-dimethyl-5-phenyl-1,3- oxazolidine, during a pseudoephedrine urinalysis, Journal of Forensic Sciences, (2000), 45(4): 898-901.
[24] Sasaki, T., Makino, Y., Effective injection in pulsed splitless mode for impurity profiling of methamphetamine crystal by GC or GC/MS, Forensic Science International, (2006), 160(1): 1-10.
[25] Jung, S.H., Ahn, J.H., Park, S.K., Choi, J.-K., A practical and convenient procedure for the N-formylation of amines using formic acid. Bulletin of the Korean Chemical Society, (2002), 23(1): 149-150.
Formylmethamphetamine is a by-product in methamphetamine produced by either the
Leuckart or Nagai methods
Acetylmethylamphetamine is also a by-product of the Nagai reaction but it can also be an analytical artefact
Phenyl-2-propanone was observed to react with methamphetamine to form acetylmethylamphetamine
MULTIPLE ORIGINS OF N-ACETYLMETHAMPHETAMINE IN METHAMPHETAMINE
N
O
