Quantitative Silylation Speciations of Primary Phenylalkyl Amines, Including Amphetamine and 3,4-Methylenedioxyamphetamine Prior to Their Analysis by GC/MS

Article abstract of DOI:10.1021/acs.analchem.5b02599

A novel, quantitative trimethylsilylation approach derivatizing 11 primary phenylalkyl amines (PPAAs), including amphetamine (A) and 3,4-methylenedioxyamphetamine (MDA), was noted. Triggering the fully derivatized ditrimethylsilyl (diTMS) species with the N-methyl-N-(trimethylsilyl)-trifluoroacetamide (MSTFA) reagent, a new principle was recognized followed by GC/MS. In the course of method optimization, the complementary impact of solvents (acetonitrile, ACN; ethyl acetate, ETAC; pyridine, PYR) and catalysts (trimethylchlorosilane, TMCS; trimethyliodosilane, TMIS) was studied: the role of solvent and catalyst proved to be equally crucial. Optimum, proportional, huge responses were obtained with the MSTFA/PYR = 2/1-9/1 (v/v) reagent applying catalysts; A and MDA needed the TMIS, while the rest of PPAAs provided the diTMS products also with TMCS. Similar to derivatives generated with hexamethyldisilazane and perfluorocarboxylic acid (HMDS and PFCA) (Molnár et al. Anal. Chem. 2015, 87, 848'852), the fully silylated PPAAs offer several advantages. Both of our methods save time and cost by allowing for direct injection of analytes into the column; this is in stark contrast with the requirement to evaporate acid anhydrides by nitrogen prior to their injection. Efficiences of the novel catalyzed trimethylsilylation (MSTFA) and our recently introduced (now, for A and MDA extended) acylation principle were contrasted. Catalyzed trimethylsilylation led to diTMS derivatives resulting in on average a 1.7 times larger response compared to the corresponding acylated species. Catalyzed trimethylsilylation of PPAAs, A, and MDA were characterized with retention, mass fragmentation, and analytical performance properties (R², LOQ values). The practical utility of ditrimethylsilyation was shown by analyzing A in urine and mescaline (MSC) in cactus samples.

Full text of DOI:10.1021/acs.analchem.5b02599

Technical Note
pubs.acs.org/ac
Quantitative Silylation Speciations of Primary Phenylalkyl Amines,
Including Amphetamine and 3,4-Methylenedioxyamphetamine Prior
to Their Analysis by GC/MS
Borbala Molnar,†,§,⊥ Blanka Fodor,†,§ Imre Boldizsar,‡,∥ and Ibolya Molnar-Perl*,†,§
́
́
́
́
‡
§
∥
†Institutes of Chemistry and Biology and Departments of Analytical Chemistry and Plant Anatomy, L. Eotvos University, 1117,
̈
̈
Paz
́
́ ́ ́ ́
many Peter setany 1/A-C, Budapest, Hungary
⊥
̈
Doctoral School of Pharmaceutical Sciences, Semmelweis University, 1085, Ullo
̋
i ut 26, Budapest, Hungary
́
S
* Supporting Information
ABSTRACT: A novel, quantitative trimethylsilylation ap-
proach derivatizing 11 primary phenylalkyl amines (PPAAs),
including amphetamine (A) and 3,4-methylenedioxyamphet-
amine (MDA), was noted. Triggering the fully derivatized
ditrimethylsilyl (diTMS) species with the N-methyl-N-
(trimethylsilyl)-trifluoroacetamide (MSTFA) reagent, a new
principle was recognized followed by GC/MS. In the course of
method optimization, the complementary impact of solvents
(acetonitrile, ACN; ethyl acetate, ETAC; pyridine, PYR) and
catalysts (trimethylchlorosilane, TMCS; trimethyliodosilane,
TMIS) was studied: the role of solvent and catalyst proved to
be equally crucial. Optimum, proportional, huge responses
were obtained with the MSTFA/PYR = 2/1−9/1 (v/v)
reagent applying catalysts; A and MDA needed the TMIS, while the rest of PPAAs provided the diTMS products also with
TMCS. Similar to derivatives generated with hexamethyldisilazane and perfluorocarboxylic acid (HMDS and PFCA) (Molnar et
́
al. Anal. Chem. 2015, 87, 848−852), the fully silylated PPAAs offer several advantages. Both of our methods save time and cost by
allowing for direct injection of analytes into the column; this is in stark contrast with the requirement to evaporate acid
anhydrides by nitrogen prior to their injection. Efficiences of the novel catalyzed trimethylsilylation (MSTFA) and our recently
introduced (now, for A and MDA extended) acylation principle were contrasted. Catalyzed trimethylsilylation led to diTMS
derivatives resulting in on average a 1.7 times larger response compared to the corresponding acylated species. Catalyzed
trimethylsilylation of PPAAs, A, and MDA were characterized with retention, mass fragmentation, and analytical performance
properties (R2, LOQ values). The practical utility of ditrimethylsilyation was shown by analyzing A in urine and mescaline
(MSC) in cactus samples.
Through our studies, in 2 weeks old MSTFA/PYR (2/1, v/v)
derivatized standard solution, again to our delight, the A-diTMS
and MDA-diTMS products were identified: surprisingly, for A-
diTMS based on the NIST library. However, the source and
preparation background of A-diTMS spectrum were not
available. A-diTMS spectrum was found in the software of
“Mass spectra of designer drugs”,10 while A-diTMS tentative
formation was learned via an e-mail received from authors
(Supporting Information).11
To investigate the characteristic properties of this novel
trimethylsilylation process, in order to optimize it as quantitative
method, a many-sided research approach was undertaken.
Reagent composition, solvent, and catalyst were selected as
well as time and temperature for optimum analytical conditions
ecently, in order to trialkylsilylate PPAAs with the HMDS
Rand PFCA reagents, to our delight, a new derivatization
reaction was recognized: instead of the expected trimethylsilyl
derivatives, their trifluoroacylated products were obtained. The
principle’s advantages and practical utility along with the
comparison of traditional methods were shown.1 Now,
continuing our interest in the classical trimethylsilylation,
pioneered by Pierce,2 a novel, selective, and quantitative method
was realized again, applying MSTFA derivatization. This
experience opened the second new approach in PPAAs analysis,
including A and MDA. In the course of these studies, along with a
recent one,1 premises related PPAAs’ MSTFA derivatizations,
prior to their GC/MS analysis, were summed up.3−9 To
quantitate A, MDA, and MSC, the use of MSTFA was reported
in solvent free medium.3−9 A and MSC were trimethylsilylated
also in the catalyzed version (MSTFA containing 1% TMCS),9
leading to A-monoTMS,3−9 while MSC resulted in diTMS9
species.
Received: July 10, 2015
Accepted: September 26, 2015
© XXXX American Chemical Society
A
DOI: 10.1021/acs.analchem.5b02599
Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry
Technical Note
were determined. In parallel with GC/MS studies, on the basic
research level, the derivatization process’s stoichiometry,
derivatives’ proportionality, stability, reproducibility, and prac-
tical utility were documented. Analytical advantages along with
the recently published HMDS and PFCAs technique,1 extended
now to the analysis of A and MDA, were compared. Proposal’s
practical utility was demonstrated with the quantitation of A in
urine and MSC in cactus samples.
RESULTS AND DISCUSSION
■
Introductory experiences with our 2 weeks old MSTFA/PYR (2/
1, v/v) derivatized standard solution called our attention to the
diTMS transformation of PPAAs, including A and MDA, in PYR
medium. This observation, meaning the existence of A-diTMS,
was confirmed.11 Thus, further studies seemed to be promising
to derivatize PPAAs both in solvent free and in TMCS catalyzed9
media as well as varying the MSTFA/PYR (v/v) reagent
composition and the impact of TMCS and TMIS12 catalysts.
These approaches were performed for the first time (Figures 1
and 2 and Table 1).
Results, based on separate monoTMS/diTMS GC/MS
evaluations (monoTMS responses were obtained for all PPAAs
in MSTFA/ETAC = 2/1, v/v reagent composition; Supporting
Information Figure S1), proved the crucial role of PYR and
catalysts. Performing solvent free MSTFA derivatization resulted
in the diTMS species (Figure 1, dotted green columns); except
the cases of A and MDA, both eluted as the monoTMS
derivatives (Figure 1, dotted red columns), in accordance with
the literature.3−9
Applying PYR in the MSTFA/PYR = 2/1−9/1, v/v range
(data in Figure 1) led to commensurable responses as obtained
with the solvent free MSTFA (striped green columns, 2-PEA,
OMBA, 2-(3,4-DiM)PEA, MSC) or to somewhat larger ones,
consisting both of diTMS and monoTMS products (striped
green + striped red columns, BA, MMBA, PMBA, 2-MMPEA, 2-
PMPEA). A and MDA provided monoTMS species (striped red
columns) only. Using TMCS catalyst, varying reagent
composition in 150 μL of total volume (MSTFA/PYR/TMCS
= from 100/48/2 up to 100/40/10, v/v), resulted in
considerably increased diTMS responses (Figure 1, plain green
columns, Table 1, diTMS species). Reactivities of A and MDA,
even under these conditions, proved to be insufficient. To obtain
A-diTMS and MDA-diTMS species, additional studies are
needed. Further derivatizations were performed with specially
activated MSTFA (MSTFATMIS).12 We applied this reagent, at
first, in the MSTFATMIS/PYR = 2/1 (v/v) medium leading to A-
diTMS and MDA-diTMS in a quantitative/proportional manner
(Figure 2, red peaks’ line, spectra 2A, 4A).
In order to gain further insight into the ditrimethylsilylation
process, parameters, affecting the reaction (solvent, reaction
time, and temperature), had to be defined. Changing the proton
acceptor PYR both for ETAC and ACN (MSTFATMIS/ETAC or
ACN = 2/1, v/v), it turned out that applying ETAC primarily the
monoTMS products were formed (Figure 2, green peaks’ line,
spectra 1A, 2A, 3A, 4A), while ACN favors the diTMS
production (Figure 2, blue peaks’ line, spectra 2A, 3A, 4A).
Also exclusively the diTMS species were obtained in the solvent
free medium (orange peaks’ line, spectra 2A, 4A), however,
resulting in ∼1/10th of the responses compared to optimum
conditions (Figure 2, red line, spectra 2A, 4A).
EXPERIMENTAL SECTION
■
MSTFA (≥97.0), MSTFA activated I, catalog no. 50994,
containing ammonium iodide and ethanethiol, forming in situ
the MSTFA/TMIS = 1000/2 (v/v) reagent (further on
MSTFATMIS), TMCS (≥99.0), HMDS (99.9), TFA (99.5),
ACN (≥99.9), ETAC (≥99.9), PYR (≥99.9), benzylamine (BA,
≥ 99.5), 2-phenylethylamine (2-PEA, ≥ 99), D-amphetamine
sulfate (A, ≥ 99), o-methoxybenzylamine (OMBA, 98), m-
methoxybenzylamine (MMBA, 98), p-methoxybenzylamine
(PMBA, 98), 2-(m-methoxyphenyl)ethylamine (2-MMPEA,
97), 2-(p-methoxyphenyl)ethylamine (2-PMPEA, ≥ 98),
( )-3,4-methylenedioxyamphetamine hydrochloride (MDA, ≥
99), 2-(3,4-dimethoxyphenyl)ethylamine (2-(3,4-DiM)PEA,
homoveratrylamine, ≥ 98), and 2-(3,4,5-trimethoxyphenyl)-
ethylamine (mescaline, MSC, 99) were of highest analytical
grade, used as received (percent purity in parentheses): all
products of Sigma-Aldrich, St. Louis, MO.
Model compounds (10−12 mg/10 mL), weighed with 0.01
mg uncertainty, were dissolved in distilled water, neutralized with
hydrochloric acid, and further diluted into a unified stock
solution providing finally in 1 μL of derivatized solution 25−
2000 pg of PPAAs, including A and MDA of each. Model
solutions and/or cactus and urine extracts (detailed sample
preparation in the Supporting Information), in triplicate, were
rotary evaporated to dryness at 30−40 °C. Residues for
trimethylsilylation were treated with (i) 40 μL of PYR, 100 μL
of MSTFA, and 10 μL of TMCS; or (ii) 50 μL of PYR was mixed
with 100 μL of MSTFATMIS; or (iii) 150 μL of MSTFA alone;
then heated in an oven (90 °C, 60 min). Acylation were
performed with 70 μL of HMDS, 30 μL of TFA, and 100 μL of
ETAC, heated in an oven at 80 °C for 20 min. Thereafter
derivatized solutions were transferred into the autosampler vial
and 1 μL was injected into the GC/MS setup.
The apparatus consisted of a Varian 240 GC/MS/MS system
(Varian, Walnut Creek, CA). The analyses were carried out using
a Varian CP-8400 autosampler, and a septum programmable
injector (SPI). The column used was a product of SGE (Victoria,
Australia); SGE forte capillary BPX5: 30 m × 0.25 mm; df = 0.25
mm. The temperatures of the transfer line, ion trap, and manifold
were in order of listing 300, 210, and 80 °C, respectively. Under
optimized temperature programs, injections were made at 280
°C and held at 280 °C for 3 min, then cooled down to 100 °C
(100 °C/min); the column temperature profile was 100 °C, held
for 1.00 min, then to 145 °C for 10 °C/min, then to 230 °C for 5
°C/min, and finally to 280 °C for 50 °C/min with a 1.00 min
hold (total elution time were 24.50 min). Helium (purity, 6.0,
99.9999%) was used. The column flow rate was 1 mL/min. The
general MS parameters were Fil/Mul delay, 3.00 min; electron
energy, 70 eV. Statistical analysis was performed with Student’s
two-tailed t-test, and p < 0.05 was considered significant.
Varying time (10, 20, 30, 60, 90 min) and temperature (70, 80,
90, 100 °C) of ditrimethylsilylations, in each case, led to
optimum conditions (90 °C, 60 min).
Concerning advantages of ditrimethylsilylation compared to
monotrimethylsilylation methods,3−9 based on response rela-
tions, undoubtedly diTMS derivatization is to be preferred:
Table 1, responses, IU/pg, A-diTMS vs A-monoTMS, 1.63 × 104
vs 0.64 × 104, and MDA-diTMS vs MDA-monoTMS, 2.03 × 104
vs 0.57 × 104.
Turning to the derivatization reproducibility, characterized
with relative standard deviation percentages, (RSD %, details in
B
DOI: 10.1021/acs.analchem.5b02599
Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry
Technical Note
Figure 2. Peak profiles and mass spectra of A (spectra 1A, 2A) and MDA
(spectra 3A, 4A), obtained with MSTFATMIS only (orange lines) or
applying MSTFATMIS/solvents in 2/1 (v/v) ratios: ETAC (green lines),
ACN (blue lines), PYR (red lines).
role of TMCS (Figure 1, red and green striped versus green plain
columns). Uncertainty in stability is associated with monoTMS
formation (Figure 1, red striped columns, BA, MMBA, PMBA, 2-
MMPEA, 2-PMPEA).
Contrasting our recent acylation1 and the just now developed
ditrimethylsilyation principles, both are quantitative, reprodu-
cible, and comparable (Table 1). Characterizing acylation and
ditrimethylsilylation analyses they are identical in any sense in
their simplicity and advantages over classical derivatization for
GC (direct injectability onto the column, time, work, and cost
efficiency). As to their derivatization dependence, the specified
response behavior, particularly in cases of A, MDA, and MSC, the
use of catalyzed trimethylsilylation should be preferred, resulting
in 1.9 (A), 2.7 (MDA), and 1.6 (MSC) greater responses,
compared to their, by HMDS and TFA provided, acylated
species. Regarding analytical performance characteristics, R2 and
LOQ values both benefit ditrimethylsilylation (Table 1, data in
6th and 7th vertical columns). R2 varied between 0.9991 and
0.9999 for ditrimethylsilylation and 0.9986 and 0.9999 for
acylation. LOQ values fall in the ranges of 4.8−16 ng/mL
(average 7.2 ng/mL) for ditrimethylsilylation and 6.1−31 ng/mL
(average 12.4 ng/mL) for acylation, respectively.
Practical Utility of Proposal. The advantages of ditrime-
thylsilylation contrasting with the just now extended acylation
principle, A from urine, MSC from cactus tissues, were summed
up in parallel determinations (A, Figure 3; MSC, Figure 4).
Linearity, proportionality, and reproducibility of this method is
indicative of a 100% yield of analyte derivatization. This is
evidenced by using model solutions (Table 1, R2, LOQ values)
and natural matrixes, A in urine and MSC in cactus tissue up to 10
ng/injected amounts (A content in urine was in average 3.69 μg/
mL, with 6.7 RSD%; MSC content in cactus proved to be in
average 0.54% (w/w), with 5.5 RSD %). Regarding MSC
quantitation from four various amounts of tissue extract, now,
from a different, albeit same, cactus type (Lophophora
Willimamsii, Lem.), considerably less MSC content (0.54%)
was obtained, in comparison with the earlier measured value
(1.39%).1 However, the 0.54% falls also in the range reported by
others (0.053−4.7%).13
Figure 1. Responses in integrator units (IU/pg × 104) for PPAAs, A, and
MDA depending on reagent compositions: monoTMS (red) and
diTMS (green) columns (Supporting Information, Figure S1,
monoTMS derivatives).
Table 1, IU/pg values and Figure 1, uncertainty bars), in model
solutions they varied between 0.10% and 5.1% (for diTMS
derivatives, average 2.14 RSD %) and 0.56% and 3.12% (for
acylated species, average 1.49 RSD %). The derivative
composition and stability behavior confirm the unambiguous
C
DOI: 10.1021/acs.analchem.5b02599
Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry
Technical Note
Table 1. Structures, Mass Fragmentation, and Analytical Performance Characteristics of PPAAs Derivatives; Comparison of
Retention and Response Values as Di-TMS (Green Columns, Completed by Spectra 1A−4B) and as Acylated Species (Blue
a
Columns), Obtained by GC/MS
a
Indications, as in Figures 1 and 2, SFI = selective f+ragment ion; R2 = R squared value for linear regression in the standard curve, in the range of limit
·
of quantitation (LOQ) and 400 ng/mL values.
two times of each. Supporting Information, Figure S1, monoTMS and Figure S2, HMDS and TFA species of A and MDA. Response differences
between acylated and ditrimethylsilylated species were significant, all at the level, p < 0.05
M
= molecular ions; * = average responses obtained from three separate derivatizations, injected
D
DOI: 10.1021/acs.analchem.5b02599
Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry
Technical Note
of LOQ and 400 ng/mL concentrations) varied between
0.9991−0.9999 (ditrimethylsilylation) and 0.9986−0.9999
(acylation); LOQ proved to be in the ranges of 4.8−16 ng/mL
(average 7.2 ng/mL) for ditrimethylsilylations and 6.1−31 ng/
mL (average 12.4 ng/mL) for acylations, respectively.
Reliability and reproducibility of analyses were characterized
with the relative standard deviation percentages confirming in
average 2.14 RSD % (catalyzed ditrimethylsilylation) and 1.49
RSD % (HMDS and TFA acylation) (Table 1, IU/pg value,
responses indicated by blue and green columns, one by one). The
utility of our proposal was shown by contrasting acylation and
ditrimethylsilylation processes in the quantitation of A from
urine and MSC from cactus stem tissue.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.anal-
chem.5b02599.
■
Figure 3. Peak profiles of acylated (a) and ditrimethylsilylated (b)
derivatives from the same urine extract, corresponding to 3.36−3.88 ng
(red lines) and 1.82−1.94 ng (green lines) A, 1 ng standard (orange
lines) and blank (blue lines); sample preparation in the Supporting
Information. Mass spectra obtained from urine in Figure S3, as acylated
and as ditrimethylsilylated derivatives.
S
Source and background of A-diTMS spectrum; detailed
sample preparation; GC/MS peak profiles and mass
spectra for PPAAs as monoTMS and as, by HMDS and
TFA, acylated A and MDA; and spectra of A from urine
and MSC from cactus samples (PDF)
AUTHOR INFORMATION
Corresponding Author
*E-mail: perlne@chem.elte.hu.
■
Author Contributions
The manuscript was written through contributions of all authors.
All have given approval to manuscript’s final version.
Notes
The authors declare no competing financial interest.
REFERENCES
■
(1) Molnar
́
, B.; Csam
́
pai, A.; Molnar
́
-Perl, I. Anal. Chem. 2015, 87,
848−852.
Figure 4. Peak profiles of MSC derivatives, from the same extract, as
acylated (a) and ditrimethylsilylated (b) species (corresponding to
found MSC ng/crude tissue ng as follows: 0.84−0.86/151 (blue), 1.61−
1.70/302 (orange), 3.87−4.00/755 (green), and 7.29−8.12/1510
(red)), indicating in average 0.54% MSC content (5.5 RSD %); sample
preparation in the Supporting Information. Mass spectra obtained from
cactus in Figure S4, as acylated and as ditrimethylsilylated derivatives.
(2) Pierce, A. E. Silylation of Organic Compounds; Pierce Chemical
Company: Rockford, IL, 1968.
(3) Toennes, S. W.; Steinmeyer, S.; Maurer, H.-J.; Moeller, M. R.;
Kauert, G. F. J. Anal. Toxicol. 2005, 29, 22−27.
(4) Pujadas, M.; Pichini, S.; Civit, E.; Santamarina, E.; Perez, K.; de la
̃
Torre, R. J. Pharm. Biomed. Anal. 2007, 44, 594−601.
́
(5) Joya, X.; Pujadas, M.; Falcon, M.; Civit, E.; Garcia-Algar, O.; Vall,
O.; Pichini, S.; Luna, A.; de la Torre, R. Forensic Sci. Int. 2010, 196, 38−
42.
CONCLUSION
(6) Gonzalez-Marino, I.; Quintana, J. B.; Rodríguez, I.; Cela, R. J.
́
̃
■
Chromatogr. A 2010, 1217, 1748−1760.
A novel approach was introduced for the quantitative and
selective ditrimethylsilylation of PPAAs, including A and MDA,
applying MSTFA in PYR medium, using catalysts (TMCS,
TMIS). The optimum condition and analytical advantages of the
new principle are compiled and contrasted with the recently
recognized conditions1 in the present paper for the A and MDA
extended acylation method (Table 1, Figures 1−4, and
Supporting Information, Figures S1 and S2). Advantages,
indicating the superiority of our new acylation and ditrime-
thylsilylation principles are comparable: both approaches
provide directly injectable species onto the column, avoiding
loss of derivatives, saving time, work, and cost of preparation
process. Comparing responses of our two recent proposals it
turned out that ditrimethylsilylation, on average, resulted in 1.7
times larger responses compared to acylation. Analytical
performance characteristics were highlighted: R2 (in the range
(7) Wan Raihana, W. A.; Gan, S. H.; Tan, S. C. J. Chromatogr. B: Anal.
Technol. Biomed. Life Sci. 2011, 879, 8−16.
(8) Pelica̧ o, F. S.; Peres, M. D.; Pissinate, J. F.; De Martinis, B. S. J. Anal.
̃
Toxicol. 2014, 38, 341−348.
(9) Liu, R. H.; Canfield, D. V.; Wang, S.-M. Quantitaion and Mass
Spectromtric Data of Drugs and Isotopically Labeled Analogs; CRC Press:
Boca Raton, FL, 2009; p 16, 52, 245.
(10) Rosner, P. Mass Spectra of Designer Drugs [CD-ROM]; Wiley-
̈
VCH: Weinheim, Germany, 2015.
(11) Rosner, P. University of Kiel, Kiel, Germany. Personal
̈
communication, 2015.
(12) Derivatization of Drug Substances with MSTFA; Analytix Notes 7,
2005; p 6.
(13) Ogunbodede, O.; McCombs, D.; Trout, K.; Daley, P.; Terry, M. J.
Ethnopharmacol. 2010, 131, 356−362.
E
DOI: 10.1021/acs.analchem.5b02599
Anal. Chem. XXXX, XXX, XXX−XXX