Journal of Analytical Toxicology 2012;36:340 –344
doi:10.1093/jat/bks026
Article
Comparison of Five Derivatizing Agents for the Determination of Amphetamine-Type
Stimulants in Human Urine by Extractive Acylation and Gas Chromatography –Mass
Spectrometry
Adrienn Dobos*, Eló´d Hidvégi and Gábor Pál Somogyi
National Institute of Forensic Toxicology, H-1146 Budapest, Varannó utca 2-4, Hungary
*Author to whom correspondence should be addressed. Email: [email protected]
Five acylation reagents have been compared for use as derivatizing
agents for the analysis of amphetamine-type stimulants (ATS) in
urine by gas chromatography–mass spectrometry (GC–MS). The
evaluated reagents were heptafluorobutyric anhydride, pentafluoropropionic anhydride, trifluoroacetic anhydride, acetic anhydride (AA)
and N-methyl-bis(trifluoroacetamide). The ATS included amphetamine, methamphetamine (MA), 3,4-methylenedioxyamphetamine
(MDA), 3,4-methylenedioxymethamphetamine (MDMA) and 3,4methylenedioxyethylamphetamine (MDEA). A mixture of the ATS
was added to urine (1 mL) followed by KOH solution and saturated
NaHCO3 solution. The sample was then extracted with dichloromethane and the derivatizing agent and 2 mL were injected into the
GC–MS instrument. The derivatizing agents were compared with reference to the signal-to-noise (S/N) ratios, peak area values, relative
standard deviations (RSDs), linearities, limits of detection (LODs) and
selectivities. The acetic anhydride proved to be the best according to
the S/N ratio and peak area results for amphetamine, MA, MDMA
and MDEA. The best RSD values of peak areas and of S/N ratios at
3 mg/mL were also given by AA in cases of MDA, MDMA and
MDEA. At 20 mg/mL, the lowest RSD values of peak areas for MDA
and the lowest RSD values of S/N ratios for MA, MDA, MDMA and
MDEA were again given by AA. Additionally, the highest correlation
coefficients for MA, MDA, MDMA and MDEA and the lowest LOD
results for MA, MDMA and MDEA were produced by AA.
Introduction
Amphetamine-type stimulants (ATS) are the second most commonly abused drugs in Hungary, after cannabinoids (1). Their
popularity appears to be based on the mistaken belief that ATS
are less dangerous and addictive than other types of drug. ATS
are used either as appetite suppressants or as stimulant party
drugs, because their entactogenic effects facilitate social interactions (2). ATS abuse in Hungary has increased by a factor of
100 over the last two decades (3 –4) and faster and more reliable analytical methods are needed. On-line acylation effectively reduces the sample preparation time. Three different
techniques have been used: on-column acylation (5– 11), headspace solid-phase microextraction (HS-SPME) with on-fiber
acylation (11–16) and extractive acylation (17 –18). Extractive
acylation was selected for this study because quantitative
HS-SPME is insufficiently reproducible and on-column acylation
may cause column deterioration.
Anhydrides are the most popular derivatizing agents for
extractive acylation, but it is not obvious which would be the
most effective in the case of ATS. Four reagents were evaluated,
including heptafluorobutyric anhydride (HFBA), pentafluoropropionic anhydride (PFPA), trifluoroacetic anhydride (TFA)
and acetic anhydride (AA). In addition, N-methyl-
bis(trifluoroacetamide) (MBTFA) was also included for comparison purposes, although it is not an anhydride, because
MBTFA is currently used as an on-column derivatizing agent at
the National Institute of Forensic Toxicology for the determination of ATS in biological fluids (5 –6).
Some of the most frequently abused ATS in Hungary are amphetamine, methamphetamine, 3,4-methylenedioxyamphetamine
(MDA), 3,4-methylenedioxymethamphetamine (MDMA) and
3,4-methylenedioxyethylamphetamine (MDEA), so these were
selected as the target analytes.
Materials and Methods
Chemicals and reagents
The reference standards of (+)-amphetamine, (+)methamphetamine, (+)-MDA, (+)-MDMA and (+)-MDEA as
their hydrochloride salts were purchased from the National
Measurement Institute (Australian Government, Pymble, NSW,
Australia). Stock solutions of these standards were prepared in
distilled water at a concentration of 2 mg/mL. A mixture containing each of the ATS at a concentration of 200 mg/mL was
then prepared and this was further diluted to give a standard
with a concentration of 10 mg/mL.
HFBA, PFPA and MBTFA were obtained from Sigma-Aldrich
Chemie (Steinheim, Germany) and were of derivatization
grade. TFA and AA, purity 99.0%, were obtained from Fluka
Analytical (Sigma-Aldrich Chemie). Tert-butyl methyl ether was
from Merck (Darmstadt, Germany), dichloromethane and
hydrochloric acid were from Sharlau Chemie S.A. (Barcelona,
Spain) and potassium hydroxide and sodium hydrogen carbonate were from Reanal (Hungary).
Blank urine samples were collected with informed consent
from laboratory staff and were confirmed to be negative for
the target analytes before use by analysis with the current
procedure.
Sample preparation
ATS were extracted and derivatized at the same time in the following way: 1 mL blank urine was spiked with 100 mL of the
200 mg/mL amphetamine standard mixture and was then mixed
with 250 mL buffer (10 M KOH–saturated NaHCO3, 3:17 v:v),
1,500 mL dichloromethane and one of the derivatizing agents.
To have equimolar quantities of the derivatizing agents, the
following quantities were added to the samples after dichloromethane; 50 mL of PFPA, 62 mL of HFBA, 35 mL of TFA, 24 mL of
AA and 36 mL of MBTFA. All the samples were vortexed immediately for 2 min after adding the derivatizing agent and were
then centrifuged at 3,000 rpm for 5 min. After centrifugation,
# The Author [2012]. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected]
the lower organic layer was transferred to a clean vial and 2 mL
was injected into the gas chromatography –mass spectrometry
(GC– MS) instrument. Samples were prepared in triplicate on
four consecutive days at concentrations of 3.0 and 20.0 mg/mL.
The data obtained were used to calculate signal-to-noise (S/N)
ratios, peak areas and relative standard deviation (RSD) values.
For the linearity study, blank urine samples were prepared
in quadruplicate and spiked with ATS at five different concentrations: 0.1, 0.3, 2.0, 5.0 and 8.0 mg/mL, covering the range
in which ATS are usually measured in our institute (0.3 to
8.0 mg/mL).
For the LOD study, a pool of blank urine was spiked with
ATS at a concentration of 125 ng/mL of each analyte and the
following concentrations were achieved by dilution with blank
urine: 62.5, 31.25 and 15.63 ng/mL. The rest of the procedure
was as described previously.
For the selectivity study, interferences from ephedrine, pseudoephedrine, phentermine, ketamine and some of the newer
ATS, mephedrone, methylone, methylenedioxypyrovalerone
(MDPV) and 4-methyl-N-ethylcathinone (4-MEC) were examined. These substances and the amphetamine type analytes
were spiked in blank urine at a concentration of 20 mg/mL of
each analyte.
GC–MS conditions
Chromatographic analyses were performed on a Shimadzu
QP5000 GC –MS system with Shimadzu Class 5000 software
Version 2.23 and fitted with a Restek Rtx-5MS capillary column
(length 30 m, internal diameter 0.25 mm, film thickness
0.25 mm). The injector was operated in splitless mode
(sampling time: 0.1 min), split ratio 1:14 at a temperature of
2708C and helium carrier gas pressure of 120 kPa, total flow
0.8 mL/min. The sample volume was 2 mL. The column temperature program was: 908C (1 min) increased to 2008C at
208C/min, then to 2908C at 308C/min, with a final hold time of
16 min. The interface temperature was 2808C. The mass spectrometer was operated in EIþ full scan mode (m/z 40 –500),
except in the limit of detection (LOD) study, when selected
ion monitoring (SIM) mode was used with a sampling interval
of 0.30 s per channel. The observed fragment ions and relative
ion intensities for the different ATS derivatives are summarized
in Table I. The fragment ions in bold were used to calculate
peak areas, S/N ratios and RSD values.
Results and Discussion
The S/N ratios (s ¼ 2), peak area values and RSD values of the
analyzed samples were calculated from the three replicate measurements on four consecutive days at concentrations of 3.0 and
20.0 mg/mL with each derivatizing agent. The RSD values were
calculated from the peak areas and S/N ratios. The results are
shown in Tables II–V. Considering the results for all five ATS
derivatives, AA proved to be the most effective reagent according to the S/N ratios and their RSD values at 3.0 and 20.0 mg/mL
(Tables II–III), although the fragment ions of the derivatives are
in a range at which the noise is high due to ions originating
from the matrix. PFPA was the second best reagent and MBTFA,
TFA and HFBA were the least effective. These findings were
supported by the peak area values at 3.0 and 20.0 mg/mL
(Tables IV– V), at which acetic anhydride gave better results
Table I
Fragment ions and relative ion intensities (Fragment ions in bold were used in the calculation of peak areas and signal-to-noise ratios)
Target substances
PFPA
m/z
Relative ion intensity (%)
m/z
Relative ion intensity (%)
m/z
Relative ion intensity (%)
m/z
Relative ion intensity (%)
m/z
Relative ion intensity (%)
Amphetamine
190
118
91
204
160
118
135
162
325
204
162
135
218
162
353
100
79
42
100
28
24
100
45
16
100
64
38
100
44
12
240
118
169
254
210
118
135
162
375
254
210
162
268
240
162
100
15
7
100
26
5
100
91
36
100
40
40
100
44
40
44
86
118
58
91
100
44
135
162
58
100
162
72
114
162
100
17
9
100
37
9
100
26
8
100
23
14
100
30
16
140
118
117
154
110
118
135
162
100
49
11
100
26
19
100
41
140
118
117
154
110
118
135
162
100
40
11
100
24
19
100
41
154
162
135
168
162
140
100
61
60
100
87
75
154
162
135
168
162
140
100
61
60
100
87
75
Meth-amphetamine
MDA
MDMA
MDEA
HFBA
AA
TFA
MBTFA
Table II
Signal-to-noise (S/N) ratios and relative standard deviations (RSD) for five different amphetamine type stimulant derivatives at a concentration of 3.0 mg/mL
Signal to noise ratio at 3 mg/mL
PFPA
HFBA
AA
TFA
MBTFA
Amphetamine
Methamphetamine
MDA
Mean
RSD(%) of S/N
Mean
RSD(%) of S/N
Mean
RSD(%) of S/N
MDMA
Mean
RSD(%) of S/N
MDEA
Mean
RSD(%) of S/N
213
205
564
398
138
9,53
5,19
5,99
14,76
8,25
300
201
985
306
240
8,97
5,79
4,78
14,32
4,48
351
304
473
559
225
8,05
7,32
4,43
12,04
6,70
201
120
963
132
268
8,86
7,19
6,56
11,85
8,14
155
53
1185
54
146
11,77
10,09
6,00
12,74
6,40
Mass Spectrometry 341
Table III
Signal-to-noise (S/N) ratios and relative standard deviations (RSD) for five different amphetamine type stimulant derivatives at a concentration of 20.0 mg/mL
Signal to noise ratio at 20 mg/mL
PFPA
HFBA
AA
TFA
MBTFA
Amphetamine
Methamphetamine
MDA
Mean
RSD(%) of S/N
Mean
RSD(%) of S/N
Mean
RSD(%) of S/N
MDMA
Mean
RSD(%) of S/N
MDEA
Mean
RSD(%) of S/N
3622
988
5565
1061
1901
12,69
4,69
10,95
11,74
14,86
4175
1002
5966
627
3722
13,48
5,74
5,51
9,40
14,99
5097
1454
4927
1675
1635
11,31
7,35
6,76
10,94
12,73
2548
1054
7753
435
3281
11,44
7,68
7,30
10,18
14,39
1172
310
8278
343
1716
8,90
7,79
6,66
11,62
15,00
Table IV
Peak area values and relative standard deviations (RSD) for five different amphetamine type stimulant derivatives at a concentration of 3.0 mg/mL
Peak area at 3 mg/mL
PFPA
HFBA
AA
TFA
MBTFA
Amphetamine
Methamphetamine
MDA
Mean
RSD(%)
Mean
RSD(%)
Mean
RSD(%)
MDMA
Mean
RSD(%)
Mean
MDEA
RSD(%)
18562
13981
105405
62647
16073
8,66
5,51
6,03
11,84
5,73
17647
10251
115431
29418
24105
8,84
5,38
6,34
12,48
5,92
32977
15751
63768
52948
33238
8,00
7,58
5,20
11,06
7,96
8022
4478
84275
8296
16846
7,44
7,04
6,05
11,58
7,93
4411
1795
84412
2151
7388
7,70
6,61
6,40
11,30
7,44
Table V
Peak area values and relative standard deviations (RSD) for five different amphetamine type stimulant derivatives at a concentration of 20.0 mg/mL
Peak area at 20 mg/mL
PFPA
HFBA
AA
TFA
MBTFA
Amphetamine
Methamphetamine
MDA
MDEA
RSD(%)
Mean
RSD(%)
Mean
RSD(%)
Mean
RSD(%)
Mean
RSD(%)
475042
102827
1047965
56222
425659
13,88
4,74
8,59
8,91
16,71
435160
70257
1210692
35359
605282
13,81
4,26
8,69
10,43
16,58
723888
176925
625936
150808
538752
11,86
7,45
6,49
9,52
13,77
199837
40035
894223
28818
280108
11,83
6,16
6,65
11,01
13,75
91709
13229
950910
11331
136662
12,07
6,63
6,82
11,62
13,60
than the other derivatizing agents. MBTFA was the second best
reagent according to the peak area values, and was followed by
PFPA. HFBA and TFA had the lowest peak area values.
According to the RSD values of peak areas at 3.0 and
20.0 mg/mL (Tables IV–V), the best results were given by PFPA
and the second best by AA. All anhydrides proved to be suitable
for ATS derivatization because neither of them exceeded 15%,
which is the recommended RSD limit (19). The only exception
was MBTFA at a concentration of 20.0 mg/mL for amphetamine,
and methamphetamine for which the values were between 16
and 17%.
Linear correlation coefficients (R2) were calculated from
the quadruplicate analyses at five concentrations (0.1, 0.3,
2.0, 5.0 and 8.0 mg/mL; Table VI). All R2 values were greater
than 0.90. The best results were obtained with acetic anhydride, for which all values were higher than 0.99, followed
by PFPA, HFBA and then MBTFA. TFA had the lowest R2
values except for amphetamine, for which it gave the best
result.
The LOD was measured in SIM mode using urine spiked
with ATS in the range 15.63 to 125 ng/mL. The S/N ratio was
calculated from triplicate measurements at four different concentrations (15.63, 31.25, 62.5 and 125 ng/mL). The lowest
concentration at which the S/N ratio was greater than 3 was
considered to be the LOD. The best result was given by HFBA,
followed by acetic anhydride, MBTFA, PFPA and finally, TFA
(Table VII).
342 Dobos et al.
MDMA
Mean
Table VI
Correlation coefficient (R2) for five different amphetamine type stimulant derivatives
Correlation coefficient
Amphetamine
Methamphetamine
MDA
MDMA
MDEA
PFPA
HFBA
AA
TFA
MBTFA
0.9808
0.9833
0.9966
0.9993
0.9692
0.9838
0.9870
0.9953
0.9098
0.9714
0.9835
0.9750
0.9984
0.9519
0.9655
0.9865
0.9860
0.9959
0.9320
0.9717
0.9808
0.9623
0.9922
0.9320
0.9663
Table VII
Limits of detection for five different amphetamine type stimulant derivatives
LoD (ng/mL)
Amphetamine
Methamphetamine
MDA
MDMA
MDEA
PFPA
HFBA
AA
TFA
MBTFA
125
16
63
125
125
31
16
16
125
31
31
31
63
63
31
63
31
16
125
31
125
31
16
125
125
Eight compounds besides the five ATS were examined in the
interference study: ephedrine, pseudoephedrine, phentermine,
ketamine, mephedrone, methylone, MDPV and 4-MEC
(Figure 1). However, in the case of some derivatizing agents,
more than eight potentially interfering peaks appeared
after spiking with the interferents, due to low yields of derivatization. AA was the only reagent that did not leave behind
significant underivatized target compounds.
Figure 1. Total ion chromatograms from interference studies for five different amphetamine type stimulant derivatives. (Amph ¼ amphetamine, Phent ¼ phentermine,
MA ¼ methamphetamine, Eph ¼ ephedrine, Meph ¼ mephedrone, Pseu ¼ pseudoephedrine, Methy ¼ methylone, Keta ¼ ketamine).
Mass Spectrometry 343
The method was not sensitive for ephedrine and pseudoephedrine at the spiked concentration and their underivatized
form could not be detected in any of the samples. The cause
could be low extraction efficiency. Their acyl derivatives could
only be identified using AA, MBTFA and PFPA.
No coelution problems were observed except in one case:
MDA could not be seperated effectively from 4-MEC if they
were derivatized with AA. However, they have no common
fragment ions and can be distinguished from each other.
In conclusion, AA is the best choice from these derivatizing
agents. HFBA and PFPA can be used, but TFA and MBTFA are not
recommended for extractive acylation. In addition, AA is the least
expensive reagent, so its use is not only time-saving but costsaving. Our future aim will be to validate a method using AA.
9.
10.
11.
12.
Acknowledgments
The authors would like to thank Ms Szilvia Mrázik for her
excellent technical assistance and Ágnes Kerner, Pharm.D. and
Dr. Robert Anderson for their technical advice.
13.
References
14.
1. Hungarian National Focal Point. 2009 national report to the
EMCDDA by the Reitox National Focal Point. http://www.emcdda.
europa.eu/html.cfm/index111966EN.html (January 2011).
2. Ujváry, I. (2000) Az amfetamin-tı́pusú drogok kultúrtörténete,
kémiája, farmakológiája és toxikológiája. Psychiatria Hungarica,
15, 641–687.
3. Nagy, J. (2003) évi Jelentés a magyarországi kábı́tószerhelyzetró´l.
http://www.szmm.gov.hu/main.php?folderID=901 (January 2011)
4. Drognet. http://drognet.uw.hu/statiszt/statisztika.htm (September
2011).
5. Hidvégi, E., Fábián, P., Hideg, Zs., Somogyi, G.P. (2006) GC-MS determination of amphetamines in serum using on-line trifluoroacetylation. Forensic Science International, 161, 119– 123.
6. Hidvégi, E., Hideg, Zs., Somogyi, G.P. (2008) Different reactivities of
amphetamines with N-methyl-bis(trifluoroacetamide) in heated gas
chromatographic injectors. Pharmazie, 63, 233– 234.
7. Miki, A., Katagi, M., Zaitsu, K., Nishioka, H., Tsuchihashi, H. (2008)
Development of a two-step injector for GC-MS with on-column
derivatization, and its application to the determination of
amphetamine-type stimulants (ATS) in biological specimens.
Journal of Chromatography B, 865, 25– 32.
8. Tzing, S.H., Ghule, A., Liu, J.Y., Ling, Y.C. (2006) On-line derivatization gas chromatography with furan chemical ionization tandem
344 Dobos et al.
15.
16.
17.
18.
19.
mass spectrometry for screening of amphetamines in urine.
Journal of Chromatography A, 1137, 76 –83.
El-Haj, B.M., Al-Amri, A.M., Hassan, M.H., Ali, H.S., Bin Khadem, R.K.
(2003) The use of cyclohexanone as a “derivatizing” reagent for
the GC-MS detection of amphetamines and ephedrines in seizures
and the urine. Forensic Science International, 135, 16– 26.
Valentine, J.L., Middleton, R. (2000) GC-MS identification of sympathomimetic amine drugs in urine: Rapid methodology applicable
for emergency clinical toxicology. Journal of Analytical
Toxicology, 24, 211– 222.
Namera, A., Yashiki, M., Liu, J., Okajima, K., Hara, K., Imamura, T.
et al. (2000) Simple and simultaneous analysis of fenfluramine,
amphetamine and methamphetamine in whole blood by gas
chromatography-mass spectrometry after headspace-solid phase
microextraction and derivatization. Forensic Science International,
109, 215– 223.
Huang, M.K., Liu, C., Huang, S.D. (2002) One step and highly sensitive headspace solid-phase microextraction sample preparation approach for the analysis of methamphetamine and amphetamine in
human urine. Analyst, 127, 1203– 1206.
Lee, M.R., Song, Y.S., Hwang, B.H., Chou, C.C. (2000) Determination
of amphetamine and methamphetamine in serum via headspace
derivatization solid-phase microextraction-gas chromatographymass spectrometry. Journal of Chromatography A, 896,
265– 273.
Namera, A., Yashiki, M., Kojima, T., Ueki, M. (2002) Automated
headspace solid-phase microextraction and in-matrix derivatization
for the determination of amphetamine-related drugs in human
urine by gas chromatography-mass spectrometry. Journal of
Chromatographic Science, 40, 19– 25.
Jurado, C., Giménez, M.P., Soriano, T., Menéndez, M., Repetto, M.
(2000) Rapid analysis of amphetamine, methamphetamine, MDA,
and MDMA in urine using solid-phase microextraction, direct
on-fiber derivatization, and analysis by GC-MS. Journal of
Analytical Toxicology, 24, 11 –16.
Kataoka, H., Saito, K. (2011) Recent advances in SPME techniques
in biomedical analysis. Journal of Pharmaceutical and
Biomedical Analysis, 54, 926– 950.
Kankaanpää, A., Gunnar, T., Ariniemi, K., Lillsunde, P., Mykkänen, S.,
Seppälä, T. (2004) Single step procedure for GC MS screening and
quantitative determination of amphetamine-type stimulants and
related drugs in blood, serum, oral fluid and urine samples. Journal
of Chromatography B, 810, 57 –68.
Marais, A.A.S., Laurens, J.B. (2009) Rapid GC-MS confirmation of
amphetamines in urine by extractive acylation. Forensic Science
International, 183, 78– 86.
Paul, L.D., Mußhoff, F. (2009) Richtlinie der GTFCh zur
Qualitätssicherung bei forensisch-toxikologischen Untersuchungen.
Toxichem Krimtech, 76, 142–176.