Journal of Analytical Toxicology, Vol. 24, September 2000
Determination of Azide in Bloodand Urine by
Gas Chromatography-MassSpectrometry
Shigetoshi Kage 1, Keiko Kudo z, and Noriaki Ikeda z,*
1ForensicScience Laboratory, Fukuoka Prefectural Police Headquarters, 7-7, Higashikoen, Hakata-ku, Fukuoka 812-8576, Japan
and 2Departmentof Forensic Pathology and Sciences, Graduate School of Medical Sciences, Kyushu University, 3-1-1, Maidashi,
Higashi-ku, Fukuoka 812-858Z Japan
Abstract
A sensitive and simple method for determining azide
in blood and urine using an extractive alkylation technique was
devised. This inorganic anion was alkylated with pentafluorobenzyl
bromide usingtetradecyldimethylbenzylammonium chloride as the
phase-transfer catalyst. 1,3,5-Tribromobenzene was used as an
internal standard. The obtained derivative was analyzed by gas
chromatography-mass spectrometry using the negative ion chemical
ionization mode with isobutane as the reagent gas. The calibration
curves for azide were linear over the concentration range from 1 to
200 nmol/mL in blood and urine, and the lower limit of detection
was 0.5 nmol/mt for blood and urine. The accuracy and precision
of the method were evaluated, and the coefficients of variation were
found to be lower than 10%.
lack specificity and involve tedious procedures. Therefore, a more
specific and simpler method is needed for forensic toxicological
examination. Furthermore, the toxicity of sodium azide is
reported to be similar to that of cyanide (12,13), and a few cases
(8,9,14) in which cyanide levels have increased in the presence of
sodium azide have been reported. Therefore, it would be convenient if the method for determining azide could also detect
cyanide. A sensitive and simple method which could determine
cyanide and thiocyanate in blood by gas chromatography with
electron capture detection (GC-ECD) and gas chromatography-mass spectrometry (GC-MS) using an extractive alkylation technique was developed previously (15). Using this
technique, determination of azide in blood and urine was
attempted.
Experimental
Introduction
Reagents
Sodium azide, a highly toxic chemical, has been mainly used as
a bactericide in laboratories. Currently, however, it is also used in
industry as the propellant of automobile air bags and aircraft
escape chutes. Sodium azide is rapidly absorbed into the body
after ingestion, but the mechanism of action in the body is not
known. It produces hypotension (1). Although the biological
effects of azide are similar to those of cyanide, its lethality does
not appear to be due to inhibition of cytochrome oxidase (2). The
LDs0of this compound taken orally by white mice is 27 mg/kg (3).
Many poisoning cases have occurred within the last 3 years in
Japan in which drinks such as coffee or tea were contaminated
with azide for criminal purposes. When a clinical and/or criminological suspicion of azide poisoning exists, prompt identification
of this compound is essential.
Several methods have been reported for the determination of
azide, and these have used spectrometry (4), high-performance
liquid chromatography (5-9), and ion chromatography (10,11).
The identification of a compound by such methods, however, is
based on the retention time or the absorbance, but these methods
* Author to whom correspondenceshouFdbe addressed.E-mail: [email protected].
Oxygen-free water was used throughout this study and was prepared by bubbling nitrogen into distilled water for 15 min.
Sodium azide and internal standard (I.S.) 1,3,5-tribromobenzene (TBB)were purchased from Wako Pure Chemical Industries
(Osaka, Japan). An alkylating agent, pentafluorobenzyl bromide
(PFBBr), was purchased from Aldrich (Milwaukee, WI).
Tetradecyldimethylbenzylammoniumchloride (TDMBA) used as
the counter ion, was purchased from Tokyo Kasei Kogyo (Tokyo,
Japan).
TBB (31.5 rag) was dissolved in 100 mL of ethyl acetate to give
a concentration of lmM. A solution of I.S. was prepared by
diluting lmM TBB ethyl acetate solution with ethyl acetate to a
concentration of 10pM. PFBBr (261 mg) was dissolved in 50 mL
of ethyl acetate to give a concentration of 20mM. TDMBA(92 rag)
was dissolved in 50 mL of oxygen-free water saturated with
sodium tetraborate to give a concentration of 5raM. The other
reagents used were analytical grade.
A standard solution of azide (10 pmol/mL) was prepared by dissolving 65 mg of sodium azide in 100 mL of distilled and deionized water. This solution was further diluted with distilled and
deionized water to prepare 1-, 0.1-, and 0.01-pmol/mL solutions.
Reproduction (photocopying) of editorial content of this journal is prohibited without publisher's permission.
429
Journal of Analytical Toxicology, Vol. 24, September 2000
Preparation of whole blood and urine samples
room temperature, maintained at 60~ in a water bath for 30
rain, and then centrifuged at 1400 xg for 15 rain. A 1-~L aliquot
of the supernatant was injected into the GC-MS apparatus.
The alkylation of azide is explained by the following formula:
Samples to be tested were prepared by adding the standard
solution to blood and urine, both of which were collected from a
healthy volunteer. For the analysis of azide, 0.2-mL samples of
blood and urine were directly submitted to the extractive alkylation procedure without any deproteinization steps.
R-Br + N3- --> R-N3 + Br-
Alkylation procedure
R- = pentafluorobenzyl-
A 0.5-mL volume of 20raM PFBBr solution in ethyl acetate,
2.0 mL of 10pM I.S. solution in ethyl acetate, and 0.8 mL of 5mM
TDMBA solution in oxygen-free water saturated with sodium
tetraborate were put into a 10-mL volume glass-stoppered test
tube for extractive alkylation. To this mixture was added the
sample, after which the preparation was vortexed for 1 min at
5
10
GC-MS conditions
GC-MS was carried out on a Hewlett-Packard HP 5790A GC
(Palo Alto, CA) interfaced to a JEOL AX505AMS (Tokyo, Japan).
The column was a J&W Scientific (Folsom, CA) fused-silica capillary tube of DB-225 (30 m x 0.32-mm i.d., 0.25-Urn film thickness). A splitless injection mode was selected with
a valve off time of 1.5 rain. The initial temperature
of the column was held at 50~ for 3 min and then
programmed to rise at ]0~
to 220~ The
injection port, separator, and ion source were kept
at 220, 200, and 220~ respectively. Helium was
used as the carrier gas at a flow-rate of 2 mL/min.
The ionization energy of the positive-ion electron
ionization (EI) condition was 70 eV. The ionization energy and reagent gas for the negative-ion
chemical ionization (CI) conditions were 200 eV
and
isobutane, respectively. Scan mode was used
blood
for both identification and quantitation of azide.
15
Time (min)
B
Spiked
1
I
I'
['1'
I
5
'1
1 '
I'
I '
'
I
'
I'
10
Time (min)
I'
.,.],
I ' 1 '
15
I'
I'
Preparationof calibrationgraphs
I
'
I ''~
Figure1. Total ion chromatograms of the derivatized extracts obtained from the blank blood (A) and
the blood containing 50 nmol/mL of azide (B) using the negative-ion CI mode of GC-MS. Peak identification: 1, the derivative of azide and 2, TBB (I.S.)
181
100501
]~
Positive
El
I
Results
223 M§
Analysis by GC-MS
W
~C
100
_=
200
42 N 3
0 "J
300
1
"
400
rn/z
so
" =-/ -;
100
; :
F
:-Ill " ;
200
500
Negative CI
-if=
. . . . . . .
300
I "
400
' ' '" " I
500
m/z
Figure2. Mass spectra of the derivative of azide using the positive-ion
modes of GC-MS.
430
Blood and urine samples were prepared to contain azide at concentrations of 1-200 nmol/mL.
These samples were extracted and derivatized in
the same manner as described here previously.
Calibration curves for azide were obtained by
plotting the peak-area ratio of the base peak ion,
N3- (m/z 42), of the derivative of azide to the base
peak ion, Br- (m/z 79), of TBB against the azide
concentration using mass chromatography.
El and the negative-ion CI
Total ion chromatograms of the derivatized
extracts from the blank blood and the blood containing 50 nmol/mL of azide using the negativeion CI mode are shown in Figure 1. Sharp and
symmetrical peaks of the derivative of azide and
TBB (I.S.) were observed, with retention times of
8.82 and 14.00 rain, respectively. Mass spectra of
the derivative of azide using positive-ion El and
negative-ion CI modes are shown in Figure 2.
Using the positive-ion EI mode, the molecular ion
of the derivative of azide was observed at m/z 223,
and the fragment ion was observed at m/z 181 [MN3]+. Using the negative-ion CI mode, the base
Journalof Analytical Toxicology,Vol. 24, September 2000
peak ion of the derivative of azide was Ng- at m/z 42, and another
ion was observed at m/z 167 [M-CH2Ng]-.The mass spectral pattern indicated that the obtained derivative was pentafluorobenzyl
azide. In the case of TBB, the most abundant ion was observed at
m/z 314 [M+2]§ and a fragment ion was observed at m/z 235
[M+2-Br]§ using the positive-ion EI mode. Using the negative-ion
CI mode, the base peak ion of TBB was observed at m/z 79 [Br]-.
Mass chromatograms of the derivatized extracts from the blood
containing 10 nmol/mL of azide using the negative-ion CI mode
are shown in Figure 3.
The calibration curves were linear within the concentration
range from 1 to 200 nmol/mL for azide in blood and urine, with
correlation coefficientsof 0.996 and 0.997, respectively.
The recoveries of azide in the blood and urine at three different
concentrations, 10, 50, and 100 nmol/mL, were determined by
comparing the peak-area ratios of N3- ion of pentafluorobenzyl
azide, to Br- ion of TBB in samples with those in water samples,
using mass chromatography. The gross recovery of azide from the
blood and urine was 50% and 90%, respectively.The lower limit
of detection for azide in blood and urine, based on a concentration
giving a signal three times stronger than the average noise intensity, was approximately 0.5 nmol/mL. Within-day precisions were
obtained using two different concentrations (10 and 100
nmol/mL) by adding azide to blank blood and urine. The coefficients of variation ranged from 2.6 to 9.6% (Table I).
Discussion
GC-MS analysis is a superior technique for the identification of
this compound for forensic examinations. Because a GC-MS
method for the analysis of azide was not found, the established
method for cyanide determination to azide analysis was applied.
Azide in blood and urine was directly alkylated in 30 rain, without
any deproteinization steps. The sensitivity of the derivative in
GC-MS was 100 times higher in the negative-ion CI mode than in
the positive-ion El mode, because the derivative contains five fluorine atoms. Therefore, both identification and quantitation of
azide were carried out using the negative-ion CI mode. In this
mode, no interferences were observed even when the ions of low
molecular weight were used for quantitation. The detection limit
of azide in blood was 0.5 nmol/mL with this method, which was
better than with other methods (8,9,11). The reported detection
limits of azide in blood using other methods were 80 ng/mL (1.9
nmol/mL) and 200 ng/mL (4.8 nmol/mL) with high-performance
liquid chromatography (8,9) and 30 ng/mL (0.7 nmol/mL) with
ion chromatography (11). Several cases of intoxication with
sodium azide have been reported in the literature (8,9,
12,14,16--21). The concentrations of azide in blood in four fatal
cases were reported to range between 7.4 (176 nmol/mL) and 262
tJg/mL (6240 nmol/mL) (9). The concentrations of azide in blood
and urine in nonfatal cases have not been reported; however, the
concentration of azide in antemortem urine 3.5 h after ingestion
in a fatal case was reported to be 135 ng/mL (3.2 nmol/mL) (20).
Therefore, the present method showing a lower detection limit of
0.5 nmol/mL would be able to detect azide in blood and urine
even in nonfatal cases.
Because this method is a modification of a previous method
designed for the analysis of cyanide and thiocyanate (15), the posm/z
sibility of detecting these compounds using the present method
was examined. Thiocyanate in blood and urine was well alkylated
and showed a sharp peak at a retention time of 15.18 min on the
[
"1.0
79
chromatogram. The base peak ion of the derivative of thiocyanate
..-1
was observed at m/z 58 [SCN]- using the negative-ion mode.
Therefore, the determination of azide and thiocyanate in blood
and urine can be carried out simultaneously. Azide and thiocyanate were also derivatized after precipitating protein in blood
with acetone or acetonitrile, but trichloroacetic acid did not work
as well. Derivatization of cyanide, however, was successful only
19
5
10
15
after precipitation with trichloroacetic acid. Therefore, detection
Time (rain)
of cyanide was carried out after protein was precipitated with 10%
ice-cold trichloroacetic acid and the sample submitted to the preFigure 3. Masschromatogramsof the derivatizedextractobtained from blood
sent procedure. A peak of cyanide was observed with a retention
containing 10 nmol/mL of azide usingthe negative-ionCI mode of GC-MS.
time of 12.57 min, and the base peak ion of the derivative of
Peak identification: 1, the derivative of azide and 2, TBB (I.S.).
cyanide was at rn/z 206 [M-H]- using the negative-ion C] mode.
Therefore, this method has an additional advantage in that cyanide and thiocyanate can also be
Table I. Accuracy and Precision of Azide Determination in Blood and Urine
detected.
N 3 added
Detected (nmol/mt) Coefficient of
In conclusion, a sensitive and simple method
variation (%)
Sample
(nmol/mt)
n
Mean
SD
for analyzing azide in blood and urine using an
extractive alkylation technique combined with
Blood
10
5
9.2
0.9
9.6
GC-MS using the negative-ion CI mode was
100
5
109.2
7.3
6.7
devised. Because this method can also determine
Urine
10
5
9.8
0.6
6.1
cyanide and thiocyanate in blood and urine, it
100
5
103.7
2.7
2.6
should be useful for diagnosing azide and/or
,,_l_L
~ -
42
,
,
'
'
- . d ~ _ _ ~ J
1
'
'
'
, z - L . ~ , l l
'
I
. . .
h.L
.
a..|--.d,,J~...I,.~-~!
'
'
'
't
'
431
Journal of Analytical Toxicology,Vol. 24, September2000
cyanide poisoning in nonfatal cases as well as in fatal cases.
Acknowledgments
We thank Ms. K. Miller (Royal English Language Centre,
Fukuoka, Japan) for revising the English used in this manuscript.
References
1. M.M. Black, B.W. Zweifach, and F.D. Speer.Comparison of hypotensive action of sodium azide in normotensive and hypertensive
patients. Proc. Soc. Exp. Biol. A4ed. 85:11-16 (1954).
2. R.R Smith, C.A. Louis, R. Kruszyna, and H. Kruszyna. Acute neurotoxicity of sodium azide and nitric oxide. Fundam. AppL Toxicol. 17:
120-127 (1991).
3. J.D.P. Graham. Actions of sodium azide. Br. J. Pharmacol. 4:1-6
(1949).
4. E.A. Terpinski. Spectrophotometric determination of sodium azide.
Analyst 110:1403-1405 (1985).
5. S.J.Swarin and R.A. Waldo. Liquid chromatographic determination
of azide as the 3,5-dinitrobenzoyl derivative. J. Liq. Chromatogr. 5:
597-604 (1982).
6. J. Vacha, M. Tkaczykova, and M. Rejholcova. Determination of
sodium azide in the presenceof proteins by high-performance liquid
chromatography. J. Chromatogr. 488:506-508 (1989).
7. M.C. Gennaro, C. Abrigo, E. Marengo, and A. Liberatori. A selective
determination of azide by ion-interaction reversed-phase HPLC.
J. Liq. Chromatogr. 16:2715-2730 (1993).
8. W.E. Lambert, M. Piette, C. Van Peteghem, and A.P. De Leenheer.
Application of high-performance liquid chromatography to a fatality
involving azide. J. Anal. Toxicol. 19:261-264 (1995).
432
9. P. Marquet, S. Clement, H. Lotfi, M.F. Dreyfuss, J. Debord,
D. Dumont, and G. Lach,~tre. Analytical findings in a suicide
involving sodium azide. J. Anal. Toxicol. 20:134-138 (1996).
10. P.L.Annable and L.A. Sly. Azide determination in protein samples by
ion chromatography. J. Chromatogr. 546:325-334 (1991).
11. R. Kruszyna, R.P.Smith, and H. Kruszyna. Determining sodium azide
concentration in blood by ion chromatography. J. Forensic Sci. 43:
200-202 (1998).
12. I.E. Albertson, S. Reed, and A. Siefkin. A case of fatal sodium azide
ingestion. Clin. Toxicol. 24:339-351 (1986).
13. R.P. Smith and D.E. Wilcox. Toxicology of selected nitric oxidedonating xenobiotics, with particular reference to azide. Crit. Rev.
ToxicoL 24:355-377 (1994).
14. W. Klein-Schwartz, R.L. Gorman, G.M. Oderda, B.P. Massaro, T.L.
Kurt, and J.C. Garriott. Three fatal sodium azide poisonings. Med.
ToxicoL Adverse Drug. Exp. 4:219-227 (1989).
15. S. Kage,T. Nagata, and K. Kudo. Determination of cyanide and thiocyanate in blood by gas chromatography and gas chromatography-mass spectrometry. J. Chromatogr. B 675:27-32 (1996).
16. S.G.N. Richardson, C. Giles, and C.H.J. Swan. Two cases of sodium
azide poisoning by accidental ingestion of Isoton. J. Clin. PathoL 28:
350-351 (1975).
17. E.A. Emmett and J.A. Ricking. Fatal self-administration of sodium
azide. Ann. Intern. Med. 83:224-226 (1975).
18. O.R Edmonds and M.S. Bourne. Sodium azide poisoning in five laboratory technicians. Br. J. Ind. Med. 39:308-309 (1982).
19. J. Abrams, R.S. EI-Mallakh, and R. Meyer. Suicidal sodium azide
ingestion. Ann. Emerg. Nled. 16:1378-1380 (1987).
20. J.D. Howard, K.J. Skogerboe, G.A. Case, V.A. Raisys, and E.Q.
Lacsina. Death following accidental sodium azide ingestion.
J. Forensic Sci. 35:193-196 (1990).
21. N. Singh, C.P. Singh, and G.K. Brat. Sodium azide~a rare poisoning.
J. Assoc. Physicians India 42:755-756 (1994).
Manuscript received December 3, 1999;
revision received February 28, 2000.