Journal of Analytical Toxicology, Vol. 30, October 2006
Cyanideand Thiocyanatein Human Salivaby
Gas Chromatography-MassSpectrometry
Buddha D. Paul and Michael t. Smith
Division of Forensic Toxicology, Office of the Armed Forces Medical Examiner, Armed Forces Institute of Pathology,
Rockville, Maryland 20850
r Abstract ]
A method is described for simultaneous determination of cyanide
(CN) and thiocyanate (SCN) in human saliva, or oral fluid. SCN
concentrations in body fluids appeared to be important in
classifying patients as smokers or nonsmokers, in determining
some clinical conditions, and in specimen validity testing in
forensic drug testing. The human saliva samples were diluted and
the anions were separated by an extractive alkylation technique.
Tetrabutylammonium sulfate was used as phase-transfer catalyst
and pentafluorobenzyl bromide as the derivatizing agent. The
products were analyzed by a gas chromatography-mass
spectrometry (GC-MS) with selected ion monitoring method.
2,5-Dibromotoluene was used as internal standard for quantitation
of CN and SCN in saliva. The calibration plot was linear over the
concentration range from 1 to 100 pmol/t (0.026-2.60 pg/mL) for
CN (R = 0.9978) and 5 to 200 pmol/L (0.29-11.6 pg/mL) for SCN
(R = 0.9996). The method was used to examine 10 saliva
specimens. The concentration ranged from 4.8 to 29 pmol//
(0.13-0.75 pg/mL) for CN and 293 to 1029 pmol/L (17-59.7
pg/mL) for SCN. The SCN results were similar to those obtained
from a method using oxidation of SCN to CN with colorimetric
detection (R = 0.9882). The proposed GC-MS confirmatory
method was found useful when the concentrations of CN and SCN
in saliva needed to be accurately determined.
surements of CN and SCN are of growing importance in monitoring some clinical conditions and distinguishing tobacco
smokers from nonsmokers.
Many colorimetric methods have been reported for determination of SCN (1,3,6,13-16) and CN in blood and plasma
(17,18). The colorimetric methods for detecting SCN are based
on reaction between SCN and ferric nitrate as a chromogenic
reagent (16) and the Konig reaction (19). Ion chromatography
(IC) (20,21), gas chromatography (GC) (2,22-25), and highperformance liquid chromatography (HPLC) (26) were also
used to detect SCN and CN in blood. SCN in saliva has been investigated by color reaction (4), HPLC (27), flow injection
(28), IC (29,30), fourier transform infrared spectrometry (31),
and capillary zone electrophoresis (32,33). In one procedure,
CN and SCN were separated by a gas-phase diffusion technique, transforming the ions to the corresponding chlorides by
chloramine-T, and detecting the products by a GC-MS method
(34). The procedure required careful handling of volatile components and had several reaction steps (CN-HCN-CNC1).We report here a GC-MS procedure for simultaneous determination
of CN and SCN in saliva, or oral fluid, after an extractive alkylation using a phase-transfer catalyst. We also present the effect of the phase-transfer catalysts on conversion of SCN to CN.
Materialsand Methods
Introduction
Thiocyanate (SCN) is usually present in serum, saliva, and
urine in low concentration (1--4). It is the principal metabolic
product of cyanide (CN) metabolism. High concentration may
be due to HCN ingestion from tobacco smoke and metabolic
conversion to SCN (5,6). When sodium nitroprusside is used to
treat a variety of cardiovascular conditions, SCN in plasma
may increase to a potentially toxic level (7). In contrast, a low
concentration of 8CN in plasma of a tobacco smoker may confirm a toxic amblyopia (8,9). Other clinical disorders like ataxic
neuropathy (10), tropical diabetes (11), and endemic goiter and
cretinism (12) may be due to elevated level of SCN in body
fluids from ingestion of cyanogenic glycosidesin cassava. Mea-
Chemical, reagents, and supplies
Pentafluorobenzyl bromide (PFB-Br), 2,5-dibromotoIuene
(2,5-DBT, internal standard), tetrabutylammonium sulfate
(TBAS), sodium cyanide, and ammonium thiocyanate were
purchased from Sigma-Aldrich (Milwaukee,WI). All solvents
and reagents were of analytical or HPLC grade.
Equipment
A GC-MS system, consisting of a 6890N GC and 5975 mass
selective detector (MSD)from Agilent Technologies (Palo Alto,
CA) was used. Helium was used as the carrier gas. The head
pressure on the capillary column (15 m x 0.25-ram i.d. x 0.20
]Jm, 5% phenyl polysiloxane,J&W Scientific, Rancho Cordova,
Reproduction(photocopying)of editorialcontentof thisjournal is prohibitedwithoutpublisher'spermission.
5'[ I
Journal of Analytical Toxicology, Vol. 30, October 2006
CA)was 8-10 psi. The instrument was operated in splitless and
temperature program modes. To avoid peak tailing, the purge
valve was turned on at 0.50 rain after sample injection. The
MSD was operated in electron impact mode at 70 eV. The detector was initially off and was turned on after a strong reagent
peak of PFB-Br (approximately3.0 rain).
Preparation of standards and reagents
Stock solutions of CN and SCN were prepared at a concentration of 10 mmol/L in water. Working solutions of CN/SCN
at concentrations of 0, 25, 50, and 100 IJmol/L in water were
used for specimen analysis. An SCN control (200 IJmol/L in
water) was used to monitor artifact formation of CN. The
phase-transfer catalyst (TBAS, 10 mmol/L) was prepared in a
saturated solution of sodium borate (pH 9.2). Derivatizing
agent (PFB-Br) and internal standard (2,5-DBT)were prepared
in ethyl acetate at concentrations of 20 mmol/L and 50 ]Jmol/L,
respectively. All solutions were stored at 2-5~ except the
TBAS solution, which was left at room temperature. The
reagents were stable for at least six months. For K6nig color
reaction, 0.6 g of barbituric acid was dissolved in 10 mL of a
mixture of pyridine and 1M hydrochloric acid (3:7, v/v). The
solution had a short shelf-life and was prepared before the
analysis.
Preparation of saliva samples
Ten unstimulated saliva (oral fluid) specimens were collected from three male (M1, M2, M3) and two female (F1, F2)
nonsmoking subjects who expectorated directly into
polypropylene collection tubes. The waiting time was at least
a week when collected from the same subject. Approximately
3 mL of saliva was centrifuged at RCF 7300 xg for 30 rain. For
analysis, 0.5 mL of the clear solution was diluted to 5 mL with
water for a 10-fold dilution.
K~nig color test for saliva specimens
The procedure was the same as that used to test CN and SCN
in blood with minor modifications (18). A solution of 0.8 mL
of perchloric acid (0.33 tool/L) was added to 0.2 mL of standards (SCN 0, 25, 50, 100, 150 tJmol/L) and saliva specimens.
After 2 rain of oxidation of SCN to CN, 0.2 mL of sodium acetate (1M) and 60 IJL of sodium hypochlorite (50 mmol/L)
were added to the solutions. The so]ution was mixed, and
within a minute, 0.2 mL of barbituric acid-pyridine reagent
was added. After 5 rain, the absorbance was recorded at 583
nm. The instrument was set to zero using a water blank prior
to analysis of specimens.
Resultsand Discussion
In extractive alkylation, CN and SCN in an aqueous solution
reacted poorly with PFB-Br in ethyl acetate. But with the use
of TBAS as a phase-transfer catalyst the reaction began immediately (CN ~ PFB-CN, SCN ~ PFB-SCN).After 30 min of
reaction, the total ion chromatogram (TIC)areas were greater
than 177,000 and 25,000 after an on-column injection of 5 picomol (0.29 ng) of SCN and 12.5 picomol (0.33 ng) of CN, respectively.In both cases the backgroundwas less than 1%. The
reactions were studied over a period of 120 rain. When monitored by ion/IS ratios, the reactions were almost complete
after 90 rain for both CN and SCN.Becauseof the excellentoncolumn sensitivity,we ran the reaction for only 30 rain. At this
reaction time, the yield was 55-65% for both CN and SCN. In
GC-MS analysis, three ions from PFB-CN in one group and
Table I. Cyanide and Thiocyanate (pmol/L) in Ten Saliva
Specimens Collected from Three Male (M) and Two
Female (F) Nonsmoking Subjects*
Extractive alkylation procedure
Solutions of 0.3 mL of PFB-Br (20 mmol/L), 0.2 mL of 2,5DBT (50 lJmol/L), and 0.8 mL of TBAS (10 mmol/L) were
added to 0.2 mL of standards (CN/SCN0, 25, 50, 100 ]Jmol/L),
control (SCN200 IJmol/L),and saliva specimens in 5-mL glass
centrifuge tubes. The tubes were capped, vortex mixed for 1
min, and heated at 55~ for 30 rain in a water bath. The solutions were centrifuged at RCF 1700 x g for 5 rain, and approximately 100 IJL of the clear ethyl acetate solutions was
transferred into screw-capped injection vials.
GC-MS analysis
The oven temperature was increased from 60~ (held for 1.0
rain) to 90~ at 10~
and then to 150~ at 30~
Both injector and detector were set at 200~ Ions monitored
were as follows:m/z 207, 188, and 157 for PFB-CN; m/z 239,
181, and 161 for PFB-SCN; m/z 250 and 169 for 2,5-DBT (IS).
Electron multiplier was set 200 V above autotune. To enhance
sensitivity, PFB-CN in one group and PFB-SCN and IS in another group were monitored. Approximately 1-3 IJL of the
ethyl acetate solution was injected onto the instrument. The
retention times for PFB-CN, PFB-SCN,and 2,5-DBTwere 3.37,
5.16, and 5.41 rain, respectively.
512
Colorimetric
Method
GC-MS Method
Saliva
CN
SCN
(CN + SCN)
pmol/L
CN/SCN (CN + SCN)
%
pmol/L
M1-1
M1-2
M1-3
M1-4
M1-5
M2-1
M3-1
F1-1
F1-2
F2-1
29
11
12
8.2
5.7
4.8
8.8
6.9
15
17
794
442
741
1029
595
293
463
726
771
573
823
453
753
1037
601
298
472
733
786
590
3.7
2.5
1.6
0.8
1.0
i .6
1.9
1.0
1.9
3.0
929
555
825
1056
634
323
494
797
815
615
Min
Max
Median
Average
Std
4.8
29
9.9
11.8
7.2
293
1029
661
643
213
298
1037
667
655
215
0.8
3.7
1.8
1.9
0.9
323
1056
716
704
220
* Method comparison for (CN + SCN), intercept = -26.7, slope = 0.9673,
r = 0.9882.
Journal of Analytical Toxicology, Vol. 30, October 2006
three ions from PFB-SCN and two ions from IS in another
group were monitored. The identification of the compounds
was based on comparing retention times (• 2%) and relative
ion abundances (• 20%) with the reference compounds. Excellent linearities were observed over the concentration range
of i to 100 lamol/L (0.026-2.60 pg/mL) for CN and 5 to 200
lamol/L (0.29-11.6 lag/mL) for SCN. The intercept, slope, and
correlation coefficientwere -0.0073, 0.9933, and 0.9978 for CN
and 0.0695, 0.0.9989, and 0.9996 for SCN, respectively. Interrun variations of ion/IS ratios were less than 11% (N = 5).
The procedure was used to examine 10 saliva specimens
collected from 3 male (M) and 2 female (F) subjects. Saliva in
our study is used synonymouslywith oral fluid to remain consistent with terminology in past publications. The specimens
were initially centrifuged, and the clear solutions diluted 10-
fold before testing. The results are recorded in Table I. The
method was verified by testing the same specimens by a colorimetric method used to test SCN and CN in blood (18). In
the procedure the SCN was oxidized to CN and tested by the
Kgnig procedure. Therefore, the result indicated total of CN
and SCN. The total CN and SCN values between the two
methods compared favorably (r = 0.9882). In all cases, the
values in GC-MS were less than those of the colorimetric
method (intercept = -26.7 pmol/L). This differencemay be due
to an increase in background absorbance from the saliva matrix. The SCN concentrations found in our experiments are
similar to those published in the literature (TableII). Considerable variation was observed in specimens collected from the
same subject (M1, 442-1029 pmol/L). Generally, the amount
of CN in all specimens was low (average 11.8 pmol/L) and
varied in the range of 0.8-3.7% of SCN. The ion chromatogram
of specimen M1-5 is presented in Figure 1.
Table II. Literature Reported Thiocyanate Concentrations
in Saliva from Nonsmokers
Linearity and precision of CN and SCN were verified in
saliva. A sample (F2-1)was tested at 5, 10, 15, 20-fold dilutions.
SCN mmol/L
The corresponding final concentrations were 20.0, 16.5, 15.4,
and 15 t~mol/L (average 16.7 _+ 14%) for CN and 600, 573,
Reference
Methods
Average-+ SD
Range
N
597, and 607 lJmol/L (average594 • 2.5%) for SCN. Two more
4
Colorimetric*
0.54 + 0.41
20
specimens also gave similar results (M3-1, CN 7.9 + 11%, SCN
30
IC
0.51 + 0.31
10
477 + 3.9%, and F1-2, CN 11.4 + 1.5%, SCN 827 ___9.2%). For
31
FT-IR
0.83 + 0.42
0.41-1.25
25
precision, M1-5 was tested in five replicates. The average +
32
CZE
0,23-1.20
48
CV% were 5.7 • 11.6% for CN and 595 • 4.3% for SCN. It apPresent method
GC-MS
0.66 + 0.22
0,30-1.04
10
peared that salivaryconcentrations of CN and SCN determined
Present method Colorimetric
0.70 + 0.22
0.32-1,06
10
by this method were linear and reproducible.
Cyanide was reported as an artifact when SCN was deriva* For colorimetricmethod, the amounts represent(CN + SCN).
tized under a basic condition (35). A basic condition was also
used to determine CN and SCN in blood (25).
In this method, SCN was derivatized in presence of benzyldimethyltetradecylammonium
5.16 PFB-SCN
chloride (BDM-TDA)as a phase-transfer cata3+§
to.~9.~
l,ge+0
lyst. We studied the BDM-TDAmethod and
l+Fo
monitored the artifact formation of CN from
1.6e+0
SCN. Solutions of SCN (0.8 mL, 25-200
4,+
1o. t l t ~
pmol/L) were mixed with 0.3 mL of PFB-Br
1.4r
,.+..*++
9
(20 retool/L), 0.2 mL of 2,5-DBT (IS, 50
1Jmol/L), and 0.8 mL of BDM-TDA(5 mmol/L
5.41 I5
1.2e+0
in saturate sodium borate in water, pH 9.2)
and heated at 55~ for 30 min. In the GC-MS
lc+07
chromatogram, a peak of CN was observed.
The
artifact CN concentrations were 7.2-9.2%
. . . . .
72
~c+06
....
t ...... t+
of SCN (average 8.5 _+9%), compared to only
~lxii
t 1+4
n.,,r
0-0.34% by the TBAS method. Moreover,
6c+06
when tested by the BDM-TDAmethod, the
amount of CN increased with increasing time
4e+06
of reaction. No significant change was ob%
served
with the TBAS method. It indicated
3.1s i';-/ '~s ' '3':C L~i
that the small amount of CN detected by the
3.37 PFB-CN
TBASwas not an artifact but an impurity pre+
A
sent in the commercial SeN. Artifact forma3.20
3.40
3.60
3.80
4.00
4.20
4,40
4.60
4.80
5.00
5.20
5.40
5.60
5,80
tion of CN from SeN was also monitored in
Time (min)
specimen analysis. When specimen F2-1 was
Figure 1. Total ion chromatogram (TIC)and selected ion monitoring (SIM)of CN as PFB-CN (RT
tested by the BMD-TDAmethod, the CN and
3.374 rain, 5.7 +stool/L)and SCN as PFB-SCN (RT 5.16 rain, 595 tJmol/L)from a saliva specimen
SeN concentrations were 86 and 514 pmol/L,
(M1-5). 2,5-Dibromotoluene was used as internal standard (IS, RT 5.41 rain).
respectively,compared to 17 and 573 pmol/L,
! /
+i ..........I
.......
.
.~.
+,
.
2e+05
513
Journal of Analytical Toxicology,Vol. 30, October 2006
respectively, by the TBAS method (Table I). Standard solutions
of CN and SCN were stable in water for at least 2 months. In
different batch analysis, the ratios of CN/IS or SCN/IS were
within + 20%. Both CN and SCN were stable in saliva stored at
2--4~ for 7 days. When six specimens were tested again, the results were within _+20% of the original values.
Three previously published studies found concentrations of
SCN in saliva of smokers to be 1655 +_841 IJmol/L (N = 20),
3620 _+ 1720 tJmol/L (N = 5), and 2050 _+450 lJmol/L (N = 3)
(4,28,33). These concentrations are much higher than those
found for nonsmokers in this study, 655 _+215 IJmol/L. In a
study of smokers with toxic amblyopia, the investigators found
that plasma SCN concentrations returned to the nonsmoker
range due to an inability to convert CN to SCN (8,9,15). Other
clinical abnormalities have also been reported in individuals
who ingest cassava and have elevated plasma levels of SCN
(10-12). It has also been proposed that identifying endogenous components of saliva, such as CN and SCN, could be
used to prove that specimens collected in drug testing programs are valid (36). One would expect that these various
groups could be distinguished based on CN and SCN concentrations in their saliva.
Conclusions
Both CN and SCN are present in body fluids in different
amounts. The validated GC-MS procedure presented here detects CN and SCN in saliva as low as 1.0 and 5.0 IJmol/L, respectively. In 10 specimens, the CN and SCN concentrations
ranged from 4.8 to 29 lJmol/L and 293 to 1029 IJmol/L, respectively. The CN concentration was 0.8-3.7% of the SCN
concentration. The saliva SCN concentrations found in this experiment were comparable with other literature procedures,
supporting validity of the GC-MS method. The procedure may
be useful in forensic drug testing when specimen validity
testing is required and also in classifying patients as smokers
or non-smokers.
References
1. G.E. Boxer and J.C. Rickards. Determination of thiocyanate in
body fluids. Arch. Biochem. Biophys. 39:292-300 (1952).
2. J.C. Valentour, V. Aggarwal, and 1. Sunshine. Sensitive gas chromatographic determination of cyanide. Anal. Chem. 46:924-925
(1974).
3. R Lundquist, J. Martensson, B. Sorbo, and S. Ohman. Method for
determining thiocyanate in serum and urine. Clin. Chem. 25:
678-681 (1979)
4. K. Tsuge, M. Kataoka, and Y Seto. Cyanide and thiocyanate in
blood and saliva of healthy adult volunteers. J. Health Sci. 46:
343-350 (2000)
5. RM. Densen, B. Davidow,H.E. Bass, and E.W.Jones. A chemical
test for smoking exposure. Arch. Environ. Health 14:865-874
(1976).
6. W.C. Butts, M Kuehneman, and G.M. Widdowson. Automated
method for determining serum thiocyanate to distinguish smokers
from nonsmokers. Clin. Chem. 20:1344-1348 (1974).
514
7. R.R Smith and H. Kruszyna. Nitroprusside propduces cyanide poisoning via a reaction with hemoglobin. J. Pharmacol. Exp. Ther.
191:557-563 (I974).
8. J. Wilson. Liber's hereditary optic atrophy: a possible defect of
cyanide metabolism. Clin. Sci. 29:505-515 (1965).
9. W.S. Foulds, J.M. Bronte-Stewart, and I.A. Chisholm. Serum thiocyanate concentration in tobacco amblyopia. Nature 218:586
(1968)
10. B.O. Osuntokun. Cassava diet, chronic cyanide intoxication and
neuropathy in Nigerian Africans. World Rev. Nutr. Diet 36:
141-173 (1981).
11. D.E. McMillan and P.J.Geevarghese. Dietary cyanide and tropical
malnutrition diabetes. Diabetes Care 2.' 202-208 (1979)
12. F. Delange, V.D. Velden, and A.M. Ermans. Evidence of an antithyroid action of cassava in man and in animals. In Chronic Cassava Toxicity, B. Nestel and R.M. Maclntyre, Eds. Int. Devel. Res.
Centre, Ottawa, Canada, 1973, pp 147-151.
13. C.J. Vesey and C.J.C. Kirk. Two automated methods for measuring plasma thiocyanate compared. Clin. Chem. 95:270-274
(1985)
14. N. Grgurinovich. A colorimetric procedure for the determination
of thiocyanate in plasma. J. Anal. Toxicol. 6:53-55 (1982)
15. A.R. Pettigrew and G.S. Fell. Simplified colorimetric determination of thiocyanate in biological fluids, and its application to investigation of the toxic amblyopias. Clin. Chem. 18:996-1000
(1972)
16. R.G. Bowler. The determination of thiocyanate in blood serum.
Biochem. J. 38:385-388 (1944)
17. P. Lundquist, H. Rosling, and B. Sorbo. Determination of cyanide
in whole blood, erythrocytes, and plasma. Clin. Chem. 31:
591-595 (1985)
18. P. Lundquist and B. Sorbo. Rapid determination of toxic cyanide
concentrations in blood. Clin. Chem. 35:617-619 (1989)
19. W.J. K6nig. Untersuchungen aus dem organischen Jaboratorium
der technischen hochschule zu dresen. LXlX. Uber eine neue,
vom pyridin deriveirende, klassevon farbstoffen. Prakt. Chem. 69'
105-137 (1904)
20. Y. Michigami, T. Takahashi, F. He, Y. Yamamoto, and K. Ueda. Determination of thiocyanate in human serum by ion chromatography. Analyst 113:389-392 (1988)
21. S. Chinaka, N. Takayama, Y. Michigami, K. Ueda. Simultaneous
determination of cyanide and thiocyanate in blood by ion chromatography with fluorescence and ultraviolet detection. J. Chromatogr. B 713:353-359 (1998)
22. J. Zamecnik and J. Tam. Cyanide in blood by gas chromatography
with NP detector and acetonitrile as internal standard. Application
on air accident fire victims. J. Anal. Toxicol. 11 : 47-48 (1987)
23. C. Maseda, K. Matsubara, and H. Shiono. Improved gas chromatography with electron-capture detection using a reaction precolumn for the determination of blood cyanide: a higher content
in the left ventricle of fire victims. J. Chromatogr. 490:319-327
(1989)
24. Y. Seto, N. Tsunoda, H. Ohta, and T. Shinohara. Determination of
blood cyanide by headspace gas chromatography with nitrogenphosphorus detection and using a megabore capillary column.
Anal. Chim. Acta 276:247-259 (1993)
25. S. Kage, T. Takeaki, 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)
26. A. SanD, N. Takimoto, and S. Takitani. High-performance liquid
chromatographic determination of cyanide in human red blood
cells by pre-column fluorescence derivatization. ]. Chromatogr.
582:13t-t35 (1992)
27. S. Tanabe, M. Kitahara, M. Nawata, and K. Kawanabe. Determination of oxidizable inorganic anions by high-performance liquid
chromatography with fluorescence detection and application to
the determination of salivary nitrite and thiocyanate and serum
thiocyanate. J. Chromatogr. 424:29-37 (1988)
Journal of Analytical Toxicology,VoL 30, October 2006
28. A.B. Bendtsen and E.H. Hansen. Spectrophotometric flow injection determination of trace amounts of thiocyanate based on
its reaction with 2-(5-bromo-2-pyridylazo)-5-diethylaminophenol and dichromate: assay of the thiocyanate level in
saliva from smokers and non-smokers. Analyst 116:647-651
I1991)
29. Y. Michigami, K. Fuiii, K. Ueda, and Y. Yamamoto. Determination
of thiocyanate in human saliva and urine by ion chromatography. Analyst 117:1855-I 858 (I 992)
30. Z.F. Chen, B.W. Darvell, and V.W.H. Leung. Human salivary anionic analysis using ion chromatography. Arch. Oral Biol. 49"
863-869 (2004)
31. C.P. Schultz, M.K. Ahmed, C. Dawes, and H.H. Mantsch. Thiocyanate levels in human saliva: Quantitation by fourier transform infrared spectroscopy. Anal. Biochem. 240" 7-I 2 (I 996)
32. u Tanaka, N. Naruishi, H. Fukuya, J. Sakata, K. Saito, and
33.
34.
35.
36.
S. Wakida. Simultaneous determination of nitrite, nitrate, thiocyanate, and uric acid in human saliva by capillary zone electrophoresis and its application to the study of daily variations.
J. Chromatogr. A 1051 : 193-197 (2004)
Z. Glatz, S. Novakova, and H. Sterbova. Analysis of thiocyanate
in biological fluids by capillary zone eJectrophoresis. J. Chromatogr. A 9T6:273-277 (2001)
I. Thomson and R.A. Anderson. Determination of cyanide and
thiocyanate in biological fluids by gas chromatographymass spectrometry. ]. Chromatogr. 188" 357-362 (1980)
H.L. Wu and W.R. Hwang. Electron-capture gas chromatographic
determination of cyanide as pentafluorobenzyl cyanide. J. Chromatogr. 284" 193-200 (1984)
Department of Health and Human Services. Proposed revisions to
mandatory guidelines for federal workplace drug testing programs. Fed. Regist. 69:19673-19731 (2004).
515