Journal of Analytical Toxicology, Vol. 16, January/February 1992
Solid-Phase Extractionand GC/MS Confirmationof Barbituratesfrom Human Urine R u g g e r o Pocci, V a n d a n a Dixit, and V y a s M. Dixit
Varian Samp/e Preparation Products, 24201 Frampton Avenue, Harbor City, Ca/ifomia 90710
Abstract A highly selective and sensitive procedure has been
developed for isolating and identifying barbiturates In human urine. With a new disposable bonded silica gel solid-phase extraction (SPE) column and hexobarbital as an internal standard (IS), amobarbital, butabarbital, pentobarbital, phenobarbital, secobarbital, and methaqualone were selectively isolated from endogenous urine components. Capillary gas chromatography/Ion trap mass spectrometry (GC/MS) analysis of the extracts generated a full mass spectrum for the detection, identification, and quantitation of barbiturates. Linear quantitative response curves for the drugs have been generated over a concentration range of 20-500 ng/mL. Overall extraction efficlencles for drugs averaged greater than 90%, and the quantitative response curves exhibited correlation coefficients of 0.996 to 0.999.
This article presents a sensitive, qualitative, and quantitative extraction procedure for amobarbital, butabarbital, pentobarbital, phenobarbital, secobarbital, and methaqualone using Bond Elut Certify IITM, a disposable solid-phase extraction column. Bond Elut Certify II is a chemically modified silica gel material bearing three different types of interactions: hydrophobic, polar, and ion exchange. The silica gel is specially designed for the extraction of acidic and neutral molecules from biological fluid and provides clean extracts of the barbiturates. This cleanliness is the result of the removal of some of the interferences during the column rinse process, while most of the impurities present in the urine are irreversibly retained on the column.
M a t e r i a l s and M e t h o d s
Introduction
Barbiturates, tile derivatives of urea, are formed by replacement of both hydrogen atoms on the carbon at position 5 by alkyl, aryl, or alicyclic groups. The potential clinical significance of barbital led to the extensive development of the barbituric acid class of drugs (1) in 1903. A definitive publication by Doran (2) in 1959 appears to be the only monograph providing useful information, including details of syntheses, chemical and physical properties, reactions, and pharmacology. Their action on the central nervous system (CNS) and the extent of CNS depression is dependent on the particular barbiturate. At low doses, they induce a feeling of relaxation and tranquility and can impair motor functions. Their rapid action tbr anesthetic purposes and intermediate action for sedative or sleeping purposes make them attractive to drug abusers (3). Overdoses of barbiturates depress the respiratory centers of the brain and can cause death. Several analytical techniques for the isolation of barbiturates have been reviewed by Pillai and Dilli (4) and Jain and Cravey (5). Traditionally these techniques involve liquid-liquid extraction of the drugs from biological fluids followed by immunoassay (6,7), ultraviolet spectrophotometry (8), infrared spectrophometry (9), nuclear magnetic resonance spectroscopy ( 10,11), mass spectroscopy (12,13), thinlayer" chromatography (14,15), or high-pressure liquid chromatography (16).
Equipment. Vac Elut | vacuum manifolds (AI 6000) were provided by Varian Sample Preparation Products. A vortex mixer was obtained from Scientific Industries, Inc. A Reacti Therm TM heating module and a Reacti VapTM evaporator were purchased from Pierce. A Varian Saturn GC/MS/DS system used for the data analysis was equipped with a Varian Model 3400 gas chromatograph, a short transfer line, an ion trap quadrupole analyzer, a COMPAQ DESKPRO 386/20e Model 40 personal computer, and a Varian 8100 autosampler. The GC was equipped with a Varian SPI injector and a 30-m • 0.25-ram x 0.25-Hm film thickness DB-5 capillary column. The injector was programmed at 40~ for 0.2 min, heated to 260~ at 150~ and held at this temperature for 25 rain. The initial column temperature was 50~ which was held for 3 min, and then heated to 260~ at 10~ The mass range from 50 to 300 ainu was scanned at a rate of 1 s/scan. Materials. Bond Elut Certify II extraction columns were provided by Varian Sample Preparation Products. Amobarbital, butabarbital, pentobarbital, phenobarbital, secobarbital, methaqualone, and hexobarbital were purchased from Alltech-Applied Science. HPLC grade methanol, hexane, and ethyl acetate were purchased from EM Science. Certified negative human urine was obtained from Fisher Scientific, and urine specimens were spiked with known amounts of barbiturates. All other chemicals were reagent grade puchased from Fisher Scientific. Extraction procedure. 5 mL urine spiked with barbiturates was added to a large test tube followed by 2 mL of 100raM sodium ac-
Reproduction (photocopying)of editorialcontentof this journal is prohibited without publisher's permission.
45
Journal of AnalyticalToxicology,Vol. 16, January/February1992
etate buffer (pH = 7.0). The pH of the specimen was adjusted to between 5 and 7. The Bond Elut Certify II columns were connected to a Vac Elut and conditioned with 2 mL of methanol. Excess methanol was removed by washing with 2 mL of 100mM sodium acetate (pH = 7.0) buffer. The urine specimen containing barbiturates was applied to the column and passed through the bed at a slow flow rate by applying vacuum at approximately 2-3 in. Hg. The column was washed with 1 mL of 100mM sodium acetate buffer (pH = 7.0) and dried under full vacuum (15 in. Hg) for 5 min. The colunm was further washed with 2 mL of hexaneethyl acetate mixture (95:5). The tips of the Vac Elut delivery needles were wiped and a rack with labeled collection tubes was placed in the Vac Elut. The drugs were eluted with 2 mL of hexane-ethyl acetate (75:25) mixture. The Vac Elut was disassembled and the test tubes were removed. To each test tube, 100 /aL of hexobarbital (0.2 mg dissolved in 10 mL ethyl acetate) was added as an internal standard and used to calculate the absolute recoveries of drugs. The test tubes were placed in the Reacti Therm evaporator. The solvent was evaporated to dryness under a slow stream of nitrogen at room temperature and reconstituted in 100 tJL of ethyl acetate. A l-gL volume of the sample was injected into the GC/MS for analysis.
ng/mL (in triplicate). Figure 1 represents the standard curve generated from the analysis of spiked urine samples. The linear plot is a representation of secobarbital over a concentration range of 20-500 ng/mL. Our experimental procedure and test results have established that the technique provides the detection of the drugs as low as 10 ng/mL, with greater than 80% recovery of each drug and linearity over the range 20-500 ng/mL when using a 5mL urine sample. The top trace of Figure 2 shows the total ion chromatogram of barbiturates at 40 ng/mL. The next two traces in Figure 2 are the additive plots of ions (I 41, 156) and (167, 168) representing butabarbital, amobarbital, pentobarbital, and secobarbital. The fourth trace, ion 204, represents phenobarbital; ion 221 is the internal standard (trace five); and the last trace is the additive plot of ions 235 and 250 representing methaqualone. Figure 3 is the full scan El spectrum of secobarbital at 40 ng/mL, demonstrating the purity of the extract. Figure 4 demonstrates the applicability and superiority of this method by analyzing and comparing the real life sample with liquid-liquid extraction. Greater recovery and better sensitivity for phenobarbital than with the conventional liquid-liquid extraction procedures are demonstrated.
2. see .0
Results and Discussion 9
2.[Jeie -
The solid-phase extraction procedure described provides a rapid, reliable, and reproducible isolation of barbiturates from spiked humau urine samples. The Bond Elut Certify II selectively retains and elutes the drugs by a mixed mode interaction mechanism. Table I provides the absolute mean % recoveries of the drugs listed. The within-run coefficient of variation (CV) was found to be 2.0-12. 1%. Table II represents the linearity data for barbiturate extractions over the concentration range 20-500
,/.,/
./,"
i
,,/I(9
l.eee 84
9
J
/; /,/o
R .5~1e
..o
Table I. Within-Run Precision*
rTT]]TI,
I I I,
I I I r]Ti
1~.1~
Compound
Absolute % Recovery (Mean)
SD
CV (%)
90 90 99 82 100 100
3,0 3.0 12.0 8.0 2.0 2.0
3,3 3.3 12.1 9.8 2,0 2.0
Amobarbital 8utabarbital Methaqualone Pentobarbital Phenobarbital Secobarbital
9Obtained tram 25-ng/mL concentration analyzed in triplicate.
i i i i i l'Frl
288.086
i i i i i , iTrl
30e.eee Concentrat
i i i i i,
4ee.see
i fl-r'rrT~]a-
sea
.088
Ion (z~j/mt)
Figure 1. The response curve for secobarbital at concentration of 20, 40. 100, 200, 300,400, and 500 ng/mL.
141t
~
o
Table II. Barbiturates Assay and Analysis*
Compound Amobarbital 8utabarbital Methaqualone Pentobarbital Phenobarbital Secobarbital
Correlation coefficients (r ~)
Standard deviation
0.998 0.999 0.996 0.997 0.999 0,999
0.031 0.044 0.036 0.049 0.065 0,027
9Concentrations of 20, 40, 100, 200, 400, and 500 ng/mL analyzed in triplicate. 46
550
18:~
lO:5R
7~
11:48
7~l IZ:31J
Scan I~mber Sc~, T l ~ e
Figure 2. Total ion chromatogram and mass chromatograms of barbiturates.
Journalof AnalyticalToxicology,Vol. 16, January/February1992
Conclusions
References
A solid-phase extraction procedure has been developed for barbiturates from human urine using bond Elut Certify II columns. The method is fast, clean, and allows multiple samples to be processed at the same time.
1. E. Fischer and J.R. von Mering. A new class of sleeping medicine. Therap. Gengenwart. 44:97-101 (1903). 2. W.J. Doran. Barbituric acid hypnotics. Med. Chem. VoL IV: p. 328, Wiley, New York (1959). 3. Members of the Substance Abuse Testing Committee, Division of Therapeutic Drug Monitoring and Clin. Toxicol., Am. Assoc. for Clin. Chem. "Critical Issues in Urinalysis of Abused Substances: Report of the Substance-Abuse Testing Committee." Clin. Chem. 34:605-32 (1988). 4. D.N. Pillai and S. Dilli. Analysis of barbiturates by gas chromatography. J. Chromatogr. 220:253-74 (1981 ). 5. N.C. Jain and R.H. Cravey. The identificatrion of barbiturates from biological specimens. J. Chromatogr. ScL 10:228-31 (1974). 6. E.J. Flynn and S. Spector. Detemination of barbiturate derivatives by radioimmunoassay. J. PharmacoL Exp. Ther. 181:547-54 (1972). 7. R. Cleeland, R. Davis, J. Heveran, and E. Grunberg. A simple and rapid Iodine-125 radioimmunoassay for the detection of barbiturates in biological fluids. J. Forens. ScL 20:45-57 (1975). 8. G.B. Schumann, K. Lauenstein, D. Le Fever, and J. Bernard. Ultraviolet spectrophotometric analysis of barbiturates. Am. ,1. Clin. PathoL 66:823-30 (1976). 9. S. Goenechea. Infrared spectroscopic investigation of barbiturates. Z. Anal Chem. 218:415-26 (1966). 10. H.W. Avdovich and G.A. Neville. Application of proton magnetic resonance spectroscopy to pharmaceuticals. II. Specific identification of barbiturates. Can. J. Pharm. ScL 4:51.-64 (1969). 11. H. Lackner and G. Doting. Identification of barbiturates by NMR spectroscopy. I. Efficiency of the method. Arch. Toxicol. 26: 220-36 (1970). 12. W. Arnold and H. Grutzmacher. Mass spectrometric detection of drug metabolites in forensic analysis. Z. AnaL Chem. 247: 179-88 (1969). 13. E.F. AbdeI-Bary, G. Bohn, and G. Ruecker. Mass spectrometrical identification of mixtures of two or more barbiturates in autopsy material. Z. Anal Chem. 266:361-83 (1973). 14. H.V. Street and C. McMartin. Quantitative estimation and identification of barbiturates in blood in emergency cases. Nature (London) 199:456-59 (1963. 15. J.A. Vinson, E.J. Hooyman, H. Koharcheck, and M.M. Holmes. Sensitive thin layer chromatographic method for urine screening of barbiturates. J. Chromatogr. 140:71-76 (1977). 16. W. Dunges, G. Naundorf, and N. Seller. High pressure liquid chromatographic analysis of barbiturates in picomole range by fluorometry of their DANS-derivatives. J. Chromatogr. ScL 12: 655-57 (1974).
Acknowledgment The authors thank Mr. Philip Dimson of Varian Sample Preparation Products for his valuable suggestions. 18H•
168
IRT
SHP
153
t24
1~;
1~8
141
181
j~ 188
227 128
148
168
188 . / = Z88
228
262 248
268
2?7 287 208
38fi
Figure 3. El spectrum of secobarbitalat 40 ng/mL.
i
(A)
ti
Figure 4. Comparisonof (A) liquid-liquid extractionof phenobarbital with (B) solid-phaseextractionon Certify II using real life sample.
Manuscript received November 19, 1990; revision received April 9, 1991.
47
Solid-Phase Extractionand GC/MS Confirmationof Barbituratesfrom Human Urine R u g g e r o Pocci, V a n d a n a Dixit, and V y a s M. Dixit
Varian Samp/e Preparation Products, 24201 Frampton Avenue, Harbor City, Ca/ifomia 90710
Abstract A highly selective and sensitive procedure has been
developed for isolating and identifying barbiturates In human urine. With a new disposable bonded silica gel solid-phase extraction (SPE) column and hexobarbital as an internal standard (IS), amobarbital, butabarbital, pentobarbital, phenobarbital, secobarbital, and methaqualone were selectively isolated from endogenous urine components. Capillary gas chromatography/Ion trap mass spectrometry (GC/MS) analysis of the extracts generated a full mass spectrum for the detection, identification, and quantitation of barbiturates. Linear quantitative response curves for the drugs have been generated over a concentration range of 20-500 ng/mL. Overall extraction efficlencles for drugs averaged greater than 90%, and the quantitative response curves exhibited correlation coefficients of 0.996 to 0.999.
This article presents a sensitive, qualitative, and quantitative extraction procedure for amobarbital, butabarbital, pentobarbital, phenobarbital, secobarbital, and methaqualone using Bond Elut Certify IITM, a disposable solid-phase extraction column. Bond Elut Certify II is a chemically modified silica gel material bearing three different types of interactions: hydrophobic, polar, and ion exchange. The silica gel is specially designed for the extraction of acidic and neutral molecules from biological fluid and provides clean extracts of the barbiturates. This cleanliness is the result of the removal of some of the interferences during the column rinse process, while most of the impurities present in the urine are irreversibly retained on the column.
M a t e r i a l s and M e t h o d s
Introduction
Barbiturates, tile derivatives of urea, are formed by replacement of both hydrogen atoms on the carbon at position 5 by alkyl, aryl, or alicyclic groups. The potential clinical significance of barbital led to the extensive development of the barbituric acid class of drugs (1) in 1903. A definitive publication by Doran (2) in 1959 appears to be the only monograph providing useful information, including details of syntheses, chemical and physical properties, reactions, and pharmacology. Their action on the central nervous system (CNS) and the extent of CNS depression is dependent on the particular barbiturate. At low doses, they induce a feeling of relaxation and tranquility and can impair motor functions. Their rapid action tbr anesthetic purposes and intermediate action for sedative or sleeping purposes make them attractive to drug abusers (3). Overdoses of barbiturates depress the respiratory centers of the brain and can cause death. Several analytical techniques for the isolation of barbiturates have been reviewed by Pillai and Dilli (4) and Jain and Cravey (5). Traditionally these techniques involve liquid-liquid extraction of the drugs from biological fluids followed by immunoassay (6,7), ultraviolet spectrophotometry (8), infrared spectrophometry (9), nuclear magnetic resonance spectroscopy ( 10,11), mass spectroscopy (12,13), thinlayer" chromatography (14,15), or high-pressure liquid chromatography (16).
Equipment. Vac Elut | vacuum manifolds (AI 6000) were provided by Varian Sample Preparation Products. A vortex mixer was obtained from Scientific Industries, Inc. A Reacti Therm TM heating module and a Reacti VapTM evaporator were purchased from Pierce. A Varian Saturn GC/MS/DS system used for the data analysis was equipped with a Varian Model 3400 gas chromatograph, a short transfer line, an ion trap quadrupole analyzer, a COMPAQ DESKPRO 386/20e Model 40 personal computer, and a Varian 8100 autosampler. The GC was equipped with a Varian SPI injector and a 30-m • 0.25-ram x 0.25-Hm film thickness DB-5 capillary column. The injector was programmed at 40~ for 0.2 min, heated to 260~ at 150~ and held at this temperature for 25 rain. The initial column temperature was 50~ which was held for 3 min, and then heated to 260~ at 10~ The mass range from 50 to 300 ainu was scanned at a rate of 1 s/scan. Materials. Bond Elut Certify II extraction columns were provided by Varian Sample Preparation Products. Amobarbital, butabarbital, pentobarbital, phenobarbital, secobarbital, methaqualone, and hexobarbital were purchased from Alltech-Applied Science. HPLC grade methanol, hexane, and ethyl acetate were purchased from EM Science. Certified negative human urine was obtained from Fisher Scientific, and urine specimens were spiked with known amounts of barbiturates. All other chemicals were reagent grade puchased from Fisher Scientific. Extraction procedure. 5 mL urine spiked with barbiturates was added to a large test tube followed by 2 mL of 100raM sodium ac-
Reproduction (photocopying)of editorialcontentof this journal is prohibited without publisher's permission.
45
Journal of AnalyticalToxicology,Vol. 16, January/February1992
etate buffer (pH = 7.0). The pH of the specimen was adjusted to between 5 and 7. The Bond Elut Certify II columns were connected to a Vac Elut and conditioned with 2 mL of methanol. Excess methanol was removed by washing with 2 mL of 100mM sodium acetate (pH = 7.0) buffer. The urine specimen containing barbiturates was applied to the column and passed through the bed at a slow flow rate by applying vacuum at approximately 2-3 in. Hg. The column was washed with 1 mL of 100mM sodium acetate buffer (pH = 7.0) and dried under full vacuum (15 in. Hg) for 5 min. The colunm was further washed with 2 mL of hexaneethyl acetate mixture (95:5). The tips of the Vac Elut delivery needles were wiped and a rack with labeled collection tubes was placed in the Vac Elut. The drugs were eluted with 2 mL of hexane-ethyl acetate (75:25) mixture. The Vac Elut was disassembled and the test tubes were removed. To each test tube, 100 /aL of hexobarbital (0.2 mg dissolved in 10 mL ethyl acetate) was added as an internal standard and used to calculate the absolute recoveries of drugs. The test tubes were placed in the Reacti Therm evaporator. The solvent was evaporated to dryness under a slow stream of nitrogen at room temperature and reconstituted in 100 tJL of ethyl acetate. A l-gL volume of the sample was injected into the GC/MS for analysis.
ng/mL (in triplicate). Figure 1 represents the standard curve generated from the analysis of spiked urine samples. The linear plot is a representation of secobarbital over a concentration range of 20-500 ng/mL. Our experimental procedure and test results have established that the technique provides the detection of the drugs as low as 10 ng/mL, with greater than 80% recovery of each drug and linearity over the range 20-500 ng/mL when using a 5mL urine sample. The top trace of Figure 2 shows the total ion chromatogram of barbiturates at 40 ng/mL. The next two traces in Figure 2 are the additive plots of ions (I 41, 156) and (167, 168) representing butabarbital, amobarbital, pentobarbital, and secobarbital. The fourth trace, ion 204, represents phenobarbital; ion 221 is the internal standard (trace five); and the last trace is the additive plot of ions 235 and 250 representing methaqualone. Figure 3 is the full scan El spectrum of secobarbital at 40 ng/mL, demonstrating the purity of the extract. Figure 4 demonstrates the applicability and superiority of this method by analyzing and comparing the real life sample with liquid-liquid extraction. Greater recovery and better sensitivity for phenobarbital than with the conventional liquid-liquid extraction procedures are demonstrated.
2. see .0
Results and Discussion 9
2.[Jeie -
The solid-phase extraction procedure described provides a rapid, reliable, and reproducible isolation of barbiturates from spiked humau urine samples. The Bond Elut Certify II selectively retains and elutes the drugs by a mixed mode interaction mechanism. Table I provides the absolute mean % recoveries of the drugs listed. The within-run coefficient of variation (CV) was found to be 2.0-12. 1%. Table II represents the linearity data for barbiturate extractions over the concentration range 20-500
,/.,/
./,"
i
,,/I(9
l.eee 84
9
J
/; /,/o
R .5~1e
..o
Table I. Within-Run Precision*
rTT]]TI,
I I I,
I I I r]Ti
1~.1~
Compound
Absolute % Recovery (Mean)
SD
CV (%)
90 90 99 82 100 100
3,0 3.0 12.0 8.0 2.0 2.0
3,3 3.3 12.1 9.8 2,0 2.0
Amobarbital 8utabarbital Methaqualone Pentobarbital Phenobarbital Secobarbital
9Obtained tram 25-ng/mL concentration analyzed in triplicate.
i i i i i l'Frl
288.086
i i i i i , iTrl
30e.eee Concentrat
i i i i i,
4ee.see
i fl-r'rrT~]a-
sea
.088
Ion (z~j/mt)
Figure 1. The response curve for secobarbital at concentration of 20, 40. 100, 200, 300,400, and 500 ng/mL.
141t
~
o
Table II. Barbiturates Assay and Analysis*
Compound Amobarbital 8utabarbital Methaqualone Pentobarbital Phenobarbital Secobarbital
Correlation coefficients (r ~)
Standard deviation
0.998 0.999 0.996 0.997 0.999 0,999
0.031 0.044 0.036 0.049 0.065 0,027
9Concentrations of 20, 40, 100, 200, 400, and 500 ng/mL analyzed in triplicate. 46
550
18:~
lO:5R
7~
11:48
7~l IZ:31J
Scan I~mber Sc~, T l ~ e
Figure 2. Total ion chromatogram and mass chromatograms of barbiturates.
Journalof AnalyticalToxicology,Vol. 16, January/February1992
Conclusions
References
A solid-phase extraction procedure has been developed for barbiturates from human urine using bond Elut Certify II columns. The method is fast, clean, and allows multiple samples to be processed at the same time.
1. E. Fischer and J.R. von Mering. A new class of sleeping medicine. Therap. Gengenwart. 44:97-101 (1903). 2. W.J. Doran. Barbituric acid hypnotics. Med. Chem. VoL IV: p. 328, Wiley, New York (1959). 3. Members of the Substance Abuse Testing Committee, Division of Therapeutic Drug Monitoring and Clin. Toxicol., Am. Assoc. for Clin. Chem. "Critical Issues in Urinalysis of Abused Substances: Report of the Substance-Abuse Testing Committee." Clin. Chem. 34:605-32 (1988). 4. D.N. Pillai and S. Dilli. Analysis of barbiturates by gas chromatography. J. Chromatogr. 220:253-74 (1981 ). 5. N.C. Jain and R.H. Cravey. The identificatrion of barbiturates from biological specimens. J. Chromatogr. ScL 10:228-31 (1974). 6. E.J. Flynn and S. Spector. Detemination of barbiturate derivatives by radioimmunoassay. J. PharmacoL Exp. Ther. 181:547-54 (1972). 7. R. Cleeland, R. Davis, J. Heveran, and E. Grunberg. A simple and rapid Iodine-125 radioimmunoassay for the detection of barbiturates in biological fluids. J. Forens. ScL 20:45-57 (1975). 8. G.B. Schumann, K. Lauenstein, D. Le Fever, and J. Bernard. Ultraviolet spectrophotometric analysis of barbiturates. Am. ,1. Clin. PathoL 66:823-30 (1976). 9. S. Goenechea. Infrared spectroscopic investigation of barbiturates. Z. Anal Chem. 218:415-26 (1966). 10. H.W. Avdovich and G.A. Neville. Application of proton magnetic resonance spectroscopy to pharmaceuticals. II. Specific identification of barbiturates. Can. J. Pharm. ScL 4:51.-64 (1969). 11. H. Lackner and G. Doting. Identification of barbiturates by NMR spectroscopy. I. Efficiency of the method. Arch. Toxicol. 26: 220-36 (1970). 12. W. Arnold and H. Grutzmacher. Mass spectrometric detection of drug metabolites in forensic analysis. Z. AnaL Chem. 247: 179-88 (1969). 13. E.F. AbdeI-Bary, G. Bohn, and G. Ruecker. Mass spectrometrical identification of mixtures of two or more barbiturates in autopsy material. Z. Anal Chem. 266:361-83 (1973). 14. H.V. Street and C. McMartin. Quantitative estimation and identification of barbiturates in blood in emergency cases. Nature (London) 199:456-59 (1963. 15. J.A. Vinson, E.J. Hooyman, H. Koharcheck, and M.M. Holmes. Sensitive thin layer chromatographic method for urine screening of barbiturates. J. Chromatogr. 140:71-76 (1977). 16. W. Dunges, G. Naundorf, and N. Seller. High pressure liquid chromatographic analysis of barbiturates in picomole range by fluorometry of their DANS-derivatives. J. Chromatogr. ScL 12: 655-57 (1974).
Acknowledgment The authors thank Mr. Philip Dimson of Varian Sample Preparation Products for his valuable suggestions. 18H•
168
IRT
SHP
153
t24
1~;
1~8
141
181
j~ 188
227 128
148
168
188 . / = Z88
228
262 248
268
2?7 287 208
38fi
Figure 3. El spectrum of secobarbitalat 40 ng/mL.
i
(A)
ti
Figure 4. Comparisonof (A) liquid-liquid extractionof phenobarbital with (B) solid-phaseextractionon Certify II using real life sample.
Manuscript received November 19, 1990; revision received April 9, 1991.
47
