Anal. Chem. 1999, 71, 1918-1921
Subzero-Temperature Liquid-Liquid Extraction of Benzodiazepines for High-Performance Liquid Chromatography Manabu Yoshida* and Atsushi Akane
Department of Legal Medicine, Kansai Medical University, Moriguchi 570-8506, Japan
On the basis of the phenomenon that hydrophilic acetonitrile is separated from the aqueous phase at -20 °C, we employed a novel extraction method, “subzero-temperature liquid-liquid extraction”, to extract benzodiazepines (estazolam and triazolam) from serum or aqueous solution for liquid chromatography. A 1:1 mixture of acetonitrile and the specimen was cooled at -20 °C for 20 min to separate the acetonitrile and aqueous phases. The acetonitrile phase was directly injected into a highperformance liquid chromatograph. Recovery rates of the drugs following the first subzero-temperature liquidliquid extraction were 50.3 ( 0.6-54.0 ( 0.9%, which were lower than those (73.9 ( 3.3-80.6 ( 0.6% and 81.6 ( 4.7-96.1 ( 2.6%) of the first conventional liquid-liquid extraction using diethyl ether and solidphase extraction using a Sep-Pak C18 column, respectively. However, three to four repeated subzero-temperature liquid-liquid extractions and conventional liquidliquid extractions resulted in recovery of almost 100% of the drugs. In the chromatogram of the benzodiazepines recovered from serum by the subzero-temperature extraction, no coextracted component interfered with determination of the drugs. Detection limits of the drugs were 0.02-0.08 µg/mL, and coefficients of variance were 1.14-2.17% suggesting high reproducibility. To detect drugs in biological materials, drugs should be extracted, purified, and concentrated before the assay. Generally, drugs are extracted by liquid-liquid extraction methods using organic solvents1-3 or solid-phase extraction methods using SepPak,4,5 Bond Elut,4,6-8 Extrelut columns9,10 or others.11 Recently, solid-phase microextraction-gas chromatography (-mass spec* Corresponding author: (e-mail)
[email protected]; (fax) +81-6-69926582. (1) Hailey, D. M. J. Chromatogr. 1974, 98, 527-568. (2) Brodie, R. R.; Chasse, L. F.; Taylor, T. J. Chromatogr. 1978, 150, 361366. (3) Coassolo, C. D.; Aubert, C.; Coassolo, P.; Cano, J. P. J. Chromatogr., B 1989, 487, 295-311. (4) Good, J. S.; Andrews, J. S. J. Chromatogr. Sci. 1981, 19, 562-566. (5) Seno, H.; Suzuki, O.; Kumazawa, T.; Hattori, H. J. Anal. Toxicol. 1991, 15, 21-24. (6) Black, D. A.; Clark, G. D.; Haver, V. M.; Garbin, J. A.; Saxon, A. J. J. Anal. Toxicol. 1994, 18, 185-188. (7) Verweij, A. M.; Hordijk, M. L.; Lipman, P. J. J. Chromatogr., B 1996, 686, 27-34. (8) Akerman, K. K. Scand. J. Clin. Lab. Invest. 1996, 56, 609-614.
1918 Analytical Chemistry, Vol. 71, No. 9, May 1, 1999
trometry),12,13 supercritical fluid extraction,14-16 affinity extraction,17-20 on-line solid-phase extraction,21,22 and automatic extraction-liquid chromatography using a column-switching technique19,23-25 were reported.26-28 In this study, a novel subzero-temperature liquidliquid extraction method was examined, based on the phenomenon that hydrophilic acetonitrile is separated from the aqueous phase at subzero temperatures (∼-20 °C). This technique was applied to extract benzodiazepines (estazolam and triazolam) from serum or aqueous solution for high-performance liquid chromatography. EXPERIMENTAL SECTION Subzero-Temperature Liquid-Liquid Extraction. Acetonitrile (0.25, 0.3, 0.4, 0.5, or 0.75 mL) was mixed well with 0.5 mL of human serum or aqueous solution containing 0.5 or 1.0 µg/ mL of standard estazolam and triazolam. The mixture of acetonitrile and serum was centrifuged at 1200g for 5 min. Each sample was cooled at -20 °C for 20 min to separate acetonitrile and aqueous phases. The acetonitrile phase was transferred to a new tube, and 10 µL of the recovered phase was directly injected into a high-performance liquid chromatograph composed of a SCL(9) Hattori, H.; Suzuki, O.; Sato, K.; Mizutani, Y.; Yamada, T. Forensic Sci. Int. 1987, 35, 165-179. (10) Verweij, A. M.; Lipman, P. J.; Zweipfenning, P. G. Forensic. Sci. Int. 1992, 54, 67-74. (11) Plumb, R. S.; Gray, R. D.; Jones, C. M. J. Chromatogr., B 1997, 694, 123133. (12) Zhang, Z.; Yang, M. J.; Pawliszyn. J. Anal. Chem. 1994, 66, 844A-852A. (13) Krogh. M.; Johansen, K.; Tonnesen, F.; Rasmussen, K. E. J. Chromatogr., B 1995, 673, 299-305. (14) Kalinoski, H. T.; Udseth, H. R.; Wright, B. W.; Smith, R. D. Anal. Chem. 1986, 58, 2421-2425. (15) Hawthorne, S. B.; Miller, D. J. J. Chromatogr. 1987, 403, 63-76. (16) Randolph, T. W. Trends Biotechnol. 1990, 8, 78-82. (17) Rule, G. S.; Mordehai, A. V.; Henion, J. Anal. Chem. 1994, 66, 230-235. (18) Cai, J.; Henion, J. D. Anal. Chem. 1996, 68, 72-78. (19) Cai, J.; Henion, J. D. J. Chromatogr., B 1997, 691, 357-370. (20) Bean, K. A.; Henion, J. D. J. Chromatogr., A 1997, 791, 119-126. (21) Bowers, G. D.; Clegg, C. P.; Hughes, S. C.; Harker, A. J.; Lambert, S. LCGC 1997, 15, 48-53. (22) McLoughlin, D. A.; Olah, T. V.; Gilbert, J. D. J. Pharm. Biomed. Anal. 1997, 15, 1803-1901. (23) Nedved, M. L.; Habibi, G. S.; Ganem, B.; Henion, J. D. Anal. Chem. 1996, 68, 4228-4236. (24) Deinl, I.; Angermaier, L.; Franzelius, C.; Machbert, G. J. Chromatogr., B 1997, 704, 251-258. (25) Yoshida, M.; Akane, A.; Okii, Y.; Yoshimura, S.; Tokiyasu, T.; Watabiki, T. Advances in Legal Medicine 3; Yoyodo: Osaka, 1997; pp 532-534. (26) McDowall, R. D. J. Chromatogr. 1989, 492, 3-58. (27) Majors, R. E. LC-GC 1998, (Suppl. May), S8-S15. (28) Henion, J.; Brewer, E.; Rule, G. Anal. Chem. 1998, 70, 650A-656A. 10.1021/ac981276g CCC: $18.00
© 1999 American Chemical Society Published on Web 03/13/1999
Figure 1. Effects of added amounts of acetonitrile on the concentration and yield of the first benzodiazepine extracts by the subzero-temperature liquid-liquid extraction method. To 0.5 mL of the specimen, 0.4 (solid bar), 0.5 (shaded bar) or 0.75 mL (open bar) of acetonitrile was added.
6B system controller (Shimadzu, Kyoto, Japan), a LC-6A pump (Shimadzu), a SIL-6B autoinjector (Shimadzu), a Capcell Pak C18 column (4.6 mm i.d. × 150 mm, Shiseido, Tokyo, Japan) and a SPD-6AV UV-visible spectrophotometric detector (Shimadzu). The mobile phase was 0.1% acetic acid-40% acetonitrile, flowing at 1.0 mL/min. The drugs were detected by optical density at 254 nm. As samples have frequently been buffered to pH 9-10 for conventional liquid-liquid extraction of benzodiazepines,1-3 the (1:1) mixture of acetonitrile and aqueous solution containing estazolam and triazolam was mixed well with 20 µL of 28% ammonia, and the effect of alkalization on subzero-temperature liquid-liquid extraction was investigated. To study the recovery rate and reproducibility, 0.5 mL of serum or aqueous solution containing 1 µg/mL of standard drugs was mixed with 0.5 mL of acetonitrile, and the acetonitrile phase was recovered by the subzero-temperature liquid-liquid extraction method. To the aqueous phase, 0.5 mL of acetonitrile was added and recovered. This extraction procedure was repeated three times, and the drugs in each extract were quantitated. To assess the detection limit and to investigate whether some serum components were co-purified and interfere with the detection of benzodiazepines, 0.5 mL of acetonitrile was mixed with 0.5 mL of serum containing no drugs or 0.5 mL of serum containing 0.020.08 µg/mL of the standard drugs, and subzero-temperature liquid-liquid extraction procedures were performed. To obtain calibration curves, 0.5 mL of acetonitrile was mixed with 0.5 mL serum or aqueous solution containing 0.08-2.5 µg/mL of the standard drugs. Conventional Liquid-Liquid Extraction. Human serum or an aqueous solution (0.5 mL) containing 1 µg/mL of the standard drugs was mixed with 20 µL of 28% ammonia, 2 mL of diethyl ether was added, and the mixture vortexed. Following centrifugation at 1200g for 5 min, the ether phase was recovered and dehydrated by adding anhydrous sodium sulfate. The sample was dried under a nitrogen stream, and the residue was dissolved in 200 µL of the mobile phase for high-performance liquid chromatography. These liquid-liquid extraction procedures were re-
peated four times for one specimen, and the concentration of the drugs in each extract was determined. Solid-Phase Extraction. According to the method of Good and Andrews,4 a Sep-Pak C18 (100 mg/mL) column was washed with 1 mL of 99% methanol twice and with 1 mL of distilled water twice. Human serum or aqueous solution (0.5 mL) containing 1 µg/mL of the standard drugs was mixed with 100 µL of 0.1 M sodium carbonate and applied to the column. Following washing with 1 mL of distilled water twice and with 50 µL of 99% methanol once, the drugs were eluted with 200 µL and then 100 µL of 99% methanol. The eluate was concentrated to 225 µL under a nitrogen stream, 10 µL of which was injected into the high-performance liquid chromatograph. A second elution with 200 µL of 99% methanol was performed, and the eluate was also analyzed. RESULTS AND DISCUSSION When not less than 0.4 mL of acetonitrile was added to 0.5 mL of the specimens, the upper acetonitrile and lower aqueous phases were separated after cooling at -20 °C for 20 min. The volume of recoverable acetonitrile phase was 0.1, 0.25, or 0.5 mL when 0.4, 0.5, or 0.75 mL of acetonitrile was added for the first time to 0.5 mL of the specimen, respectively, suggesting that part of the added acetonitrile was mixed in the aqueous phase even after cooling. When the mixture of acetonitrile and specimen was cooled at -20 °C for more than 20 min, aqueous phase was frozen and the intermediate layer appeared between the acetonitrile and frozen aqueous phases. Since the boundary between the intermediate and acetonitrile layers was unclear, it was difficult to transfer the whole amount of the acetonitrile phase to another tube. Therefore, the cooling time was set at 20 min in this study. Centrifugation of the mixture of acetonitrile and the specimen at -20 °C may improve separation of the phases for extraction of the drugs, but the usual refrigerated centrifuges cannot be used at -20 °C. Added amounts of acetonitrile were inversely correlated with concentrations of drugs in the recovered acetonitrile solution (Figure 1), but were correlated to the recovered amounts (yields) of drugs due to the positive correlation between the added and Analytical Chemistry, Vol. 71, No. 9, May 1, 1999
1919
Figure 2. High-performance liquid chromatograms of subzerotemperature liquid-liquid extracts from sera. Upper, serum containing 1 µg/mL of each benzodiazepine. Lower, serum containing no drugs.
recovered amounts of acetonitrile. The addition of 0.4 mL of acetonitrile resulted in the highest concentration of the drugs. So, the highest sensitivity of the liquid chromatography would be expected when 10 µL of recovered acetonitrile was directly injected into the chromatograph. Because of the lowest yield at 0.4-mL extraction volume, however, the added amount of acetonitrile was set at 0.5 mL in the following recovery study. Serum components extracted from sera of healthy individuals by the subzero-temperature liquid-liquid extraction technique appeared within 3.5 min after injection in the chromatogram (Figure 2), which did not interfere with detection of estazolam or triazolam, the retention times of which were 4.1 and 4.9 min, respectively. Addition of 28% ammonia to specimens before the extraction did not improve the recovery or detection of the drugs, but resulted in the appearance of a large peak in the chromatogram 1.3 min after the injection. The area of this peak was not correlated with the added amount of 28% ammonia, suggesting that ammonia induced elution of impure contents from silica gel in the column. On the basis of this result, ammonia was not added to the specimen thereafter. The recovery rates of the first extraction by the subzerotemperature liquid-liquid extraction method were 50.3-54.0%, which were lower than those (73.9-80.6 and 81.6-96.1%) of the first extractions by conventional liquid-liquid and solid-phase extraction methods (Figure 3). When extraction was repeated three to four times, 99.3-103.1% of the drugs was recovered by the subzero-temperature liquid-liquid extraction technique as well as the conventional liquid-liquid extraction method. The second extract of the solid-phase extraction method did not contain any drugs. When 0.5 mL of acetonitrile was added, the detection limits of the subzero-temperature liquid-liquid extraction technique of estazolam and triazolam were 0.02 and 0.08 mg/mL, which corresponded to 0.2 and 0.8 ng of the absolute amounts, respec1920 Analytical Chemistry, Vol. 71, No. 9, May 1, 1999
Figure 3. Comparison of accumulated recovery rates of benzodiazepines from serum (top) or aqueous solution (bottom) among the subzero-temperature liquid-liquid (ST), conventional liquid-liquid (LL), and solid-phase (SP) extraction methods. Solid bar, recovery rate for the first extraction. Shaded, dotted, and open bars, accumulated recovery rates for two, three, and four extractions, respectively.
tively. Coefficients of variance for recovery rates of the drugs from serum containing 1.0 µg/mL estazolam and 1.0 µg/mL triazolam were 1.30 and 1.64%, and those from the aqueous solution were 1.25 and 1.69%, respectively, indicating sufficiently high reproducibility of the extraction method. Linear calibration curves for 0.082.5 µg/mL estazolam and triazolam in serum or aqueous solution were obtained with correlation coefficients of 0.9999 and 0.9994, or 0.9999 and 0.9999, respectively. Acetonitrile is an organic solvent with a melting point of -45 °C. As acetonitrile is hydrophilic and miscible with water, acetonitrile never separates in 1:1 mixtures with water at room temperature. However, the miscibility of acetonitrile decreases at subzero temperatures, resulting in some separation of acetonitrile from the mixture. On the basis of this phenomenon, human serum or aqueous solution containing benzodiazepines was mixed with acetonitrile and cooled at -20 °C, and the drugs in the separated acetonitrile phase were recovered and directly injected into a liquid chromatograph using acetonitrile solution as a mobile phase. The result that the first extract contained ∼50% of the drugs of the original specimen was likely because part of the added acetonitrile remained mixed in the aqueous phase. Decreasing
the added amounts of acetonitrile resulted in reduced recovered amounts of the acetonitrile phase and the drugs but increased concentrations of the drugs in the recovered phase and improved sensitivity of the assay, as described above. On the contrary, increasing the added amount of acetonitrile resulted in increased amounts of the recovered acetonitrile phase and the drugs but decreased concentrations of the drugs in the recovered phase. Thus, the drugs could be concentrated or diluted, and assay sensitivity or recovery could be improved, by changing the added amount of acetonitrile. Following three to four repeated procedures of subzerotemperature or conventional liquid-liquid extraction, almost 100% of the drugs was recovered. The recovery rate by solid-phase extraction was lower than 100%, probably because retention of the drugs in the column was less than 100% or because some of the drugs were eluted when the column was washed before the elution. Subzero-temperature liquid-liquid and solid-phase extraction methods were simple because the recovered acetonitrile or methanol solution containing the drugs could be directly injected
into the liquid chromatograph. On the contrary, conventional liquid-liquid extraction was troublesome because the drugs recovered in diethyl ether had to be dried and redissolved in the mobile phase. Thus, the subzero-temperature liquid-liquid extraction method has the advantages of both conventional liquid-liquid and solid-phase extraction methods. Furthermore, conventional liquid-liquid extraction of benzodiazepines requires the addition of 28% ammonia to specimens to increase the extraction efficiency. However, ammonia may corrode the silica gel in the chromatographic column, so the subzero-temperature liquid-liquid extraction method, in which the addition of ammonia was not needed, was more suitable for liquid chromatography. If a refrigerated autosampler is used, subzero-temperature liquid-liquid extraction and injection of the samples can be performed automatically.
Received for review November 18, 1998. Accepted February 6, 1999. AC981276G
Analytical Chemistry, Vol. 71, No. 9, May 1, 1999
1921
