Capillary Gas Chromatography with Cryogenic Oven Temperature for

for chloroform measurements, and vice versa), the procedure was the same as described above. GC Conditions. GC analyses were carried out on an HP 6890...
3 downloads 0 Views 276KB Size
Anal. Chem. 1997, 69, 5178-5181

Capillary Gas Chromatography with Cryogenic Oven Temperature for Headspace Samples: Analysis of Chloroform or Methylene Chloride in Whole Blood Kanako Watanabe,* Hiroshi Seno, Akira Ishii, and Osamu Suzuki

Department of Legal Medicine, Hamamatsu University School of Medicine, 3600 Handa-cho, Hamamatsu 431-31, Japan Takeshi Kumazawa

Department of Legal Medicine, Showa University School of Medicine, 1-5-8 Hatanodai, Shinagawa-ku, Tokyo 142, Japan

A new and sensitive gas chromatography (GC) method for measurement of chloroform or methylene chloride in whole blood is presented. Trace levels of these analytes present in the headspace of samples were cryogenically trapped prior to on-line GC analysis. After heating of a blood sample containing chloroform and methylene chloride (internal standard, and vice versa) in a 7.0-mL vial at 55 °C for 20 min, 5 mL of the headspace vapor was drawn into a glass syringe. All vapor was introduced into an Rtx-Volatiles middle-bore capillary column in the splitless mode at -30 °C oven temperature to trap the entire analytes, and the oven temperature was programmed up to 280 °C for detection of the compounds and for cleaning of the column. The present conditions gave sharp peaks for both chloroform and methylene chloride and very low background noises for whole blood samples. As much as 11.5 and 20.0% of chloroform and methylene chloride, respectively, which had been added to whole blood in a vial, could be introduced into the GC column. The calibration curves showed linearity in the range of 0.05-5.0 µg/0.5 mL of whole blood. The detection limit was estimated to be about 2 ng/0.5 mL. The coefficients of intraday and interday variations were 1.31 and 8.90% for chloroform and 1.37 and 9.03% for methylene chloride, respectively. The data on chloroform or methylene chloride in rat blood after inhalation of each compound were also presented. Accidental, suicidal, or homicidal cases of death due to chloroform or methylene chloride poisoning are occasionally encountered in forensic science practice.1-3 Exposure to methylene chloride for industrial workers is also a problem from a hygienic point of view.4 There have been many reports dealing with analysis of chloroform or methylene chloride by gas chromatography (GC) with the headspace method.3,5-10 In some of these reports, conventional packed columns, which give relatively (1) Forrest, A. R. W.; Marsh, I.; Usher, A. Bull. Int. Assoc. Forensic Toxicol. 1990, 20 (3), 26-27. (2) Ryall, J. E. Bull. Int. Assoc. Forensic Toxicol. 1990, 20 (3), 28-29. (3) Kim, N.; Park, S.; Suh, J. J. Forensic Sci. 1996, 41, 527-529. (4) Ghittori, S.; Marraccini, P.; Franco, G.; Imbriani, M. Am. Ind. Hyg. Assoc. J. 1993, 54, 27-31.

5178 Analytical Chemistry, Vol. 69, No. 24, December 15, 1997

low sensitivity and poor separation, were used.3,5,8 With widebore capillary columns, only a 0.1-0.5-mL volume of the headspace vapor can be injected;6,7 with middle-bore capillary columns, split injection has to be used.9 Purge-and-trap sample concentration followed by capillary GC is probably the most sensitive technique to detect volatile organic compounds from a large volume of water11 or solid12,13 samples. However, this technique is not suitable for biological samples, such as blood and tissue homogenates, because it causes serious bubbling. Recently, a microcomputer controlling cryogenic oven temperatures below 0 °C has become available for new types of GC instruments. It was originally designed for rapid cooling of the oven to reduce time for analysis. In this paper, we describe a new technique which allows us to inject as much as a 5-mL volume of the headspace vapor without any loss into a middle-bore capillary column by use of a cryogenic oven temperature, resulting in much higher sensitivity. EXPERIMENTAL SECTION Materials. Chloroform, methylene chloride, and methanol were obtained from Wako Pure Chemical Industries (Osaka, Japan). An Rtx-Volatiles fused silica middle-bore capillary column (30 m × 0.32 mm i.d., film thickness 1.5 µm) was purchased from Restek Corp. (Bellefonte, PA). Human whole blood was obtained from a local Red Cross Center or from healthy subjects. Procedure. Stock solutions (10 µg/mL) of chloroform or methylene chloride were prepared in distilled water with shaking. To a 7.0-mL screw-cap vial containing 0.5 mL of whole blood was added 0.5 mL of distilled water containing 5 µg or less of the above compounds (either of them as internal standard, IS). The vial (5) Di Pasquale, G.; Capaccioli, T. J. Chromatogr. 1981, 206, 589-593. (6) McNeal, T. P.; Hollifield, H. C. J. Assoc. Off. Anal. Chem. 1990, 73, 328331. (7) Morgan, D. L.; Cooper, S. W.; Carlock, D. L.; Sykora, J. J.; Sutton, B.; Mattie, D. R.; McDougal, J. N. Environ. Res. 1991, 55, 51-63. (8) Rao, K. N.; Virji, M. A.; Moraca, M. A.; Diven, W. F.; Martin, T. G.; Schneider, S. M. J. Anal. Toxicol. 1993, 17, 99-102. (9) Seto, Y.; Tsunoda, N.; Ohta, H.; Shinohara, T. J. Anal. Toxicol. 1993, 17, 415-420. (10) Page, B. D.; Conacher, H. B. S.; Salminen, J.; Nixon, G. R.; Riedel, G.; Mori, B.; Gagnon, J.; Brousseau, R. J. AOAC Int. 1993, 76, 26-31. (11) Wampler, T. P.; Bowe, W. A.; Levy, E. J. J. Chromatogr. Sci. 1985, 23, 64-67. (12) Ju ¨ ttner, F.; Wu ¨ rster, K. J. Chromatogr. 1979, 175, 178-182. (13) Saferstein, R.; Park, S. A. J. Forensic Sci. 1982, 27, 484-494. S0003-2700(97)00468-X CCC: $14.00

© 1997 American Chemical Society

was rapidly sealed with a silicone septum cap and heated at 55 °C on an aluminum block heater. After heating for 20 min, a 24gauge needle of a glass syringe (5-mL volume) was passed through the septum. A 5-mL volume of the headspace vapor was drawn into the syringe and injected into the GC port in the splitless mode at -30 °C of oven temperature. For analysis of chloroform or methylene chloride in whole blood of rats after exposure of either solvent, the blood was diluted 40-50-fold with distilled water, because of high concentrations of the compounds. After addition of 5 µg of IS (methylene chloride for chloroform measurements, and vice versa), the procedure was the same as described above. GC Conditions. GC analyses were carried out on an HP 6890 series gas chromatograph equipped with flame ionization detection (FID) and with a cryogenic oven temperature device (HewlettPackard Co., Palo Alto, CA). An electrically operated solenoid valve introduces liquid carbon dioxide at a rate appropriate to cool the oven to desired temperatures. The GC conditions were column temperature -30 to 280 °C (1 min hold at -30 °C, 10 °C/min from -30 to 100 °C, 20 °C/min from 100 to 280 °C); injection temperature 250 °C; detection temperature 280 °C; and helium flow rate 3 mL/min. The vapor samples were injected in the splitless mode, and the splitter was opened 1 min after completion of the injection. Animal Experiments. Male Sprague-Dawley rats, weighing about 200 g, were used. Each animal was put into a glass container (space volume 4050 cm3), where gauze soaked with 10 mL of chloroform or methylene chloride had been placed at a corner of the container bottom. After the top of the container was covered, we waited until the animal fell into deep anaesthesia by inhalation of vapor of either chloroform or methylene chloride. The time required for deep anaesthesia was 3-10 min for chloroform and 5-20 min for methylene chloride. Under the anaesthesia, the animal was rapidly subjected to laparotomy, and about 5 mL of arterial blood was drawn from the abdominal aorta in the presence of heparin and NaF; the animal was thus killed by exsanguination. The tube containing the blood was tightly capped, and the analysis of the solvents by GC was made within the same day of the samplings. RESULTS AND DISCUSSION Analytical Conditions. Various conditions for the headspace extraction of chloroform and methylene chloride from whole blood were tested. We heated the vials at 55 °C for 10, 20, and 30 min; it was found that optimal extraction into the headspace was attained at 20 min of heating. We have tested various initial oven temperatures for trapping chloroform and methylene chloride vapor (Figure 1). At 0 °C, the peaks of both compounds were relatively broad and became sharper upon lowering the oven temperature to -40 °C; such trend was more marked for methylene chloride than that for chloroform. At -40 °C, however, some background noises appeared. In order to maximize peak sharpness and minimize background noises, -30 °C was selected for trapping both compounds. We have used an Rtx-Volatiles column for the cryogenic oven temperature trapping at -30 °C. According to the manufacturer’s note, the temperature range should be 0-280 °C. In our experience, however, this column was found to be resistant to the low temperature at -30 °C; it could be used for at least 3 months with good reproducibility.

Figure 1. Headspace capillary GC for chloroform (2) and methylene chloride (1) as a function of various initial oven temperatures. Five micrograms of each compound was added to 0.5 mL of human whole blood for headspace extraction.

Reliability of the Method. Figure 2 shows gas chromatograms for nonextracted authentic chloroform and methylene chloride (5 µg on column) dissolved in methanol and for headspace extracts from 0.5 mL of human whole blood in the presence (5 µg each) and absence of the compounds. The retention times for methylene chloride and chloroform were about 8.0 and 10.0 min, respectively. The backgrounds gave very small impurity peaks; no interfering peaks appeared around the test peaks (Figure 2, bottom panel). Analytical Chemistry, Vol. 69, No. 24, December 15, 1997

5179

Figure 2. Capillary GC chromatograms with cryogenic trapping at -30 °C for the authentic chloroform (2) and methylene chloride (1) with direct injection (top panel), for whole blood spiked with 5 µg of each compound in 0.5 mL (middle panel), and for whole blood in the absence of the compounds (bottom panel). The vertical scale of the authentic chloroform (top panel) is not the same as those of the middle and bottom panels.

The net recovery of chloroform and methylene chloride was determined. Peak areas of blood spiked with a known amount (5 µg/0.5 mL) of each compound (after cryogenic trapping of the headspace prior to GC analysis) were compared with peak areas obtained by direct GC injection of the authentic compounds. The net recovery was 11.5 ( 1.31% (mean ( SD, n ) 5, CV ) 11.4%) for chloroform and 20.0 ( 1.80% (mean ( SD, n ) 5, CV ) 9.0%) for methylene chloride. The different recovery values are probably due to differences in partitioning of each compound between gas and liquid phases in heated vials.14 Calibration curves for chloroform and methylene chloride in human whole blood were drawn by plotting eight concentrations (14) Seto, Y. J. Chromatogr. A 1994, 674, 25-62.

5180 Analytical Chemistry, Vol. 69, No. 24, December 15, 1997

Figure 3. Headspace capillary GC with cryogenic oven temperature for human whole blood (0.5 mL) which had been spiked with 0.01 µg of chloroform (A) or 0.01 µg of methylene chloride (B) as indicated with arrows; a chromatogram for blank blood without addition of either compound is also shown (C). The amount of each IS was 5 µg/0.5 mL of blood.

against 5 µg of IS (for measurements of chloroform, methylene chloride was IS, and vice versa). They were linear in the range of 0.05 - 5.0 µg/0.5 mL of blood. The equations and r values for the curves were y ) 0.120 x + 0.003 54 and r ) 0.998 for chloroform; y ) 0.360 x + 0.107 and r ) 0.993 for methylene chloride. Figure 3 shows gas chromatograms obtained from headspace extracts of human whole blood, to which 5 µg of chloroform (IS) together with 0.01 µg of methylene chloride (panel A), and 5 µg of methylene chloride (IS) together with 0.01 µg of chloroform (panel B), had been added; the chromatogram for the headspace extract without addition of either compound (negative control) is also presented (panel C). As can be seen in the figure, 0.01 µg

Figure 4. Headspace capillary GC with cryogenic oven temperature for chloroform (B) and methylene chloride (A) in rat whole blood after inhalation of each compound until a deep anaesthetic state of the animals.

of each compound gave a clear peak with a signal-to-noise ratio of more than 10. The detection limit (signal-to-noise ratio ) 3) for both compounds was estimated to be about 2 ng/0.5 mL of blood. Rao et al.8 reported that the detection limit measured by GC-FID with a packed column for chloroform was 0.9 µg/mL plasma. Di Pasquale and Capaccioli5 reported the detection limit of methylene chloride at 30 ng/mL of sunflower oil by GC/ electron capture detection (ECD) with a packed column. If the present cryogenic method is coupled with ECD, the sensitivity is likely to be improved. Certainly, upon coupling the method with the use of selected ion monitoring in mass spectrometry, sensitivity and selectivity would be improved. To check the reproducibility of the present method, we have added 5 µg of each compound to 0.5 mL of whole blood and analyzed each calibration curve. The coefficients of intraday and interday variations were 1.31 and 8.90% for chloroform and 1.37 and 9.03% for methylene chloride, respectively (n ) 5 each). Chloroform and Methylene Chloride in Rat Blood. In addition to the above spiked human whole blood samples, we have measured the levels of chloroform or methylene chloride in rat whole blood under an anaesthetic state after inhalation of each

compound (Figure 4). Since the levels were too high to be measured by the present method, the rat whole blood was diluted 40-50-fold with distilled water. The levels of chloroform and methylene chloride were 794 ( 590 and 1270 ( 354 µg/mL, respectively (mean ( SD, n ) 5 for both compounds). Neither conversion of chloroform to methylene chloride nor vice versa was observed in rats after inhalation of each compound. CONCLUSION To our knowledge, this is the first report dealing with GC with cryogenic trapping for volatile compounds in biological samples. The present method is simple and very sensitive and shows low background noise. The approach described here appears to have potential for analysis of a wide range of volatiles in complex biological matrices. Received for review May 7, 1997. Accepted September 18, 1997.X AC970468T X

Abstract published in Advance ACS Abstracts, November 1, 1997.

Analytical Chemistry, Vol. 69, No. 24, December 15, 1997

5181