T
inn
150
120
90
6C
15
Tim. m , n 5
I
100
I
150
120
90 Timc.mins
60
30
Figure 10. Chromatograms of the 5 components of Figure 4 at window compositions indicated in Figure 8 using mechanically-mixed packings
0
Figure 9. Chromatograms of the 5 components of Figure 4 at the three best column compositions indicated in Figure 8: ( a ) 4 s = 0.9200, I#JDNNP = 0.0388,~ D P T C= 0.0412; (b)4 s = 0.9000, #JDNNP = 0.0715, ~ D P T C= 0.0285;( c ) 4 s = 0.8800,4 ~ ~ j = - p0.1038, ~ D P T C= 0.0162
tions, e.g., alkane/alkane/alkane as well as to supposedly interacting systems, e.g., electron donor/electron acceptor/ base solvent, it seems that ideas on weak molecular complexing also need re-examination. From the analytical point of view, new horizons are quite clearly presented. Very complex analyses may now be contemplated on a single column-single run basis. In addition, this new method makes it possible to state precisely what is required in terms of column length and performance. T h e widespread adoption of our method, however, has certain implications. In particular, its use demands that, in future, d a t a be reported as partition coefficients, specific retention volumes (and densities), or as retentions relative to some simple and easily available standard solute. Reporting of d a t a in the form of one or another of the several retention indices will clearly be of little value to those interested in multisolvent work, since, a t worst, the d a t a will be useless and, a t best, will demand much arithmetic revision. Finally, since, as we have shown, mixed packings work as well as mixed solvents, it should be possible to attack a
wide range of problems with a stock of a dozen or so prepared packings, which will only require mixing in the correct proportions. We hope, in fact, soon to report on an approach to developing a computerized library storage of GC data, such t h a t the analyst need eventually only tell the computer which compounds he wishes to separate; the computer will then respond with the exact composition of stationary phases necessary, the length of column needed, t h e temperature, and the order of elution of the components.
LITERATURE CITED (1) J. H. Purnell and J. M. Vargas de Andrade, J. Am. Chem. SOC.,97, 3585 (1975). (2) J. H. Purnell and J. M. Vargas de Andrade, J. Am. Chem. SOC.,97, 3590, 11R75) . - . _,. ~
. . isi (7) (8) (9)
R. J. Laub and J. H. Purnell, J. Am. Chem. SOC.,98, 30 (1976). R . J. Laub and J. H. Purnell, J. Am. Chem. SOC.,98, 35 (1976). R. J. Laub and J. H. Purnell, J. Chromatogr.. 112, 71 (1975). R. J. Laub and J. H. Purnell, paper presented at 10th International Advances in Chromatography Symposium, Munich, Germany, 1975. H. Miyake, M. Mitooka, and T. Matsumoto, Bull. Chem. SOC.Jpn, 38, 62 (1965). M. Mitooka, Jpn Anal., 21, 189 (1972). S.H. Langer and J. H. Purnell, J. Phys. Chem., 70, 904 (1966).
RECEIVEDfor review November 17, 1975. Accepted January 16, 1976. R.J.L. acknowledges the Foxboro Company for financial support.
Determination of Trace Hazardous Organic Vapor Pollutants in Ambient Atmospheres by Gas Chromatography/Mass Spectrometry/Computer E. D. Pellizzari," J. E. Bunch, R. E. Berkley,
and J. McRae
Chemistry and Life Sciences Division, Research Triangle Institute, P.O. Box 72794, Research Triangle Park, N.C. 27709
Tenax GC cartridges were used to collect organic vapors in the ambient air of Houston, Tex., the Los Angeles, Calif. Basin, and the Raleigh, N.C. area. The vapors were thermally desorbed and analyzed by a capillary gas-liquid chromatograph coupled to a mass spectrometer. An on-line computer recorded data on magnetic tape and generated
normalized mass spectra and mass fragmentograms. The ubiquitous background of hydrocarbons from automobile exhaust were substantially resolved from each other, and 21 halogenated hydrocarbons were detected, including the carcinogens vinyl chloride and trichloroethylene, as well as numerous oxygen, sulfur, nitrogen, and silicon compounds. ANALYTICAL CHEMISTRY, VOL. 48, NO. 6, MAY 1976
0
803
We have described (1) a technique for collection of organic vapor phase pollutants from ambient air by concentrating them on a sorbent medium. We have also reported (2) development of a thermal desorption inlet manifold for quantitative recovery and transfer of the trapped vapors to an analytical system, and we have examined ( 3 ) the performances of several commercially available sorbent media under different environmental sampling conditions. We now report the incorporation of this method into an analytical system featuring use of a capillary gas chromatographmass spectrometer interfaced with a computer. T h e result is a substantial improvement on previously reported methods (4-18) because it combines the versatility and established reliability of collection on a sorbent (1-3, 19) with the great sensitivity of capillary GC/MS/computer analysis. We have collected samples from several geographical areas which were characterized by the presence of large scale industrial activity and/or the occurrence of photochemical smog formation. Since we had little specific information as to which compounds might be found in the atmosphere, we began by attempting to collect and identify whatever vapors were present, with special emphasis on any which could be considered biologically hazardous. Samples were taken in the Houston, Tex. metropolitan area, the Los Angeles, Calif. Basin, and the Raleigh, N.C. area.
EXPERIMENTAL Preparation of Sampling Cartridges. Tenax GC (2,6-diphenyl-p-phenyleneoxide polymer) was obtained from Applied Science, State College, Pa. The sorbent was purified by extracting with methanol for 18 h in a Soxhlet apparatus. Sampling cartridges were Pyrex glass tubes (1.5 cm i.d. X 10 cm long) containing a 6-cm depth of sorbent bed (36/60 mesh) supported by 1-cm plugs of glass wool. Prior to use, glass tubes, Corex centrifuge tubes, and glass wool were heated in a furnace to 500 "C for 2 h in order to remove traces of organic material. Tenax GC sampling cartridges were conditioned at 270 OC for 20 min under He flow (ca. 50 ml/min) in a thermal desorption chamber (19) to remove traces of background vapors, and then transferred directly to Corex tubes with Teflon lined screw caps for cooling to ambient temperature. Additional glass wool provided a cushion for transportation of cartridges in centrifuge tubes to and from field sampling sites. Field Sampling. Ambient air was drawn through the sampling cartridges using a Universal Sampler 5-1068 equipped with a Teflon multiport head (20). A glass fiber filter (Type E, Gelman Inst. Co., Ann Arbor, Mich.) removed particulates before the air entered the sorbent bed. Unused reference control cartridges were also analyzed after each sampling trip. Gas Chromatography/Mass Spectrometry. Collected vapors were recovered using a thermal desorption inlet manifold (2, 20) interfaced to a Varian MAT CH-7 gas chromatograph/mass spectrometer/620i computer. In a typical desorption cycle a sampling cartridge was placed in the thermal desorption chamber (270 "C), and He was used to purge the vapors into a liquid Nz cooled Ni capillary trap (0.25 cm i.d. X 0.5 m long). After thermal desorption was complete (ca. 4 min) the six-port valve (180 OC) was rotated and the temperature on the capillary loop was rapidly raised to 180 OC whereupon the He GLC carrier gas (4 ml/min) carried the vapors onto support-coated open-tubular (SCOT) columns. SCOT capillary columns (200 f t OV-17, 200 ft OV-101, and 400 ft OV-101, Perkin-Elmer Corp., Norwich, Conn.) were programmed from 20 to 220 OC at 4 "C/min. The carrier gas stream entered the mass spectrometer through a single stage glass jet separator (Model No. 0512-42158, Finnigan Corp., Palo Alto, Calif.) which was pumped with an ED40 Welch vacuum pump. The jet separator and quartz microprobe inlet (0.5 cm o.d., 0.3 mm i.d. X 13 cm long) were maintained at 200 "C. The ion source vacuum Torr. The trace of water which was trapped during was 2-4 X sampling eluted as a broad peak at approximately 10-20 min but did not interfere with the identification of constituents. Mass spectra were continuously obtained at a scan rate of 1 s/ decade from m / e 30 300 throughout the chromatographic development and automatically accumulated along with retention time
-
804
ANALYTICAL CHEMISTRY, VOL. 48, NO. 6, MAY 1976
Table I. Sampling Protocol for Houston, Tex., Los Angeles, Calif., and Raleigh, N.C. Areas Sampling Sampling rate, time,a h l./min
Site May St. Houston, Tex. Stuebner Airline Rd. Houston, Tex. Shaw Dr. Pasadena, Tex. Texas City, Tex.
24
7200 Bayway Dr. Baytown, Tex. Arizona Ave. Santa Monica, Calif. Phillips Ave. West Covina, Calif. East Laurel Glendora, Calif. West Broadway Anaheim, Calif US.lN, 11 miles North of Raleigh, N.C. U.S. 401N, 8 miles North of Raleigh, N.C.
24
Environmental conditions
15 11/6-11/7/74,4:30-4:30 p.m., heavy overcast, rain, 50-70 O
F
24
15 1118-11/9/74, 10:30-10:30 a.m., cloudy, 70 O F
24
15 11/11-11/12/74, 11:30-11:30 a.m., clear, northwest wind 10-12 mph, 51-64 OF 15 11/12-11/13/74, 2:15-2:15 p.m., clear, northwest wind 8-10 mph, 48-64 O F 15 11/13-11/14/74, 4:30-4:30 p.m., clear, south wind 10-12 mph, 45-64 "F
24
8
6.25
15 3/31/75, 8:50 a.m.-5:00 p.m., clear, west wind 10 mph, 55 OF 15 4/1/75, 1O:OO a.m.-4:15 p.m., partly cloudy, south wind 5
mph, 62 OF 11
15 4/2/75, 1O:OO a.m.-9:00 p.m., sunny, no wind, 70 O F
7.5
15 4/3/75, 9:30 a.m.-5:OO p.m., sunny, no wind, 70 O F
4.25
40 10/3/74, 12:05-4:22 p.m., clear, north wind 10 mph, 54-59 O F
48
22
11/25/74-11/27/74, 3:OO p.m.1O:OO a.m., 4 h light rain, then clear 20-55 O F
Four replicate cartridges were simultaneously used on the multiport sampling head during field sampling. data on magnetic tape for further processing by an on-line Varian 620i computer. Acquired spectra were converted by computer software programs into a sequence of normalized mass spectra which were correlated to peak retention times on total ion current plots. Listings of mass spectrum numbers along with the number of peaks in each cracking pattern, the total maximum and minimum peak intensities, and standard deviations from calibration were also generated. After mass conversion, mass fragmentograms of selected masdions were plotted. Identification of resolved components was achieved by comparing cracking patterns with an index of mass spectra (21) and/or computer library searching using the Cyphernetics MSS System (Ann Arbor, Mich.). In several cases, the identification was confirmed by comparison with the mass spectrum and the retention time on two different columns of an authentic sample. Particular note was made of the relationship of the boiling point to elution temperature and to elution order, since the OV-101 stationary phase capillaries separated homologous series in order of boiling point.
RESULTS AND DISCUSSION Sampling locations and prevailing conditions have been summarized in Table I. Figure 1 shows a total ion current chromatogram for a sample taken in the Los Angeles, Calif. Basin. Mass fragmentograms were obtained by computer manipulation of the data which were coljected and stored
TEMPERATURE
('Cl
10
-
I
so
32
44
56
68
80
92
IC4
116
.. 128
140
1
152
B Figure 1. (A) Total ion current chromatogram of ambient air sample from Glendora, Calif. A 400 ft OV-101 SCOT programmed from 20 to 230 OC at 4 OC/min was used. (B) Total ion current chromatogram of Tenax GC cartridge blank. A 400 ft OV-101 SCOT was used (1) Difluorochloromethane; (2) 1-butene; (3) isobutane; (4) unknown; (5) unknown; (6) isopentane, trichlorofluoromethane; (7) 1-pentene, C5Ha; (8) +pentane; (9) isoprene; (10) methylene chloride; (11) propanal; (12) acetone; (13) unknown; (14) unknown; (15) 2-methylpentane, 2-fluoro-2-methylpropane; (16) 3-methylpentane; (17) C6H12; (18) 3-methyl-2-pentene 4- n-hexane, 2-methylfuran; (19) chloroform; (20) C6H14; (21) CsH12: (22) methyl vinyl ketone (tent.), methyl ethyl ketone; (23) 1,l,l-trichloroethane. ethyl acetate; (24) benzene; (25) carbon tetrachloride; (26) C6ii12; (27) 2,3-dimethylpentane; (28) 1,1.3,3-tetramethylcyclopentane; (29) cyclohexenol isomer; (30) unknown; (31) l-trans-2-dimethylcyclopentane, C7H16. C7H12.trichloroethylene; (32) +heptane; (33) C7H12; (34) C6H14; (35) 2,2,3,34etramethyIbutane; (36) 4,4-dlmethyl-2-pentene; (37) 2,5-dimethylhexane; (38) 3-heptene; (39) CaH16; (40) 1-octene; (41) 2,3-dimethylhexane; (42) toluene; (43) 2,4-dlmethylhexane; (44) 2,2-dimethyl-3-ethylpentane; (45) trimethylcyclopentane isomer; (46) n-octane; (47) hexamethylcyclotrisiloxane; (48) tetrachloroethylene; (49) unknown; (50) 2,3,4-trimethylhexane; (51) CaH16; (52) CgH20; (53) n-butyl acetate; (54) CgHla; (55) chlorobenzene; (56) ethylbenzene; (57) pxylene: (58) mxylene; (59) 2,2,5,54etramethylhexane; (60) CgHls; (61) unknown; (62) CgH20; (63) &xylene; (64, 65) C10H22 (isomers), CgHia; (66) ClOH22; (67) isopropyibenzene, CSH16; (68) 2,6-dimethyloctane; (69) C10H20; (70) n-propylcyclohexane; (71) 4-ethyl-3-octene; (72) unknown; (73) 2-methyl-3ethylheptane; (74) 4-methylnonane: (75) 5-methyldecane; (76) n-propylbenzene: (77) 2,6-dlmethyloctane; (78) octamethylcyclotetrasiloxane; (79) C11H24; (80) rn-ethyltoluene; (81) Cl0HZ0; (82) unknown, n-decane; (83) 1,3,5-trimethylbenzene: (84) unknown; (85) unknown; (86) isobutylbenzene; (87) sec-butylbenzene; (88) mdichiorobenzene; (89) CllH24; (90) C4-alkylbenzene; (91) 1,2,4-trlmethylbenzene; (92) unknown; (93) unknown; (94) 4-ethyl-1-octyn-3-01: (95) &diethylbenzene; (96) ppropyltoluene; (97) mdiethylbenzene; (98, 99) C11H24 (isomers); (100) sec-butylbenzene; (101) 1,4-dimethyi-2-ethylbenzene; (102) C i 1H22; (103) 1,3-dimethyi-3-ethylbenzene; (104) n-undecane; (105) Cs-alkylbenzene; (106) CllH22, CllH20; (107) C12H26, CllH20; (108) n-dodecane; (109) 2Wimethyldecane; (110) unknown; (111) CllH20, C12H24; (112) CllH20; (113) C12H24, C12H26: (114) CllH24: (115) C12H26: (116) unknown; (117) CllH24; (118) C12H26; (119) naphthalene.
MASS SPECTRUM NO.
Figure 2. Mass fragmentograms of characteristic ions representing carbon tetrachloride (m/e 117), tetrachloroethylene (m/e 166), and m d i chlorobenzene ( m / e 146). See Figure 1 for total ion plot
during this chromatographic separation. T h e single ion plots in Figure 2 demonstrate distinct detection of carbon tetrachloride (mle 117), tetrachloroethylene (mle 1661, and
m-dichlorobenzene (mle 146). No strenuous effort was made to identify all of the individual hydrocarbon isomers which were resolved, although it would have been possible ANALYTICAL CHEMISTRY, VOL. 48. NO. 6, MAY 1976
805
Table 11. Halogenated Hydrocarbons Identified in Ambient Air
Table 111. Oxygen-, Nitrogen-, Sulfur-, and SiliconContaining Organic Vapors in Ambient Air
Geopraohical area Compound Difluorochloromethane Trichlorofluoromethane Vinyl chloride Methyl chloride Ethyl chloride 1,2-Dichloroethane Fluorochloromethane Methylene chloride Chloroform Carbon tetrachloride Trichloroethylene l,l,l-Trichloroethane Tetrachloroethylene
Houston, Tex. and vicinity
Los Angeles Basin X
X X X
X X X X X
X X X
X
X X X
X
X
X
X X
1,2,4)
Trichlorobenzene (? isomer) Bromoform 1,2,3,3-Tetrachloro~ro~ene
X
X
X
X
X
in principle to do this by comparison with authentic samples. Aliphatic hydrocarbon isomers (77) were observed in t h e Los Angeles Basin and 101 in t h e Houston, Tex. area. Also, 37 specific aromatic hydrocarbons were identified in Los Angeles and 36 in Houston, in addition t o 13 C5-alkylbenzene isomers in each area. Of much greater interest were the potentially hazardous compounds which were found in the atmosphere in each area. These included carcinogens, potential mutagens, and toxic substances. A total of 21 halogenated hydrocarbons which were detected have been listed in Table 11. Among them were the carcinogens vinyl chloride (22-25) and trichloroethylene (26, 27). A number of toxic compounds were found t o be widely distributed. Bromoform and tetrachloroethylene, for example, were virtually ubiquitous in both Houston and Los Angeles. A variety of oxygen-, nitrogen-, sulfur-, and silicon-containing materials were also detected (Table 111) despite the fact t h a t the columns used were not especially suitable for chromatography of the more polar molecules. T h e fact t h a t polar, reactive materials such as alcohols, esters, and amines were eluted from stainless steel SCOT capillary columns in quantities detectable against a large hydrocarbon background suggests their presence in rather substantial concentrations. Several of t h e silicon compounds would be plausible artifacts from the stationary phase in the chromatographic column but no evidence for this has been found in blank runs. Furthermore, there are potential sources of silicon-containing pollutants (28). Air samples were also collected near Raleigh, N.C. in a n attempt to establish t h e presence of biphenyl in ambient air. Most of the compounds collected were components of automobile exhaust, b u t an additional peak appeared, the mass spectrum of which was virtually identical with t h a t of a n authentic sample of biphenyl. T h e concentration of biphenyl was estimated at 1-3 ppm 0.5 mile down wind from the suspected source. This estimate was based on a comparison of the response of the GLC-MS system to measured quantities of biphenyl. 806
Compound
X
2-Fluoro-2-methylpropane
Chlorobenzene rn-Dichlorobenzene o-Dichlorobenzene Trichlorobenzene (probably
Geographical area
ANALYTICAL CHEMISTRY, VOL. 48, NO. 6, MAY 1976
Acetaldehyde Acetone Ethyl acetate Propanal Amyl acetate Furan 2-Methylfuran 2-Propanol Acetyl acetone Styrene oxide (tent.) Cyclohexanol p-Tolualdehyde Ethyl hexyl ketone Methyl isobutyl ketone Acetophenone Diethyl ether Methyl ethyl ketone Methyl n-propyl ether n-Propyl acetate Ethanol (tent.) Phenylacetaldehyde Benzaldehyde Methyl vinyl ketone (tent.) n-Butyl vinyl ether 4-Phenyl butan-2-one p -Methylanisole 2-Methylbenzo(B)furan Methyl phthalide 2,6-Ditertiarybutyl-p-cresol Dimethyl disulfide n-Butylamine (tent.) Dimethylfuran isomer (tent.) Sulfur dioxide Hexamethylcyclotrisiloxane Octamethylcyclotetrasiloxane (tent.) Tetradecamethylhexasiloxane
Houston, Tex. and vicinity
Los Angeles Basin
X X X
X X X
X X
X
X X X X
X X X
X
X
X X X
(tent.) Tetramethylsilane (tent.) Diethylfuran (tent.) Benzonitrile a-Cyanopyridine
X
ACKNOWLEDGMENT Richard Flannery and Lloyd Stewart of the Texas Air Quality Board, Walter Crider of the National Environmental Research Center, Research Triangle Park, N.C., and J i m Davis and Bob Hoshide of Rockwell International Science Center, Thousand Oaks, Calif. are thanked for help in site selection and the use of facilities. T h e helpful suggestions of E. Sawicki of the Environmental Protection Agency, Research Triangle Park, N.C., and M. E. Wall and J. Bursey of RTI are appreciated.
LITERATURE CITED (1)E. D. Pellizzari, J. E. Bunch, E . H. Carpenter, and E. Sawicki, Environ. Sci. Technol., 9,552 (1975). (2)E. D. Pellizzari, B. H. Carpenter, J . E. Bunch, and E. Sawicki, Environ. Sci. Technol.. 9,556 (1975). (3)E . D, Pellizzari, J. E. Bunch, R. E. Berkley, and J. McRae, Anal. Lett., in
press. (4)Methods in Air Sampling and Analysis, American Public Health Association, Washington, D.C., 1972,p 480. (5) D. L. Brooman and E. Edgeley, J. Air Pollut. Control Assoc., 16, 25 (1966). (6)F. W. Williams and M. E. Umstead, Anal. Chem., 40,2232 (1968). (7)D. C. Leggett. R . P. Murrmann, T. J. Jenkins, and R. Barriera, "Method of Concentrating and Determining Trace Organic Compounds in the Atmosphere", Cold Regions Res. Eng. Lab., Hanover, N . H . . U.S. Nat. Tech. AD Rep. No. 745125,1972,p 14. (8)I . H. Williams, Anal. Chem., 37, 1723 (1965).
(9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19) (20)
J. W. Russell, Environ. Sci. Technol., 9, 1175 (1975). J. S. Parsons and S. Mitzner, Environ. Sci. Technol., 9, 1053 (1975). D. Ellgehausen, Anal. Lett., 8, 11 (1975). M. Novotny, M. L. Lee, and K. D. Bartle, Chromatographia, 7, 333 (1974). A. Zlatkis, H. A. Lichtenstein, and A. Tishbee. Chromatographia, 6 , 67 (1973). D. Schuetzle, A. L. Crittenden, and R . J. Charlson, J. Air follut. Control Assoc., 23, 704 (1973). P. W. Jones, "Analysis of Nonparticulate Organic Compounds in Ambient Atmospheres", 67th Air Pollution Control Association Meeting, Denver, Colo.. Paper No. 74-265, June 1973. K. Grob and G. Grob, J. Chromatogr., 62, 1 (1971). W. Bertsch, R. C. Chang, and A. Zlatkis, J. Chromatogr. Sci., 12, 175 (1974). A. Raymond and G. Guiochon, Environ. Sci. Techno/., 8, 143 (1974). E. D. Pellizzari, "Development of Analytical Techniques for Measuring Ambient Atmospheric Carcinogenic Vapors", Publication No. EPA60012-75-076, Contract No. 68-02-1228, November 1975, p 187. E. D. Pellizzari, "Development of Method for Carcinogenic Vapor Anal-
(21) (22) (23) (24) (25) (26) (27) (28)
ysis in Ambient Atmospheres", Publication No. EPA-650/2-74-12 1, Contract No. 68-02-1228, July 1974, p 148. Eight Peak Index of Mass Spectra, Vol. I (Tables 1 and 2) and II (Table 3), Mass Spectrometry Data Centre, AWRE, Aldermaston, Reading, RG74PR, UK, 1970. J. L. Cresch, Jr., and M. N. Johnson, J. Occup. Med., 16, 150 (1974). I. R . Tabershaw and W. R. Garrey, J. Occup. Med., 16, 509 (1974). R . R . Monson. J. M. Peterson, and M. N. Johnson, Lancet, 397 (1974). H. J. Martsteller, W. K. Lelbach. R . Muller, and P. Gedigke, presented at the Working Group, Toxicity of Vinyl Chloride-Polyvinyl Chloride, New York Academy of Sciences, New York, May 10-11, 1974. Chem. Eng. News. 53,41 (May 19, 1975). Chem. Eng. News, 53,6 (May 5 , 1975). R. West and E. Carberry. Science, 189, 179 (1975).
RECEIVEDfor review September 22, 1975. Accepted February 5, 1976. Research supported by EPA Contract No. 6802-1228 from the Environmental Protection Agency, Health, Education, and Welfare.
Determination of Debrisoquin and Its 4-Hydroxy Metabolite in Plasma by Gas Chromatography/Mass Spectrometry Susan L. Malcolm and Timothy R. Marten* Department of Biochemistry, Roche Products Ltd, Broadwater Road, Welwyn Garden City, Herts. England
An analysis of debrlsoquin and its 4-hydroxy metabolite In plasma has been developed. After derivatization with hexafluoroacetylacetone, samples, containing deuterated internal standard, were examined by gas chromatography/mass spectrometry. Linear responses for the drug and metabolite to 1 ng/ml arld 5 ng/ml plasma, respectively, were observed. The method has been applied to the measurement of plasma levels after therapeutic dosing.
Metabolic studies ( I ) on the anti-hypertensive drug, debrisoquin sulfate (Ro 05-3307) (I), have shown t h a t the major metabolite is the 4-hydroxy compound (11) (Ro 037594). This compound has also been shown t o exhibit some hypotensive activity in anaesthetized cats (I. L. Natoff and T. C. Hamilton, personal communication), and it is therefore desirable t h a t an analytical method for debrisoquin in blood should also be capable of measuring the metabolite. Two methods for the determination of debrisoquin in blood have been described, but neither is suitable for the estimation of the metabolite. The method of Medina et al. (2) involves the direct extraction of the drug a t high p H and measurement of the fluorescence of the ninhydrin derivative. The sensitivity of the method which is only 50 ng/ml plasma, is not sufficient t o measure the levels found in patients receiving therapeutic doses of the drug. I t also suffers from the disadvantage t h a t the 4-hydroxy metabolite cannot be extracted quantitatively into organic solvents a t any p H ( 1 ) . The alternative method ( 3 ) relies on the hydrolysis of the amidino group with strong base, followed by gas chromatography with an electron capture detector or gas chromatography/mass spectrometry (GC/MS) of the derivatized tetrahydroisoquinoline. However, under the hydrolysis conditions employed, the 4-hydroxy compound is unstable. The method described here is based on the extraction procedure ( I ) used to isolate metabolites from biological samples. By condensing the amidino group with acetylace-
tone to form a pyrimidino compound, the polarity was reduced sufficiently for the drug and metabolites t o be extracted into organic solvents. The use of hexafluoroacetylacetone in a two-phase reaction mixture as described by Erdtmansky and Goehl ( 4 ) made it possible to analyze low levels of debrisoquin and its 4-hydroxy metabolite by monitoring for single ions characteristic of the bis(triflu0romethy1)pyrimidines (111) and (IV) on a GC/MS system, using derivatized decadeuteriodebrisoquin (V) as an internal standard. In all cases, the monitored peaks corresponded t o the base peaks in the mass spectra of the individual compounds (Figure 1). The method has been successfully applied t o the measurement of drug and metabolite levels in plasma taken from a patient treated with a single dose of debrisoquin sulfate. R
&N-CZ"
R
"2
I
R = H
111
R = H
11
R
=
IV
R=OH
v
o,olll
OH
EXPERIMENTAL General. A F i n n i g a n 1015D gas chromatographiquadrupole mass spectrometer f i t t e d w i t h a glass c o l u m n (1.5 m, i.d. 2 m m ) packed w i t h 3% OV-17 o n Gas C h r o m Q was used. H e l i u m was used as carrier gas (flow rate 20 m l / m i n ) a n d t h e oven was p r o grammed f r o m 150-190 O C a t 4 " l m i n . A single stage glass j e t separ a t o r a t 210 "C was employed. T h e m a n i f o l d temperature was 60 O C a n d t h e electron voltage, 70 eV. I o n currents a t m / e 344, 347, a n d 356 (see Figure 1) were m o n i t o r e d w i t h a 4-channel automatic peak selector w i t h a channel w i d t h o f 0.5 a m u a n d a sampling t i m e o f 10 mdchannel. T h e integrated signals were o u t p u t o n t o a rnultichannel recorder. T h e derivatives were characterized o n a Varian CH7 mass spectrometer (Figure 1). ANALYTICAL CHEMISTRY, VOL. 48, NO. 6, M A Y 1976
807
