by Giddings ( I ) . In gas chromatography, however, we see no additional peak dispersion when the post-column connecting tubing is equal to or less than the diameter of the column. In addition, as was pointed out by Deininger and Halasz ( I @ , the extra column effect diminished as the retention volume (or the partition coefficient) increased, This is expected to hold true both in liquid and gas chromatographic systems. CONCLUSIONS
Systematic study of connecting tubes in gas chromatography indicates some important practical conclusions. The effect of the connecting tubes seems, in general, to be worse when in front of the column. However, as the diameter of the connecting tube decreases, or as the solute becomes more retained (higher partition coefficient), the contribution of the dead volume diminishes. When the diameter of the dead volume is larger than the column, the efficiency of the system will be adversely affected whether the connecting tubes are in the front or the back end of the column. In this connection our results with connecting tubings having equal volume but different diameter should be noted. It is important to realize the effect of the number of plates generated in the connecting tubes. The velocity dependence of the dead volume effects should also be noted. The practical implications of this study are important. When attaching the chromatographic system to post-column instrumentation, the requirements on the connecting tubes
are not too stringent. As long as that tube diameter is at most equal to the chromatographic column’s, the efficiency of the system will be maintained. This supports the contention of many workers who maintain that their system showed no loss of efficiency when connected to, say, a mass spectrometer. On the other hand, the connection between pre-column reaction chamber and the column can present a much greater hazard. This is especially true when the connecting tubes are of the same diameter as the column and when the net retention time of the solutes is small, as often occurs with pyrolyzers. Similarly, large volume injection valves can be detrimental to the efficiency. Much more careful attention must be given to pre-column contribution to the plate height. In general, it can be said that the connecting tubes on the front end of the column should be as small as possible (while keeping the pressure drop in mind) if the researcher is to obtain an efficient gas chromatographic system. ACKNOWLEDGMENT
We would like to thank Stanley Bruckenstein for the use of his Wang 720 calculator which was purchased with an AFOSR grant. We would also like to thank Jane Maynard who did many of the tedious manual measurements of chromatograms. RECEIVED for review January 17, 1972. Accepted March 13, 1972. The support of the Research Foundation of the State University of New York is also gratefully acknowledged.
Quantitative Determination of 5-Hydroxyindole-%AceticAcid in Cerebrospinal Fluid by Gas Chromatography-MassSpectrometry Leif Bertilsson, Arthur J. Atkinson, Jr., James R. Althaus, Ase Harfast, Jan-Erik Lindgren, and Bo Holmstedt Department of Toxicology, Swedish Medical Research Council, and Department of Pharmacology, Dioision of Clinical Pharmacology, Karolinska Institutet, S-104 01 Stockholm 60, Sweden A highly sensitive and specific method for the quantitative determination of 5-hydroxyindole-3-acetic acid (5-HIAA) in human cerebrospinal fluid (CSF) has been developed. The 5-HIAA diheptafluorobutyryl methyl ester derivative has been analyzed by the combination of gas chromatography and mass spectrometry. By the technique called mass fragmentography, the two major ions (m/e 538 and 597) of the mass spectrum of the 5-HIAA derivative were recorded after elution from the chromatographic column. Dideuterium-labelled 5-HIAA has been synthesized and used as an excellent internal standard for the quantitation of 5-HIAA in CSF. The mass fragmentographic analysis takes less than 2 minutes and allows an accurate determination of 5HlAA in the range of 2-50 ng/ml of CSF, when 2 ml of CSF is used for the analysis. The standard deviation of the method is less than 7% in the 8-20 ng/ml range and 1-276 at higher concentrations of 5-HIAA. THESTRIKING ABILITY of certain chemical compounds to alter mood has provoked interest in the biochemistry of mental disease. A derangement of tryptophan metabolism in affective disorders was first suggested by the discovery that reserpine, a drug that causes profound depression in some patients, markedly lowers the concentration of 5-hydroxytryptamine in 1434
ANALYTICAL CHEMISTRY, VOL. 44, NO. 8, JULY 1972
the brain (I). Because 5-hydroxytryptamine is converted by monoamine oxidase in brain tissue to 5-hydroxyindole-3acetic acid (5-HIAA), a number of clinical studies have been conducted in order to determine whether the concentration of this acid in cerebrospinal fluid (CSF) is lower in depressed patients than in patients not suffering from an affective disorder. In some investigations, significantly less 5-HIAA was found in the CSF of depressed patients (2-4) but in others the results were equivocal (5-7). In each of these studies, fluorometric (1) A. Pletscher, P. A. Shore, and B. B. Brodie, J . Pharmacol. Exp. Ther., 116, 84 (1956). (2) G. W. Ashcroft and D. F. Sharman, Nature, 186, 1050 (1960). (3) S. J. Dencker, U. Malm, B-E. Roos, and B. Werdinius, J. Neurochem., 13, 1545 (1966). (4) H. M. van Praag and J. Korf, Psychopharmacologia, 19, 148
(1971).
(5) K.Fotherby, G. W. Ashcroft, J. W. Affleck,and A. D. Forrest, J . Neurol. Neurosurg. Psychiat., 26,71 (1963). (6) M. B. Bowers, Jr., G. R. Heninger, and F. Gerbode, Int. J. Neuropharmacol., 8, 255 (1969). (7) B-E. Roos and R. Sjostrom, Pharmacol. Clitz., 1, 153 (1969).
100
80
538 [loo%)
143
ae
4e
-
197 COC3F7
60-
169 C3F7
.'-=P
341
40-
69 CF3
er al
20 -
Molw. 597 5-HIAA-Methyl ester-HFBp
----
x 2 7,
(325.197)
59
\
128 (538.213) 325
0
\
384
I I
400
I
I
I\,
405*
\
I
\
600
.
100
80
I C3F7
8
E
-
60
e
Molw. 599
I
P = -
40
al
I
,LL (327-197)
20
a
100
200
400
300
500
600
Ve
Figure 1. Mass spectra and proposed fragmentation patterns for the derivatives prepared from reference 5-HIAA (upper panel) and 5-HIAA-d2(lower panel) used as an internal standard in the analysis of 5-HIAA. Asterisks indicate metastable peaks techniques were relied upon for the analysis of 5-HIAA. Certainly, the cyclic nature of depression and the possibility that other neurotransmitters may be involved added to the difficulty of such attempts to elucidate a biochemical basis for depressive illness. Another source of uncertainty lies in the lack of specificity of the analytical methods that have been used. A gas chromatographic method for the measurement of 5-HIAA in CSF also has been proposed (8), but little evidence was given that the peak shown on the chromatogram was indeed a derivative of 5-HIAA, and the CSF concentration of 3.1 pg/ml that was reported exceeds the usual result of fluorometric analysis by one hundred times. During the preparation of the present paper, a gas chromatographic method for the analysis of 5-HIAA in mouse brain has been reported (9). In the present investigation a quantitative mass spectrometric technique, called mass fragmentography (IO), has been used for the determination of 5-HIAA in human CSF. In this method characteristic ions of the mass spectrum of material with the chromatographic retention time of 5-HIAA were (8) R. H. Leonard, J. Gas Chromatogr., 5 , 323 (1967). (9) Y . Maruyama and A. E. Takemori, Biochem. Pharmacol., 20, 1833 (1971). (10) C-G. Hammar, B. Holmstedt, and R. Ryhage, Anal. Biochem., 25, 532 (1968).
recorded and used for the quantitative analysis of each sample. It is possible that with this enhanced specificity the conflicting results of earlier studies will be resolved. EXPERIMENTAL
Reagents and Reference Compounds. Diazomethane was prepared in diethyl ether from N-methy1-N-nitroso-N'nitroguanidine (Koch-Light Laboratories Ltd., Colnbrook, Bucks., England) and stored over potassium hydroxide pellets at -15 "C. Before use it was distilled, its concentration determined by titration, and diluted to a final concentration of 0.05M with distilled ether (7 7). N-heptafluorobutyrylimidazole (HFBI) (Pierce Chemical Co., Rockford, Ill.) was obtained in sealed 1.0-ml glass ampoules and stored at $4 "C. The ethyl acetate used for the CSF extractions was of Nanograde quality (Mallinckrodt Chemical Works, St. Louis, Mo.). 5-HIAA (Schuchardt, Miinchen, Germany) (recrystallized from water, mp 162-4 "C), 5-methoxyindole-3-acetic acid (Aldrich Chemical Co., Milwaukee, Wis.), 5-hydroxyindole-3propionic acid (Regis Chemical Co., Chicago, Ill.), and 4benzyloxyindole-3-acetic acid (Koch-Light) were stored at +4 "C. The last compound was debenzylated by catalytic hydrogenation to 4-hydroxyindole-3-acetic acid. As this is (11) C . M. Williams and C. C. Sweeley in: "Biomedical Applications of Gas Chromatography," H. A. Szymanski, Ed., Plenum Press, New York, N.Y., 1964, p 247. ANALYTICAL CHEMISTRY, VOL. 44, NO. 8, JULY 1972
1435
D
I538
I540 (100%)
1597 (50%)
1538 (19%)
-
1597
(loo%!
2
ao%) 1
2
0
1
2
1
2
Minutes
Figure 2. Mass fragmentograms obtained from derivatives of: ( A ) reference 5-HIAA; ( B ) reference 5-HIAA-d2; (C) material extracted from CSF; ( D ) material extracted from CSF to which the internal standard solution containing 5HIAA-d2, 4-hydroxyindole-3-acetic acid (CHIAA), and 5hydroxyindole-3-propionic acid (5-HIPA) had been added. The CSF sample contained 5-HIAA in a concentration of 31 ng/ml. The mass spectrometer was set to detect m/e 538, 540, and 597 unstable as an oil and not crystalline (12), it was dissolved in water and stored frozen at - 15 "C in a concentration of about 100 pg/ml. 5-Hydroxyindole-3-acetic-1- 4c acid (6.45 mCi/mmole) was purchased from New England Nuclear Corp., Boston, Mass., and stored at -15 "C. 5-Hydroxyindole-3-acetic-2-*H2acid (5-HIAA-d2) was prepared by a procedure previously described for the synthesis of 5-HIAA (13). Deuterated formaldehyde (30 % C D 2 0 in D20, isotopic purity 99%, from Merck, Darmstadt, Germany) was used to synthesize 5-benzyloxygramine-1-d2. All reagents and solvents with active hydrogens were replaced by their deuterated analogs (isotopic purity 99 %). Before crystallization the final product was extracted from a "-water solution at acid pH to remove active deuterium atoms. The 5-HIAA-d2 thus prepared formed light salmon-pink crystals from chloroform (mp 163-5 "C). Radio-Gas Chromatography. Radio-gas chromatography was performed on a Barber-Colman Model 5000 gas chromatograph equipped with a model 5190 radioactive monitoring system (Barber-Colman Co., Rockford, Ill.). The column was a 1.5 m X 6 mm (i.d,) glass U-tube packed with 3.3% OV-17 on Gas Chrom P, 100/120 mesh (Applied Science Lab. Inc., State College, Pa.). Analyses were made at a column temperature of 170 "C, with a nitrogen carrier gas flow of 60 ml/min. The temperature of the vaporizer and flame detector was 220 OC. The combustion chamber was kept at 800 "C. The flow of the argon-propane quench gas through the radiation detector was 6 ml/min. Gas Chromatography-Mass Spectrometry. An LKB Model 9000 gas chromatograph-mass spectrometer (LKB-Produkter, Bromma, Sweden) was used. The separations were made on a 1.0 m x 3 mm (i,d.) silanized glass column, packed with 3 % XE-60 on Gas Chrom P (Applied Science), maintained at a temperature of 160 "C. The flow rate of helium carrier gas was 25 ml/min. The ionizing potential and trap current were 40 eV and 60 pA, respectively. The temperature of the flash heater was 200 "C and the ion source was kept at 250 "C. (12) F. Kalberer, W. Kreis, and J. Rutschmann, Biochem. Pharmacol., 11, 261 (1962). (13) A. Stoll, F. Troxler, J. Peyer, and A. Hofmann, Helu. Chim. Acta, 38, 1452 (1955). 1436
0
ANALYTICAL CHEMISTRY, VOL. 44, NO. 8, JULY 1972
This instrument was used to obtain conventional spectra of reference derivatives that were prepared in milligram amounts, but was modified by adding a new multiple ion detector (MID) (14) to permit quantitation of the nanogram amounts of 5HIAA that were present in the CSF samples. For quantitative mass spectrometry, the MID served as an ion-specific detector for the gas chromatograph and was adjusted to record exclusively the intensity of three of the ions generated by the mass spectrometer. In this case mje 597 (molecular ion) and 538 (base peak) from 5-HIAA and mje 540 (base peak) from 5-HIAA-dz were selected. Careful focusing was critical to ensure that the proper ion signals were recorded and was done by adjusting the magnetic field manually to record mje 538 with the accelerating voltage set at 3.5 kV. A computer (PDP-12, Digital Equipment Corp., Maynard, Mass.) interfaced with the MID was set to adjust the accelerating voltage so that mje 540 and 597 were brought into focus. Narrow slits of the mass spectrometer were needed to avoid contributions from ions of adjacent mje. Final focusing and resolution were checked by chromatographing derivatives prepared from 5-HIAA and 5-HIAA-d2 and comparing the intensity of the conventional mass spectra (Figure l), with the m/e 538, 540, and 597 signals (Figures 2A and B). Drift of the magnetic field shown by the mass marker (giving one decimal of m/e) was manually corrected. The accelerating voltage was stable during one day's analysis. The intensity of m/e 538 fragment of the 5-HIAA derivative was measured at different ionizing potentials and had a maximum at 40 eV. This potential was used both for regular mass spectra and for mass fragmentography. Procedure. CSF samples were obtained by lumbar puncture and stored in glass tubes at - 15 OC until analyzed. A 2.0-ml aliquot of each sample was placed in a 15-ml silanized, glass-stoppered centrifuge tube and 40 pl of the internal standard solution (see below), 0.5 ml 2N HCl, 2 g NaC1, and 4 ml ethyl acetate were added. The tube was shaken and centrifuged at 4000 x g for 5 min. The organic phase was transferred to a silanized, pear-shaped 10-ml flask and evaporated to dryness at 15 mm Hg in a 40 "C water bath. The residue was transferred by 0.3 ml methanol to a methanol-washed 3-ml glass tube and evaporated in a stream of nitrogen at 40 "C to dryness and until the smell of acetic acid was no longer present. The remaining material was dissolved in 0.1 ml methanol and cooled for 10 min at -15 "C. Then 0.5 ml diazomethane solution was added and the solution was mixed and reacted for 10 min at - 15 "C. The solvent was then evaporated to dryness with nitrogen. The residue was dissolved in 50 p1 of HFBI and reacted for 2 hours at 80 "C. After cooling to 4 "C the resulting product was washed with 1.0 ml of 1N H2S04for 5 min and extracted into 0.5 ml nhexane. The hexane phase was transferred to a methanolwashed 3-ml glass tube and evaporated in a stream of nitrogen. The final product was dissolved in 50 p1 n-hexane and stored at 4 OC; 1-4 p1 was then analyzed by mass fragmentography. The internal standard solution added to the CSF sample contained 5-HIAA-d2 (80 ng), 4-hydroxyindole-3-acetic acid (2 pg), and 5-hydroxyindole-3-propionic acid (300 ng) in 40 p1 water. This solution was prepared by diluting stock solutions of the compounds in water (100 pg/ml) stored at - 15 "C, the same day it should be used. The standard curves for the determination of 5-HIAA in CSF were prepared by treating in the same way as described for CSF specimens, a series of standard solutions containing known amounts of 5-HIAA in artificial CSF (15). This artificial CSF contained NaH2P04.2H20 (78 mg), NazHP04. 2H20(45 mg), MgC12.6H20(81 mg), CaClz (72 mg), KC1 (222 (14) C-G. Hammar and R. Hessling, ANAL.CHEM., 43,298 (1971). (15) J. R. Pappenheimer, S. R. Heisey, and E. F. Jordan, Amer. J. Physiol., 200, 1 (1961).
mg), NaCl (7.19 g), NaHC03 (2.10 g), and human serum albumin (250 mg) per liter. A typical mass fragmentographic analysis of 5-HIAA in CSF is shown in Figure 2 0 . The ratio of the heights of the 5-HIAA (m/e = 538) and the 5HIAA-d2 (m/e = 540) peaks was calculated, and the concentration of 5-HIAA determined from a standard curve (Figure 3) in which the peak height ratio (m/e 538/mle 540) was plotted against the known 5-HIAA concentration of the standard solutions. A 1.0-ml aliquot of each CSF sample was analyzed without an internal standard to make sure that the CSF did not contain materials that would interfere with the internal standard peak at m/e 540. This is shown in Figure 2C. The yields of the extraction and derivativization procedures were estimated by adding 400 ng of IF-labelled 5-HIAA to human CSF specimens and carrying them through the extraction, derivativization, and clean-up steps. Four hundred nanograms (about 10 times the endogenous amount) was needed because of the relatively low specific radioactivity. The efficiency of the extraction was more than 90% and the fmal hexane phase contained 60 of the initially added radioactivity, Radio-gas chromatography was used to check the purity of the derivative formed from radioactive 5-HIAA. Analogous derivatives were prepared from unlabelled 5HIAA and 5-methoxyindole-3-acetic acid and used as references. Radioactivity corresponding to formation of the HFB derivative of methyl 5-methoxyindole-3-acetate due to methylation of the phenolic hydroxy group of labelled 5HIAA amounted to less than 10% of the total product. No radioactivity could be located in other portions of the chromat ogram. RESULTS AND DISCUSSION
The preparation of the heptafluorobutyryl derivatives of indole amines and alcohols by an acyl transfer reaction with heptafluorobutyrylimidazole has been described recently (16). We have found that formation of analogous derivatives from indole acids requires prior esterification, and have used diazomethane as a methylating agent. Under the experimental conditions, methylation of the 5-hydroxy substituent on the indole ring of 5-HIAA accounted for less than 10% of the final methylated product, in agreement with previous work (11). The final derivative was stable in n-hexane at $4 “C for more than 4 days and showed excellent gas chromatographic properties. The structure of the derivative of 5-HIAA prepared on a milligram scale, was deduced by means of gas chromatography-mass spectrometry. The mass spectrum of this derivative (Figure l , upper spectrum) showed prominent peaks for the molecular ion (m/e 597) and for the indole fragment resulting from loss of the carbomethoxy group (M - 59). This cleavage of the bond of the side chain p to the indole nucleus is characteristic of indole-3-acetic acid derivatives (17). The metastable peak at m/e 485 for 5-HIAA is compatible with this transition (538.02/597.0 = 484.8). The technique of mass fragmentography was used initially to identify materials in biological fluids with more specificity than with conventional gas chromatography and with more sensitivity than when a full mass spectrum was obtained by gas chromatography-mass spectrometry (10). Mass fragmentography has recently been applied to quantitative problems (16) J. Vessman, A. M. Moss, M. G. Horning, and E. C. Horning, Anal. L e f t . ,2, 81 (1969). (17) H. Budzikiewicz, C. Djerassi, and D. H. Williams, “Structure Elucidation of Natural Products by Mass Spectrometry, Vol. I, Alkaloids,” Holden-Day lnc., San Francisco, Calif,, 1964, p 42.
Figure 3. Standard curve for the quantitative determination of 5-HIAA in CSF. The curve was prepared by analyzing standard solutions of 5-HIAA in artificial CSF by the entire procedure described in the experimental section and has been used to measure the tricyclic antidepressant drug nortriptyline in plasma (18). For the analysis of 5-HIAA the molecular ion (m/e 597) and the base peak (m/e 538) of the prepared derivative were monitored. For material in CSF to be measured as 5-HIAA, it was required to have the same gas chromatographic retention time and relative intensities of these two ions as reference material prepared from 5-HIAA (cf. Figures 2A and C). The ideal internal standard for quantitative analysis by gas chromatography is one that has physical and chemical properties that are as similar as possible to those of the compound being measured. It has recently been shown feasible to use as an internal standard for mass fragmentography the compound to be measured, modified by the introduction of the stable isotopes 15Nand deuterium (19, 20). In the present work 5-HIAA-d2 was chosen as the internal standard for 5-HIAA analysis. Other compounds as similar to 5-HIAA as 5-hydroxyindole-3-propionic acid (5-HIPA) and 4-hydroxyindole-3-acetic acid (4-HIAA) previously had been tried and found unsatisfactory in that analyses of standard 5-HIAA samples gave results that were not sufficiently reproducible. Although not useful as internal standards, these two compounds were added to the CSF samples. The addition of 4-HIAA gave higher yields of the 5-HIAA and 5HIAA-d2 derivatives, probably because of its antioxidant properties. The 4-HIAA derivative showed a mass spectrum almost identical to that of 5-HIAA, but the two compounds were completely separated on the mass fragmentogram (Figure 20). The 5-HIPA derivative was used to monitor channel 1 of the MID. It is cleaved /3 to the indole nucleus in the mass spectrometer giving mje 538 as a prominent fragment (Figure 20). 5-HIAA-dzwas synthesized by introducing formaldehyd-d2 by a Mannich reaction. The mass spectrum of its diheptafluorobutyryl methyl ester derivative is shown in Figure 1, (18) 0. BorgB, L. Palmer, A. Linnarsson, and B. Holmstedt, Anal. Lett., 4, 837 (1971). (19) B. Samuelsson, M. Hamberg, and C. C. Sweeley, Anal. E o chern., 38, 301 (1970). (20) T. E. Gaffney, C.-G. Hammar, B. Holmstedt, and R. E. McMahon, ANAL.CHEM., 43, 307 (1971). ANALYTICAL CHEMISTRY, VOL. 44, NO. 8, JULY 1972
1437
Table I. Five Samples from Each of Five Pools of CSF from Different Sources Were Analyzed for Their Content of 5-HIAA Mean 5-HIAA in Std dev in Pool ng/ml CSF i std dev of mean Range 1 8.2 i 0 . 5 6.1 7.6-8.8 2 17.2 =t1 . 1 6.4 16.3-19 .O 3 26.4 i 0 . 5 1.9 26 .&26.9 4 28.7 i 0.3 1.o 28.2-29.0 5 32.0 i 0.5 1.6 31.3-32.5
fragmentogram (Figure 2B) of the 5-HIAA-d2 derivative, The standard curve was linear over the range studied. The precision of this analytical method for 5-HIAA was determined by analyzing five pools of human CSF from different sources five times each. The pools contained between 8 and 32 ng 5-HIAA/ml and the standard deviation was less than 7% below 20 ng/ml and 1-2Z above this level (Table I). The CSF pool 3 was analyzed for its content of 5-HIAA after having been stored frozen at -15 “C for two months. The loss of 5-HIAA during this period was less than 5 %. ACKNOWLEDGMERTT
lower spectrum. The compound contains not only 5-HIAAd2,but also mono- and nondeuterated 5-HIAA (62 and 19 %, respectively, of the dideuterated). The mass spectrometer could be focused exactly on mle 538 and 540 on channels 1 and 2, respectively, without contributions from ions of adjacent m / e , when mass fragmentography was used. This is shown by the agreement between the relative proportion of the fragments in the mass fragmentograms (Figures 2A and B ) and the mass spectra (Figure 1). The standard curves for the quantitative determination of 5-HIAA in CSF (Figure 3) were prepared using 5-HIAA-d2 as the internal standard. The intercept of the ordinate at 0.20 is in agreement with the ratio between m / e 538 and 540 in the mass spectrum (Figure 1, lower spectrum) and in the mass
We appreciate the help of Birgitta Sjoqvist in performing the radio-gas chromatographic studies and for the skillful technical assistance we are indebted to Eva Dufva. D. Efron, NIMH, kindly supplied the 5-HIPA. The project has been cleared by the ethical committee at the Karolinska Institute. RECEIVED for review November 15,1971. Accepted February 8, 1972. This project was supported by grants from the Tri-Centennial Fund of the Bank of Sweden to B. Holmstedt (68153) and to B. Cronholm and F. Sjoqvist (68/90), the National Institutes of Health, Bethesda, Md., (GM 13 978), the National Institute of Mental Health, Chevy Chase, Md., (Grant MH 12007), the Wallenberg Foundation, Swedish Medical Research Council B 72-40 Y-2375-05, and by funds from the Karolinska Institute.
Hexafluoroacetone Ketals as Derivatives for Positional and Geometrical Characterization of Double Bonds Bruce M. Johnson and James W . Taylor1 Department of Chemistry, University of Wisconsin, Madison, Wis. 53706
The characterization of double bond position and geometry has been improved through conversion of n-al kenes to hexafluoroacetone ketals. Derivatives are synthesized by stereospecifically forming bromohydrins from the alkenes and converting them to ketals via base-promoted reaction with hexafluoroacetone in sealed glass reaction tubes. The addition of hexafluoroacetone proceeds with at least 97% trans specificity. Mass spectral fragmentations which indicate the original double bond position are discussed as are the stereochemical effects. The uses of GLC, NMR, and IR for derivative characterization are also examined. The combination of GC and MS on the derivative provides complete characterization of the original olefin.
MASS SPECTROMETRIC ANALYSIS of long chain unsaturated compounds is complicated by geometrical and positional isomerization which may occur before fragmentation of the molecular ion (1-3). Since olefin isomerization is a facile process, the mass spectra of a series of positional and geometrical isomers are generally very similar and extremely Author to whom correspondence should be directed. (1) J. H. Beynon, “Mass Spectrometry and its Applications to Organic Chemistry,” Elsevier, Amsterdam, 1960, p 262. (2) D. S.Weinberg and C. Djerassi,J. Org. Chem., 31,115 (1966). (3) B. J. Millard and D. F. Shaw, J. Chem. SOC.B, 1966,664. 1438
ANALYTICAL CHEMISTRY, VOL. 44, NO. 8,JULY 1972
difficult to use for structural assignments. The most successful approach to this problem has been to form a derivative of the double bond which has a fragmentation pattern reflecting the original configuration of the unsaturated molecules. The first such alkene derivative was produced by reduction of the double bond with deuteriohydrazine ( 4 ) . Use of this derivative, however, was complicated by partial H-D exchange (5, 6). Other approaches have included oxidation of the double bond to an epoxide used directly (7) or followed by conversion to ketones (6) or N,N-dimethylamino alcohols (8). Methoxymercuric adducts have been de-mercurated to a pair of isomeric methoxy compounds (9). Double bonds have also been oxidized to 1,2-diols and converted to dimethyl (4) N. Dinh-Nguyen, R. Ryhage, and S. Stallberg-Stenhagen, Ark. Kemi, 15,433 (1960). (5) N. Dinh-Nguyen, R. Ryhage, S. Stallberg-Stenhagen, and E. Stenhagen, ibid., 18,393 (1961). (6) G. W. Kenner and E. Stenhagen, Acta Chem. Scand., 18, 1551 (1964). (7) R. T. Aplin and L. Coles, Chem. Commun., 1967,858. (8) H. Audier, S. Bory, M. Fetizon, P. Longevialle, and R. Toubiana, Bull. Soc. Chim. Fr., 1964,3034. (9) P. Abley, F. J. McQuillin, D. E. Minnikin, K. Kusamran, K. Maskens, and N. Polgar, Chem. Commun.,1970,348.
