metals studied. Trace amounts of chromic acid were dissolved in 5.OM HC1-2.5M HC104 and in 5 M HC104 and spectra run immediately. In both cases, the spectra indicate formation of a mixture of HCr04- and HzCr04. On standing for 10 hr, the spectrum in perchloric acid remains unchanged whereas in the acid mixture almost complete conversion to chromium(II1) has occurred with only minor amounts of chromium(V1) remaining, presumably containing Cr03C1- (28). The spectrum of ruthenium(1V) was run after diluting the stock solution in 10M HC1. The spectrum obtained was the same as obtained by Wehner and Hindman, indicating that the solution contains a mixture of R u ( O H ) Z C ~ and ~ ~ - a t least one other complex of tetravalent ruthenium. The presence of trivalent ruthenium is therefore doubtful especially due to the ability of this ion
to reduce perchlorate or dissolved oxygen (29). Bearing in mind the procedure employed in preparation of these ions, there then can be no ambiguity about oxidation state. Partial oxidation of the resin forming chromium(II1) must be also ruled out because of the large molar absorptivity of the eluted peak and absence of separated peaks for chromium(II1) and chromium(VI) in 10M HC1. I t must be concluded that a t least one detectable complex of these two ions is poorly sorbed to this resin.
ACKNOWLEDGMENT The authors gratefully acknowledge the help of Louise Goodkin with modification of the chromatograph. Received for review September 14, 1972. Accepted January 18, 1973.
(28) G . P. Haight, Jr., D. C. Richardson, and N. H. Coburn, Inorg.
Chem., 12,1777 (1964).
(29) P. Wehner and J. C. Hindman, J. Phys. Chem.. 56, 10 (1952).
Identification of Phenylcycloparaffin Alkanes and Other Monoaromatics in Green River Shale by Gas ChromatographyMass Spectrometry E. J. Gallegos Chevron Research Company. Richmond, Calif. 94802
Combined gas chromatography-mass spectrometry, GC-MS, is used in the provisional identification of 23 mono- and 1 diaromatic hydrocarbons isolated from Green River shale. These include 10 phenyl(cyc1ohexyl)alkanes, 5 alkylbenzenes, 2 benzomonocycloparaffins, 3 benzodicycloparaffins, 2 benzocycloparaffin(cyc1ohexyl)alkanes, 1 benzotricycloparaffin, and 1 napthalene monocycloparaffin. This work represents part of a continuing effort on the analysis of hydrocarbons in Green River shale.
Much has been reported in the literature on the analysis of the branched cyclic saturate fraction of Green River shale ( I , 2 ) . However, the literature is sparse on the aromatics ( 3 )from this source. This paper covers the first detailed GC-MS analysis of the monoaromatic fraction of Green River shale.
EXPERIMENTAL Green River shale taken from the vicinity of Grand Valley, Colo., was crushed to approximately 2-3 mm diameter particle size. The organic material was extracted over a period of a week with a 50/50 benzene-methanol mixture at reflux temperature. The predominantly monoaromatic cut from Green River shale was obtained in two different ways, both giving similar total ion monitor traces, TIM, mass chromatograms, and GC-MS data. Tswett chromatography on silica gel with n-hexane as the eluting solvent was used in one way. This was done on a 130 X 1.25 ( 1 ) E. J. G a l l e g o s , A n a / . Chem.. 43, 115 (1971). ( 2 ) D. E. Anders and W . E. Robinson, Geochim.
Cosmochim. Acta, 35,
661 (1971). (3) F. G . Doolittle, D. E. Anders, and W . E. Robinson, Pittsburgh Con-
ference on Analytical Chemistry and Applied Spectroscopy, Cleveiand, Ohio, March 8.1972, Paper 167.
cm i.d. column packed with silica gel Grade 923, 100-200 mesh, from Grace Davison Chemical Co. UV was used to determine the break between the first cut, i . e . , saturates, and the second cut, i.e., the monoaromatic fraction used in this study. Aluminum oxide, Woelm Basic Activity Grade 1, was used for the other separation. Cyclohexane was used as the eluting solvent to obtain Cut 1, i.e., the saturate fraction ( I ) , and benzene was used to elute Cut 2, i . e . , the fraction used in this study. A 200-ft 0.02-in. i.d. Dexsil-coated capillary column was coupled to a Suclide 12-90-G spectrometer for the GC-MS analysis of Cut 2. The temperature was programmed from 100 to 350 "C at a rate of 2 "C/min. Typically, 0.2 p1 was injected using a carrier gas flow rate of approximately 2 ml/min. Mass spectra were gathered a t 4-sec intervals using 70 eV ionizing voltage. Negligible column bleed was observed. Partial mass spectra of individual components were obtained by background subtraction of mass spectra recorded before and after that taken under the GC peak of interest and are the spectra reproduced in Figure 1. These data are a compilation of several GC-MS runs made on the predominantly monoaromatic fraction of Green River shale. High resolution mass measurements were made of important fragment and parent ions of the total sample to confirm empirical formula assignments reported in this paper ( 4 ) . The high-resolution measurements were made on an AEI MS-9 mass spectrometer at a resolution of 15,000. Selected fragment ion mass chromatograms were obtained using an automatic peak intensity monitoring device designed at this laboratory. (R. F. Klaver, Chevron Research Co., Richmond, Calif.)
-
RESULTS The individual mass spectra of 24 components are shown in Figure 1. They are listed in order of elution time. Also included are proposed structures, i.e., provisional carbon skeletons which are most consistent with (4) W . K. Seifert, R. M . Teeter, W . G . Howells. and M . Anal. Chem., 41, 1638 (1969)
J. R .
A N A L Y T I C A L CHEMISTRY, VOL. 45, NO. 8, J U L Y 1973
Cantow.
1399
1
I
2
10
I
133
11
158
119
145
105 159
C16H24
-CHI 213
4 12
I
lo5
-
Mo
200
100
MW
2w MW
5
13 !19
6 19
19
14
A
qy2 119
Mi 1244
131 ZH3
C18H28
7 15
8
I
L
71
I
16
I
‘20H32
132
M W-
Figure 1. Mass spectra and proposed structure of monoaromatics in Green River shale
MW
17
18
M+ 286
19
20
I
I
1W
Mo
200 MW
21
22
1170
I119
23
24
I
1W
their mass spectra and history. The total ion monitor GC-MS trace and selected mass chromatograms are shown in Figure 2. The per cent of aromatic types was calculated from total ion intensities and represents the fraction of each type in the 24 aromatics discussed here. Phenyl(cyc1ohexy)alkanes (55%) C16H24, C17H26Two Isomers, C18H28, C19H30, C20H32, C21H34, C22H36, C26H44. These mass spectra appear in Figure 1 as Spectra 3, 5,6, 10, 12, 14, 16, 18, 20, and 24, respectively. The mass spectral feature common to five of these monoaromatic hydrocarbons is the presence of a base peak a t mle 119. Spectra 3, 12, 18, and 24 show a base peak at mle 105. Spectra 10 shows a base peak a t mle 133. Ions a t m / e 133, 119, and 105 are all typical alkylbenzene fragments. However, these hydrocarbons belong to the CnHzn-8 type which means that, in addition to an aromatic ring, they must have a saturated cyclic ring, fused or nonfused. The presence of strikingly intense alkylbenzene fragments for these 2 = -8 compounds clearly implies that the saturate ring is not fused to the benzene ring but is attached to it by a single bond or alkyl chain as in I.
200 MW
Mo
A search through the Aldermaston files ( 5 ) reveals three phenylmonocyclic-like compounds which show base peaks corresponding to alkylbenzene fragments. These are: cyclohexylphenylmethane, MSDC 3601; l-cyclopentyl-3(2,4-dimethyl)phenylpropane,MSDC 113; and l-phenyl3-cyclopentylpropane, MSDC 3517. The alkylbenzene fragment ion intensities of these three 2 = -8 compounds represent on the average 50% of the total ionization. This is also true of the alkylbenzene fragment ion intensities of the 10 phenyl(cyc1ohexyl)alkanes reported in this work. Further, the spectra of those compounds with fused structures are given in the same file. They are 1,1,6-trimethyltetralin (MSDC 3759), 1-n-hexadecylindan (MSDC 3518), and 1,1,5-trimethylindan (MSDC 101). None produce alkylbenzene fragments in major proportions, i. e., the alkylbenzene fragment ion intensities represent 5% or less of the total ionization. Phenyl( cyclohexy1)alkane hydrocarbons should on electron impact show 2 = -8 phenylcyclohexylalkane parent ions, 2 = -6 alkylbenzene, and 2 = 0 one ring saturate fragment ions. The mass spectra presented for this series of compounds are consistent with this statement. Mass chromatograms of m / e 119 and 83 were obtained in order to highlight this proposed requirement. These are shown in Figure 2. Peaks which occur simultaneously on the two mass chromatograms can be taken as due to phenyl(cyclohexy1)alkanes. If a peak occurs only on the m / e 119 mass chromatogram, it is due to an alkylbenzene. Speculation as to the origin of these monoaromatics serves to further support the phenyl(cyclohexy1)alkane carbon skeleton. The most likely progenitors of the monoaromatic phenylcyclohexanes are the tricyclic and pentacyclic terpenoids or even steroids, maturation of which could result in aromatization of one ring and cleavage of the center or B-ring bond resulting in many possibilities, including 11-VI. Bond cleavage is presumed to occur only adjacent to the ring which has dehydrogenated to give the monoaromatic. I11 is the carbon skeleton most consistent with the mass spectra obtained for the phenyl(monocyc1ic)alkanes. 111 on ( 5 ) File of Mass Spectra from the Mass Spectrometry Data Centre, AWRE, Aldermaston. England.
A N A L Y T I C A L CHEMISTRY, VOL. 45, NO. 8, J U L Y 1973
1401
@ @@ k ; ?
MONOAROMATICS I N GREEN R I V E R S H A L E BY I GC-MS D E X S I L 2 0 0 - F T COLUMN
, .
20
Legend S p e c t r u m Number M o l e c u l a r Weight
Ot
Lompouna
I
/
m/e 131
Figure 2.
Monoaromatics in Green River shale by GC-MS Dexsil 200-ft column
II
-
Route2
Ill
@cn V
IV
flCn VI
electron impact should produce an important m / e 119 fragment ion. Further argument will be presented a t the conclusion of this paper to defend the choice of the tri-, tetra-, or pentacyclics as progenitors to the phenyl(cyc1ohexy1)alkanes. There is no evidence to support the cyclohexyl over cyclopentyl. Cyclohexyl is chosen because of the preponderance of six-membered rings in terpenoid systems. Evidence has been reported previously for the monoand diaromatization of terpenoid and steroid systems to produce hydrocarbons (6). (6) B Tissot, J L Oudin, and R Pelet, "Advances in Organic Geochemistry, 1971, Pergamon Press. Oxford. Braunschweig 1972, pp 113134
1402
Benzornonocycloparaffins (1%) C13H18, Cl&b& Within that same CnH2n-8 series there are two compounds containing the fused-ring structure. Their spectra are given in Figure 1, spectra 1 and 9. The presence of a large (Mf - CH3) fragment ion, spectrum 1, suggests a gemdimethyl group on the saturated ring. Assuming a dicyclic terpane progenitor, the following scheme could produce such a compound from VI1 upon maturation, which on electron impact would produce a relatively large (Mf CH3) a t m / e 159.
* ANALYTICAL CHEMISTRY, VOL. 45, NO. 8, JULY 1973
VI I
mle 159
The mass spectrum 1 in Figure 1 corresponds very closely to that of ionene, 1,2,3,4-tetrahydro-l,1,6-trimethyl naphthalene, m / e 174, C13H18, MSDC 3759. The second benzocyclo component, spectrum 9, Figure 1, suggests a c6 side chain to account for the important ( M + - C&3) fragment ion. Attachment of the other methyl groups is consistent with a biological progenitor. Benzodicycloparaffins (12%) C17Hz4, C1&6-Two Isomers. The benzodicyclo compounds are shown in Fig-
ure 1, spectra 4, 7, and 11. The common feature of spectra 4 and 7 is the presence of a large parent minus a methyl fragment ion, suggesting gem-dimethyls probably on ring A. A maturation route from terpenoids and steroids to these benzodicyclo compounds is shown in the following scheme.
This monoaromatic structure will provide a reasonable explanation for the observed mass spectra. Benzotricycloparaffin C22H32, Naphthalene Cycloparaffin C2lH28 < 1%. Benzotricycloparaffin and naphthalene cycloparaffin, C22H32 and C21H28, are shown as mass spectra 21 and 22 in Figure 1. The carbon skeletons proposed best fit the mass spectra data. C2lH28 is the only diaromatic of importance in this fraction. CONCLUSIONS
The monoaromatic hydrocarbon produced would most certainly have a large parent minus a methyl. Indeed, a 1,1,4,8-tetramethyl-s-hydrindacene MW 214, C16H22, MSDC 111, has a gem-dimethyl group and shows a base peak a t the ( M + - CH3) position. A similar argument can be used to defend the ring structure proposed for spectrum 11, Figure 1. The ( M + CH3) fragment ion is much less intense than in the previous two, suggesting the absence of a gem-dimethyl. Alkylbenzenes (17%) C15H24, C18H30-Two Isomers, C20H34, C25H44. These alkylbenzenes, CnHzn-,3 are shown in Figure 1, spectra 2, 8, 13, 17, and 23, respectively. Three of these compounds have base peaks a t m l e 119 in their mass spectra. Another, spectrum 8, shows its base peak a t m l e 133, 14 mass units higher. Those having base peaks a t m l e 119 very possibly came from aromatized carotanes uia a mechanism somewhat like that which follows.
Substituted phenyl(cyclohexy1)- or (cyclopenty1)alkane hydrocarbons make up better than 50% of the total monoaromatics identified in Green River shale. Five of these have a base peak a t mle 119. The remaining four show a base peak a t mfe 105, 14 amu lower and one with a base peak a t m l e 133. There are five alkylbenzenes, -16% of the total, three of which also have a base peak a t m l e 119. Of the two remaining, one shows a base peak a t m l e 105, the other a t m l e 133. The dominance of compounds having large m l e 119 fragments is very evident from the m l e 119 mass chromatogram, Figure 2. Mass 119 is consistent with the isoprenoid structure following dehydrogenation. For example
n
1
1
@ LF+
mle 119
This corresponds to compounds having the same spectra shown in spectra 2, 13, and 23, respectively, Figure 1. The alkylbenzene with the base peak a t m l e 133 probably comes from slightly different precursors or it has followed a different maturation path from a similar precursor. Indan(cyclohexy1)alkanes (2%) C20H30, C21H32. There is evidence for two indan cyclopentyl- or cyclohexyl-like hydrocarbons in significant amounts in this aromatic fraction of Green River shale. The mass spectra are shown in Figure 1, spectra 15 and 19. These components may well result from an aromatization of steranes involving ring cleavage similar to that proposed to explain the formation of the phenyl( cyclohexy1)-alkane compounds. This is shown in the following scheme.
mle 119
Either way structure D can be produced. Route 2 is preferred because tricyclic terpanes, C, have been identified in Green River shale; whereas, there has not been any evidence for the presence of compounds such as B in the saturate fraction of this shale. These monoaromatics can perhaps be considered second-order biological marker compounds. ACKNOWLEDGMENT The author would like to thank L. P. Lindeman and R. M . Teeter for discussions and encouragement during the course of this work and W. K. Seifert for the Hexane Cut 2. Received for review October 18, 1972. Accepted January 22, 1973.
A N A L Y T I C A L CHEMISTRY, VOL. 45,
NO. 8,
J U L Y 1973
1403
