Stable carbon isotope ratios and chemical properties of kerogen and extractable organic matter in pre-Phanerozoic and Phanerozoic sediments — Their interrelations and possible paleobiological significance

Stable carbon isotope ratios and chemical properties of kerogen and extractable organic matter in pre-Phanerozoic and Phanerozoic sediments — Their interrelations and possible paleobiological significance

Chemical Geology, 21 (1978) 335--350 335 © Elsevier Scientific Publishing Company, Amsterdam - - Printed in The Netherlands STABLE CARBON ISOTOPE RAT...

901KB Sizes 0 Downloads 18 Views

Chemical Geology, 21 (1978) 335--350 335 © Elsevier Scientific Publishing Company, Amsterdam - - Printed in The Netherlands

STABLE CARBON ISOTOPE RATIOS AND CHEMICAL PROPERTIES OF KEROGEN AND EXTRACTABLE ORGANIC MATTER IN PREPHANEROZOIC AND PHANEROZOIC SEDIMENTS - THEIR INTERRELATIONS AND POSSIBLE PALEOBIOLOGICAL SIGNIFICANCE

TOGWELL A. JACKSON l , PETER F R I T Z 2 and ROBERT DRIMMIE 2

Department of the Environment, Inland Waters Branch, Freshwater Institute, Winnipeg, Man., R 3 T 2N6 (Canada) 2Department of Earth Sciences, University of Waterloo, Waterloo, Ont., N2L 3G1 (Canada) (Received July 26, 1976; accepted for publication March 28, 1977)

ABSTRACT Jackson, T.A., Fritz, P. and Drimmie, R., 1978. Stable carbon isotope ratios and chemical properties of kerogen and extractable organic matter in pre-Phanerozoic and Phanerozoic sediments - - their interrelations and possible paleobiological significance. Chem. Geol., 21: 335--350. Kerogen and the polar (or " h u m i c " ) fraction of benzene/methanol extractable organic matter in various pre-Phanerozoic and Phanerozoic sedimentary rocks were analyzed for stable C isotope content. Variations in the 613C values as functions of the geologic age and chemical properties of the organic substances were investigated. The ~ 13C values of both kerogen and extractable polar organic matter in mudstones showed a tendency to decrease with increasing geologic age, possibly indicating secular decrease in atmospheric and hydrospheric CO 2 . However, the secular trend of the polar extracts has a much higher statistical significance and much less scatter than the secular trend of the kerogen. This suggests that the heterogeneity and complexity of kerogen tend to mask systematic variations, whereas the polar extract, being a specific small fraction of the organic matter, varies less erratically. Bulk isotope analyses of kerogen have limited usefulness for investigation of systematic.variations; the various components of the kerogen molecule should be separated by fractionation procedures and analyzed individually. The 813C values of all kerogen samples, regardless of rock type, age, or depositional environment, showed a highly significant negative correlation with the ratio of aliphatic to condensed aromatic components of their respective polar extracts, and 6 ~3C of kerogen in carbonate rocks and cherts showed a significant positive correlation with the aromaticity of the kerogen; other data indirectly suggest that carbonyl-type groups are enriched in ~3C. These results demonstrate that the aliphatic components of organic matter are isotopically lighter than the associated aromatic components, suggesting that the aliphatic chains are derived from lipids of algae. Thus, fractionation of kerogen prior to isotope analysis should at least separate the C atoms of aliphatic, aromatic and carbonyl-type groups. ~ 13C of kerogen in carbonate rocks and most cherts showed significant negative correlation with concentration of OH groups in polar extracts, suggesting loss of phenolic

336

OH groups with increase in condensed-aromatic character. The observed relationships between polar extracts and kerogen confirm previously reported evidence that kerogen and its associated extractable polar substances are chemically similar and are largely derived from the same source materials. Possibly paleobiological and paleo-environmental information coded in the molecular structure of kerogen may be obtained more easily, rapidly, and economically from polar extracts. The existence of secular variation in polar extracts and a relationship between polar extracts and kerogen supports the conclusion that the polar extracts are indigenous to the rock in which they occur, and thus are valid chemical fossils. Comparisons between ~ 13C values of kerogen and polar extracts from mudstones are tentatively interpreted, as indicating that terrigenous organic matter, and hence land life, originated between 1.05.109 and 1.3.109 years ago as a manifestation of the early evolution of eucaryotic algae. INTRODUCTION

Stable C isotope ratios of organic matter in sedimentary rocks are potentally valuable for research on the evolution and ecology of ancient life, but their usefulness is severely limited by the heterogeneity and complexity of the organic matter. Kerogen (the insoluble organic fraction), which comprises more than 95% of the organic matter in sedimentary rocks (Degens, 1965), is formed from various biological starting materials made up of many different biochemical compounds, and is formed in practically every kind of sedimentary environment. Considering that the stable C isotope ratios of fossil organic matter vary as functions of source material, environment of formation, and post-depositional alteration {Degens, 1969), it is not surprising that bulk analyses of kerogen or total organic matter in sedimentary rocks of all ages have shown considerable dispersion and little systematic variation through geologic time (cf. Oehler et al., 1972). Important information is probably coded in the distribution of C isotopes in fossil organic matter but is largely obscured by the heterogeneity of the organic matter. Bulk analysis of such complex materials has limited usefulness for investigation of systematic variations. The various components of the organic matter should be separated by standardized fractionation procedures and analyzed individually, or bulk isotope data should be normalized to compensate for chemical heterogeneity. In our research on organic matter in various ancient sedimentary rocks, we dealt with this problem by: (1) analyzing a specific small fraction of the organic matter, and (2) investigating the relationship between the C isotope composition and chemical composition of organic matter. A brief announcement of our findings has been presented elsewhere {Jackson et al., 1976). Chemical characteristics of the organic matter have been described by Jackson {1973, 1975) and by Jackson and Moore (1976). MATERIALS AND METHODS

The rock samples consisted of cherts, carbonates, and an assortment of

337

shales, mudstones, argillites and tillites (all collectively termed "mudstones" for convenience) ranging in age from early Archean to Miocene (Table I). These rocks are unmetamorphosed, except for samples 14, 15 and 44, which represent the lower greenschist facies of regional metamorphism (M. Viljoen, pers. commun., 1972). Only unweathered portions of each specimen were used. The rocks were pulverized in a tungsten carbide disc mill and extracted with benzene--methanol (9:1). The extracts were evaporated to dryness, redissolved in hexane, and fractionated on a silica gel column by elution with hexane, benzene and methanol in that order. The methanol eluate, which comprised the polar fraction of the extract, contained brown or yellow material that has been described as a type of humic substance (Jackson, 1973, 1975). Chemical similarities between this "humic" matter and the kerogen associated with it have been noted (Jackson and Moore, 1976). The polar fraction of the extract and the non-extractable organic matter, or "kerogen," were the organic fractions used for C isotope determination. Most of the solvent extracts had been used up previously for chemical analyses or fell short of the minimum quantity needed for 5 ~3C determination, but isotope data were obtained for extracts from seven mudstone samples ranging in age from Archean to Ordovician. Detailed information on the rock samples, methods of sample preparation, and chemical analysis of the organic matter are given elsewhere (Jackson, 1975; Jackson and Moore, 1976). Samples of pulverized, pre-extracted rock were digested with 10% HC1 to remove carbonates, after which organic C in both the rock samples and the extractable polar material (hereafter called "polar extract") was converted to CO2 by combustion at 900°C in 02 • Following purification by removal of sulfate, nitrogen and halogens, the CO2 was analyzed in an MAT-GD-150 double-collecting mass spectrometer. The data are reported in 5 '3C per mil units normalized with respect to the PDB belemnite standard. The 5 ~3C values were plotted against geologic age and against chemical data obtained by spectrophotometric analysis of the polar extracts (Jackson, 1975). Regression lines, correlation coefficients (r), significance probabilities (p), and standard error of estimate were computed by normal statistical procedures. RESULTS AND DISCUSSION

~3C values of kerogen plotted against geologic age (Fig. 1; Table I) show excessive scatter and only a suggestion of systematic variation. The data for mudstone kerogen suggest a weak negative correlation with geologic age (i.e., the older kerogens tend to be isotopically lighter), but this trend is quite insignificant from a statistical standpoint (r = --0.263 for a semi-log plot; p > 0.1). Comparable results have been reported by other workers (Degens, 1969; Oehler et al., 1972; Welte et al., 1975). The ~ ~3C values of

TABLE I

00

5 ~3C and CH 2/A66sr~n values of the organic matter, together with information about the rock samples Sample Name of rock no.

Age (. 109 yrs.)

Location

Rock type

Comments

6'3c (°/00) kerogen

1

Gowganda Fro.

2.288 _+0.087

2

Hector Fro.

0.7

3

Nonesuch Shale

1.05

4 6

Beck Spring Dolomite Kingston Peak Fm.

Ontario, Canada Alberta, Canada Michigan, U.S.A.

argillite shale shale

glacio-lacustrine marine;near-shore(deltaic? ) marine tidal flat or deltaic topset

27.4 --30.4--30"6 I -~ = --30.5

}

CH2/A665nm polar extract --

53.3

--27.0

82.8

1.3

California, U.S.A.

dolomite

marine; littoral

--33.1 --33.0 I --33.4 "~ = --33.1 --32.9 --15.2

--31.1 ~ ~ -~ = --29.8 --28.4J

161

1.1 (?)

California, U.S.A.

tillite or pebbly mudstone chert

glacio-marine (?)

--19.5

--

27.7

marine; near-shore, shallow-water

--15.6

--

3i.7 62.7 I -~ ffi 47.2

--

12.3

l

--

32.4

11

Gunflint Iron Formation

2.0

Ontario, Canada

12+13

Bitter Springs Fm.

0.8

Australia

chert

marine; shallow-water, partly intertidal

--15.8

14

Kromberg Fro. (Onverwacht Group, Swaziland System)

3.4

South Africa

chert

marine;eugeosynclinal basin?

--26.1 t --26.2 ~ X" = --26.2

--

15

Swartkoppie Fro. (Onverwacht Group, Swaziland System)

3.3

South Africa

chert

marine; eugeosynclinal basin?

--26.9

--

17

Paradise Creek Fm.

1.6

Australia

chert

marine; supra- to infra-tidal

--28.9 --28.6 ( X = 28.8

X = 407

653

730

36.9

82.8

--25.6 --28.4 --20.9 --34.8 f --35.9 .~ = - - 3 5 . 3 --29.1

marine; euxinic; climate glacial marine; shallow embayment probably marine; shallow inland sea marine; shallow-water, near-shore;"black shale" e n v i r o n m e n t marine (?); near-shore

argillite shale (with dolomite) limestone shale

shale (with carbonate minerals)

Norway

Michigan, U.S.A.

Germany

Minnesota, U.S.A.

Scotland

0.700 (?)

0.350 (Devonian)

0.250 (Permian)

2.0

0.935

Bjoranas Shale

Antrim Shale

Kupferschiefer

Rove Shale

Stoer Bay Fm.

28

29

30

32

27

--33.5

marine tidal flat or deltaic topset

argillite (calcareous)

Michigan, U.S.A.

--27.0

marine;

shale

Maryland, U.S.A.

0.370 (Devonian)

R o m n e y Shale

35

offshore

--22.6

marine; deep basin

shale (calcareous)

0.016 (Miocene)

34

(bathyal?)

--30.1

marine (?); near-shore

shale

Scotland

California, U.S.A.

0.751

Diabaig Shale

Monterey Shale

33

--

--27.7

--

--

--

--27.4

--28.1

1.05

Nonesuch Shale

26

~ 2 5 ~8

--27.5 ~ --26.2 I ~" --24.4 } --25.0

probably glaciomarine

tillite or pebbly mudstone

Utah, U.S.A.

0.827 ± 0.030

Mineral Fork Tillite

25

marine tidal flat or deltaic

argillite

Arizona, U.S.A.

1.3

Dripping Springs Quartzite

22

--21.4 } .Y = --21.5 ---21.1 --21.8 M d = --21.4 --27.5 } X = 26.5 --28.0 --27.0 --24.9 } Md = --27.0

marine "black shale" environment

mudstone (calcareous)

Utah, U.S.A.

0.827 ± 0.030

(no name yet)

21

--26.2

r

marine; shallow-water

Namibia

limestone

0.680 (Ediacarian)

Kuibus Fro.

20

--27.8

shale (calcareous)

British Columbia, Canada

marine; euxinic

O.540 (Cambrian)

Burgess shale

18

(continued)

I

TABLE

I

.~ = - - 2 7 . 8

166

98

683

788

580

142

225

10.2

381

135

82.6

241 178 366

183

28.8

53.0

X=109

I 2 = 262 Md =241

co cD

Green River Shale Urquhart Shale

Fig Tree Group (Swaziland System)

44

43

3.3

1.6

0.040 (Eocene)

0.075 (Cretaceous)

(Devonian)

0.360

0.510 (Cambrian)

42

41

40

39

0.475 (Ordovician)

Chambersburg Limestone Conococheague (or Elbrook) Limestone Givetian Limestone Pierre Shale

38

Shale

0.490 (Ordovician)

(. 10 ~ yrs.)

Age

Didy m ograp tus

6

Sample N a m e of rock no.

TABLE I (continued)

shale

limestone

limestone

limestone

shale

Rock type

South Africa

Australia

(siliceous)

argillite (dolomitic; siliceous) argillite

western U.S.A. dolomite

Wyoming, U.S.A.

Belgium

Virginia, U.S.A.

Maryland, U.S.A.

Norway

Location

marine; eugeosynclinal basin?

marine; deep euxinic basin

saline lake

marine; offshore, fairly deep water

marine

marine; similar to Bahama Banks

marine ;similar to Bahama Banks

marine; eugeosynclinal trench; poorly aerated or euxinic

Comments

(%°)

--28.6 1 -~ = --28.8 --29.5 --28.3 Md = --28.6

--23.9

--29.2

--28.1 1 X = --27.9 --27.6

--29.3

--30.9

--29.9

--29.7

kerogen

'+c

--31.1

--27.1

polar extract

59.7

125

224

74.9

1,090

675

102

393

CH2/A6,snm

c~

341

-15

12+13

~

II

I ° tl

~-2-

\\ \

2~AI25 2

A4I ~ 2 5

T

14

~2

~



28 4o '8

-35 . . . .

I

05

. . . .

[

I0

. . . .

I

15

. . . .

I

. . . .

20

I

25

. . . .

I

30

. . . .

l

35

AGE (×109YEARS)

Fig. 1 . 5 '3C values o f k e r o g e n a n d p o l a r e x t r a c t s p l o t t e d against geologic age. T h e exp o n e n t i a l regression line r e p r e s e n t s p o l a r e x t r a c t s f r o m m u d s t o n e s . Explanation of symbols: • = polar extracts from mudstones; • = kerogen from muds t o n e s ; = -- k e r o g e n f r o m c h e r t s ; A = k e r o g e n f r o m c a r b o n a t e rocks.

polar extracts from mudstones also indicate a negative correlation with geologic age {Fig. 1; Table I); but, unlike the kerogen data, they show relatively little scatter, and so the trend is statistically very significant despite the regretably small number of samples (r = --0.881 for a semi-log plot; p < 0.01). Note that the dispersion of the polar extract data is considerably lower than the dispersion of the kerogen data for the s a m e seven rock samples from which the extracts were taken. Thus, the standard error of estimate is 2.22 for the kerogen plot b u t only 0.679 for the polar-extract plot. Furthermore, the kerogen associated with the extracts gave an insignificant secular trend (r = 0.0949; p > 0.1). It would seem, then, that the observed contrast between the polar extracts and the kerogen is genuine, although more data are admittedly needed to confirm the secular trend of the polar extract data. These preliminary results suggest that the extractable polar material, being a specific small fraction of the total organic matter, represents a more restricted range of biochemical precursors than does the kerogen. The polar fraction of the extract has been interpreted as a diagenetic alteration product

342 of sedimentary bitumen derived principally from algal pigments and fatty acids (Jackson, 1975), whereas kerogen is formed chiefly by diagenetic alteration of sedimentary humic matter (Degens, 1965, 1967), a more complex and heterogeneous mixture of components. We tentatively conclude that the scatter in the kerogen data is the result of kerogen's being composed of different molecular subunits that differ from each other in isotopic composition and vary in relative abundance. If this interpretation is correct, it may be possible to reveal hidden patterns of variation in the isotope composition of kerogen by fractionating a set of kerogen samples and analyzing the component parts separately or by normalizing the bulk 613C values with respect to chemical composition so that any differences between the normalized data represent variations in the isotopic composition of one particular subunit or class of subunits in the kerogen molecule. The observed tendency of organic matter to become isotopically lighter with increasing age has been noted by other workers (Degens, 1969; Oehler et al., 1972; Welte et al., 1975), and is thought to reflect a long-term secular decrease in the CO2 content of the atmosphere and hydrosphere (Degens, 1969; Calder and Parker, 1973; Cloud, 1974) or an increase in photosynthetic activity (Welte et al., 1975). The trend cannot be attributed to greater post-depositional alteration of older organic matter, because maturation processes cause organic matter to become isotopically heavier, not lighter (Degens, 1969; McKirdy and Powell, 1974). The existence of secular variation in the extracts would seem to indicate that most, if not all, of the extractable organic matter is indigenous to the rock in which it occurs (cf. Jackson, 1973, 1975; Jackson and Moore, 1976), even though some geochemists have warned that the extractable organic matter of ancient sedimentary rocks may include appreciable quantities of contaminants which migrated into the rocks from younger formations (Hoering, 1965; Smith et al., 1970; Nagy, 1970). Possibly extractable organic matter could even be used for the dating of sedimentary rocks. Now let us examine the relationship between the isotopic composition and chemical properties of kerogen. Kerogen molecules consist of two major structural elements: an aliphatic component and a more or less condensed aromatic component. The ratio of aliphatic to aromatic components, which varies with the biological source material, depositional environment, and degree of post-depositional alteration, is a useful index for semi-quantitative comparison of different kerogens (Forsman, 1963; Degens, 1967). A logical beginning, then, in any attempt to unravel the complexities of the isotope distribution in kerogen would be to try to differentiate between the respective contributions of the aliphatic and aromatic components to the overall 13C value of kerogen. Lacking direct measurements of the aliphatic/aromatic ratios of the kerogens, we tried using the more easily estimated aliphatic/aromatic ratios of the polar extracts instead, on the basis that there is a measureable chemical similarity between kerogen and the extractable polar material associated with it (Jackson and Moore, 1976). To compute

343

the ratio, we divided the intensity of the IR absorption band at 2910 cm -1 (representing aliphatic --CH2 -- groups) by visible-light absorption at 665 nm (representing condensed aromatic material) (Jackson, 1975). The values of this ratio, symbolized as CH2/A66srfn, are given in Table I. Fig. 2 shows 513C of kerogen plotted against the CH2/A66sv~n ratio of its corresponding polar extract. There is a highly significant negative correlation between the two variables (r = --0.596 for a semi-log plot; p < 0.001), and all samples tend to cluster about the same regression line regardless of age, rock type, or depositional environment, indicating that the relationship is a universal attribute of sedimentary organic matter. These data confirm previously reported evidence that the polar extracts are chemically similar to their associated kerogen, are largely derived from the same biological source materials, and can thus be regarded as valid chemical fossils (Jackson and Moore, 1976). The graph reveals that isotopically lighter kerogens tend to be associated with more highly aliphatic soluble polar compounds. This is the result expected on the theory that the aliphatic chains are mostly derived from the lipids of algae. Planktonic algae are regarded as the most important source of the organic matter in marine sediments (Bordovskiy, 1965; Nissenbaum and Kaplan, 1972), and the lipid fraction of these algae tends to be isotopically lighter than the algal biomass as a whole (Degens, 1969). Note that this interpretation provides a plausible explanation for the otherwise puzzling observation that kerogens formed in supposedly glacial environments (samples 1, 6, 25 and 27) are somewhat enriched in ~3C corn-

-i~1 •

• •

-201 U -25 <5, ~

-30-

~-~

© -55

C)

-4o

'

I0

I

. . . .

I

'

I

. . . .

I

I

50 100 500 I,O00 CH2/A665nm (POLAR EXTRACT)

Fig. 2. ~ 13C o f kerogen plotted against CH~ IA66snm ratio of polar extract on a semi-log scale.

Explanation of symbols: glacial rnudstones;

• = calcareous mudstones; • = cherts; • = carbonate rocks.

o = non-calcareous

mudstones;

~ =

344

pared with kerogens in other mudstones (Fig. 1). One might expect "glacial' organic matter to be anomalously rich in 12C, not ~3C, because organisms growing under frigid conditions are enriched in ~2C (Degens, 1969}; accelerated erosion of continental areas resulting in accelerated delivery of isotopically light soil humus to the sea during glacial episodes would further increase the concentration of ~2C in the organic matter. Under glacial conditions, however, the lipids of organisms are more highly unsaturated and consequently more labile than under warm climatic conditions (Shorland, 1962; Martin et al., 1963; Blumer, 1965; Abelson, 1967). Preferential destruction of these unstable lipids during diagenetic alteration of "glacial" organic matter would result in preferential loss of ~2C, thus causing the residual organic matter preserved in the sediments to be enriched in ~3C. This would explain why the glacial kerogens have a relatively high '3C content and why the associated polar extracts are poor in aliphatic components (Jackson, 1975). If 12C concentration in kerogen varies with degree of aliphatic character, as we have demonstrated, then it follows that the aromatic components of the kerogen should be enriched in '3C compared with the aliphatic components. Data obtained from carbonates and cherts appear to support this assumption. Fig. 3 shows that ti ~3C values of kerogen in these rocks tend to increase with an increase in degree of condensed-aromatic character

-15-

-20-

-to

-25-

m m

I

0

r

t

I

I

i

I

I

0.5

I

~

I

CR/C T (KEROGEN) Fig. 3. ~ 13C v a l u e s o f kerogen in c a r b o n a t e r o c k s a n d c h e r t s p l o t t e d a g a i n s t

Explanation of symbols: • = c h e r t s ; A ffi c a r b o n a t e r o c k s .

CR/C T r a t i o .

345

as measured by the " C R / C T " ratio (ratio of "residual" C to " t o t a l " C), which is the proportion of the insoluble organic C remaining as a nonvolatile residue after pyrolysis (Jackson and Moore, 1976). The correlation between 8 ~3C and CR/CT ratio is quite significant (r = 0 . 7 4 4 ; p < 0.01). Thermal metamorphism causes kerogen to become increasingly aromatic and at the same time increasingly rich in '3C, presumably by "cracking" off the ~2C-rich aliphatic chains to form volatile hydrocarbons, leaving behind a more stable, '3C-rich non-volatile aromatic residue. In the present case, however, the relationship between '3C content and aromaticity is probably primary, inasmuch as the samples which have undergone the most severe post-depositional alteration -- namely the Swaziland cherts (samples 14 and 15), representing the lower greenschist grade of metamorphism (M. Viljoen, pers. commun., 1972) -- are not the ones that yielded the highest 8 ~3C values (Table I) and CR/CT values (Jackson and Moore, 1976). Note that the kerogens with the highest 5 ~3C values (samples 4, 11 and 12+13) are all of Proterozoic age (Table I), and those with the highest CI:t/CT values are specifically Middle to Late Proterozoic (Jackson and Moore, 1976). Similarly, certain chemical data representing the aromatic components of the polar extracts show a Late Proterozoic peak, and this has been interpreted as having paleobiological significance (Jackson, 1973, 1975). We do n o t know w h y the relationship between 5 ~3C and CR/C T w a s n o t observed in the mudstone kerogens, b u t a possible explanation is that the mudstones represent a greater diversity of depositional environments and source materials than do the carbonates and cherts, or that clay-sized minerals and metal ions in the mudstones had various u n k n o w n catalytic effects during pyrolysis. The carbonate rocks yielded additional information (Fig. 4), namely a strong negative correlation between 5 ~3C of kerogen and concentration of OH groups in the polar extracts [as measured by the intensity of the IR absorption band at ~ 3 4 0 0 cm -~ , in absorbance units, per mg of sample (Jackson, 1975)] ; r = --0.903, and p < 0.01. This implies that kerogens with a greater degree of condensed-aromatic character are poorer in phenolic OH groups [most of the OH groups are probably phenolic, as pointed out by Jackson ( 1 9 7 5 ) ] . Such a relationship is to be expected, because condensation of aromatic rings in humic matter or kerogen is accompanied by loss of polar groups and side-chains (Degens, 1967). The cherts tend to fall on the same regression line as the carbonates (Table I of this paper; table 3 of Jackson, 1975), except for the Gunflint chert (sample 11), whose organic matter is anomalously rich in ~3C or in OH groups (5 ~3C = --15.6; OH-absorbance/mg = 0.119). The anomalous composition of the Gunflint organic matter could be explained by the fact that the polar extract (and, therefore, probably the kerogen) of this sample is exceptionally rich in carboxyl or other carbonyl-type groups (Jackson, 1975), which, in turn, are probably enriched in '3C (Degens, 1969). We have demonstrated that each of the two major components of the

346

-15-

-20LLJ ~9

o

PC LLI "d:

-25(.P u3

-30i

Q05 OH/mg

i

L

i

i

i

i

QI 0.15 ( POLAR EXTRACT)

Fig. 4.613C values of kerogen in carbonate rocks plotted against concentration of OH groups in polar extract.

kerogen molecule has its own particular C isotope composition. Aliphatic groups tend to be enriched in 12C, whereas aromatic components, and probably carbonyl-type groups as well, tend to be enriched in ~3C. Variations in the distribution of isotopes within each class of components could well introduce further complications. The overall 513C value of kerogen is a composite of the 513C values of the various components of the molecule. Thus, it is not difficult to understand why bulk isotope analyses of kerogen yield relatively little paleo-biological or paleo-environmental information: any systematic patterns of variation would be largely obscured owing to the heterogeneity, complexity, and diverse origins of the kerogen. We attempted to find a partial solution to this problem by normalizing the raw 513C values of kerogen with respect to degree of aliphatic character (as measured by the CH2/A66snm ratio and other parameters) on the assumption that any variations in the normalized data would, to a first approximation, represent variations in the aromatic components of the kerogen. The normalized data were then plotted against geologic age. Although this procedure is valid in principle and did reveal systematic secular variations, it proved to be inapplicable in the present case, because the degree of aliphatic character is orders of magnitude more variable than the C isotope composition. Stepwise linear regression analysis using 5 ~3C, CH~/A66sn m ratio, and age as the variables confirmed this conclusion. Thus, the normalized isotope data reflected the inherent variability of the CH2/A66snm ratio (cf. Jackson, 1977), and the observed patterns of secular variation could have been as-

347

certained from the CH2/A66snm data alone. Therefore, we abandoned this approach. We conclude that the most fruitful approach to this problem would be to fractionate the kerogen physico-chemically and then to determine the 513C value of each fraction. We are currently engaged in further research along these lines. Our demonstration of significant secular variation in the polar extract, which may be regarded as a soluble fraction of the kerogen, augurs well for such an approach. Our data support previously reported evidence that the polar fraction of a benzene--methanol extract of sedimentary rock is, in essence, a solventextractable fraction of the kerogen with which it is associated, and that its properties are more or less related to those of the kerogen as a whole (Jackson, 1975; Jackson and Moore, 1976). These results could have interesting ramifications: the investigation of polar extracts may yield important new paleo-biological and paleo-environmental information (Jackson, 1973, 1975) and information of this kind stored in the molecular structure of kerogen may be obtained more easily, rapidly, and economically from polar extracts. In addition, analysis of polar extracts may lead to a better understanding of kerogen. One of the major obstacles in research on kerogen has been the fact that the substance is insoluble. It is a tedious, time-consuming, and expensive task to separate kerogen from its matrix rock, the separation is at best incomplete, and research on the chemical properties of kerogen depends heavily on specialized, expensive, and destructive analytical techniques such as pyrolysis and ozonolysis (Forsman, 1963). In contrast, soluble polar substances can be completely freed of their rock matrix easily, rapidly, and relatively inexpensively by solvent extraction, and can be analyzed by simple, rapid, non-destructive, widely available conventional methods such as spectrophotometry. Detailed investigation of the molecular structure of polar extracts, as with kerogen, would of course require more highly specialized techniques, but a modestly equipped, modestly financed laboratory could make meaningful quantitative comparisons between the polar extracts of a large number of rock samples in a relatively short time. Ideally, extractable organic matter and kerogen should be studied concurrently. Finally, let us consider whether the isotope data support the hypothesis that a soil microflora existed in pre-Phanerozoic times (Jackson, 1967, 1971, 1973, 1975). A brief review of the distribution of C isotopes in the organic matter of Recent marine sediments will demonstrate why C isotopes should be of key importance in determining the validity of this idea. In Recent marine sediments, "allochthonous" humic matter eroded or leached from soil and introduced into the sea by rivers is isotopically lighter than "autochthonous" humic matter formed by decay of marine phytoplankton -- the major source material (Bordovskiy, 1965; Nissenbaum and Kaplan, 1972) -and other organisms inhabiting the basin of deposition. The difference in isotope composition results from the fact that land plants, being subaerially exposed, utilize more CO2 relative to HCO~ for photosynthesis than do submerged marine algae (Degens, 1969). Therefore, even a pre-Phanerozoic

348 terrestrial flora c o m p o s e d of algae instead of vascular plants would presumably have been isotopically lighter than marine algae of the same age. Allochthonous humic matter, which is largely sorbed on terrigenous clay particles and is partially flocculated on entering the sea, is concentrated in mud deposited near the shore --especially at the mouths of rivers -- whereas offshore muds are richer in autochthonous humic matter (Bordovskiy, 1965; Degens, 1969; Nissenbaum and Kaplan, 1972). This distribution of allochthonous and a u t o c h t h o n o u s components of sedimentary humic matter is detectable in the kerogen, or diagenetically altered humic matter, of ancient sedimentary rocks (Breger and Brown, 1963). Furthermore, the lipid fraction of Recent marine p h y t o p l a n k t o n tends to be isotopically lighter than the phytoplankton biomass as a whole (Degens, 1969). Such differences in primary isotopic composition tend to persist after burial of organic matter in sediments, and can be used for the interpretation of fossil organic matter; according to Degens (1969), diagenesis generally causes only minor changes in the initial isotope ratio. On the generalization that: (1) kerogen is essentially a decay product of aquatic algae combined to a greater or lesser extent with terrigenous organic matter, and (2) sedimentary bitumen (benzene/methanol extractable organic matter) is mostly a decay product of the lipid fraction of these same aquatic algae (Jackson, 1973, 1975), we can make the following predictions. (1) Purely autochthonous kerogen should have a 5 ~3C value less negative than or equal to that of its associated bitumen: (2) Kerogen with an appreciable allochthonous c o m p o n e n t could have a 13C value more negative than that of its associated bitumen: (3) All other conditions being equal, kerogens richer in allochthonous c o m p o n e n t s should have more highly negative 5 '3C values. N o w let us apply these criteria to the seven mudstones for which we have isotope data representing both kerogen and the polar fraction of the bitumen (Table I). The data are scanty and preliminary, b u t they are at least suggestive. In samples 2, 3, 33 and 36, comprising four of the five samples that are 1.05 - 109 years old or younger, the kerogen is isotopically lighter than the associated polar extract, suggesting the presence of allochthonous organic matter (the exception is sample 25, which may be anomalous because it was laid d o w n under glacial conditions). But in samples 22 and 44, which are 1.3 • 109 and ~ 3 . 3 • 109 years old, respectively, the kerogen is isotopically heavier than the polar extracts, as would be expected if the kerogen were purely autochthonous. If our interpretation is correct, the data indicate that land life was in existence at least as long ago as 1.05 - 109 years B.P. Samples 2, 3 and 33 all represent shallow near-shore marine environments where one might expect to find appreciable concentrations of allochthonous organic matter. This is particularly true of sample 3, which was laid down in a deltaic or tidal-flat environment and is thought to be partly non-marine. Yet sample 22, which was laid d o w n in a very similar type of environment, gave no evidence for the presence of terrigenous organic matter. These data,

349

though far from conclusive, suggest that land life arose between 1.3 • 109 and 1.05 • 109 years B.P., the ages of samples 22 and 3, respectively. The earliest known eucaryotic algae appeared in the fossil record 1.3 • 109 years ago, in the mid-Proterozoic (Cloud et al., 1969), and a natural consequence of their subsequent adaptive radiation would be colonization of the land. However, we will need a much larger body of isotope data from suites of sedimentary rocks about which we have reliable paleo-environmental, paleogeographical and geochronological information in order to substantiate this very tentative interpretation. Further research on this topic is being undertaken. ACKNOWLEDGEMENTS

The isotope determinations were financed by National Research Council grant No. A-9754 (awarded to P. Fritz) and by the Canadian Department of the Environment (Inland Waters Branch). Investigation of the spectral properties of polar extracts was performed under the auspices of Dr. Preston Cloud (Biogeology Clean Laboratory, University of California, Santa Barbara), and the carbon analyses and pyrolysis of kerogen were done in the laboratory of Dr. Carleton B. Moore (Center for Meteorite Studies, Arizona State University, Tempe). Most of the chemical work was supported by grants from NASA and the NSF (awarded to P. Cloud and C.B. Moore). P. Cloud furnished most of the rock samples and most of the background information abouth them. We thank Dr. D.P. Scott for help and advice concerning statistical treatment of the data, and Dr. M. Viljoen for information on Swaziland cherts. Other relevant acknowledgments are given by Jackson (1975) and Jackson and Moore (1976).

REFERENCES Abelson, P.H., 1967. Conversion of biochemicals to kerogen and n-paraffins. In: P.H. Abeison (Editor), Researches in Geochemistry~2. Wiley, New York, N.Y., pp. 63--86. Blumer, M., 1965. Organic pigments: their long-term fate. Science, 149: 722--726. Bordovskiy, O.K., 1965. Accumulation and transformation of organic substances in marine sediments. Mar. Geol., 3 : 1 - - 1 1 4 . Breger, I.A. and Brown, A., 1963. Distribution and types of organic matter in a barred marine basin. N.Y. Acad. Sci. Trans., Ser. II, 25: 741--755. Calder, J.A. and Parker, P.L., 1973. Geochemical implications of induced changes in C 13 fractionation by blue-green algae. Geochim. Cosmochim. Acta, 37: 133--140. Cloud, P., 1974. Evolution of ecosystems. Am. Sci., 62: 54--66. Cloud, P., Licari, G.R., Wright, L.A. and Troxel, B.W., 1969. Proterozoic eucaryotes from eastern California. Natl. Acad. Sci. Proc., 62: 623--630. Degens, E.T., 1965. Geochemistry of Sediments. Prentice-Hall, Englewood Cliffs, N.J., 342 pp. Degens, E.T., 1967. Diagenesis of organic matter. In: G. Larsen and G.V. Chilingar, (Editors), Diagenesis in Sediments. Elsevier, Amsterdam, pp. 343--390.

350

Degens, E.T., 1969. Biogeochemistry of stable carbon isotopes. In: G. Eglinton and M.T.J. Murphy (Editors), Organic Geochemistry. Springer, Heidelberg, pp. 304--329. Forsman, J.P., 1963. Geochemistry of kerogen. In: I.A. Breger (Editor), Organic Geochemistry. Pergamon, New York, N.Y., pp. 148--182. Hoering, T.C., 1965. The extractable organic matter in Precambrian rocks and the problem of contamination. Carnegie Inst. Washington Yearb., 64: 215--218. Jackson, T.A., 1967. Fossil actinomycetes in Middle Precambrian glacial varves. Science, 155: 1003--1005. Jackson, T.A., 1971. Carbonaceous inclusions, sulfides, and "fossil gas bubbles" of presumably biologic origin associated with rafted erratics in Huronian (Precambrian) glacial-lake argillites. J. Sediment. Petrol., 41: 313--315. Jackson, T.A., 1973. "Humic" matter in the bitumen of ancient sediments: variations through geologic time. Geology, 1: 163--166. Jackson, T.A., 1975. "Humic" matter in the bitumen of pre-Phanerozoic and Phaneoroic sediments and its paleobiological significance. Am. J. Sci., 275: 906--953; 276: 560. Jackson, T.A., 1977. A relationship between crystallographic properties of illite and chemical properties of extractable organic matter in pre-Phanerozoic and Phanerozoic sediments. Clays Clay Miner., 25: 187--195. Jackson, T.A. and Moore, C.B., 1976. Secular variations in kerogen structure and carbon, nitrogen and phosphorus concentrations in pre-Phanerozoic and Phanerozoic sedimentary rocks. Chem. Geol., 18: 107--136. Jackson, T.A., Fritz, P. and Drimmie, R., 1976. Carbon isotope ratios and chemical properties of kerogen and extractable organic matter in pre-Phanerozoic and Phanerozoic sediments. Geol. Soc. Am., Annu. Meet., Nov. 8--11, 1976, Denver, Colo., Abstr. Programs, p. 938 (abstract). Martin, R.L., Winters, J.C. and Williams, J.A., 1963. Distributions of n-paraffins in crude oils and their implications to origin of petroleum. Nature (London), 199: 110--113. McKirdy, D.M. and Powell, T.G., 1974. Metamorphic alteration of carbon isotopic composition in ancient sedimentary organic matter: new evidence from Australia and South Africa. Geology, 2: 591--595. Nagy, B., 1970. Porosity and permeability of the Early Precambrian Onverwacht chert: origin of the hydrocarbon content. Geochim. Cosmochim. Acta, 34: 525--527. Nissenbaum, A. and Kaplan, I.R., 1972. Chemical and isotopic evidence for the in situ origin of marine humic substances. Limnol. Oceanogr., 17: 570--582. Oehler, D.Z., Schopf, J.W. and Kvenvolden, K.A., 1972. Carbon isotopic studies of organic matter in Precambrian rocks. Science, 175: 1246--1248. Shorland, F.B., 1962. The comparative aspects of fatty acid occurrence and distribution. In: M. Florkin and H.S. Mason (Editors), Comparative Biochemistry 3, Academic Press, New York, N.Y., pp. 1--102. Smith, J.W., Schopf, J.W. and Kaplan, I.R., 1970. Extractable organic matter in Precambrian cherts. Geochim. Cosmochim. Acta, 34: 659--675. Welte, D.H., Kalkreuth, W. and Hoefs, J., 1975. Age-trend in carbon isotopic composition in Paleozoic sediments. Naturwissenschaften, 62: 482--483.