Chapter 3

Bias in the concentration and composition of biomarkers present in

all the fractions of the Jet Rock total sedimentary organic matter

3.1 Introduction and review of previous bound biomarker studies

Bound biomarkers (biomarkers which are covalently bound into macromolecular

organic matter) have been the subject of study for the last two decades in organic

geochemistry, although few such studies have considered investigating the biomarker

compounds present in more than one bound fraction of the total sedimentary organic

matter. Of the studies that have considered more than one bound fraction, only a few

have considered all the fractions and of those that have, most have been non-

quantitative.

3.1.1 Studies without the fractionation of sedimentary organic matter

Pyrolysis (Mukhopadhyay et al., 1995), hydropyrolysis (Eglinton and Douglas, 1989)

and chemical degradation (Schaefer et al., 1995) studies have used whole rock and

whole bitumen fractions as a substrate for performing pyrolysis or chemical

degradation investigations. The products from these studies, while accounting for all

of the biomarkers of a given type that are present in the total sedimentary organic

matter, do not distinguish which of the organic geochemical fractions (e.g., resin,

asphaltene and kerogen fractions) contained the bound biomarkers. Additionally,

unless a labeling technique is used, it will not be possible to discern free phase

biomarkers from their covalently bound counterparts. However, by comparing the

yields of hydrocarbon biomarkers obtained by solvent extraction to the yields of

biomarkers obtained from the pyrolysis or chemical degradation of another fraction

separated from the sedimentary organic matter, it has been possible to discern that

considerable quantities of biomarkers can be present in the bound phases, outside of

the most commonly considered solvent extractable analytical window (Eglinton and

Douglas, 1989).

3.1.2 Pyrolysis correlation studies

Pyrolysis studies have previously analysed asphaltene and kerogen fractions to release

bound biomarker constituents to construct molecular profiles for oil-oil and oil-source

67

rock correlation purposes (Behar et al., 1984; Garg and Philp, 1984; Behar and Pelet,

1985; Philp and Gilbert, 1985; Pelet et al., 1986; Del Rio et al., 1996; Mukhopadyay

et al., 1995). In most instances this has involved comparing the molecular profiles of

asphaltene or kerogen pyrolysates to those of the hydrocarbon fraction of bitumens or

asphaltene pyrolysates. Frequently in pyrolysis investigations (though not in all

cases), the resin fraction biomarkers (i.e. biomarkers covalently held in compounds

that elute in the most polar part of a maltene fraction) are overlooked. This is probably

because asphaltenes are thought to have a greater homogeneity in terms of molecular

weight and composition than resins (Pelet et al., 1986) and when the need to

streamline analytical methods arises, asphaltenes and not resins are perceived to be a

more reliable organic substrate allowing better use of resources. As well as a general

trend to only cover part of the total sedimentary organic matter, most pyrolysis bound

biomarker studies, particularly those using flash pyrolysis Py-GCMS, have been non-

quantitative.

3.1.3 Chemical degradation studies

Time-intensive chemical degradation studies produce their highest yields of

analysable products from low-molecular-weight polar molecules. Thus, chemolysis

studies have tended to focus on resin-type compounds instead of kerogen-sized

compounds, as degradation reactions are unlikely to go to completion in kerogen

fractions due to steric hindrance of potentially reactive binding sites (Rullkotter and

Michaelis, 1990). The practical consequence of this has been that asphaltene-bound

biomarkers (Richnow et al., 1991 and Koopmans et al., 1999; 1997; 1996b), kerogen-

bound biomarkers (Kohnen et al., 1991) or both (Koster et al., 1997) have often been

overlooked in otherwise comprehensive investigations detailing biomarker

compositions and binding mechanisms.

3.1.4 Studies covering all organic matter fractions

Of the studies that have examined biomarkers bound into all geochemical fractions

some have been non-quantitative and have suffered from pyrolysis artefact in a

number of instances (Jenisch et al., 1990; Michels et al., 2000; Murray, 2001).

Additionally, many studies have used multiple degradation methods, typically mixing

a chemical degradation scheme with a pyrolysis technique for analysing kerogen (van

Kaam-Peters et al., 1997; Hold et al., 1999; Koopmans et al. 1999). While this greatly

68

broadens the bound biomarker analytical window it complicates the comparison of

products, as different methods release biomarkers with different compositions. In all,

at the time of writing, only one paper has documented the complete bound biomarker

content of only 3 samples from the Huqf formation in Oman (Hold et al., 1999). In

this study, the samples had all their biomarker pools (kerogen, asphaltene, resin and

free fraction) quantified in their entirety. Clearly the bound biomarker content of the

various fractions of organic matter is still relatively poorly represented in geochemical

literature.

3.1.5 This study

In this study, bound biomarkers have been released by hydropyrolysis, a degradation

method with a unique ability to generate reproducibly high yields of structurally

unaltered bound biomarkers from all types of macromolecular organic fractions (Love

et al., 1995; 1998; Bishop et al., 1998). Three of the most widely studied groups of

aliphatic polycyclic biomarkers (tricyclic terpanes, hopanes and steranes) have been

quantified within the four constituent organic geochemical fractions of sedimentary

organic matter (free hydrocarbons, resins, asphaltenes, kerogens) in sediments from

the Jet Rock formation (Yorkshire, Jurassic, UK). By comparing the compositions of

the four organic geochemical fractions the way in which different biomarker pools

relate to each other, and the inherent biomarker bias of different fractions is

highlighted.

3.2 Geology and sampling

The Jet Rock is the most organic-rich part of the Whitby Mudstone Formation (Pye

and Krinsley, 1986), an early Jurassic (lower Toarcian) deposit laid down in the distal

part of an epicontinental sea that covered much of NW Europe. From its coastal

outcrop in Port Mulgrave (Yorkshire, UK), a series of fourteen mudstone samples

were collected that covered an 8m vertical section including beds 41, 38, 37 and 35

(for bed numbering details, see Howarth, 1962). Whether sampling foreshore or cliff,

care was taken to obtain samples not effected by any apparent surface weathering or

oxidation (i.e. all were sampled from greater than 5 cm depth). All the samples are

part of Morris’s (1979) restricted bituminous facies, and are characterized by an

absence of infauna and low diversity bivalve assemblages interpreted to have been

69

deposited in periods of seabed anoxia. Despite its name and the presence of small

carbonaceous flakes and jet, Jet Rock organic matter has been previously shown to be

chiefly derived from marine algae with relatively minor and limited bacterial and

terrestrial contributions (Farrimond et al., 1989; Ibrahim, 1995; Saelen et al., 2000).

Rock-Eval, and bulk elemental data were obtained for all 14 samples, and measured quantities of the

14 samples were extracted via soxhlet apparatus as described in Chapter 2. Eight samples were then

selected for further study. The bitumen of the samples selected for further study was separated into

aliphatic, aromatic, resin and asphaltene fractions using the method described in Chapter 2. Resin,

asphaltene and kerogen fractions were then subjected to hydropyrolysis and the products separated into

aliphatic, aromatic and polar components following the procedures and methods described in Chapter

2. Hydrocarbon products were analyzed by GC and GC-MS as described in Chapter 2, with compound

identification being based on relative retention times, mass spectra and comparisons to published work.

3.3 Aliphatic biomarkers: Results and discussion

3.3.1 Hand specimen description

A cursory hand specimen description was undertaken on a fragment of each sample.

In general, the results show that the Jet Rock is a homogenous black mudstone.

Exceptions to the general homogeneity of the Jet Rock samples are the amount of

carbonaceous material (jet) present, subtle variations in colour and the type of

sedimentary texture visible with a hand lens. Although some variation in colour

exists, the differences are not stark, and the data in Table 3.1 was obtained by

arranging the samples in order of blackness. Pyrite is present in all samples and

evidenced by iron staining of weathering surfaces at the outcrop, although not all

samples had visible pyrite. It is interesting to note that in samples where the pyritic

replacement of fossils had occurred, ammonite fossil fragments weren’t identified.

The fragments are probably present but their identifying features have probably been

destroyed by re-mineralisation.

In summary, the top of the section (Bed 41) is typically gray/black or gray in colour

with visible disseminated pyrite and small flakes of carbonaceous jet. Beds 35 and 38

are generally blacker in colour, have a less diverse content of fossil fragments and jet

or carbonaceous material is uncommon.

70

Table 3.1, Matrix showing geological features identified in hand specimen

Csample no. bed no. G GB B A B J UF N R D W L N O S J

1 412 413 414 415 416 387 388 389 38

10 3711 3512 3513 3514 35

LaminaeColour Fossils Pyrite Texture

Colour: samples were ordered and grouped: G, gray. GB, gray/black. B, black.Fossils: A, ammonite clasts (unspecified). B, bivalve. J, jet. UF, unidentified.Pyrite (Visible with hand-lens): N, non-visible. R, replacement of fossils . D, disseminated.Texture: W, wavy or discontinuous. L, laminated. N, none or very thick laminationsUnusual Laminae: O, other. S, shell fragments. J, jet.C concreations

Table 3.2 Bulk chemical, and Rock-Eval data

Bed + depth TOC** HI** EOM PI** Tmax**sample m* % mg/g mg/g oC

1 41 3 3.6 459 6.1 0.12 433

2 41 2 3.5 468 5.5 0.15 4313 41 1.3 3.8 514 5.8 0.14 433

4 41 0.8 4.2 523 8.0 0.16 4315 41 0.2 6.8 564 13.1 0.17 430

6 38 -0.2 3.9 585 5.7 0.15 431

7 38 -0.8 4.9 556 9.6 0.16 430

8 38 -1.4 5.6 546 13.1 0.18 4309 38 -1.7 5.1 543 11.7 0.18 431

10 37 -1.9 5.9 524 14.6 0.20 42811 35 -3.4 11.2 549 20.7 0.16 431

12 35 -3.6 12.4 523 17.7 0.15 43013 35 -3.8 8.9 545 18.7 0.16 429

14 35 -4.2 8.9 519 18.1 0.15 427

mean 6.34 530 12.0 0.16 430

SD 2.92 34.3 5.4 0.02 1.6

+ According to Howarth (1962)*distance from base of bed 41**duplicate averagesamples in bold were subjected to hydropyrolysis experiments

71

Tab

le 3

.3.

Yie

ld o

f pr

oduc

ts p

pt (

mg/

g T

OC

init

ial s

ed)

obta

ined

by

solv

ent e

xtra

ctio

n an

d hy

drop

yrol

ysis

of

kero

gen,

asp

halt

ene

and

resi

n fr

acti

ons.

Res

in

pol

0.7

nd 2.6

4.2

0.5

7.5

5.5

nd nd

aro

0.1

nd 1.3

1.5

0.2

2.0

2.3

nd nd

ali

1.0

nd 5.1

7.2

0.3

14 3.1

nd nd

tot

4.8

nd 9.3

18 4.4

18 12 nd nd

init

14 23 53 58 23 55 31 50 nd

Asp

halt

ene

pol

1.9

0.6

0.6

0.3

2.0

2.2

7.4

0.7

nd

aro

1.2

0.6

0.3

0.4

1.8

1.2

0.7

1.0

nd

ali

5.8

1.3

0.2

0.1

2.9

12 0.7

0.8

nd

tot

10 5.2

0.7

2.6

8.7

12 8.7

3.5

nd

init

15 10 8 10 18 29 15 22 nd

Ker

ogen

pol

78 34 172

90 112

146

191

187

132

ali =

ali

phat

ic; a

ro =

aro

mat

ic; p

ol =

pol

ar

aro

107

132

181

200

124

127

171

137

127

init

= in

itia

l am

ount

rec

over

ed f

rom

DC

M e

xtra

ct

ali

147

146

128

157

132

131

166

82 192

nd =

not

det

erm

ined

tot =

tota

l rec

over

ed a

fter

HyP

y

tot

424

383

545

451

463

445

465

449

640

DC

M e

xtra

ct

aro

46 10 30 42 16 37 42 54 nd

ali

25 37 61 61 30 80 72 65 nd

EO

M

169

152

193

145

199

223

185

211

nd

Sam

ple 1 3 5 6 7 9 11 13

dup

11

72

3.3.2 Bulk chemical data

Rock-Eval maturity parameters suggest that the Jet Rock samples are at the threshold

of oil generation (Peters, 1986; Production Index (PI) >0.1 but relatively low Tmax ~

430 oC; see Table 3.2). This agrees with previously measured vitrinite reflectance

values of %Ro = 0.55 to 0.69 (Ibrahim, 1995). Tmax varies little through the studied

section; the short depth interval (7.2 m) and its relative geological homogeneity of

which suggest that the samples collected are effectively isomature. The average

Hydrogen Index (HI) (determined according to Langford and Blanc-Valleron (1990)

to avoid matrix effects) of 554 mg/g and generally high TOC values are typical of an

organic matter-rich Type II marine source rock (Peters, 1986).

3.3.3 Bitumen bulk composition and hydropyrolysis yields

The yield of extractable organic matter (EOM) and the composition of the bitumen

phase (wt. %) is shown in Table 3.3. The bitumen comprises only a minor part (< 20

%) of the Jet Rock organic matter, whilst the kerogen constitutes between 80 and 90

wt. % (based on Rock-Eval data, and HyPy recovery data). Of the bitumen, typically

around 40 wt. % or less is in the polar fractions (resin and asphaltene). The greatest

amount of aliphatic material is in the kerogen fraction (~140 mg/g TOC), the next

largest amount is in the DCM extracted aliphatic fraction (~75 mg/g TOC). The resin

and asphaltene fraction’s hydropyrolysis aliphatics constitute a small part, typically

less than the free aliphatic fraction, of the total sedimentary organic matter. The

largest fraction overall is the aromatic fraction of the kerogen hydropyrolysate (~145

mg/g TOC), although though in a few instances the aliphatic material of the kerogen

hydropyrolysate constitutes the largest fraction.

3.3.4 Biomarker abundance

Biomarker concentrations are reported as both ppm (g/g TOC) for total steranes,

hopanes and tricyclic terpanes (Table 3.4) and for C29

15(H)(H)(H)20R) sterane, C30 17(H)(H) hopane and C23

13(H)(H)tricyclic terpane (Table 3.5). These specific biomarkers have been

reported so as to allow a comparison between this study and the work of Hold et al.

(1999) who quantified these specific biomarkers in four geochemical fractions in the

Huqf formation Oman, and Murray et al. (1998) who performed a comparable

73

hydropyrolysis study of the Kimmeridge Clay focusing on the distribution and

abundance of free and kerogen bound hopanes and steranes across the oil window.

Based on this early oil window data set, it is apparent that the free aliphatic

hydrocarbon fraction is relatively enriched in biomarkers, and most notably steranes,

compared to the macromolecular fractions (Table 3.4, note sterane concentrations

include diasteranes). Steranes are likely enhanced in the aliphatic hydrocarbon

fraction by the formation of diasteranes and diasterenes early in diagenesis (de Leeuw

et al., 1989; van Kaam-Peters, 1998), but also the terminal position and single A-ring

binding site of steranes (see Chapter 1) leads to their preferential release from

macromolecules (Kohnen et al, 1991) and higher abundance in the free fraction. The

resin and asphaltene fractions do not appear to be quantitatively significant biomarker

pools, which is in accord with other data (Hold et al., 1999). Similarly, when the

proportion of these polar bitumen fractions constituting the organic matter is much

greater, presumably their quantitative significance as biomarker pools is also greater

(Schaeffer et al., 1995). Within the Jet Rock data set the resin fraction in general

appears to contain a slightly higher proportion of 5(H)(H)(H) (20R) C29

sterane, 17(H)(H) C30 hopane and 13(H) (H) C23 tricyclic terpane (Table

3.5) than the asphaltene fraction, but this is probably because it also constitutes a

greater proportion of organic matter (Table 3.3). The quantities of these 3 biomarkers

seen in the polar fractions of the Huqf formation (Hold et al., 1999), although of a

similar order of magnitude, do not have the same distribution between the polar

fractions, and show that the resin fraction is the quantitatively less significant polar

bound biomarker pool. This difference may in part be due to the two comparable but

not identical sample fractionation methods.

Table 3.4. Total biomarker concentrations ppm (g/g TOC)

Sample Steranes ppm TOC Tricyclic terpanes ppm TOC Hopanes ppm TOCFree Res Asph Ker Free Res Asph Ker Free Res Asph Ker

1 140 0.5 0.2 35 17 0.4 0.9 15 85 0.8 0.5 883 220 nd 0.3 22 30 nd nd nd 150 nd 0.7 505 220 1.1 0.2 27 83 2.4 0.1 36 210 1.6 0.3 906 150 1.3 0.5 25 17 1.0 0.8 18 93 1.1 1.3 657 130 0.8 0.4 31 30 0.3 1.6 16 99 1.1 0.9 659 270 3.0 0.3 30 66 2.8 0.4 15 240 6.0 1.0 88

11 260 0.8 0.2 32 42 0.8 0.02 16 260 1.4 0.5 10013 180 nd 0.2 26 30 nd 0.2 12 180 nd 0.7 50

Free = solvent extract

74

Res = Resin; Asph = Asphaltene; Ker = Kerogen hydropyrolysates

The kerogen fraction is by far the dominant bound biomarker pool, and appears to be

enriched in hopanes (Table 3.4) relative to the other two compound types. It still

contains a slightly lower proportion of total biomarkers than the free aliphatic

hydrocarbon fraction, but these still account for approximately 30 to 50 % of the total

biomarkers depending on biomarker type. Hopanes are bound into macromolecules at

a larger number of sites than steranes (Hofmann et al., 1991; Richnow et al., 1991;

Rohmer et al., 1993; Chapter 1) and are, therefore, bound in greater quantities into

kerogen further into the oil window (see Hofmann et al. (1991) for a description of

the Serpiano kerogen). The data of Hold et al. (1999) also suggests that the kerogen

fraction contains more biomarkers than either of the other macromolecular fractions,

however, kerogen can even contain more biomarkers than the free fraction at the

beginning and end of the oil window (Murray et al., 1998) – although the data of this

study do not fall within these regions of the oil window.

Table 3.5. Individual biomarker yields ppm (g/g TOC)

C30 Hopane ppm TOC C29 R Sterane ppm TOC C23 Tricyclic ppm TOCFree Res Asph Ker Free Res Asph Ker Free Res Asph Ker

1 25 0.1 0.1 11 8.4 0.1 0.05 12 2.6 0.1 0.2 4.23 41 nd 0.1 5.8 15 nd 0.1 6.4 4.2 nd nd nd5 57 0.2 0.04 11 16 0.4 0.1 9.6 11 0.5 0.02 5.96 26 0.1 0.2 8.1 8.9 0.4 0.1 6.6 2.5 0.2 0.1 2.87 27 0.1 0.1 7.7 9.1 0.3 0.1 8.9 3.9 0.1 0.3 3.09 66 1.0 0.2 11 18 0.7 0.1 9.1 8.5 0.4 0.1 2.611 75 0.1 0.1 12 16 0.2 0.1 8.6 6.2 0.1 < 0.01 2.813 52 nd 0.04 8.5 11 nd 0.1 7.1 4.0 nd 0.04 1.7

Free = solvent extractRes = Resin; Asph = Asphaltene; Ker = Kerogen hydropyrolysatesC29 R Sterane = 5(H) (20R) C29 sterane; C30 Hopane = 17(H) C30 hopane; C23 Tricyclic = 13 (H) C23 Tricyclic terpane

3.3.5 n-Alkanes

Figure 3.1 shows the total ion chromatogram (TIC) of the free aliphatic hydrocarbon

fraction and the aliphatic fractions of the resin, asphaltene and kerogen

hydropyrolysates of sample 5. The n-alkanes range up to C38 and beyond in all

fractions, but the hump associated with the polycyclic biomarkers (hopanes and

steranes) is only prominent in the TIC trace of the free aliphatic hydrocarbon fraction.

n-Alkenes are only observed in significant amounts in the kerogen hydropyrolysates

as previously observed in hydropyrolysis studies (Love et al., 1995, 1998; Murray et

75

al., 1998). Of particular note in Figure 3.1, are the pronounced C16, C18 and C22 n-

alkanes in the hydropyrolysis products, a feature which has also been observed in

chemical degradation products (Richnow et al. 1991). This suggests that even-carbon

number n-alkanes predominate as geomacromolecule sub-units, and that

hydropyrolysis is a relatively mild degradation technique capable of liberating and

preserving carbon number distributions in a similar manner to chemical degradation

techniques and methods. Hydropyrolysis experiments (Love, personal communication

2002) with carboxylic acid model compounds show that the complete reduction of

carboxyl is favoured over decarboxylation (which typically occurs when performing

pyrolysis in an inert gas atmosphere e.g. He or N2 – see also section 1.5 Chapter 1),

this may help to preserve carbon number distributions.

76

asphaltene

d4 cholestane standard

30

1622

18

34

kerogen

30

16

22

18

34

resin

16

2218

3034

free Biomarker region16

22

18

elution time

Figure 3.1. TIC (Total ion current) of aliphatic hydrocarbons released from each of the organic matter fractions in a Jet Rock sediment (sample 5). There is only a very slight stratigraphic variation in the n-alkane traces.

77

free

resin

asphaltene

kerogen

C27

R

C29

R

C28

R

C30

R

C27

S di

a

C29

C27

S

C28

S

C29

S

C29

R+

S

C29

R

C28

R

C30

R

C29

R+

S &

C27

R

C27

S

C 29

C29

R

C30

R

C28

R

C29

S

C29

R

C30

R

C28

RC

27

R

C27

S

&C

27

C27

S

elution time

C27

R d

ia

C29

S di

a

C29

R d

ia

R S

Figure 3.2. m/z 217 Ion chromatograms for sample 5. C27 S dia = C27 13(H), 17(H) (20S) diasterane; C27 R dia = C27 13(H), 17(H) (20R) diasterane; C29 S dia = C29 13(H), 17(H) (20S) diasterane; C29 R dia = C29 13(H), 17(H) (20R) diasterane; C27 S = C27 5(H), 14(H), 17(H) (20S) sterane; C27 R = C27 5(H), 14(H), 17(H) (20R) sterane; C27 R+S = C27

5(H), 14(H), 17(H) (20S) + (20R) steranes; C27 = C27 5(H), 14(H), 17(H) sterane

3.3.6 Steranes

All fractions contain C27 to C30 steranes dominated by 5(H)(H)(H)

configurations. In the bound fractions the (20R) isomers are significantly more

abundant than the (20S) isomers (Table 3.6). (H)(H)(H) steranes are

found in the pyrolysates of the kerogen, resin and asphaltene fractions but not in the

free aliphatic hydrocarbon fraction, which contains the greatest proportion of

5(H)(H)(H) steranes (Figure 3.2 and 3.3). The sterane isomers are also

78

present to lesser degrees in the resin, asphaltene and kerogen fractions but they co-

elute with isomers and are less clearly resolved. As reported previously (Peters et

al., 1990; Love et al., 1995; Murray et al., 1998), sterane isomerisation is generally

more developed in the free aliphatic hydrocarbon biomarkers compared with those of

the bound fractions (Table 3.6). This is interpreted to be a consequence of the steric

protection offered to the bound biomarkers by their covalent binding within bulky

macromolecular host networks.

C27

R

C29

R

C28

R

C30

R

C27

S

dia

C29

C29

S

C27

S

C27

R

C29

R

C28

R

C30

R

C29

S

C27

S

C27

R

C29

R

C28

R

C30

R

C29

S

C27

S

C27

R

C29

R

C28

R

C30

R

C29

S

C27

S &

C27

free

resin

asphaltene

kerogen

elution time

C27

R

dia C29

S

dia

C29

R

dia

C29

R+

SC

29

R

+S

&

C 29

R S

Figure 3.3.m/z 217 Ion chromatograms for sample 11. For key see Figure 3.2.

79

Diasteranes are present in the free aliphatic hydrocarbon fractions (Figure 3.2 and

3.3), but only trace amounts are found in some (and not all) of the bound biomarker

fractions. The very low concentrations of diasteranes present in these bound fractions

probably represent free diasteranes that have co-eluted with the resin and asphaltene

fractions during sample preparation and not diasteranes that are covalently bound into

macromolecules. An instance of apparently-bound diasteranes in a source rock extract

maltene fraction was reported by Peng et al. (1999), but most previous work suggests

that diasteranes are not bound into asphaltene or kerogen fractions (Seifert, 1978;

Philp and Gilbert, 1985; Murray et al., 1998; Strausz et al., 1999). Rather, they are

formed from exclusively free phase rearrangement reactions, perhaps involving clay

minerals and free sterenes (De Leeuw et al., 1989).

It is likely that diasteranes do not become bound into the macromolecular fractions

due to the diagenetic reactions that lead to their formation producing non-reactive

species that are unable to bind with macromolecules, e.g. sterane backbone

rearrangement catalyzed by acidic clays creates diasterenes that have their double

bond in a hindered position, and are subsequently reduced to form diasteranes (de

Leeuw et al., 1989; van Kaam-Peters et al., 1998). These reactions may effectively

determine that a large proportion of free steroids remain in the free aliphatic

hydrocarbon fraction, partially explaining the relative low proportions of steranes in

the kerogen fraction.

Free hydrocarbon biomarkers (often trapped in micropores) can contaminate

hydropyrolysis products by co-precipitating and co-eluting with the resin, asphaltene

and kerogen fractions during sample preparation. Based on the restricted occurrence

of diasteranes to predominantly the free hydrocarbon fractions, when they are present

in m/z 217 traces it can be inferred that contamination by free phase biomarkers has

occurred. This relies on some of the first principles established for bound biomarkers

(Seifert, 1978; Rubinstein et al., 1979):

1) Compounds not covalently bound into macromolecules should not be present in

hydropyrolysis products.

2) Maturity parameters in bound fractions should show evidence of steric protection.

80

Applying these criteria to the data in Figure 3.4 reveals that the resin hydropyrolysates

of samples 1 and 9 have anomalously high maturity parameter values (their values are

off-trend); these samples also contain significantly more diasterane signal in the m/z

217 trace than the hydropyrolysis products of the other resin fractions (see sample 1

and 9 Figures 4.3b and 4.3c in Chapter 4). On review of laboratory data this could be

linked to overloading of silica columns during maltene fractionation and consequently

poor separation by column chromatography, although the possibility exists that a

small proportion of the trapped free hydrocarbon species are impossible to remove in

any situation.

0

0.1

0.2

0.3

0.4

0.5

free resin asphaltene kerogen

C29

ste

rane

S/S

+R

sample 1 and 9

sample 11

Figure 3.4. Plot of sterane maturity parameters measured in the hydropyrolysis products and DCM extracts of the Jet Rock. Open circles denote parameters influenced by occluded or chelated free steranes that have not been sterically protected. Sterane C29 S/S+R = C29 sterane (20S) / C29 sterane (20S) + C29 sterane (20R)

The asphaltene hydropyrolysate of sample 11 contains notably more diasteranes than

any of the other asphaltene fraction hydropyrolysates, and its C29 sterane

(20S)/(20S+20R) value is off trend in Figure 3.9. Again, this would seem to suggest

that free biomarkers are occluded by or chelated in the asphaltene fraction and have

mixed with the bound biomarkers released by hydropyrolysis.

Ternary plots of sterane carbon number are often used to facilitate oil-oil and oil-

source rock correlation and to provide a general comparison of depositional

environments (Peters and Moldowan, 1993). Significantly however, the steranes of

the different organic fractions within a particular sample have different carbon

81

number distributions (Figure 3.5). In general the bound steranes appear to be slightly

enriched in the C27 homologues compared with the free aliphatic hydrocarbon

fraction. This has been previously observed in both hydrous pyrolysis products

(Fowler and Brooks, 1987) and chemical degradation products (Richnow et al., 1991).

This difference is significant at the 95 % confidence interval (based on ANOVA

results (ANalysis Of VAriance); however this would probably not impact the

correlation and distinguishing of source rocks from vastly different depositional

environments by use of biomarkers from different organic fractions. Additionally, the

presence of other key compounds indicative of specific depositional environments

such as C30 steranes would be used to supplement parameters based on the regular

steranes.

resin

asphaltene

kerogen

free

C27 R sterane

C 28

R

ster

ane C

29 R sterane

0

40 20

602060 40

20 40

Figure 3.5. Ternary plot showing the sterane 5(H)(H)(H) (R) composition of the various organic fractions, that the bound fractions are generally enriched in C27 steranes compared to the free sterane.

3.3.7 Hopanes

All fractions contain C27 to C35 (excepting C28) hopanes with 17(H)(H) and

17(H)(H) isomers being the dominant stereochemistry. However, the relative

proportions of these compounds differ in each fraction (Figures 3.6 and 3.7). The free

fraction hopane profile is always dominated by C30 hopane. Hopanes in the resin

and asphaltene hydropyrolysis products (Figure 3.6 and 3.7) show more varied carbon

number distributions in which generally C32 (resin) or C31 (asphaltene) hopanes

dominate, although this varies between samples. As noted in a previous study (Bishop

et al., 1998), hopanes released from the kerogen phase of ancient sediments by

82

hydropyrolysis are often dominated by the C29 hopane homologue; the kerogen

fraction displays a degree of uniformity in hopane composition similar to that of the

free hydrocarbons.

C29

C30

C30

C31

S&R

C32

S&R

C33

S&R

C34

S&R

C35

S&R

C27

Ts

C27

Tm

TT

C23

TT

C24

TT

C25

TT

C26

S&

R

TT

C28

S

&R

TT

C29

S

&R

TT

C30

S

&R

TT

C22

}TT

C 3

3

}TT

C34

TT

C31

S

&R

free

resin

asphaltene

kerogen

C29

C30

C30

C31

S&R

C32

S&R

C33

S&R

C34

S&R

C35

S&R

TT

C23

TT

C24

TT

C25

TT

C26

S

&R

TT

C22

C27

Tm

C29

C30

C

30

C31

S&R

C32

S&R

C33

S&R

C34

S&R

C35

S&R

TT

C23

TT

C24

TT

C25

TT

C26

S

&R

TT

C22

C27

Tm

C29

C29

elution time

SR S

RS

RS

R S R

S R S RSR

S R S R

SR

S R

SR

SR S R

TT

C23

TT

C24

TT

C25

Figure 3.6. m/z 191 Fragmentograms TT C23 = C23 13(H), 14(H) tricyclic terpane (note various tricyclic terpane isomers are not individually labeled); C27Ts = C27 18a(H)-22, 29, 30 trisnorneohopanes; C27 Tm = 17a(H)-22, 29, 30 trisnorhopane; C29 =C29 17(H), 21(H) hopane; C29 =C29 17(H), 21(H) hopane; C31 S = C31 17(H), 21(H) (22S) hopane; C31 R = C31

17(H), 21(H) (22R) hopane. Tricyclic terpanes are more prominent in the resin and asphaltene fractions than in the free and kerogen fractions. Sample 5 contains greater amounts of tricyclic terpanes in all fractions than sample 9 (Figure 3.7).

The rearranged C27 and C29 18(H)-trisnorneohopanes are significant only in the free

fractions and the resin and asphaltene hydropyrolysis products of some of the

samples. In the Jet Rock, their partitioning is broadly similar to that of the diasteranes,

i.e. only present in hydropyrolysates due to the partial trapping of free hydrocarbons

with the resin fraction during preparative column chromatography. It is likely that

only negligible quantities of 18(H)-trisnorneohopanes, are actually bound into the

resin, asphaltene and kerogen fractions of the Jet Rock sedimentary organic matter.

83

free

resin

asphaltene

kerogen

C29

C30

C30

C31

S&R

C32

S&R

C33

S&R

C34

S&R

C35

S&R

C27

Ts

C27

Tm S

RS

RS

RS

R S RT

T C

23

TT

C24

TT

C25

TT

C23

TT

C24

TT

C25

TT

C26

S&

R

TT

C28

S

&R

TT

C29

S

&R

TT

C30

S

&R

TT

C22

C27

Tm

C29

C30

C30

C31

S&R

C32

S&R

C33

S&R

C34

S&R

C35

S&R

SR

SR

SR S

R S R

C29

C30

C30

C31

S&R

C32

S&R

C33

S&R

C34

S&R

C35

S&R

SR

SR

SR

SR S R

C27

Tm

TT

C24

TT

C25

TT

C26

S&

R

TT

C22

TT

C23

TT

C28

S

&R

TT

C23

C29

C30C

30

C31

S&R

C32

S&R

C33

S&R

C34

S&R

C35

S&R

SR S

RS

RS R S R

C27

Tm

C29

elution time

Figure 3.7.GC-MS m/z 191 fragmentograms of sample 9 showing the distribution of tricyclic terpanes and hopanes. Sample 9 contains much smaller amounts of tricyclic terpanes in it’s fractions than sample 5 (Figure5). For key, see Figure 3.6.

The data in Table 3.6 show that the C30 hopane parameter decreases in

steps between fractions, and thus the biomarkers bound into progressively larger

macromolecules contain greater proportions of isomers that are most abundant at

lower thermal maturities (e.g. moretanes with the configuration as opposed to

hopanes with the configuration). This could be interpreted as larger

macromolecules offering greater degrees of steric-protection (Seifert, 1978; Murray et

al., 1998), which in turn suggests that the C30 hopane parameter is

primarily controlled by isomerisation that is occurring in both the bound and free

biomarkers. If this is the case, than the preferential release of one isomer over another

during oil (and biomarker) generation is not the controlling process for the C30 hopane

parameter, at least at this stage in the oil window and for this type of a

source rock. This interpretation differs from many previous studies of the mechanisms

controlling other biomarker maturity parameters (list).

84

Table 3.6. Biomarker parameters measured for the free solvent extracts and resin, asphaltene and kerogen hydropyrolysis products for Jet Rock sediments.

free resin asphaltene Kerogen free resin asphaltene kerogen1Sterane C29 S/S+R 2Hopane C31 S/S+R

1 0.43 0.40 0.24 0.30 0.57 0.61 0.56 0.503 0.43 nd 0.23 0.33 0.56 nd 0.59 0.515 0.46 0.24 0.27 0.32 0.58 0.67 0.60 0.536 0.46 0.25 0.28 0.34 0.58 0.67 0.64 0.517 0.45 0.28 0.31 0.34 0.57 0.64 0.59 0.519 0.47 0.36 0.30 0.36 0.59 0.60 0.59 0.49

11 0.44 0.30 0.34 0.36 0.58 0.56 0.56 0.5013 0.44 nd 0.24 0.37 0.59 nd 0.60 0.50

3Sterane C27 R / C29 R 4HopaneC30

1 0.54 1.11 0.41 1.17 0.15 0.18 0.28 0.363 0.69 nd 0.70 1.06 0.15 nd 0.27 0.405 0.86 1.16 1.03 1.24 0.14 0.20 0.27 0.376 0.68 0.99 1.00 0.90 0.16 0.22 0.28 0.377 0.81 2.40 1.19 1.05 0.14 0.20 0.27 0.379 0.83 1.08 1.23 1.17 0.14 0.17 0.25 0.39

11 0.70 1.12 1.24 0.88 0.15 0.24 0.26 0.3813 0.72 nd 0.78 0.89 0.14 nd 0.27 0.38

5C27-29 R Steranes/ C30 Hopane 6C30 Hopane/C23 Tricyclic terpane

1 1.1 1.9 2.2 1.6 11.0 1.8 0.6 4.13 1.1 nd 2.0 1.8 11.3 nd nd nd5 0.7 3.6 2.7 1.2 5.9 0.5 2.0 3.06 1.0 6.1 1.3 1.4 12.3 0.9 2.3 4.77 0.9 3.5 2.1 1.9 8.2 2.2 0.6 4.09 0.7 1.5 0.9 1.2 9.0 2.7 3.3 6.7

11 1.1 3.3 1.1 1.2 11.0 1.3 33.2 7.113 0.6 nd 3.3 1.4 14.9 nd 1.4 8.1

1Sterane C29 S/S+R = C29 sterane (20S)/ C29 sterane (20S) + C29 sterane (20R)2Hopane C31 S/S+R = C31 hopane (22S)/C31 hopane (22S) + C31 hopane (22R) 3Sterane C27 R / C29 R = C27 sterane (20R) / C29 sterane (20R) 4HopaneC30 = hopaneC30 17(H) / hopaneC30 17(H) + hopaneC30 517(H) C27-29 R Steranes/ C30 Hopane = C27 sterane (20R) + C28 sterane (20R) + C29 sterane (20R) / hopaneC30 17(H) + hopaneC30 17(H) 6C30 Hopane/C23 Tricyclic Terpane = hopaneC30 17(H) / C23 13, 14 (H) tricyclic terpane

The C31 hopane (22S)/(22S+22R) parameter, aside from showing exposure to near

oil window temperatures, can not be interpreted with any degree of certainty for two

reasons. 1) The parameter is at or very close to equilibrium value in most instances,

and 2) the exact cause of the higher values in the resin fraction is uncertain. Co-

eluting C33 extended tricyclic terpanes, the defunctionalisation in hydropyrolysis of

free fraction hopanoic acids that co-eluted with the macromolecular component of the

resin fraction during preparative column chromatography or even analytical error

caused by applying the integration algorithm used to measure peak areas to vastly

85

different sized and shaped peaks could all be cited as the cause of the difference

between the observed values.

3.3.8 Tricyclic terpanes

C20 to C30 13, 14 (H) tricyclic terpanes are found in all fractions of all the samples.

They are less abundant than the hopanes in m/z 191 mass chromatograms of the free

aliphatic hydrocarbon fractions (Figure 3.6, and Table 3.5), but are of comparable

abundance relative to the hopanes in the resin and asphaltene fractions. Due to the

small proportion of the total organic matter comprised by the resin and asphaltene

fractions (Table 3.2), the total tricyclic terpane content is less than the hopane content

(Table 3.4). Extended (>C30) tricyclic terpanes are most visible in the chromatograms

of the asphaltene and resin fractions where they range up to at least C34, although they

can also be seen to be very minor constituents of the free fraction. Their apparent

absence from the kerogen fraction may be due to dilution by the more abundant

hopanes. Extended tricyclic terpanes (ETT) have previously been found to be

relatively enriched in resin chemical degradation products of early oil window (0.83

% Ro) coals (Jenisch et al., 1990), as well as the resin chemical degradation products

of a Tasmanites oil shale (McCaffrey et al., 1994). Deuterium incorporation suggests

that tricyclic terpanes, like hopanes, are bound into macromolecules at multiple

positions (2 or more) on their side chain (Peng et al 1997; Strausz et al., 1999),

Although biological precursors have functional groups on their side chain, as well as

in their ring system (McCaffrey et al., 1994) these have not been reported as sites of

attachment in geomacromolecules. These previous observations as well as the data of

this study suggest that the polar and particularly the resin fraction, could be greatly

enriched in extended tricyclic terpanes (ETT) at the onset of oil generation because of

the potentially higher number of binding sites present in the biomarker structure that

could lead to tight binding into macromolecules (as is observed for hopanes see

Figure 1.23).

However, not all the samples have resin fractions that are as greatly enriched in

extended tricyclic terpanes as sample 5, sample 11, for example, does not (compare

Figures 3.6 and 3.7). This variation in the prominence of extended tricyclic terpanes

could represent stratigraphic variation as these samples are from different horizons,

but more likely represents dilution by free hopanes that co-eluted with the resin

86

fraction during sample preparation (preparative column chromatography) and then

swamp the extended tricyclic terpanes. Maturity data in Table 3.6 partially support

this as sample 11, whose resin hydropyrolysis tar contains what is relatively a large

proportion of hopane for a resin fraction, has slightly higher hopane C31

(22S)/(22S+22R) and sterane C29 (20S)/(20S+20R) parameters and lower C30

( parameters (e.g. maturity parameters are close to those of the free

aliphatic hydrocarbon fraction due to the presence of free aliphatic hydrocarbons in

the resin fraction hydropyrolysate).

3.4 Summary of aliphatic bound biomarkers: Implications for

biomarker studies

3.4.1 Implications for correlation studies

The differences in biomarker abundance and distribution that are observed between

the different fractions in a particular sediment sample will impact biomarker

interpretation in a number of ways. Firstly, the different organic matter pools (free,

resin, asphaltene and kerogen) have different biomarker compositions. For example,

the m/z 191 traces of the different fractions in Figures 3.6 and 3.7 look very different

since the proportions of tricyclics to hopanes and hopane carbon number distributions

produce vastly different GC-MS fingerprints (Peters and Moldowan, 1993). The

sterane profiles of different fractions of sedimentary organic matter are also

significantly different, perhaps even enough to prevent correlation of oils and source

rock extracts if using biomarkers from the different fractions of sedimentary organic

matter. Both of these facts ought to raise a cautionary note when correlating oils of

similar thermal maturities using biomarkers from different fractions of sedimentary

organic matter. The fact that C30 regular sterane, an indicator of marine

environmental conditions, was found in all organic matter fractions could prove very

useful in distinguishing oils from marine and non-marine sources using biomarker

fingerprints obtained from the different fractions of sedimentary organic matter.

4.2 Quantitatively-significant kerogen-bound biomarker pool

The samples of this study are marginal oil window maturity and despite the

expectation that many biomarkers would have been released into the free fraction,

significant quantities of biomarkers are still present in the kerogen fraction. The

87

obvious implication of this finding is that the kerogen fraction merits attention in

geochemistry biomarker studies as it represents a major biomarker pool even at the

onset of oil generation, an observation that accords with a previous hydropyrolysis

study of bound biomarkers across the oil window (Murray et al., 1998). In fact, only

in very immature and high-organic-sulphur sediments and rocks, where resin and

asphaltene molecules comprise a greater proportion of sedimentary organic matter

(e.g. the Green River Formation see Eglinton and Douglas, 1988; Bishop et al 1998),

would it appear that their cracking during thermal maturation will significantly alter

the concentration and composition of biomarkers in the free extractable hydrocarbon

fraction. In general, it would seem that only the kerogen fraction contains sufficient

quantities (Table 3.4 and 3.5) of bound biomarkers at this stage of the oil window (the

onset of oil generation) to alter the composition of the free aliphatic hydrocarbon

fraction biomarkers in source rocks. An additional observation that should be borne in

mind when performing correlations is the difference between resin and asphaltene,

and kerogen and free fraction biomarker populations.

3.4.X Differences in the biomarker parameters measured in different fractions

Figure X. Conceptual models of the C30 hopane C29 sterane (20S)/(20S+20R) and C31

hopane (22S)/(22S+22R) parameters. The dashed lines represent the distribution of parameter values that would be expected if the C29 sterane (20S)/(20S+20R) was governed by the same processes as the C30 hopane parameter, and the literal interpretation of the data measured for the C31 hopane (22S)/(22S+22R) parameter. For the reasons discussed in the text such a literal interpretation of the C31 hopane (22S)/(22S+22R) parameter data is not possible in this instance.

Figure X shows three conceptual models of the biomarker parameters reported in

Table 3.6 and discussed in section 3.3. The C29 sterane (20S)/(20S+20R)

parameter differs from the C30 hopane parameter in that the points do not

lie along a straight line. The kerogen bound steranes do contain greater proportions of

the (20R) isomer than the free steranes, consistent with the steric protection of bound

biomarkers (Seifert, 1979; Murray et al., 1998). However, the resin and asphaltene

parameter values do not show the same pattern that was observed for the C30 hopane

88

parameter, e.g. a progressive decrease in the proportion of the thermally

less mature isomers with decreasing macromolecule size. The opposite trend is

observed for the C29 sterane (20S)/(20S+20R) parameter and the resin and

asphaltene bound steranes appear to be more thermally protected than kerogen bound

steranes. This is different to the observations of the C30 hopane parameter

which suggested that the existence of differening rates of isomerisation in the bound

and free fractions is an important process effecting the value of maturity parameters It

could even be considered to contradict the position of steranes as pendant units

bonded to the exterior of macromolecular networks (Kohnen et al., 1991), where they

might be expected to receive less steric protection than the precursors of C30 hopanes

that would expected to be cross linked into the centre of macromolecules. Another

explanation is needed, and is best supplied by the preferential release of the (20S)

isomer during petroleum (and biomarker) generation. Preferential release of the

Sterane (20S) isomer over the (20R) isomer has been has been reported in a

number of studies (Bishop and Abbot, 1993; Farrimond et al., 1998). The resin and

asphaltene bound steranes are most altered by this mechanism because in organic

matter of oil window thermal maturity resins and asphaltenes predominantly represent

moieties that have been cracked from kerogens at high temperatures, and thus

represent the recalcitrant products of the back reactions that have created free

hydrocarbons (Pelet et al., 1986), vis-à-vis if the free hydrocarbon fraction is enriched

in the Sterane (20S) isomer the bound fractions will be further enriched in the

Sterane (20R) isomer. Thus the C29 sterane (20S)/(20S+20R) parameter is

controlled by the preferential release of the (20S) isomer over the (20R) isomer, at

least at this stage in the oil window and for the type of organic matter in the Jet Rock.

89

3.4.3 Factors controlling the bound biomarker composition of the resin and

asphaltene fractions at the onset of catagenesis

Preferential release ofbiomarkers bound byweaker (-S-S-, ester, C-Setc.) or fewer bonds e.g.steranes.

Outside of kerogen particleis most effected duringearly stages of petroleumgeneration, e.g., moreinteraction with claymineral catalyst.

When resins andasphaltenes are generatedthey are enriched inrecalcitrant biomarkersbound by strong ormultiple linkages e.g.extended tricyclic terpanesand hopanes.

Figure 3.8. Showing the generation of the resin and asphaltene fractions enriched in the most tightly bound biomarkers such as C31 to C35 hopanes and extended tricyclic terpanes.

At the onset of catagenesis, the resin and asphaltene fractions contain only minor

quantities of biomarkers that are compositionally distinct in their proportions of

hopane and tricyclic terpanes. One possible interpretation of this is that at this stage

in the oil window resin and asphaltene molecules are unique end members in

biomarker generation, enriched in tricyclics and particularly extended tricyclic

terpanes and C31 hopane homologues which are behaving in a recalcitrant way. The

resin and asphaltene fractions contain relatively higher proportions of these particular

biomarkers because they are covalently bound by strong linkages (C-C or C-O) or

more than one bond – e.g. C31+ hopanes and extended tricyclic terpanes. This could be

interpreted as the resin and asphaltene macromolecules being formed from the

cleavage of weaker covalent cross-linkages, with the result that their structures are

likely to be thermodynamically very stable but impoverished in biomarkers that are

weakly bound into macromolecules.

These biomarker pools are not simply lower-molecular-weight versions of kerogen

though, and their biomarker composition is so radically different that it would

probably not be possible to directly correlate resin and asphaltene fractions to kerogen

fractions by a straight comparison of their biomarker content. This observation is at

90

odds with the traditional view that resin and asphaltene molecules are simply smaller

sub-units cracked from their parent kerogen (Pelet et al., 1986), but supports previous

work that noted compositional differences between the biomarkers present in the

different fractions of sedimentary organic matter (Richnow et al., 1991).

3.4.4 Sterane composition

The difference in sterane composition between the fractions presents an enigma; does

the difference in sterane composition represent sterane cracking during pyrolysis or a

true compositional difference between the fractions? The proportion of C27 steranes

increases in oils with maturity (Peters and Moldowan, 1993), and this could be

envisaged to arise from the cracking of the alkyl side chain of higher number steranes

(C28-C30). Similarly, laboratory pyrolysis techniques might induce some side chain

cracking, and thus, the hydropyrolysis products of this study and the hydrous-

pyrolysis products of Fowler and Brooks (1987) might have become enriched in C27

sterane by this route. Alternatively, the kerogen phase has previously been reported as

being enriched in algal derived lipids and biomarkers, whilst the extractable phase has

been noted to be enriched in higher plant lipids and biomarkers (Love et al., 1998).

The quantities of kerogen-bound steranes are significant enough that their release

during oil generation would enrich the free aliphatic hydrocarbon fraction in C27

steranes, consistent with the known increase in the proportion of C27 steranes with

thermal maturity mentioned previously (Peters and Moldowan, 1993). Thus, it is

possible that the kerogen-bound biomarker signal is naturally enriched in C27 steranes

from the early incorporation of biomarkers into the kerogen phase during kerogen

formation, and that it simply overwrites or alters the free fraction biomarker signal

during catagenesis as biomarker hydrocarbons are released into the free phase. The

potential change in sterane composition caused by the release of kerogen bound

steroids would be easy enough to calculate using a simple model, but the relative rates

of cracking are harder to determine being dependant on so many factors e.g. heating

rate, lithology, formation water composition and the catalytic properties of minerals.

It also seems reasonable to propose that both processes might occur.

91

3.5 Aromatic hydrocarbons: Results and discussion

Aromatic hydrocarbons are a significant component of the DCM extracts and the

hydropyrolysates (Table 3.3), but due to time constraints, they were not examined in

as much detail as the aliphatic biomarkers. GC-FID data was collected for the

aromatic hydrocarbons of all the fractions of samples 1 and 11 (these two samples

were chosen because they showed the most stratigraphic difference), as well as for 7

samples of the DCM extracts, and for 8 samples of the kerogen hydropyrolysates. In

addition to the GC-FID data, full scan GC-MS data was obtained for a few samples to

aid peak identification. Other authors have reported that isorenieratane and its aryl

and diaryl derivatives (Koopmans et al., 1996a; Saelen et al., 2000) are largely absent

from both the free and bound fractions of the Jet Rock. This seems surprising given

the depositional conditions and high organic matter content of the Jet Rock, and so a

single ion monitoring program was devised for GC-MS to specifically target tetra-

alkylated aryl isoprenoids and C40 diaryl isoprenoids to check for the presence in Jet

Rock aromatic fractions.

Before presenting and discussing the results of the analysis of the aromatic fractions

of the solvent extracts and hydropyrolysates it is important to highlight that there is

considerable stratigraphic variation in the molecular composition of the aromatic

fractions of the different samples. This is described more fully in Chapter 4 which

deals with the molecular stratigraphy of the Jet Rock. The aromatic GC-FID traces in

Figures 3.9 and 3.10 represent two stratigraphic end-members; the majority of the

other samples having a distribution between these two end members but more similar

to that of sample 11.

92

free

resin

asphaltene

kerogen

Py

MPy

C1C

& M

B[c

]P

P

MP

DM

P B[g

hi]P

e

CoB

[?]F

aB

[?]P

y

DM

BP

& E

BP

MB

P

P MP

DM

PPy M

Py

C1C

& M

B[c

]P

P DM

P

B[?

]Fa

B[?

]Py

DM

BP

& E

BP

MP

C &

B[c

]P

B[g

hi]P

e

Triaromaticsteroids

inc MTAS

MN

’s

END

MN

’s

TMN

’s

FaFa

elution time

Figure 3.9. Aromatic GC-FID traces of sample 1, the most terrestrially influenced sample. MN = methyl napthalenes; EN = ethylnapthalenes; DMN = dimethylnapthalenes; TMN = trimethylnapthalenes; EBP = ethylbiphenyls; P = phenanthrene; MP = methylphenanthrenes; DMP = dimethylphenanthrene; Py = pyrene; Fa = fluoranthene; C = chrysene; B[c]P = benzo[c]phenanthrene; C1C = methylchrysene; MB[c]P = methyl-benzo[c] pyrene; B[?]Fa = benzo-fluoranthenes; B[?]Py = benzo-pyrenes; B[ghi]Pe = benzo[g,h,i] perylene; Co = coronene; TAS = triaromatic steroids; MTAS = methyl-triaromatic steroids.

3.5.1 Aromatic composition

The GC-FID traces in Figures 3.9 and 3.10 compare the distribution of aromatic

compounds between the four fractions. The two figures correspond to samples with

the most stratigraphic difference. The free hydrocarbon aromatic fractions of both

samples comprise a range of methyl- and ethyl- napthalenes, phenanthrenes and

polymethylated phenanthrenes and a slight aromatic UCM. Triaromatic steroids (and

monoaromatic steroids) are present and elute latter, methylated triaromatic steroids

might also be present as a series of compounds with a m/z 245 ion that elute amongst

the triaromatic steroids (m/z 231). Triaromatic steroids are not visible on the GC-FID

93

traces of any of the other fractions which suggests, that much like the partitioning of

aliphatic biomarkers (Figure 3.1), the DCM-extractable free fraction is particularly

enriched in steroid biomarkers, and that triaromatic steroids are present at a much

lower relative abundance in the hydropyrolysates of the resin, asphaltene and kerogen

fractions. Much like the GC-FID traces of the free aromatic hydrocarbon fraction, the

resin fraction traces of both samples are similar to each other. The polymethylated and

ethylated napthalenes and phenanthrenes, as well as the aromatic UCM, that are

characteristic features of the free fraction, are present in the GC-FID traces of the

resin hydropyrolysates. An additional feature of the GC-FID traces in Figures 3.9 and

3.10 is the presence of biphenyls as well as PAHs with greater than four rings (e.g.

pyrene and other later eluting PAHs).

The GC-FID traces of the kerogen and asphaltene fractions differ from the resin and

free fractions in showing a significant degree of stratigraphic variation. In the kerogen

fraction of sample 1 (Figure 3.9) the dominant resolvable compounds are the

biphenyls and large ring PAHs that elute after phenanthrene, but in sample 11 (Figure

3.10) the dominant resolvable compounds are earlier eluting napthalenes,

phenanthrenes and biphenyls; there is also a considerable aromatic UCM that swamps

the latter eluting PAH compounds (see Chapter 4 for details). The biphenyls are

present in the hydropyrolysis products of the resin, asphaltene and kerogen fractions

of each sample, and are mostly likely to have been derived from the

defunctionalisation of dibenzothiophenes during hydropyrolysis. Their limited

occurrence in only the resin, asphaltene and kerogen hydropyrolysates and not the

free hydrocarbon fractions further supports this origin. The distribution of aromatic

compounds in the asphaltene fraction is similar to that seen in the kerogen fraction

and displays a similar stratigraphic variation, with the exception that methylated PAH

species are far more abundant relative to their unsubstituted counterparts.

94

free

resin

asphaltene

kerogen

P

MP

DM

P Triaromaticsteroids

inc MTAS

Py

MPy

C1C

& M

B[c

]P

P

MP DM

P

B[g

hi]P

e

B[?

]Fa

B[?

]Py

DM

BP

& E

BP

MB

P

P

MN

’s

EN

DM

N’s

TM

N’s

MN

’s EN D

MN

’sT

MN

’s

MN

’s

EN D

MN

’s

TM

N’s

DM

BP

& E

BP

MB

P

elution time

Figure 3.10. Aromatic GC-FID traces of sample 11, the least terrestrially influenced sample. MN, methylnapthalenes; EN = ethylnapthalenes; DMN = dimethylnapthalenes; TMN = trimethyl-napthalenes; MBP = methylbiphenyls; DMP = dimethylbiphenyls; EBP = ethylbiphenyls; P = phenanthrene; MP = methylphenanthrenes; DMP = dimethylphenanthrene; Py = pyrene; Fa = fluoranthene; C = chrysene; B[c]P = benzo[c]phenanthrene; C1C = methylchrysene; MB[c]P = methyl-benzo[c] pyrene; B[?]Fa = benzofluoranthenes; B[?]Py = benzopyrenes; B[ghi]Pe = benzo[g,h,i] perylene; Co = coronene; TAS = triaromatic steroids; MTAS = methyl-triaromatic steroids.

3.5.2 Origin of free and bound PAH in the Jet Rock

Alkylated napthalene compounds can have a range of inputs and possible sources

including bacterial and higher plant precursors (Puttmann and Villar, 1987; Killops,

1991; Borrego et al., 1997). One study has described these types of compounds, and

1,3,6,7 tetramethylnapthalene (TeMN) in particular, as representing a constant non-

source specific microbial background input and used them to normalise data and

produce a stratigraphic signal (Jiang et al., 1998). Phenanthrene and methyl

phenanthrenes can also be derived from combustion (Killops and Massoud, 1992) and

diagenetic alteration (Puttmann and Villar, 1987; Borrego et al., 1997) of various

precursors. Both methylated napthalenes and phenanthrenes are ubiquitous in all

fractions and samples. However, it is important to note that that they are most

prominent in the free and resin fractions of all samples, and in the kerogen (and

95

asphaltene) fractions of the samples in the lowest part of the section. Another

important component of the GC-FID traces of aromatic fractions is the aromatic

unresolved compound mixture (UCM) visible in the free DCM extracts. This is well

developed in most of the GC-FID traces of the aromatic fractions of the kerogen

hydropyrolysates of the samples from the lowest part of the section. Aromatic UCM’s

normally comprise alicyclic aromatic compounds with complex alkylation patterns

(Simoneit and Fetzer, 1996), and are often found in the solvent extracts and natural

hydrothermal pyrolysis products of marine (type I) sedimentary organic matter. In the

following chapter these observations are interpreted in the context of changes in the

type of sedimentary organic matter reaching the centre of deposition brought about by

changes in relative sea level.

Large PAH compounds comprising 4 or more benzene rings are ubiquitous in all of

the resin hydropyrolysates, but are only prominent in the kerogen hydropyrolysis

products of the samples from the top of the section. The distribution of kerogen bound

PAHs can easily be explained as a greater input of unsubstituted parent PAHs derived

from the combustive oxidation of terrestrial organic matter (Killops and Massoud,

1992) to the samples at the top of the section, and greater inputs of the polymethylated

napthalenes and phenanthrenes, and UCM components (presumably derived from a

marine source) to the kerogens from the rest of the section.

3.5.3 Summary and interpretation of PAH

In the case of the aliphatic biomarkers (described in section 3.4) it is relatively

straight forward to link the distribution of compounds to binding mechanisms and

other chemical factors; this is a lot more difficult for the aromatic compounds due to

the stratigraphic variation observed between the samples. Figure 3.9 suggests a trend

of decreasing methyl-substitution and increasing aromatization and condensation of

macromolecular structures moving from the free through the resin, asphaltene and

kerogen fraction. For example, there is a strong prominence of benzo[g,h,i]perylene

and pyrene over methyl substituted PAHs in the kerogen fraction, an increased

prominence of methylpyrenes and methylphenanthrenes compared to non-substituted

pyrene and phenanthrene in the asphaltene fraction and an almost equal prominence

of methylphenanthrenes, pyrene, methylpyrenes and benzo[g,h,i]perylene in the resin

fraction. This is in accord with the observations of numerous authors concerning a

96

trend of increased aromaticity, condensation and decreased aliphatic and methyl

substitution with molecular weight and size in the non-hydrocarbon macromolecular

fractions (Koots and Speight, 1975; Bandurski, 1982, Pelet et al., 1986). However, it

is sample 11 (Figure 3.10) that is most typical of the rest of the Jet Rock samples that

were collected for this study. Sample 11’s GC-FID traces in Figure 3.10 do not show,

to the same extent, the trends that were observed in the distribution of PAHs between

the resin, asphaltene and kerogen fractions of sample 1 (Figure 3.9). This is for two

reasons. First, any trends are swamped by the very prominent aromatic UCM,

although GC-MS data shows that the non-methyl substituted PAH are less prominent

relative to their methylated counterparts in these samples (Figure 4.13, Chapter 4).

Second, sample 1 has received a head start in terms of thermal maturity in the form of

a depositional input of oxidised PAHs that were predominantly unsubstituted and

derived from oxidation or combustion of organic matter (Killops and Massoud, 1992).

However, some of the trends described for sample 1 can be observed in the GC-FID

traces in Figure 3.10, such as the proportion of methylphenanthrenes to phenanthrene.

Yield data would also generally support the resin (and asphaltene) fraction being

considerably less aromatic in character than the kerogen fraction (Table 3.3). It would

be expected that if the thermal evolution of the sedimentary organic matter in the Jet

Rock had continued, that the distribution of the PAH in the GC-FID traces of the

resin, asphaltene and kerogen fractions of sample 11 would have more closely come

to have resembled those of sample 1. This essentially reflects the different patterns of

thermal evolution that are observed for the H/C ratios of terrestrial and marine organic

matter on a Van Krevelen diagram, e.g. lesser changes in type III than in type II or I

organic matter as the proportion of hydrogen decreases relative to carbon (Tissot et

al., 1974).

3.5.4 Isorenieratane and derivatives

Only the extract of a single sample of the Jet Rock has previously been reported to

contain C40 isorenieratane or one of it’s derivatives (Saelen et al., 2000 citing personal

communication with J. Sinninghe Damste), and aryl-isoprenoids have been reported

to be absent in both extracts and chemical degradation and pyrolysis products (Saelen

et al., 2000). This is remarkable given the anoxic environment, high TOC sediment

and probable highly euxinic depositional conditions of the Jet Rock (Raiswell and

Berner, 1985), and its early oil window thermal maturity (Tmax ~ 430 oC) which

97

would be expected to be conducive to the preservation of diaryl and aryl isoprenoids

(Kohnen et al., 1989; Requejo et al., 1992).

Time-->

28 29 30 31 32 33 34 35 36 37 38 39

C13

C14 C15 C16

1

3

9

11

13

6

5

7

Time-->

29 30 31 32 33 34 35 36 37

C13 C14

C15C16

1

3

9

11

13

free fraction kerogen fraction

Figure 3.11, m/z 133 traces of the free extracts and kerogen hydropyrolysates. C13 – C16 homologues of 1-isoalkyl 2, 3, 6-trimethylbenzene are labeled and their spectra are show in Figure 3.12.

This study, in contradiction to previous work, has found aryl isoprenoids in the free

and kerogen fractions of every sample of the Jet Rock examined by GC-MS, and

evidence for the presence of diaryl isoprenoids ( likely to be isorenieratane) in most of

the free aromatic hydrocarbon fractions. The C13 to C16 homologues of 1-isoalkyl-

2,3,6-trimethylbenzene (Figure 3.11) were identified in the free and kerogen fractions

by comparison to relative retention times presented in other published work

(Summons and Powel, 1987; Requejo et al., 1992) and their mass spectra (Figure

3.12). The aryl isoprenoids are most prominent in the free fraction, where a clear

series of doublets ranging upto C23 can be seen (the latter eluting compounds are

thought to be the 1-isoalkyl-2,3,4-trimethylbenzene isomer). In the kerogen fraction

many other compounds with a 133 m/z fragment can be seen and these probably

correspond to many of the other alkylated benzene compounds, not necessarily

derived from green sulphur bacteria, that are often found in kerogen and asphaltene

pyrolysates (Hartgers et al., 1992).

98

The spectra in Figure 3.12 are from the GC-MS full-scan dataset initially acquired for

identifying the PAH GC-FID peaks, and were measured on aromatic fractions as

opposed to the monoaromatic fractions used in many other studies. No quantification

was carried out but the C14 homologue was found to have a peak height equal to that

of trimethylnaphthalenes that eluted adjacent to it in gas chromatography. The relative

clarity of the spectra also attests to the quantitative abundance of the aryl isoprenoids

in the Jet Rock.

Figure 3.12, Mass spectra for C13 to C16 1 alkyl 2, 3, 6, isoprenoid homologues. Important features are the 133 m/z base fragment and the 176, 190, 204 and 218 molecular ions that correspond to the C 13, C14, C15 and C16 homologues.

The series extends up to C23 and beyond, although the earliest eluting peaks produce

the clearest spectra. The kerogen traces (Figure 3.13) are the noisiest at the top of the

section. This could be due to lesser relative concentrations of aryl isoprenoids in these

samples and greater concentrations of other similar compounds possessing m/z 133

fragments. The overall envelope of the m/z 133 chromatograms in the kerogen and

free fraction traces is very different, the kerogen trace is the noisiest (Figure 3.13) – a

99

feature normally associated with the m/z 133 traces of mature source rock extracts.

This suggests that the release of kerogen-bound compounds other than aryl

isoprenoids during petroleum generation might initially be responsible for the

observed relative decrease in concentration of (1-isoalkyl-2,3,6-trimethylbenzene)

aryl isoprenoids in extracts with rising maturity (Requejo et al., 1992).

Figure 3.13. Full m/z 133 ion chromatograms for the free and kerogen fractions. 1-isoalkyl-2,3,6-trimethylbenzenes are labeled on the free and kerogen traces. The insert shows C32 and C33 diaryl isoprenoids that are derived from isorenieratene.

No spectra with a strong base peak at m/z 133, and clear molecular ion at m/z 547 that

are diagnostic of isorenieratane could be obtained. However, diagnostic spectra for

C32 and C33 diaryl isoprenoids that are derived from isorenieratene via the formation of

100

an eight membered ring transition state during diagenesis (Koopmans et al., 1996a)

were obtained and these are shown in Figure 3.13.

These compounds are derived from isorenieratene, which is an undisputed biomarker

for chlorobiaceae (a green sulphur bacteria) and can be a potential source of aryl

isoprenoids (Summons and Powell, 1986). However, aryl isoprenoids can also be

derived from other sources such as a -carotene which is derived from algae

(Koopmans et al., 1996b). If isorenieratene was present, than the co-occurrence of

compounds that are derived from isorenieratene with the aryl isoprenoids strongly

suggests that the latter compounds were derived from isorenieratene, but to what

extent can not be determined without isotopic data to show that the compounds are

enriched in 13C and were formed via a reverse TCA cycle. Their presence in all the

sections of the Jet Rock sequence studied suggests that euxinic conditions were a

common feature of the Toarcian water column in the area of deposition of the Jet

Rock. A previous study that failed to identify aryl and diaryl isoprenoids in the Jet

Rock (Saelen et al., 2000) interpreted their absence to be the result of either a

chemocline that was permanently below the photic zone (which would contradict the

TOC and total Sulphur content evidence of Raiswell and Berner (1985), or a stable

water column in which frequent kill-off events did not occur and lead to the mass

deposition, and preservation of large quantities of aryl and diaryl isoprenoids (Tyson

et al., 1979). The model that was favoured by Saelen et al. (2000) was the latter, and

this was taken to be evidence that a highly stratified water column did not persist for

extended periods of time and that primary inputs of aryl isoprenoids were low, or that

overturning of the water column was infrequent and that rates of degradation were

always in balance to rates of deposition.

The kerogen traces in Figure 3.11 show that the proportion of 1-isoalky-l,2,3,6,-

trimethylbenzenes (assumed to be predominantly derived from green sulphur bacteria)

to other alkylbenzenes (assumed to be derived from other sources) is greatest in the

samples from the base of the section deposited when sea level was highest, when the

water column was most likely to be stratified and seasonal euxinic conditions to

persist. Samples from the rest of the section show generally poorer preservation of

aryl- and diaryl- isoprenoids relative to other alkylbenzenes, this could be interpreted

as a limited amount of variability in the position of the chemocline in the water

101

column, and that overall only the samples from the base of the section were deposited

in kill-off events caused by a cyclically varying chemocline. Thus the Jet Rock was

likely to have been persistently rather than cyclically euxinic and that for great

periods of time the rates of diaryl- and aryl- isoprenoid deposition and preservation

via sulphur binding into macromolecules (Kohnen et al., 1989), whilst exceeding the

rates of degradation and destruction, appear to have been constant and resulted in only

the poor preservation of 1-isoalkyl-2,3,6-trimethylbenzenes.

3.6 Conclusions

Based on hydropyrolysis yield data, the bulk of the sedimentary organic matter in the

Jet Rock comprises hydrocarbons that are covalently bound into the kerogen fraction.

Much greater quantities of sterane, hopane and tricyclic terpane biomarkers are bound

into the kerogen fraction than are bound into the resin and asphaltene fractions.

Although both the resin and asphaltene fractions constitute a only a small part of the

sedimentary organic matter at the near oil window thermal maturity of the Jet Rock.

Not only are the quantities of biomarkers in the resin and asphaltene fractions

considerably smaller than those in the kerogen fraction, they were found to be

compositionally very distinct. This was most apparent on ternary plots of sterane

composition but could also be seen by visual inspection of the m/z 191 ion

chromatogram which showed that the resin and asphaltene fractions contain a greater

proportion of extended tricyclic terpanes than the free and kerogen fractions.

The free aliphatic hydrocarbon fraction is enriched in biomarkers at the onset of oil

generation (the thermal maturity of samples in this study), and steranes appear to be

particularly enriched in the solvent extracted free aliphatic fraction. This could be due

to a number of reasons including the binding of the steranes into the macromolecule

by only a single bond that is cleaved in early catagenesis, and the early formation of

diasteranes in the free fraction. The early formation of diasteranes creates non-

reactive sterane species for which binding into macromolecules is an unfavorable

reaction during early diagenesis and kerogen formation. This effectively sequesters

diasteranes into the free fraction.

102

The distribution of aromatic compounds varies both stratigraphically and between

fractions. The most abundant resolvable compounds on the GC-FID traces of the free

hydrocarbon fraction and kerogen hydropyrolysates are the polymethylated

napthalenes and phenanthrenes, although there is also a significant aromatic UCM

present which swamps the later eluting large ring PAH, that are only prominent in the

kerogen hydropyrolysates of few samples from the top of the section

In terms of the differences in the distribution of PAH in the four fractions, this creates

a split between the samples from the top of the section that contain the greatest

amounts of terrestrial organic matter, and samples from the base which contain more

marine organic matter. In the samples from the top of the section a clear progression

though the kerogen, asphaltene and resin fractions can be observed in which

methylated PAH become more prominent in the smaller lower molecular weight non-

hydrocarbon fractions. In the majority of the samples from the lower parts of the

section this trend is less clear, suggesting that the type of sedimentary organic matter

greatly influences the evolution of its molecular composition with increasing thermal

maturity, in a similar manner to that which is observed on a Van Krevelen diagram for

H/C ratios, e.g. the H/C ratios of type I kerogen change more than the H/C ratios of

type III kerogen with thermal maturity.

Aryl isoprenoids are present in the Jet Rock in addition to compounds that form from

isorenieratene early in diagenesis. Further evidence to support that the aryl

isoprenoids were derived from green sulphur bacteria via the reverse TCA cycle could

only be supplied by obtaining compound specific isotope data that demonstrated that

the compounds were isotopically heavy. To do this it might first be necessary to

remove as much of the aromatic UCM as possible by preparing a mono-aromatic

fraction, this could also aid the identification of any diaryl isoprenoids (including

isorenieratane) that are present in the Jet Rock.

103