Molecular analysis of sulphur-rich brown coals by flash pyrolysis-gas chromatography-mass spectrometry: The type III-S kerogen

Show preview

Hide preview

Journal of Chromatography, 607 (1992) 361-376 Elsevier Science Publishers B.V., Amsterdam
CHROM. 24 401
Molecular analysis of sulphur-rich brown coals by flash pyrolysis-gas chromatography-mass spectrometry
The Type III-S kerogen*
Jaap S. Sinninghe Dam& *, F. Xavier C. de las Heras** and Jan W. de Leeuw Organic Geochemistry Unit, De@ University of Technology, de Vries van Heystplantsoen 2, 2628 RZ De@ (Netherlands)
ABSTRACT
The molecular composition of five brown coals from three different basins (Maestrazgo, Mequinenza and Rubielos) in Spain was investigated by flash pyrolysis-gas chromatography and flash pyrolysis-gas chromatography-mass spectrometry. In these techniques, the macromolecular material is thermally degraded in an inert atmosphere and the compounds formed are on-line separated, identified and quantified. This information provided insight into the macromolecular structure of the coals which was inaccessible by other means. The composition of the pyrolysates is described in detail with emphasis on the distributions and relative abundance of n-alkanes, n-1-alkanes, (alkyl)phenols, sulphur compounds [(alkyl)thiophenes and (alkyl)benzothiophenes], (alkyl)benzenes and (alkyl)naph- thalenes. These compound classes represent the major pyrolysis products of the samples analysed and were used to assess the contribu- tions of specific biomacromolecules mainly originating from higher plants. One of the five brown coal samples investigated is so rich in organic sulphur (one sulphur atom for every 9-15 carbon atoms as determined by elemental analysis) that a new kerogen type (Type III-S) describing the kerogen contained in this coal is defined. Type III-S kerogen is defined as a kerogen with high atomic S,,,/C (> 0.04) and O/C (> 0.20) ratios. Two of the five brown coals samples investigated contain a series of long-chain alkylbenzenes with an unprecedented carbon number distribution pattern with a second maximum at C,,. This unusual distribution pattern is thought to originate from the presence of long-chain alkylbenzene moieties bound via a heteroatom (presumably an ether bond) to the macro- molecular coal matrix preferentially at position 12 in the alkyl side-chain of these moieties.
INTRODUCTION
Coal is predominantly a macromolecular organic substance mainly derived from specific tissues of higher plants such as woody tissue, cuticles, spores, pollen, seeds and corkified cell walls which have un- dergone chemical alterations by the coalification process [l]. To some extent remains of these tissues can still be recognized by light microscopic investi- gation of coal. This type of recognition has led to the maceral concept: macerals are defined as micro-
* Delft Organic Geochemistry Unit Contribution 277. ** Present address: Escola Universitaria Politecnica de Manre-
sa, UPC, Av. Bases de Manresa 61-73, 08240 Manresa, Spain.
Apart from this microscopic approach coal has also been analysed chemically for more than a cen- tury [2]. Elemental analysis and other bulk chemical analyses have been and still are important analyt- ical techniques for characterizing coals. There is, however, increasing interest in the characterization of the structure of coal at the molecular level. Be- cause of the macromolecular, insoluble nature of coal such a molecular characterization is more diffi- cult than that of the other major fossil fuels, pet- roleum and natural gas. Spectroscopic techniques such as Fourier transform infrared and solid-state 13C NMR spectroscopy have been used but do not
0021-9673/92/$05.00 0 1992 Elsevier Science Publishers B.V. All rights reserved
scopically recognizable entities in the coal matrix and shed light on the composition of coal and its original precursors.

362 J. S. SINNINGHE DAMSTB et al.
provide information on the arrangement of atoms. Specific chemical degradation reactions have also been applied to coals to identify the structures of released moieties (e.g., [3,4]). However, the yields are relatively low or the information is limited ow- ing to less specific chemical reagents. Analytical py- rolysis (controlled thermal degradation in an inert atmosphere) in combination with gas chromatogra- phy, mass spectrometry and gas chromatography- mass spectrometry can supply detailed molecular information on coal [5-S] and isolated maceral frac- tions [9-l 11. These pyrolysis approaches in combi- nation with spectroscopy have led to the recogni- tion of resistant, selectively enriched biomacromol- ecules in coal derived from plant cuticles (cutan [ 12- 15]), corkified cell walls (suberan [16]), spores and pollen (sporopollenin [ 17,1 S]), seed coats [193, woo- dy tissues [20] and resins [21].
Organic sulphur in coal is not derived from the biomacromolecules present in plant tissues. It is formed by syndepositional incorporation of re- duced inorganic sulphur species formed by micro- bial reduction of sulphate into the organic matrix (for reviews see [22,23]) which leads to the forma- tion of organically bound sulphur. It is therefore that marine-influenced depositional environments generate coals which generally have a higher sul- phur content. Since the presence of organic sulphur is of major environmental concern in the utilization of this fossil fuel resource, a better understanding of the formation, forms and distribution of organic sulphur in coal is required [22-241.
In this paper, the results of analysis by flash py- rolysis-gas chromatography (Py-GC) and flash py- rolysis-gas chromatography-mass spectrometry (Py-GC-MS) of five sulphur-rich brown coals from Spain are reported. The results indicate that under specific conditions the organic sulphur content can become so high (one sulphur atom for every 9-14 carbon atoms) that it is appropriate to define a new type of kerogen, Type III-S.
EXPERIMENTAL
Coal samples Five brown coal samples were selected from three
different sedimentary basins in Spain. Three sam- ples were taken from the Maestrazgo basin, which is located in the Iberian mountain chain and the
southern sector of the Catalan coastal range in NE Spain. Two of these samples (Estercuel and Portal- rubio) are from the Utrillas formation and were both deposited in proximal areas of a delta estuary during the middle Albian (upper Lower Cretaceous, ca. 105 Ma). These coals have high sulphur contents (see Table I). This is possibly due to an influx of sulphate resulting from weathering of gypsum of the evaporitic Keuper formation in the catchment area into the delta estuary [25]. This resulted in sig- nificant sulphate reduction which, in turn, led to reaction of organic matter with reduced forms of inorganic sulphur. It is noteworthy that these con- ditions have led to a higher sulphur content of the coals deposited in proximal areas than in coals from the same basin derived from marine influenced dep- ositional environments [25]. The Paula lignite was collected from sediments of the same basin and is thought to be of Tertiary age.
One sample was taken from the Mequinenza sub- basin which is located in the SE margin of the large Catalan Ebre (Ebro) basin. The Mequinenza basin is mainly filled with carbonate sediments deposited in an extensive, shallow, open palaeolake. The coal- bearing carbonate sequences were deposited in the open lacustrine zones closer to marginal evaporitic and marsh environments in the Oligocene (ca. 35 Ma) [26,27].
The Rubielos coal is from a basin located in the SE part of the Iberian mountain chain (NE Spain), which belongs to a Miocene (ca. 14 Ma) lacustrine system where lignites were deposited in the middle unit and are interbedded with lacustrine limestones which contain significant amounts of immature or- ganic matter [28,29].
Sample treatment Two coal samples (Mequinenza and Rubielos)
were Soxhlet extracted with dichloromethane- methanol (2:1, v/v) for 36 h. The other three were analysed as such.
Elemental analysis Elemental analysis (C, H, N, S,,,.) were perfoim-
ed on Carlo Erba Model 1106 and 1500 elemental analysers. Duplicate analyses indicated good repro- ducibility. Ash contents were determined gravimet- rically by heating the sample at 900°C for 2 h. Pyrite
and Sorg. were determined according to ASTM methods.

Py-GC-MS OF SULPHUR-RICH BROWN COALS
Curie-point pyrolysis-gas chromatography The brown coals were thermally degraded using a
non-commercial Curie-point pyrolyser and ferro- magnetic wires with a Curie temperature of 610°C. The brown coals were applied to the wire by press- ing the samples on the wire [30]. The pyrolyser was mounted on the injection port of a Varian Model 3700 gas chromatograph. On-line separation of the flash pyrolysate was accomplished by using a fused- silica capillary column (25 m x 0.32 mm I.D.) coat- ed with CP Sil-5 CB (film thickness 0.40 pm) (Chrompack, Middelburg, Netherlands). The oven of the gas chromatograph was temperature pro- grammed from 0°C to 300°C at 3°C min- ’ using a cryogenic unit. The oven was first held at 0°C for 5 min and finally at 300°C for 15 min. Helium was used as the carrier gas. Pyrolysis products were de- tected by simultaneous flame ionization detection (FID) and sulphur-selective flame photometric de- tection (FPD) using a stream splitter (SGE) at the end of the capillary column.
Curie-point p_vrolysis-gas chromatography-mass spectrometry
The coal samples were thermally degraded using a Curie-point pyrolyser (FOM-3LX [31]) and ferro- magnetic wires with a Curie temperature of 610°C. The pyrolyser was connected directly to a gas chro- matograph (Hewlett-Packard Model 5890) in tan- dem with a magnetic sector mass spectrometer (VG-70s) by direct insertion of the capillary col- umn into the ion source. The gas chromatograph was fitted with a fused-silica capillary column (25 m x 0.32 mm I.D.) coated with CP Sil-5 CB (film thickness 0.40 pm) in an oven that was temperature programmed from 0°C to 300°C at 3°C min- ‘. The oven was first held at 0°C for 5 min and finally at 300°C for 15 min. Helium was used as the carrier gas. The mass spectrometer was set at an ionizing voltage of 70 eV and operated at a cycle time of 1.8 s over the mass range m/z 40-800 at a resolution of 1000. Date acquisition was started 1 min after py- rolysis.
RESULTS AND DISCUSSION
The five selected coal samples were thermally de- graded using ferromagnetic wires with a Curie tem- perature of 610°C. The pyrolysates of the coals were
363
analysed on-line by GC-MS. Compounds were identified by comparison of mass spectral and rela- tive retention time data with literature data [32-361. The total ion currents (TICS; Figs. lA-5A) reveal the general composition of the pyrolysates. Al- though (alkyl)phenols, (alkyl)benzenes, (alkyl) naphthalenes, 1 -pristene and n-alkanes and n- 1 -al- kenes are major components in all pyrolysates, sig- nificant differences between the coals are observed. Large variations in the relative amounts of sulphur compounds (mainly alkylated thiophenes and ben- zo[b]thiophenes) are also noted. Hopanes and a se- ries of higher-molecular-mass alkylbenzenes are on- ly present in significant amounts in the Estercuel
Fig. 1. (A) Total ion current (TIC) trace and (B) summed mass chromatogram of m/z 55 + 57 of the flash pyrolysate (Curie temperature 61o’C) of the Mequinenza coal. Key for the TIC: B = (alkyI)benzenes; T = (alkyl)thiophenes; P = (alkyl)phenols; N = (alkyl)naphthalenes; Bt = (alkyl)benzothiophenes; D = 1,2-dihydroxybenzene; Pr = 1-pristene; To = tocopherols; H = hopanes. The number of carbon atoms of several members of the homologous series of n-1-alkenes and n-alkanes (series of dou- blets) in the m/z 55 + 57 mass chromatogram are indicated.

Py-CC-MS OF SULPHUR-RICH BROWN COALS 365
Fig. 4. (A) Total ion current (TIC) trace and (B) summed mass chromatogram of m/z 55 + 57 of the flash pyrolysate (Curie temperature 610°C) of the Paula coal. For the key to the symbols and numbers, see Fig. 1.
highly resistant biomacromolecule is selectively pre- served during diagenesis and is, in case of coal for- mation, concentrated in the coal maceral cutinite [9911]. In extant cuticles and in immature sediments these series of n-alkanes and n- 1 -alkenes are accom- panied by a series of mainly long-chain methyl ke- tones [14]. The latter were found in the pyrolysate of the Paula coal but were below the detection limit in the other coal pyrolysates.
Other highly aliphatic biomacromolecular frac- tions which can be present in coal are suberan [16] and algaenan [4045]. Suberan is a highly aliphatic biomacromolecule present in the outer bark tissue and explains the liptinitic nature of the coal maceral suberinite [16]. Algaenan is also a highly aliphatic biomacromolecule and is a major constituent of the outer cell walls of certain types of freshwater algae [4045]. Alginite-rich coals probably contain these
A
I B
Fig. 5. (A) Total ion current (TIC) trace and (B) summed mass chromatogram of m/z 55 + 57 of the flash pyrolysate (Curie temperature 610°C) of the Rubielos coal. For the key to the symbols and numbers, see Fig. I.
highly resistant algaenans. Recently, two other highly aliphatic biomacromolecules have been de- scribed in fossil spore walls (massulae) of water ferns [40] and in the inner coat (tegmen) of seeds of water plants [46]. These biomacromolecules can al- so explain, at least in part, the series of n-alkanes and n-1-alkenes in coal pyrolysates. Since all these different types of highly aliphatic biomacromole- cules generate the same suite of compounds with distribution patterns depending on actual precursor organisms and stage of diagenesis, it is at present not possible to relate more specifically the series of n-alkanes and n-1-alkenes to one or more of the above-mentioned plant organs.
I-Pristene is a major component in all coal pyro- lysates indicating the low stage of thermal maturity of the coals [47]. In the three unextracted coal sam- ples it was possible to determine the pristane forma-

G ~
,
T A
B L
E I

E L
E M
E N
T A
L A
N A
L Y
S IS
, O
P T
IC A
L M
E A
S U
R E
M E
N T
A N
D P
Y R
O L
Y S
IS D
A T
A F
O R
T H
E B
R O
W N
C O
A L
S S
T U
D IE
D
C o
a l
E le
m e
n ta
l a n
a ly
si s"
P
y ro
ly si
s
C
H
N
S o
t, ..
O ~
H
/C
S or
g ./
C
O /C
A
sh
S ,o
t.
P y
ri te
R
o :
T R
c S
o r~
./ C
P
F I ~
R
o (%
):
(% )
(% )
(% )
(% )
(% )
(% )
(% )
(% )
(% )
(e st
im a
te d
) (e
st im
a te
d )
M e
q u
in e
n za
E
st e
rc u
e l
P o
rt a
lr u
b io
P
a u
la
R u
b ie
lo s
6 0
.6
5 .3
1
.0
1 0
.1
2 3
.0
1 .0
5
0 .0
6 2
7
3 .2
5
.8
0 .7
3
.2
1 7
.1
0 .9
5
0 .0
1 6
7
3 .5
4
.4
0 .8
1
.3
2 0
.0
0 .7
2
0 .0
0 7
5
1 .8
5
.3
1 .2
5
.2
3 6
.5 e
1 .2
3
0 .0
3 7
4
5 .6
3
.8
1 .5
3
.2
4 5
.9 ~
1 .0
0
0 .0
2 6

0 .2
8
1 3
.9
1 1
.5
1 .2
0
.1 8
1
9 .2
4
.4
2 .3
0
.2 0
1
7 .4
3
.0
1 .7
0
.5 3
e 4
0 .0
1
0 .3
5
.0
0 .7
5 e
1 5
.4
4 .2
0
.8
0 .3
1
1 .6
0
0 .1
1
n d
n
d
0 .3
9
0 .4
4
0 .0
3
0 .1
9
0 .4
2
0 .4
7
0 .1
3
0 .0
1
0 .3
2
0 .4
6
0 .3
5
0 .5
7
0 .0
4
0 .0
9
< 0
.0 4
0
.3 4
0
.2 4
0
.0 2
n
d
n d

" O
n a
d ry
a n
d a
sh -f
le e
b a
si s e
x ce
p t fo
r a
sh , t
o ta
l s u
lp h
u r a
n d
p y
ri te
c o
n te
n ts
, w h
ic h
a re
o n
a d
ry w
e ig
h t b
a si
s.
B y
d if
fe re
n ce
. c
T R
= T
h io
p h
e n
e ra ti
o =
( 2
,3 -d
im e
th y
lt h
io p
h e
n e
)/ [(
1 ,2
-d im
e th
y lb
e n
ze n
e )+
(n -l
-n o
n e
n e
)] .
a P
F I
= P
ri st
a n
e fo
rm a
ti o
n in d
e x
= (
p ri
st a
n e
)/ [(
p ri
st a
n e
)+ (p
ri st
-l -e
n e
)+ (p
ri st
-2 -e
n e
)] .
e T
h e
se v a
lu e
s a re
e x
tr e
m e
ly
h ig
h a
n d
p ro
b a
b ly
e rr
a ti
c. In
te rp
re ta
ti o
n
o f
th e
se v a
lu e
s s h
o u
ld b e
p e
rf o
rm e
d
w it
h g
re a
t c a
u ti
o n
. :
R o
= M
e a
n ra
n d
o m
v it
ri n
it e
re fl
e ct
a n
ce
(o il)
.
o, 1
Z z z :n
>
~ q
2

l’-GC-MS OF SULPHUR-RICH BROWN COALS 367
tion index (PFI), which is defined as [pristane]/ {[pristane] + [I-pristene] + [Zpristene]) [47] and was measured from peak-height data in the m/z 55 + 57 mass chromatograms (Table I). Comparison with literature data of the Mahakam coal sequence [47] led to an estimate of the thermal maturity level in terms of vitrinite reflectance values (Table I). In the Estercuel, Portalrubio and Paula coals the calculat- ed values agree surprisingly well with the measured values of vitrinite reflectance. These data confirmed the relatively low level of thermal maturity. Inter- estingly, the coals which yielded very similar pyroly- sates (Estercuel and Portalrubio; cf, Figs. 2A and 3A; see also later) do not have the same level of thermal maturity; the Portalrubio coal is more ma- ture than the Estercuel coal.
chromatogram of m/z 191 revealing the distribution of the triterpanes in the Portalrubio coal pyrolysate. The distribution in the Estercuel coal pyrolysate is very similar to that shown in Fig. 7. All identified triterpanes belong to the hopane family; no triter- panes characteristic of higher land plants, such as oleananes or ursanes, were found. The abundance of hopanes relates to a contribution from bacteria to the coals.
Phenols and benzenediols
It is thought that 1-pristene (and 2-pristene) are derived from thermal breakdown of macromolec- ularly bound tocopherols [48]. The abundance of tocopherols in the extracted Mequinenza coal (Figs. 1A and 6) support this idea. It is not completely understood why in the other coal pyrolysates to- copherols are much lower whereas 1 -pristene is still a significant pyrolysis product (Figs. 2A-5A).
All pyrolysates contain abundant phenol and C1-CB alkylated phenols. No major differences in their distribution patterns were observed with the exception of the Rubielos coal pyrolysate where phenol is the most abundant component whereas in the other pyrolysates 3- and 4-methylphenol (which co-elute on the stationary phase used) are the most abundant. In addition 1,2-benzenediol and C-al- kylated 1,Zbenzenediols are important pyrolysis products in the pyrolysates. The well known pyroly- sis products of lignin, 4-alkyl-2-methoxyphenols and 4-alkyl-2,6-dimethoxyphenols [49], are not im- portant pyrolysis products in these brown coals.
Hopanes are significantly present in the pyroly- sates of the Estercuel and Portalrubio coals (Figs. 2A and 3A). Fig. 7 shows a partial accurate mass
Similar distributions of phenols and benzenediols have been reported for pyrolysates of fossil outer seed walls (testae) of water plants and the Beulah
3300 34ca
scan number A
Fig. 6. Partial, summed mass chromatogram of m/z 137 + 151 + 165 + 402 + 416 + 430 showing the distribution of the tocopherols in the flash pyrolysate (Curie temperature 610°C) of the Mequinenza coal. The inset shows the structures of the to- copherols.
I .%a
scan number -
Fig. 7. Partial, accurate (mass window 0.04 dalton) mass chro- matogram of m/z 191.22 revealing the distribution of the ho- panes in the flash pyrolysate of the Portalrubio coal. Key: 1 = trisnorhopene; 2 = 22,29,30-trisnorhop-17,21-ene; 3 = 17a (H)-22,29,30-trisnorhopane; 4 = 17/?(H)-22,29,30_trisnorho- pane; 5 = 17a(H),21P(H)-30-norhopane; 6 = 17B(H),21~((H)- 30-norhopane; 7 = hopene; 8 = 17a(H),21/3(H)-hopane; 9 = 17/I(H),21B(H)-30-norhopane; 10 = 17B(H),2la(H)-hopane; 11 = 17a(H),21/?(H)-homohopane; 12 = 17&H),21cc(H)-homoho- pane; 13 = unknown hopane; 14 = 17P(H),21/f(H)-homoho- pane.

368
Zap lignite [19]. At present, it is not clear whether the phenols encountered are derived from highly degraded lignin [20] or from a novel biomacromole- cule, a polyphenol [19]. The first hypothesis is deemed unlikely by Van Bergen et al. [19] since the morphology of the testae of the fossil seeds is per- fectly preserved. Such a preservation is unlikely if the underlying chemistry has been modified consid- erably. On the other hand, pyrolysates of recent an- giosperm or gymnosperm wood, the major source of lignin-like materials in coal, contain only low rel- ative amounts of phenols, which weakens the sec- ond hypothesis. Therefore, a straightforward inter- pretation of the origin of phenols in lignites and more mature coal pyrolysates cannot be made.
Sulphur compounds The sulphur compounds present in the pyroly-
sates of the coals studied are dominated by hydro- gen sulphide (as revealed by the FPD chromato- grams) and (alkyl)thiophenes, although (alkyl)ben- zothiophenes were also present in all pyrolysates. Hydrogen sulphide is formed by thermal degrada- tion of (poly)sulphide linkages in the macromolec- ular coal matrix whilst alkylthiophenes and -ben- zothiophenes are produced from sulphur-contain- ing aromatic units [50].
The dominance of alkylthiophenes over alkylben- zothiophenes is consistent with the low level of ther- mal maturity of the brown coal samples [51]. The distribution patterns of the Ci-C4 alkylthiophenes in the pyrolysates are relatively similar (Fig. 8). 2- Methylthiophene is in all cases the most abundant thiophene present. A difference is observed in the abundance of the Cz+ alkylthiophenes relative to 2-methylthiophene: in the pyrolysate of the Mequi- nenza coal the C2 + alkylthiophenes are more abun- dant than in the pyrolysate of the Rubielos coal (c$, Fig. 8A and C). In the latter pyrolysate Cq alkyl- thiophenes are difficult to identify because of their low concentrations. In the &-cluster 2,4-dimethyl- thiophene is the most dominant component. This is a characteristic pattern for coal pyrolysates [23,X& 531. In kerogen pyrolysates 2,5- and 2,3-dimethyl- thiophene (compounds 4 and 6 in Fig. 8) are always more abundant than 2,4_dimethylthiophene [XL 531. The abundance of 3-isopropyl-2-methylthio- phene (compound 14 in Fig. S), especially in the Mequinenza coal pyrolysate, is noteworthy. This is
J. S. SINNINGHE DAMSTE et al.
Fig. 8. Partial, accurate (mass window 0.02) summed mass chro- matogram of m/z 97.01 + 98.01 + 111.03 + 112.03 + 125.03 + 126.03 + 139.03 + 140.03 illustrating the distribution of the C,-C, alkylated thiophenes in the flash pyrolysates of the (A) Mequinenza, (B) Estercuel and (C) Rubielos coals. Key: 1 = 2-methylthiophene; 2 = 3-methylthiophene; 3 = 2-ethylthio- phene; 4 = 2,5_dimethylthiophene; 5 = 2,4_dimethylthiophene; 6 = 2,3_dimethylthiophene; 7 = 3,4_dimethylthiophene; 8 = 2-propylthiophene; 9 = 2-ethyL5methylthiophene; 10 = 2-eth- yl-4-methylthiophene; 11 = ethylmethylthiophene; 12 = 2,3,5- trimethylthiophene; 13 = 2,3,4_trimethylthiophene; 14 = 3-iso- propyl-2-methylthiophene; 15 = 2-methyl-5-propylthiophene; 16 = 2,5_diethylthiophene; 17 = 2-butylthiophene; 18 = meth- ylpropylthiophene; 19 = 2-ethyl-3,5_dimethylthiophene; 20 = ethyldimethylthiophene; 21 = 5-ethyl-2,3_dimethylthiophene; 22 = 2,3,4,5_tetramethylthiophene. Mass chromatograms are normalized on peak 5.
the first coal sample in which this compound is the most abundant Cq alkylthiophene. 3-Isopropyl-2- methylthiophene has also been encoutered as a ma- jor compound in flash pyrolysates of kerogens from the Monterey Formation [54]. The structure of 3- isopropyl-2-methylthiophene suggests that it is

Py-GC-MS OF SULPHUR-RICH BROWN COALS 369
formed during pyrolysis via P-cleavage of macro- molecular moieties in the coal matrix which were formed by incorporation of sulphur into 24-ethyl steroids [51]. The predominance of this compound in all coal pyrolysates can be rationalized by the fact that the major steroids biosynthesized by higher plants are compounds with an ethyl at C24 [55].
Although qualitatively no major differences were observed in thiophene composition, large variations in abundance relative to other pyrolysis products are evident (Figs. lA-5A). The Mequinenza coal pyrolysate (Fig. 1) contains the highest amounts of alkylthiophenes relative to the other pyrolysis prod- ucts. This is also reflected by the very high thio- phene ratio (1.60, Table I), which is defined as [2,3- dimethylthiophene]/{[1,2-dimethylbenzene] + [n-l- nonene]} [52]. This ratio is obtained by integration of the appropriate peaks in the FID chromatogram of the pyrolysate. The thiophene ratio can be used to obtain an idea of the organic sulphur content of the samples pyrolysed [52]. Using the plot of the thiophene ratio versus atomic S,,,./C ratios deter- mined for a whole suite of samples as reported by Eglinton et al. [52], the S,,,./C ratio of the Mequi- nenza coal can be estimated to be 0.11. Elemental analysis indicates an atomic S,,,./C ratio of 0.062 (Table I), which suggests that the concentration of organic sulphur is overestimated by this approach. However, these measurements both indicate that this coal is extremely rich in organic sulphur (for every 9-14 carbon atoms it contains one sulphur atom). In this context, it should be noted that Orr [56] has defined organic sulphur-rich Type II kero- gens as Type II-S kerogens when their atomic S,,,./C ratios were larger than 0.04. Following the same definition, we propose here to classify the Me- quinenza coal as a Type III-S kerogen characterized by a low atomic H/C and a high atomic S,,,,/C (>0.04) and O/C (>0.20) ratios. The other coal pyrolysates contain less abundant sulphur com- pounds (Figs. 2A-5A) as is also evident from their thiophene ratios (Table I). The Paula coal is, how- ever, still fairly sulphur-rich and its estimated S,,,./C ratio (Table I) indicate that it is has a com- position close to that of a Type III-S kerogen, as confirmed by the elemental analysis data (Table I). The organic sulphur data obtained by Py-GC-MS show the same trends as the elemental composition data (Table I), although some discrepancies exist in absolute values.
An interesting observation is the significant dif- ference in the abundance of alkylthiophenes in the pyrolysates of the Estercuel and Portalrubio coals, whereas the distribution of other pyrolysis products is very similar (see also the later discussion of ex- tended alkylbenzenes), suggesting a very similar contribution of organic matter to these coals. This can be explained by the availability of inorganic sul- phur species which are capable of reacting with the organic matter to form thiophene moieties in the coal matrix in the palaeodepositional environment. In the case of the depositional environment of the Portalrubio coal less inorganic sulphur species were available than in the case of the Estercuel deposi- tional environment. A second explanation is the slightly higher level of thermal maturity of the Por- talrubio coal; it is known that on increasing thermal stress organic sulphur is preferentially removed [51,57].
In the Rubielos coal pyrolysate a series of as yet unrecognized sulphur compounds were identified. Their presence was evident from the FPD chro- matogram of the pyrolysate (Fig. 9), which contains a large peak not previously noted in other analyses of coal and kerogen samples [50-521. The mass spectrum of this component (Fig. 9, inset) indicates that it is dimethyl tetrasulphide. The presence of four sulphur atoms in this molecule explains why the presence of this component in the pyrolysate leads to such a large peak in the FPD chromato- gram (note also that the FPD instrument has a quadratic response). Dimethyl trisulphide and di- methyl disulphide were also identified in the pyroly- sate. It was not possible to determine whether di- methyl sulphide was also present in the pyrolysate since the acquisition of mass spectra starts 1 min after pyrolysis. The geochemical significance of these dimethyl polysulphides in pyrolysates is as yet unknown but their formation by secondary reac- tions is unlikely because primary pyrolysis products are removed very rapidly from the heated zone. The presence of these dimethyl polysulphides may ex- plain the discrepancy between the S,,,.jC ratio as estimated by the thiophene ratio and the higher atomic S,,,./C ratio as determined by elemental analysis because the former determination takes on- ly relative thiophene abundance into consideration which in case of the Rubieios coal will lead to an underestimation owing to the presence of dimethyl polysulphides in the pyrolysate.

370 J. S. SINNINGHE DAMSTJ? et al.
retention time -
Fig. 9. Partial (O-90 min) FPD chromatogram (normalized on compound 3) of the pyrolysate of the Rubielos coal. Key: 1 = thiophene: 2 = dimethyl disulphide; 3 = 2-methylthiophene; 4 = 3-methylthiophene; 5 = 2,3_dimethylthiophene; 6 = dimethyl trisulphide; 7 = benzo[b]thiophene; 8 = dimethyl tetrasulphide. The inset shows the mass spectrum (corrected for background) of dimethyl tet- rasulphide.
Alkylbenzenes The C1-C4 alkylated benzenes are dominant py-
rolysis products in all coal pyrolysates but no major differences were observed in their distribution pat- terns. Fig. 10 illustrates this statement; it shows the distributions in two coal pyrolysates in which some variation can be observed. For example, ethylben- zene (compound 2 in Fig. lo), propylbenzene (com- pound 7) and butylbenzene (compound 21) are rela- tively higher in the Portalrubio coal pyrolysate than in the Paula coal pyrolysate. In general, the alkyl- benzene distributions are typical for those observed in coal pyrolysates [58]. The relatively low amounts of 1,2,3,4_tetramethylbenzene reveals that photo- synthetic sulphur bacteria are not major contrib- utors of organic matter to these coals [58-611. In case of the Rubielos coal pyrolysate a substantial reduction in the abundance of CZ-C4 alkylbenzenes relative to toluene is observed, an observation also made in case of the alkylthiophenes.
Striking differences between the coal pyrolysates were observed in the distributions of long-chain al- kylbenzenes. In most samples these compounds are
dominated by monoalkylbenzenes; only the pyroly- sates of Paula and Rubielos coals contain a series of 2-alkyltoluenes with a concentration of the same or- der of magnitude as the monoalkylbenzenes. The distributions of the monoalkylbenzenes in three representative pyrolysates are shown by partial mass chromatograms of m/z 9 1 + 92, the two most abundant ions in the mass spectra of monoalkyl- benzenes (Fig. 11). In case of the Mequinenza and Rubielos coal pyrolysates (Fig. 11B and C), the concentration of the higher members of this series of compounds is only a few percent of the first member (i.e., toluene). However, in case of the Es- tercuel (Fig. 11A) and Portalrubio (not shown but identical with that of Estercuel) coal pyrolysates, a second maximum is observed at C1 8. Further, a sec- ond series of compounds eluting just before the monoalkylbenzenes is apparent in Fig. 11A. This second series has a maximum at Cl7 and a slight odd-over-even carbon number predominance in the Cr6-CZ0 range, whereas the monoalkylbenzenes possess an even-over-odd carbon number predom- inance in this range. Mass spectra of the C,, mem-

Py-GC-MS OF SULPHUR-RICH BROWN COALS
Fig. 10. Partial summed mass chromatograms of m/z 91 + 92 + 105 + 106 + 119 + 120 + 133 + 134revealing thedistributions of the C,-C, alkylated benzenes in the flash pyrolysates of the kerogens of (A) Paula and (B) Portalrubio coal. Key: 1
= xylene; 6 = isopropylbenzene;
1,4_diethylbenzene; 21 = butylbenzene; 22 = 1,2-diethylben- zene; 23 = I-ethyl-3$dimethylbenzene; 24 = I-methyl-2-pro- pylbenzene: 25 = 2-ethyl-1,4_dimethylbenzene; 26 = I-eth- yl-2,4_dimethylbenzene; 27 = I-ethyl-3,4_dimethylbenzene; 28 = 2-ethyl-1,3_dimethylbenzene; 29 = I-ethyl-2,3_dimethylben- zene; 30 = 1,2,4,5_tetramethylbenzene; 31 = 1,2,3,5_tetrameth- ylbenzene; 32 = 1,2,3,4- tetramethylbenzene. Mass chromato- grams are normalized on peak 3 + 4. Semi-quantitative determi- nation of the alkylbenzene abundances indicated that the con- centration of 1,3- and 1,4_dimethylbenzene is 30% and 8% of the toluene concentration in the pyrolysates of the Paula and Portal- rubio coal, respectively.
bers of these series (Fig. 12) indicated that the sec- ond (earlier eluting) series is most likely a linear alkylbenzene with an unsaturation in the side- chain. The characteristic fragment of m/z 104, which is present in all the mass spectra of this series, indicates that the double bond is not conjugated
371
with the aromatic ring but reference mass spectra indicated that this fragment cannot be used to fur- ther assess the double bond position [62]. The for- mation of this ion is most likely induced by an a-hy- drogen transfer. Mass chromatography of m/z 91 + 92, 104 and 230 of the C17 cluster (Fig. 13) re- veals that, in addition to the major series described above, other minor components (compounds 1, 4 and 5 in Fig. 13) are present which possess similar mass spectra, suggesting that they are isomers of the major “104” component. This can be rationalized by the presence of isomers with the double bond in
‘I a
A
B
20 25
IlllIJJ~~,
C
a
15 I I J 20 25 ,1,,,1,,,,‘
1500 2500 scan number-t
Fig. 11. Partial, summed mass chromatograms of m/z 91 + 92 revealing the distributions of alkylbenzenes in the flash pyroly- sates of (A) Estercuel, (B) Mequinenza and (C) Rubielos coals. Numbers indicate total number of carbon atoms of this series of compounds. Mass chromatograms are normalized on 1,3- and 1,4-dimethylbenzene (peak a) in case of the Mequinenza and Rubielos coal pyrolysates and on dodecylbenzene in case of the Estercuel coal pyrolysate.

372
lOO-
50-
91
9 12 A im- 91
232
J. S. SINNINGHE DAMSTl? et al.
B
104
b 100 150 250
Fig. 12. Mass spectra (subtracted for background) of (A) n-undecylbenzene and (B) n-undec-1 I-enylbenzene obtained from the Py-GC- MS analysis of the Estercuel coal.
different positions and in different stereochemical configurations (cis and trans). By analogy with the retention behaviour of n-alkenes relative to n-al- kanes [63], compound 2 in Fig. 13 is tentatively identified as undec- lo-enylbenzene, compounds 3 and 4 as cis- and trans-undec-9-enylbenzene and the cluster of compounds indicated by 1 as monounsat- urated undecylbenzenes with the double bond posi- tion at ~~-0’.
To the best of our knowledge, a distribution of monoalkylbenzenes as observed in the Estercuel and Portalrubio coal pyrolysates has not yet been re- ported. This phenomenon probably indicates that the Estercuel and Portalrubio coals contain specific moieties from which these long-chain monoalkyl- benzenes are generated. Similar distributions have been observed for homologous series of pyrolysis productsgenerated from aromatic moieties (which preferentially cleave at the @-carbon-carbon bond of the alkyl side-chain): 2-alkyltoluenes, 2-alkyl-5- methylthiophenes and 2-alkyl-5ethylthiophenes in the pyrolysate of the kerogen of the Guttenberg Oil Rock [59] and for 5-alkyl-1,3-benzenediols and mono- and dimethyl-5-alkyl-1,3-benzenediols in the pyrolysate of the kerogen of the Estonian kukersite [64]. In both studies these distributions were inter- preted as caused by thermal degradation of a moie- ty (toluene, 2_methylthiophene, 2-ethylthiophene, 1,3-benzenediol and methylated 1,3-benzenediol) with a long alkyl chain bound to the macromolec- ular matrix via a heteroatom. The second maximum in the distribution can then be used to assess the position in the alkyl side-chain at which most of the moieties are bound to the macromolecular matrix. These interpretations are, to some extent, support-
ed by pyrolysis of model compounds [60]. For ex- ample, flash pyrolysis of the sodium salt of 16-(4’- methylphenyl)hexadecanoic acid generates a suite of 4-alkyltoluenes with a maximum at CZZ. Further, unsaturated counterparts are also formed and show
run number d
Fig. 13. Partial mass chromatograms of m/z 91 + 92, m/z 104 and m/z 230 of the pyrolysate of the Portalrubio coal.

Py-GC-MS OF SULPHUR-RICH BROWN COALS 373
a shift in their carbon number distribution by one carbon atom [60]. These data suggest that in case of the Estercuel and Portalrubio coal long-chain alkyl- benzenes are present and are preferentially bound at position 12 in the alkyl side-chain of these moie- ties. At present we can only speculate on the hetero- atom involved in the bonding of these moieties but on basis of the significant differences in organic sul- phur content between the two coals (see above) and the abundance of oxygen (Table I), ether bonds are not unlikely.
Alkylnaphthalenes Naphthalene and alkylated naphthalenes are im-
portant compounds in all pyrolysates, but are espe- cially abundant in the Estercuel and Portalrubio coal pyrolysates. Fig. 14 (left panels) shows the dis- tribution of naphthalene and its Cl-C3 alkylated derivatives by summed mass chromatograms of m/z 128 + 141 + 142 + 155 + 156 + 169 + 170 (the major ions in the mass spectra of these compounds) in the pyrolysates of the Mequinenza, Estercuel, Paula and Rubielos coals. The distribution of the (alkyl)naphthalenes in the Estercuel and Portalru- bio coals pyrolysates is virtually identical, illustrat- ing again the similar composition of these two sam- ples. Substantial differences are observed in the car- bon number distributions of the naphthalenes: naph- thalene is the most dominant component in the Ru- bielos coal pyrolysate whereas the Ci and the C3 alkylnaphthalenes dominate in the pyrolysates of the Mequinenza and Paula coals and Estercuel coal, respectively.
Significant differences are also observed in the in- ternal distribution patterns. For example, the distri- bution of the CZ alkylated naphthalenes (Fig. 14, middle panels) is similar for the Estercuel, Paula and Rubielos coal pyrolysates but substantially dif- ferent from that of the Mequinenza coal pyrolysate. In the latter pyrolysate, 1,5_dimethylnaphthalene (compound 12 in Fig. 14) dominates whereas in the other pyrolysates 1,7_dimethylnaphthalene and/or 1,6_dimethylnaphthalene (compounds 9 and 10) are the most abundant CZ alkylnaphthalene(s).
The distribution of the C3 alkylnaphthalenes (Fig. 14, right panels) in the Estercuel, Portalrubio and Paula coal pyrolysates is similar and is dom- inated by 1,2,5_trimethylnaphthalene (compound 27). In case of the Estercuel and Portalrubio pyroly-
sates this compound is prominent in the TIC. The C3 alkylnaphthalene distributions in the pyroly- sates of the other two coals are not dominated by a specific isomer. It is worth noting that although the C2 alkylnaphthalene distribution of the pyrolysate of the Rubielos coal is similar to that observed in the Paula and Estercuel coal pyrolysates, the C3 al- kylnaphthalene distribution is completely different (Fig. 14). In the pyrolysates where the C3 alkyl- naphthalenes are dominated by 1,2,5_trimethyl- naphthalene the C4 alkylnaphthalenes are also dominated by one specific isomer, 1,2,5,6-tetra- methylnaphthalene (not shown). The concentration of this component is ca. 50% of the concentration of 1,2,5_trimethylnaphthalene. The co-occurrence of 1,2,5_trimethylnaphthalene and 1,2,5,6-tetra- methylnaphthalene as major alkylnaphthalenes has been described before in extracts of coal and shale samples [65]. Piittmann and Villar [65] and De las Heras [66] attributed the predominance of these two specific polymethylnaphthalenes in the coal extracts to their diagenetic derivation from pentacyclic tri- terpenoids via 8,14-seco-triterpenoids. Such a deri- vation is in agreement with the abundance of ho- panes in two of the three samples (i.e., Estercuel and Portalrubio) with these characteristics. The coal samples examined by Piittmann and Villar [65] had a slightly higher level of thermal maturity (random vitrinite reflectance 0.5%0.83%) than the brown coals studied in this work. In the case of the brown coals these specific naphthalenes are probably gen- erated mainly via thermal breakdown of the macro- molecular matrix, suggesting that macromolecular- ly bound aromatized 8,14-seco-triterpenoids are the precursors of these compounds.
In the pyrolysates of the Estercuel and Portal- rubio coals cadalene (Cisopropyl- 1,6-dimethyl- naphthalene) is an important component and the dominant C5 alkylnaphthalene. Its co-occurrence with 1,6_dimethylnaphthalene as a major CZ alkyl- naphthalene can be ascribed to the presence of res- in-derived material (resinite) in the coal samples
[671.
CONCLUSIONS
The pyrolysates of the five brown coals all con- tain abundant n-alkanes, n-1-alkenes, (alkyl)phe- nols and alkyl(benzenes). The n-alkanes and n-l-

1
8 3 2
3 6
u 4 -
14 -
CBnaphthalenes 12
I
6
7
6
6
+10
1.2 16
J. S. SINNINGHE DAMSTF et al.
C3-naDhthalenes ‘61 24t25
2,4 20+21 26
24+26
26 I
)+21
k
26
27
i
Fig. 14. Partial summed mass chromatograms of m/z 128 + 141 + 142 + 155 + 156 + 169 + 170 and m/z 156 and 170 illustrating the distribution of naphthalene and C,-C, alkylated naphthalenes and the C, and C, naphthalenes, respectively, in the pyrolysates of the Mequinenza, Estercuel, Paula and Rubielos coals. Keys: 1 = naphthalene; 2 = 2-methylnaphthalene: 3 = 1-methylnaphthalene; 4 = C, naphthalenes; 5 = 2-ethylnaphthalene; 6 = 2,5_dimethylnaphthalene; 7 = 2,7-dimethylnaphthalene; 8 = 1,3_dimethylnaphthalene; 9 = 1,7_dimethylnaphthalene; 10 = 1,6_dimethylnaphthalene; 11 = 2,3- and 1,4_dimethylnaphthalene; 12 = 1,5_dimethylnaphthalene; 13 = 1,2_dimethylnaphthalene; 14 = C,-naphthalenes; 15 = 2-propylnaphthalene; 16 = I-propylnaphthalene; 17 = ethylmethylnaph- thalenes; 18 = 1,3,7_trimethylnaphthalene; 19 = 1,3,6_trimethylnaphthalene; 20 = 1,4,6_trimethylnaphthalene; 21 = 1,3,5_trimethyl- naphthalene; 22 = 2,3,6-trimethylnaphthalene; 23 = 1,2,7:trimethylnaphthalene; 24 = 1,6,7_trimethylnaphthalene; 25 = 1,2,6-tri- methylnaphthalene; 26 = 1,2,4_trimethylnaphthalene; 27 = 1,2,5_trimethylnaphthalene; 28 = 1,2,3_trimethylnaphthalene.

Py-GC-MS OF SULPHUR-RICH BROWN COALS
alkenes are derived from thermal breakdown of ali- phatic biomacromolecules such as cutan, algaenan and/or suberan whereas the (alkyl)phenols reflect the presence of highly degraded lignin or a polyphenol biomacromolecule. The short-chain al- kylbenzenes cannot be ascribed to a specific source material in the brown coals. All pyrolysates contain I-pristene as a major pyrolysis product. This prob- ably reflects the presence of macromolecularly bound tocopherols. This was confirmed by the pres- ence of significant amounts of a series of tocophe- rols in the pyrolysate in one extracted brown coal.
One of the five brown coal samples investigated is so rich in organic sulphur (one sulphur atom for every 9914 carbon atoms) that it is appropriate to define a new kerogen type describing the kerogen contained in this coal. Type III-S kerogen is defined as a kerogen with high atomic S,,,./C (> 0.04) and O/C ratios (> 0.20). The flash pyrolysates of such samples are dominated by sulphur compounds [mainly (alkyl)thiophenes and (alkyl)benzothio- phenes] and (alkyl)phenols.
Two of the five brown samples investigated con- tain a series of long-chain alkylbenzenes with an unprecedented carbon number distribution pattern with a second maximum at Cls. This unusual distri- bution pattern is thought to originate from the pres- ence of long-chain alkylbenzene moieties bound via a heteroatom (presumably an ether bond) to the macromolecular coal matrix preferentially at posi- tion 12 in the alkyl side-chain of these moieties. The origin of such moieties is unknown.
The (alkyl)naphthalene distributions in the pyro- lysates of the brown coals show a large variation both in carbon number distribution and in predom- inance of specific isomers. The predominance of 1,2,5_trimethylnaphthalene together with 1,2,5,6- tetramethylnaphthalene is ascribed to the presence of macromolecularly bound, aromatized 8,14-seco- triterpenoids. The co-occurrence of 1,6-dimethyl- naphthalene and cadalene as major naphthalene re- veals the contribution of resin-derived material in some of the brown coals.
ACKNOWLEDGEMENTS
The authors thank Drs. P. Anadon, L. Cabrera and X. Querol for samples and geological data and H. Veld for vitrinite reflectance measurements. F. X. C. d.1. H. thanks la Caixa de Manresa for the
375
fellowship “Beta Viatge de Recera a 1’Europa Co- munitaria”.
REFERENCES
1
2
3
4
5
6
I
8
9
10
11
12
13
14
15
16
17
18
19
20 21
22
E. Stach, G. H. Taylor, M. Th. Mackowsky, D. Chandra, M. Teichmiiller and R. Teichmiiller, Stachs Textbook of Coal Petrology, Gebriider Borntraeger, Berlin, 3rd ed., 1982. D. W. van Krevelen, Coal. Typology, Chemistry, Physics and Constitution, Elsevier, Amsterdam, 1961. N. C. Deno, K. W. Curry, D. Jones, K. R. Keegan, W. G. Rakitsky, C. A. Richter and R. D. Minard, Fuel, 60 (1981) 210-212. Ph. Blanc and P. Albrecht, Org. Geochem., 17 (1991) 913- 918. G. van Graas, J. W. de Leeuw and P. A. Schenck, J. Anal. Appl. Pyrol., 2 (1980) 265-216. G. van Graas, J. W. de Leeuw and P. A. Schenck, in A. G. Douglas and J. R. Maxwell (Editors), Advances in Organic Geochemistry 1979, Pergamon Oxford, 1980, pp. 21 l-217. M. Nip, J. W. de Leeuw, P. A. Schenck, H. L. C. Meuzelaar, S. A. Stout, P. H. Given and J. J. Boon, J. Anal. Appl. Pyrol., 8 (1985) 221-239. M. Nip,, W. Genuit, J. J. Boon, J. W. de Leeuw, M. Blazlo and T. Szekely, J. Anal. Appl. Pyrol., 11 (1987) 125-147. M. Nip, J. W. de Leeuw and P. A. Schenck, Geochim. Cos- mochim. Acta, 52 (1988) 637-648. M. Nip, J. W. de Leeuw, P. A. Schenck, W. Windig, H. L. C. Meuzelaar and J. C. Crelling, Geochim. Cosmochim. Acta, 53 (1989) 671-683. M. Nip, J. W. de Leeuw, P. A. Schenck and J. C. Crelling, Energy Fuels, 6 (1992) 125-136. M. Nip, E. W. Tegelaar, J. W. de Leeuw, P. A. Schenck and P. J. Holloway, Naturwissenschaften, 73 (1986) 579-585. M. Nip, E. W. Tegelaar, H. Brinkhuis, J. W. de Leeuw, P. A. Schenck and P. J. Holloway, Org. Geochem., 10 (1986) 769- 778. E. W. Tegelaar, J. W. de Leeuw, C. Largeau, S. Derenne, H.-R. Schulten, R. Miiller, J. J. Boon, M. Nip and J. C. M. Sprenkels, J. Anal. Appl. Pyrol., 15 (1989) 29-54. E. W. Tegelaar, J. H. F. Kerp, H. Visscher, P. A. Schenck and J. W. de Leeuw. Palaeohiologv. 17 (1991) 133-144. E. W. Tegelaar, Ph.D. Thesis, Delft University of Technol- ogy, Deift, 1990, pp. 115-134. P. A. Schenck, J. W. de Leeuw, G. van Graas, J. Haverkamp and M. Bouman, in J. Brooks (Editor), Organic Maturation Studies and Fossil Fuel Exploration, Academic Press, New York, 1981, pp. 225-237. W. J. Guilford, D. M. Schneider, J. Labovitz and S. J. Opel- la, Plant Physiol., 86 (1988) 134136. P. F. van Bergen, M. E. Collinson, J. S. Sinninghe Damsti: and J. W. de Leeuw, Am. Chem. Sot. Div. Fuel Chem. Prepr., 36 (1991) 698-701. P. G. Hatcher, Org. Geochem., 16 (1990) 959-968. B. G. K. van Aarssen, H. C. Cox, P. Hoogendoorn and J. W. de Leeuw, Geochim. Cosmochim. Acta, 54 (1990) 3021-3031. D. J. Casagrande, in A. C. Scott (Editor), Coal and Coal- bearing Strata: Recent Advances, Publication No. 32, Geo- logical Society London. 1987, pp. 87-105.

316 J. S. SINNINGHE DAMSTE et al.
23 J. S. Sinninghe Damste and J. W. de Leeuw, Fuel Process. Tech&., 30 (1992) 1099178.
24 L. M. Stock, R. Wolny and B. Bal, Energy Fuels, 3 (1989) 651-661.
25 X. Querol, R. Salas, G. Pardo and L. Ardevol, Int. J. Coal Geol., 11 (1989) 171-189.
26 L. Cabrera and A. Saez, J. Geol. Sot. London, 144 (1987) 451461.
27 L. Cabrera, F. Colombo and S. Robles, in M. Misla and S. Rose11 (Editors), Excursion Guide Book of the 6th European Regional Meeting of the International Association of Sedimen- tologist, Excursion No. 10, 1985, pp. 394492.
28 P. Anadon, L. Cabrera, R. Julia, E. Rota and L. Resell, Palaeogeogr. Palaeoclimat. Palaeoecol., 70 (1989) 7-28.
29 P. Anadon, L. Cabrera and R. Julia, in A. J. Fleet, K. Kelts and M. R. Talbot (Editors), Lacustrine Petroleum Source Rocks, Special Publication No. 40, Geological Society Lon- don, 1988, pp. 353-361.
30 A. Venema and J. Veurink, J. Anal. Appl. Pyrol., 7 (1985) 207-213.
31 J. J. Boon, A. D. Pouwels and G. B. Eijkel, Biochem. Sot. Trans., 15 (1987) 251-258.
32 M. L. Lee, D. L. Vassilaros, C. M. White and M. Novotny, Anal. Chem., 51 (1980) 7688774.
33 S. J. Rowland, R. Alexander and R. 1. Kagi, J. Chromatogr., 294 (1984) 407412.
34 J. S. Sinninghe Damste, A. C. Kock-van Dalen, J. W. de Leeuw and P. A. Schenck. J. Chromatogr., 435 (1988) 435- 452.
35 P. G. Forster, R. Alexander and R. I. Kagi, J. Chromatogr., 483 (1989) 384-389.
36 W. A. Hartgers. J. S. Sinninghe Damste and J. W. de Leeuw, J. Chromatogr., 606 (1992) 211-220.
37 G. Eglinton and R. J. Hamilton, in T. Swain (Editor), Chem- ical Plant Taxonomy, Academic Press, New York, 1963, pp. 187-217.
38 M. Nip, J. W. de Leeuw, P. J. Holloway, J. P. T. Jensen, J. C. M. Sprenkels, M. de Pooter and J. J. M. Sleeckx, J. Anal. Appl. Pwol.. 11 (1987) 287-295.
39 E. W. Tegelaar, J. Wattendorf and J. W. de Leeuw, Rev. Palaeobot. Palynol., in press.
40 J. W. de Leeuw, P. F. van Bergen, B. G. K. van Aarssen, J.-P. L. A. Gatellier, J. S. Sinninghe Damste and M. E. Collinson, Proc. R. Sot. London, Ser. B., 333 (1991) 329-337.
41 C. Largeau, S. Derenne, E. Casadevall, E. Kadouri and N. Sellier, Org. Geochem., 10 (1986) 1023-1032.
42 A. Kadouri, S. Derenne, C. Largeau, E. Casadevall and C. Berkaloff, Phytochemistry, 27 (1988) 551-557.
43 K. Goth, J. W. de Leeuw, W. Piittmann and E. W. Tegelaar, Nature (London), 336 (1989) 759-761.
44 E. W. Tegelaar, J. W. de Leeuw, S. Derenne and C. Largeau, Geochim. Cosmochim. Acta, 53 (1989) 3103-3106.
45 J.-P. L. A. Gatellier, J. W. de Leeuw, J. S. Sinninghe Damste, S. Derenne, C. Largeau and P. Metzger, Geochim. Cosmo- chim. Acta, submitted for publication.
46 P. van Bergen, unpublished results. 47 H. Goossens, A. Due, J. W. de Leeuw, B. van der Graaf and
P. A. Schenck, Geochim. Cosmochim. Acta, 52 (1988) 1189- 1193.
48 H. Goossens, J. W. de Leeuw, P. A. Schenck and S. C. Bras- sell, Nature (London), 312 (1984) 44&442.
49 C. Saiz-Jimenez and J. W. de Leeuw, Org. Geochem., 6 (1984) 417422.
50 J. S. Sinninghe Dams& T. I. Eglinton, J. W. de Leeuw and P. A. Schenck, Geochim. Cosmochim. Acta, 53 (1989) 8733889.
51 T. I. Eglinton, J. S. Sinninghe Damste, M. E. L. Kohnen, J. W. de Leeuw, S. R. Larter and R. L. Patience, in W. L. Orr and C. M. White (Editors), Geochemistry of Sulfur in Fossil Fuels (ACS Symposium Series, Vol. 429), American Chemical Society, Washington, DC, 1989, pp. 529-565.
52 T. I. Eglinton, J. S. Sinninghe Damste, M. E. L. Kohnen and J. W. de Leeuw, Fuel, 69 (1990) 13941404.
53 T. I. Eglinton, J. S. Sinninghe Dams& W. Pool, J. W. de Leeuw, G. Eijkel and J. J. Boon, Geochim. Cosmochim. Acta, 56 (1992) 1545-1560.
54 J. S. Sinninghe Damste, T. I. Eglinton, W. I. C. Rijpstra and J. W. de Leeuw, in W. L. Orr and C. M. White (Editors), Geochemistry of Sulfur in Fossil Fuels (ACS Symposium Se- ries, Vol. 429) American Chemical Society, Washington, DC, 1989, pp. 486528.
55 W. R. Nes and M. L. McKean, Biochemistry of Steroids and Other Isopentenoids, University Park Press, Baltimore, 1977.
56 W. L. Orr, Org. Geochem., 10 (1986) 499-516. 57 J. A. Gransch and J. Posthuma, in B. Tissot and F. Bienner
(Editors), Advances in Organic Geochemistry 1973, Editions Technip, Paris, 1974, pp. 727-739.
58 W. A. Hartgers, J. S. Sinninghe Damste and J. W. de Leeuw, Geochim. Cosmochim. Acta, submitted for publication.
59 A. G. Douglas, J. S. Sinninghe Damste, M. G. Fowler, T. I. Eglinton and J. W. de Leeuw, Geochim. Cosmochim. Acta, 55 (1991) 275-291.
60 W. A. Hartgers, J. S. Sinninghe Damste and J. W. de Leeuw, Am. Chem. Sot. Div. Fuel Chem. Prep., 36 (1991) 790-795.
61 A. G. Requejo, J. Allan, S. Creaney, N. R. Cray and K. S. Cole, Org. Geochem., in press.
62 F. W. McLafferty and D. B. Stauffer, The Wiley/NBS Regis- try of Mass Spectral Data, Wiley, New York, 1989.
63 J. S. Sinninghe Damste, F. X. de las Heras, P. F. van Bergen and J. W. de Leeuw, Geochim. Cosmochim. Acta, 55 (1991) 3379-3385.
64 S. Derenne, C. Largeau, E. Casadevall, J. S. Sinninghe Dam- ste, E. W. Tegelaar and J. W. de Leeuw, Org. Geochem., 16 (1990) 873-888.
65 W. Ptittmann and H. Villar, Geochim. Cosmochim. Acta, 51 (1987) 3023-3029.
66 F. X. de las Heras, J. 0. Grimalt, J. Albaiges, Geochim. Cos- mochim. Acta, 55 (1991) 3387-3390.
67 B. G. K. van Aarssen, J. W. de Leeuw and B. Horsfield, J. Anal. Appl. Pyrol., 20 (1991) 125-139.