Analysis of unresolved complex mixtures of hydrocarbons extracted from Late Archean sediments by comprehensive two-dimensional gas chromatography (GC×GC)

Abstract

Hydrocarbon mixtures too complex to resolve by traditional capillary gas chromatography display gas chromatograms with dramatically rising baselines or “humps” of coeluting compounds that are termed unresolved complex mixtures (UCMs). Because the constituents of UCMs are not ordinarily identified, a large amount of geochemical information is never explored. Gas chromatograms of saturated/unsaturated hydrocarbons extracted from Late Archean argillites and greywackes of the southern Abitibi Province of Ontario, Canada contain UCMs with different appearances or “topologies” relating to the intensity and retention time of the compounds comprising the UCMs. These topologies appear to have some level of stratigraphic organization, such that samples collected at any stratigraphic formation collectively are dominated by UCMs that either elute early- (within a window of C15–C20n-alkanes), early- to mid- (C15–C30n-alkanes), or have a broad UCM that extends through the entire retention time of the sample (from C15–C42n-alkanes). Comprehensive two-dimensional gas chromatography time-of-flight mass spectrometry (GC×GC–MS) was used to resolve the constituents forming these various UCMs. Early- to mid-eluting UCMs are dominated by configurational isomers of alkyl-substituted and non-substituted polycyclic compounds that contain up to six rings. Late eluting UCMs are composed of C36–C40 mono-, bi-, and tricyclic archaeal isoprenoid diastereomers. Broad UCMs spanning the retention time of compound elution contain nearly the same compounds observed in the early-, mid-, and late-retention time UCMs. Although the origin of the polycyclic compounds is unclear, the variations in the UCM topology appear to depend on the concentration of initial compound classes that have the potential to become isomerized. Isomerization of these constituents may have resulted from hydrothermal alteration of organic matter.

Full text

Page 1

Analysis of unresolved complex mixtures of hydrocarbons extracted
from Late Archean sediments by comprehensive two-dimensional gas
chromatography (GC?GC)
Gregory T. Venturaa,*, Fabien Keniga, Christopher M. Reddyb, Glenn S. Frysingerc,
Robert K. Nelsonb, Ben Van Mooyb, Richard B. Gainesc
aUniversity of Illinois at Chicago, Department of Earth and Environmental Sciences, M/C 186, 845 West Taylor Street, Chicago, IL 60607-7059, USA
bDepartment of Marine Chemistry and Geochemistry, MS#4, Woods Hole Oceanographic Institution, Woods Hole, MA 02543-1543, USA
cDepartment of Science, US Coast Guard Academy, New London, CT 06320-8101, USA
a r t i c l ei n f o
Article history:
Received 22 October 2007
Received in revised form 3 March 2008
Accepted 4 March 2008
Available online 26 March 2008
a b s t r a c t
Hydrocarbon mixtures too complex to resolve by traditional capillary gas chromatography
display gas chromatograms with dramatically rising baselines or ‘‘humps” of coeluting
compounds that are termed unresolved complex mixtures (UCMs). Because the constitu-
ents of UCMs are not ordinarily identified, a large amount of geochemical information is
never explored. Gas chromatograms of saturated/unsaturated hydrocarbons extracted
from Late Archean argillites and greywackes of the southern Abitibi Province of Ontario,
Canada contain UCMs with different appearances or ‘‘topologies” relating to the intensity
and retention time of the compounds comprising the UCMs. These topologies appear to
have some level of stratigraphic organization, such that samples collected at any strati-
graphic formation collectively are dominated by UCMs that either elute early- (within a
window of C15–C20n-alkanes), early- to mid- (C15–C30n-alkanes), or have a broad UCM that
extends through the entire retention time of the sample (from C15–C42n-alkanes). Compre-
hensive two-dimensional gas chromatography time-of-flight mass spectrometry (GC?GC–
MS) was used to resolve the constituents forming these various UCMs. Early- to mid-elut-
ing UCMs are dominated by configurational isomers of alkyl-substituted and non-substi-
tuted polycyclic compounds that contain up to six rings. Late eluting UCMs are
composed of C36–C40mono-, bi-, and tricyclic archaeal isoprenoid diastereomers. Broad
UCMs spanning the retention time of compound elution contain nearly the same com-
pounds observed in the early-, mid-, and late-retention time UCMs. Although the origin
of the polycyclic compounds is unclear, the variations in the UCM topology appear to
depend on the concentration of initial compound classes that have the potential to become
isomerized. Isomerization of these constituents may have resulted from hydrothermal
alteration of organic matter.
? 2008 Elsevier Ltd. All rights reserved.
1. Introduction
The identification of individual molecular constituents
within complex organic mixtures of petroleum and sedi-
ment solvent extracts is typically achieved by a series of
chemical and chromatographic separations. The most com-
mon technique employs capillary gas chromatography
(GC), often coupled to a mass spectrometer, in order to
facilitate the separation, quantitation, and identification
of individual molecular components in complex mixtures.
Unfortunately, the chromatographic resolution afforded
by capillary GC is insufficient to resolve some complex
mixtures, which appears as a pronounced rising baseline
or a series of rising baselines in a gas chromatogram (e.g.
Gough and Rowland, 1990). In oils, the number of uniden-
0146-6380/$ - see front matter ? 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.orggeochem.2008.03.006
* Corresponding author. Tel.: +1 508 289 5262.
E-mail address: gventura@whoi.edu (G.T. Ventura).
Organic Geochemistry 39 (2008) 846–867
Contents lists available at ScienceDirect
Organic Geochemistry
journal homepage: www.elsevier.com/locate/orggeochem

Page 2

tified compounds comprising such unresolved complex
mixtures (UCMs) may amount to 250,000 compounds (Sut-
ton et al., 2005) indicating that an enormous amount of
inaccessible, geochemical information is unexploited.
UCMs are frequently observed in petroleums and hydro-
carbon (HC) extracts that have undergone biodegradation
(Rowland and Maxwell, 1984; Killops and All-Juboori,
1990; Gough and Rowland, 1990; Frysinger et al., 2003)
or hydrothermal alteration (Rushdi and Simoneit, 2002;
Simoneit et al., 2004; Zarate-del Valle and Simoneit,
2005). UCMs of biodegraded oils typically contain a low
abundance or complete absence, of aliphatic compounds
(e.g. Connan, 1984; Rowland et al., 1986; Palmer, 1993;
Swannell et al., 1995; Bost et al., 2001; Reddy et al.,
2002). The removal of these constituents, which averages
32% of the total hydrocarbon composition of most crude
oils (Killops and Killops, 2005) concentrates a preexisting
complex mixture to yield the amplified chromatographic
baseline of a UCM (Killops and All-Juboori, 1990).
Petroleum formed by hydrothermal alteration of sedi-
mentary organic matter undergoes accelerated diagenesis
and catagenesis (Simoneit and Lonsdale, 1982) from expo-
sure to high temperature fluids that reach up to 350 ?C
(Simoneit et al., 1984, 1992; Simoneit, 1994; Kvenvolden
and Simoneit, 1990). The UCMs of these hydrothermal
petroleums contain abundant low and high molecular
weight n-alkanes (e.g. Ogihara and Ishiwatari, 1998;
Yamanaka et al., 2000; Rushdi and Simoneit, 2002a, b;
Simoneit et al., 2004). These petroleums occasionally have
homologous or pseudohomologous series of compounds
with a carbon number predominance, such as n-cycloal-
kanes and branched alkanes with quaternary carbon atoms
(BAQCs; Ogihara and Ishiwatari, 1998; Simoneit, 1994;
Rushdi and Simoneit, 2002a, b; Simoneit et al., 2004). The
UCMs of hydrothermal petroleums are thought to be com-
posed of branched and cyclic compounds (e.g. Rushdi and
Simoneit, 2002; Simoneit et al., 2004). Aside from this, their
composition and formation has yet to receive the same
level of attention as the UCMs produced by biodegradation.
Because the compositional evolution of petroleum is
strongly dependent on numerous variables such as
temperature, time, source composition, water washing,
fractionation during migration, and microbial degradation,
the large inventory of organic compounds present in UCMs
may yield critical information that can be used to under-
stand factors regulating the occurrence of petroleum in
subsurface environments.
Attempts to identify the constituents in UCMs have met
with varying degrees of success and involved chemical oxi-
dation (Gough and Rowland, 1990; Killops and All-Juboori,
1990; Revill, 1992; Warton, 1999; Warton et al., 2000), thin
layer chromatography (Liu et al., 2005), chemical oxidation
followedbytreatmentwith molecularsieves(e.g. Armanios
etal.,1994;Ellisetal.,1994;Fazeelatetal.,1994),statistical
deconvolution methods (Pool et al., 1997; Dagan 2000;
Demir et al., 2000), and field ionization mass spectrometry
(Payzant et al., 1979). These attempts have led to the initial
beliefthatUCMswerecomposedofhydrocarbonswithsim-
ilar chemical properties that include large numbers of
branched and cyclic aliphatic and aromatic isomers
(Eglinton et al., 1975; Payzant et al., 1979; Alexander
et al., 1982; Sanders and Tibbetts, 1987; Killops and All-
Juboori, 1990). The use of oxidative degradation followed
by GC enables the release of additional compounds, which
led to the notion thatUCMs mightbe mixturesof fairly sim-
ple compounds comprising unsubstituted alkyl chains such
as isometric monoalkyl-substituted ‘‘T”-branched alkanes
(Gough et al., 1992; Warton et al., 2000). ‘‘T”-branched
alkanes were shown to be resistant to biodegradation
(Goughetal.,1992).Wartonetal.(1997)demonstratedthat
3% of the alkanes from a biodegraded crude oil were ‘‘T”-
branched alkanes. Payzant et al. (1979) showed that UCM
of the urea non adduct fractions of heavy oils was domi-
nated by mono- and bicyclic compounds, with decreasing
contribution of tri- to hexacyclic compounds. However,
the structure of most of these compounds was not
determined.
Even with excellent capillary GC columns and the exten-
sive use of chemical degradation methods, some complex
mixtures of compounds cannot be resolved and the nature
of the compounds that form these UCMs remains unclear
(Gough et al., 1992; Gough and Rowland, 1990; Warton,
1999). This problem, which was latent in the study of
biodegraded crude oil and crude oil contamination (i.e. Fry-
singer and Gaines, 2001; Reddy et al., 2002; Frysinger et al.,
2003), is overcome with comprehensive two-dimensional
gas chromatography (GC?GC; Liu and Phillips, 1991).
GC?GC links two capillary columns, with different station-
ary phases, via a modulator that creates packets of analytes
by temporarily focusing the effluent leaving the first col-
umn before entering the second column. The separation
of these packets by the second column produces a chro-
matogram with a high signal-to-noise ratio. Furthermore,
the separation power of the first column is conserved into
the second column, such that compounds not resolved by
the first column may be resolved by the second column.
GC?GC was successfully used to separate and identify bio-
markers (molecular fossils) in crude oils (Frysinger and
Gaines, 2001), modern and Holocene marine sediment ex-
tracts (Johnson et al., 2003), and UCMs of crude oil contam-
inated sediments (Frysinger et al., 2003; Reddy et al., 2002).
More recently, GC?GC has been used to distinguish drilling
mud contaminants from naturally occurring compounds in
oils (Reddy et al., 2007).
Sediments from the 2.710 to 2.704 Ga Tisdale and 2.685
to 2.676 Ga Porcupine Assemblage (Ayer et al., 2002) of the
Abitibi greenstone belt of Ontario, Canada were analyzed
to assess the composition, origin, and preservation poten-
tial of organic matter from Late Archean hydrothermal
depositional environments (Ventura, 2006; Ventura et al.,
2007). Bitumens were extracted from greywackes and
argillites of the Tisdale Assemblage collected from inter-
flow sedimentary rocks deposited between tholeiitic bas-
alts and tholeiitic dacites (Brisbin, 1997). Bitumens were
also extracted from distal margin turbidites of the overly-
ing Porcupine Assemblage (Fig. 1; Rice et al., 1992). These
sediments were subjected to post burial hydrothermal
alteration (Brisbin, 1997; Kerrich and Ludden, 2000) and
lower greenschist metamorphism (Dimroth et al., 1982;
Thompson, 2002).
Thesaturatedandunsaturatedhydrocarbon(s/uHC)frac-
tions of these bitumens were observed to contain pro-
G.T. Ventura et al./Organic Geochemistry 39 (2008) 846–867
847

Page 3

nounced UCMs (Fig. 1; Ventura, 2006; Ventura et al., 2007).
These UCMs have different appearances or ‘‘topologies”
relatingtotheintensityandretentiontimeofthecompounds
comprising the UCMs. These topologies also appear to have
some level of stratigraphic organization suggesting that the
type of UCM topology likely dependents upon the specific
natureoftheinitialsourceoforganicmatteraswellasdiage-
netic, or catagenetic processes (Ventura et al., 2007). In this
study, GC?GC–MS is used to identify the hydrocarbons that
represent the dominant compounds forming the different
UCM topologies. These results are then used to explore po-
tential mechanisms of UCM formation.
2. Samples and methods
2.1. Samples
Samples were collected near Timmins, Ontario in an
area named the Porcupine Gold Camp (PGC). Samples were
Fig. 1. Stratigraphic profile of the PGC (left). Total ion current (TIC) chromatogram from the extractable hydrocarbon fraction of representative samples of
the Tisdale and Porcupine Assemblage (right). Unresolved complex mixtures (UCMs) are labeled ‘‘Type ‘I–III” in reference to their topology (see text). Filled
circles indicate n-alkanes. Number above peak indicates numbers of carbon atoms in the compound. H indicates hopanes. Open squares label acyclic
irregular isoprenoids.
848
G.T. Ventura et al./Organic Geochemistry 39 (2008) 846–867

Page 4

selected to represent the pronounced UCM topology of
each stratigraphic formation. The s/u HC fractions of sol-
vent extracts from seven samples included two greywac-
kes from the Vipond Formation, one argillite from the
Krist Formation, and four argillites from the Hoyle Forma-
tion. A complete description of the geologic setting, bulk
organic carbon characterization, and biomarker composi-
tion is provided in Ventura (2006) and Ventura et al.
(2007).
2.2. Methods
2.2.1. Bitumen extraction
The surfaces of all samples were either scrubbed first
with a wire brush and rinsed with double distilled water
before being washed with dichloromethane (DCM)/metha-
nol (MeOH) in a 7.5:1, v/v ratio or ground off using a steel
wire brush attached to a Dremmel. The samples were then
ground to a powder using a shatter box. Between 100 and
180 g of powdered sediment was Soxhlet extracted with a
mixture of dichloromethane: methanol (DCM:MeOH; 7.5:1
v/v) for 72 h. The resulting bitumen was separated into po-
lar and apolar fractions by column chromatography on alu-
mina oxide following the method described in Simons and
Kenig (2001). The apolar fraction was collected using hex-
ane/DCM (9:1 v/v) as an eluent. Polar fractions were recov-
ered using DCM/MeOH (1:1 v/v) as an eluent. Elemental
sulfur was removed from the apolar fractions by passing
the apolar fraction through a Pasteur pipette filled with
activated Cu (HCl 6 N) using hexane as the eluent. Apolar
fractions were then separated into s/u HC and aromatic
fractions by column chromatography on silica gel. S/u HC
fractions were recovered with hexane. Aromatic fractions
were recovered using DCM/MeOH (7.5:1 v/v) as an eluent.
Sub-fractions were rotary evaporated to ?1 mL volume,
transferred to tared vials, dried under a continuous low
flow of N2, and again weighed. A procedural blank was
added to each batch of samples.
2.2.2. Gas chromatography–mass spectroscopy (GC–MS)
The s/u HC and aromatic fractions were analyzed in full
scan and selected ion monitoring (SIM) modes using a
Hewlett-Packard 6890 gas chromatograph (GC) coupled
to a Hewlett-Packard5973
(MSD). The GC was fitted with an HP-5MS capillary column
(30-m; 0.25-mm ID; 0.25-lm film thickness). The MSD
was operated in electron ionization mode at 70 eV, scan-
ning a mass range of m/z 40–650 at 2.44 scans/s. Injection
was done in pulsed-splitless mode and helium was the car-
rier gas. Samples were injected at 60 ?C (held for 1.5 min).
The oven temperature was programmed to 130 ?C at
20 ?C min?1, and then at 4 ?C min?1to 315 ?C, and held at
315 ?C for 90 min.
mass selectivedetector
2.2.3. Metastable reaction monitoring gas chromatography–
mass spectrometry (MRM-GC–MS)
Metastable reaction monitoring (MRM) was performed
at the Department of Earth, Atmospheric and Planetary Sci-
ences, Massachusetts Institute of Technology (Cambridge
MA, USA) following the method described in Brocks et al.
(2003). Saturated and unsaturated fractions of 23 of the
26 samples were analyzed by GC–MS MRM using a VG
Autospec Ultima-Q coupled to a CarloErba GC (8000 series)
with a tandem high resolution, double focusing, magnetic
sector-quadrupole mass analyzer. Internal standard D4
(d4-C29-aaa-ethyl cholestane; Chiron Laboratories AS)
was added to all samples prior to injection.
2.2.4. Comprehensive two-dimensional gas chromatography–
mass spectroscopy (GC?GC–MS)
S/u HC fractions were analyzed by GC?GC–MS using a
Leco Pegasus 4D system consisting of a Hewlett-Packard
6890 GC configured with a split injector, two chromato-
graphic columns, a liquid nitrogen-cooled pulsed jet mod-
ulator, and a time-of-flight (ToF) mass spectrometer.
Samples were dissolved in 25 lL cyclohexane and 2.0 lL
was injected into a 300 ?C split less injector (2 min purge
time). The first dimension separation was performed on a
nonpolar 5% phenyl polydimethylsiloxane phase (Agilent,
DB-5, 30.0-m; 0.25-mm ID; 0.25-lm film thickness) and
temperature programmed from 120 (1 min) to 320 ?C at
2 ?C min?1. The modulation column was deactivated col-
umn (1.0-m ? 0.10-mm ID) and temperature programmed
from 200 (1 min) to 400 ?C at 2 ?C min?1. The second
dimension separation was performed on a polar 50% phe-
nyl equivalent polysiloxane phase (BPX-50, SGE, 1.5-m;
0.10-mm ID; 0.1-lm film thickness) and temperature pro-
grammed from 120 (1 min) to 320 ?C at 2 ?C min?1. A
1.0-m ToF detector transfer line of 0.10-mm ID deactivated
column was used. Hydrogen was used as the carrier gas in
constant flow mode (1.2 mL/min). The GC?GC modulator
period was 8 s. The GC?GC was coupled with a Leco ToF
mass spectrometer that collected spectra from 45 to
600 lm at 100 Hz. A detector voltage of 1800 V was used
with a solvent delay of 11 min. The similar column array
used for GC?GC-FID provided equivalent compound
separation.
Comparisons of retention times and mass spectral
data collected GC–MS, GC-MRM–MS, and GC?GC–MS bio-
marker compound classes were used to identify the pres-
ence of common biomarker compound classes such as
n-alkanes, branched alkanes, acyclic isoprenoids, and poly-
cyclic terpanes, such as hopanes and steranes (Ventura
et al., 2007). C40acyclic irregular isoprenoids were identi-
fied by mass spectral analysis using GC?GC–MS and coin-
jection of an archaeal lipid standard containing acyclic,
mono-, bi-, and tricyclic biphytane. Cycloalkanes were ten-
tatively identified by mass spectral analysis. All reported
mass spectra were obtained using the ‘‘Peak True” decon-
volution algorithm of ChromaToF, the data processing tool
of Leco’s software package. The ‘‘Peak True” algorithm is
based on an approach by Biller and Biemann (1974). ‘‘Peak
True” mass spectra have greater fragmentation resolution
of closely eluting peaks and provides a more consistent
and tractable method than conventional caliper selected
approaches.
2.2.5. Preparation of the archeael biphytane standard
A standard mixture of archaeal biphytanes was pre-
pared from membrane phospholipids of the archaeon Ther-
moplasma acidophilum (Matreya L.L.C) using protocols
adapted from Pease et al. (1992) and DeLong et al.
G.T. Ventura et al./Organic Geochemistry 39 (2008) 846–867
849

Page 5

(1998). To remove the glyco- and phospho-lipid head-
groups, the membrane lipid was hydrolyzed in 1.25 M
HCl in MeOH at 100 ?C for 2 h. The reaction was neutral-
ized with 3 M NaOH and extracted three times with
DCM. After drying over NaSO4, an aliquot of the resultant
glycerol dialkyl glycerol tetraethers (GDGTs) was exam-
ined for purity using high pressure liquid chromatogra-
phy–massspectrometry (HPLC/MS)
Hopmans et al. (2000). To cleave the ether bonds of the
of GDGTs, the extract was refluxed in concentrated HI for
4 h at 100 ?C, extracted three times with hexane and dried
via rotary evaporation. To reduce the resultant alkyl io-
dides to hydrocarbons, the extract was dissolved in 2 M
LiAlH4in THF and refluxed under N2for 2 h at 70 ?C. Ethyl
acetate was added dropwise to quench excess LiAlH4, and
the reaction mixture was extracted with three additional
aliquots of ethyl acetate, which were dried over NaSO4.
The presence of biphytanes was confirmed by GC–MS
and comparison with spectra of DeLong et al. (1998).
as describedby
3. Results
All samples collected at the PGC have s/u HC fractions
with UCMs that can be grouped in three categories based
on chromatographic baseline topology (Fig. 1). Type I
UCMs are raised baselines from early- to mid-retention
times corresponding to the elution of C15–C25n-alkanes.
Type II UCMs are dramatically rising baselines at late-
retention times that span the range of C30–C42n-alkanes.
UCMs with a type III topology have a single massive UCM
of early- to late-retention time (n-C15–n-C42).
UCM topologies vary stratigraphically and geographi-
cally (Fig. 1). The s/u HC fractions of the four samples col-
lected from the Vipond Formation have a type III UCM
topology. The two samples collected from the Gold Center
Formation samples have a type II UCM topology, and the
Krist Formation sample has a type I UCM topology. All 10
Hoyle Formation samples collected away from areas of
gold mineralization have s/u HC fractions with a type I
UCM topology and all nine Hoyle Formation samples col-
lected in areas of gold mineralization have both type I
and II topologies.
GC–MS was unable to resolve the constituents forming
the three UCM topologies. Averaged mass spectra across
these UCMs display abundant mono- and bi-unsaturated
fragment ions, such as m/z 97 and m/z 95 (Fig. 2). Due to
the abundance of aliphatic compounds within the type I
topology the most intense fragment ion is m/z 57
(Fig. 2A). Mass chromatography of the fragment ions m/z
97, m/z 95 and m/z 93 illustrates the close elution of
compounds producing mono-, bi-, and tri-unsaturated
fragment ions upon electron ionization in all UCMs
(Fig. 2A–D). Hydrogenation of these samples does not
affect the UCM topology, suggesting the UCMs are domi-
nated by cyclic compounds.
3.1. Composition of type I UCMs
Type I UCMs are operationally defined as having a
raised ‘‘baseline” in the n-C15to n-C25range (Fig. 1). The
GC?GC–MS total ioncurrent (TIC)chromatogram
(Fig. 3A) displays some well resolved compounds (e.g. n-al-
kanes). However, the mass of cyclic compounds appears
unresolved. Because compounds forming type I UCM have
similar fragmentation patterns, monitoring of typical
mono- and bicyclic fragment ions (e.g. m/z 69, m/z 83,
m/z 97, and m/z 123) does little to resolve diversity of these
compounds. Visualization of individual peaks within the
UCM is greatly enhanced by sequentially monitoring mass
chromatograms of molecular ions for compounds with dif-
ferent numbers of rings and carbon atoms over the entire
elution range of the UCM (Fig. 3B and C). For example,
for the C16and C17compounds, mass chromatography of
m/z 224 and m/z 238 (monocyclic), m/z 222 and m/z 236
(bicyclic), m/z 220 and m/z 234 (tricyclic), and m/z 218
and m/z 232 (tetracyclic) shows four sub-parallel linea-
ment of peaks, and one peak for m/z 218 (tetracyclic;
Fig. 3B and C). When this approach is applied to molecular
ions of compounds from C15to C29(Fig. 3D) the bulk of the
UCM is observed to consist of near uniformly distributed
peaks of overlapping groups of mono- (C13–C29), bi- (C15–
C25), tri- (C16–C25), tetra- (C16–C23), penta- (C19–C21), and
hexacyclic (C22–C23) unsubstituted and alkyl-substituted
compounds (Fig. 3D). The tentative identification of some
of these compounds is provided below.
3.1.1. Monocyclic alkanes
The overall distribution of monocyclic alkanes is moni-
tored by the summed mass chromatography of fragment
ions m/z 96 and m/z 110 (Fig. 4A). However, to resolve
the distribution of monocyclic alkanes at each carbon
number, the mass chromatogram of the molecular ion
was added to the summed mass chromatogram. For exam-
ple, monitoring of C16alkylcycloalkanes was done on the
summed chromatogram m/z 224 + 96 + 110 (Fig. 4B).
n-Alkylcyclohexaneand
identified on the basis of mass spectral fragmentation
and by comparison with published spectra (e.g. Rubinstein
and Strausz, 1979; Hoffmann et al. 1987; Simoneit et al.
2004). n-Alkylcyclopentane and n-alkylcyclohexane homo-
logs having 16 or less carbon atoms nearly coelute
(Fig. 4B). Homologues with greater than 16 carbon atoms
become increasingly separated in the first dimension
(Fig. 4A). Tentativelyidentified
(Fig. 5A) elutes at slightly earlier first-dimension and later
second-dimension retention times relative to n-alkylcylo-
hexane and n-alkylcyclopentane (Fig. 4A and B). The
n-alkylcyclohexanes, n-alkylcyclopentanes, and n-alkylcy-
cloheptanes of the Vipond and Krist Formation sediment
extracts have extremely pronounced carbon number pre-
dominance. A similar carbon number predominance is
not observed in the substituted cycloalkanes contributing
to the formation of type I UCM.
A group of seven peaks, tentatively identified as isomers
of nonylmethylcyclohexane (Fig. 4B), were identified by
mass spectral comparison and by the similarity of the first
dimension isomers elution pattern to published data (Fow-
ler et al., 1986; Hoffmann et al., 1987). The mass spectra of
trans 1-methyl-2-nonylcyclohexane is shown in Fig. 5B.
The mass spectra of all observed alkylmethylcyclohexane
stereoisomers were nearly identical and dominated by
the fragment ion m/z 97. Alkylmethylcyclohexanes were
n-alkylcyclopentanewere
n-alkylcycloheptane
850
G.T. Ventura et al./Organic Geochemistry 39 (2008) 846–867

Page 6

observed in the C16–C28range, which extends past the zone
of maximum intensity of the UCM.
Three groups of monocyclic compounds eluting at
slightly earlier first and second dimension times than the
cluster of alkylmethylcyclohexanes (Fig. 4A and B) are also
present. The peaks in each group contain identical
fragmentation patterns suggesting the members of each
group have stereochemical differences similar to the
alkylmethylcyclohexanes. However, slight difference in
the fragmentation patterns between each group indicates
a variation in the number of alkyl substituents bonded to
the cyclohexane or cyclopentane ring. The first group
containing three peaks that elute before the alkylmethylcy-
clohexanesistentativelyidentifiedasisomersofalkylmeth-
ylcyclopentanes (Fig. 5). The mass spectra of these peaks
display a pattern of skeletal ring decomposition which
results in the formation of 1-alkene ions (McLafferty and
Turec ˇek, 1993) that is common to cycloalkanes with less
than six-membered rings. The dominant m/z 83 fragment
ion, characteristic of a six carbon ring, is likely due to the
presence of a methyl substituent on the pentacyclic ring.
The two other groups noted unknown X1and X2in Fig. 4B
could not be tentatively identified on the sole basis of their
mass spectral fragmentation (Fig. 5E and F), but are likely
compounds with two or more methyl substituents on a
cyclohexane or cyclopentane ring.
Fig. 2. Mass chromatograms of fragment ions m/z 99, m/z 97, m/z 95 and m/z 93 (left) and 10–50 min averaged mass spectra (right) for the UCMs of samples
(A) TI-117 and (B) OC-110m of the Hoyle Formation; (C) DP-2 of the Gold Center Formation, and (D) DM-1 of the Vipond Formation.
G.T. Ventura et al./Organic Geochemistry 39 (2008) 846–867
851

Page 7

3.1.2. Bicyclic alkanes
Bicyclic alkanes above C18 occupy part of the same
GC?GC chromatographic space as monocyclic alkanes that
have one less carbon atom (Fig. 3D). Monitoring of bicyclic
alkane distributions is thus facilitated by mass chromatog-
raphy of molecular ions. In the mass chromatogram of the
molecular ion of C16bicyclic alkanes (m/z 222; Fig. 6A),
three groups of peaks with very similar mass spectra are
identified.
The first group (B in Fig. 6A) includes at least four peaks.
The mass spectra of these compounds (Fig. 6B) include the
fragment ion m/z 137, which corresponds to a cleaved C10
Fig. 3. (A) GC?GC–MS TIC chromatogram of the Krist Formation sample DMSD (Type I UCM). (B) Partial, summed mass chromatogram of the fragment ions
m/z 224 + 222 + 220 + 218 showing C16mono-, bi-, tri-, and tetracyclic compounds. (C) Partial, summed mass chromatogram of the fragment ions m/z
238 + 236 + 234 + 232 showing C17mono-, bi-, tri-, and tetracyclic compounds. (D) GC?GC–MS TIC chromatogram from the s/u HC fraction of the Krist
Formation sample DMSD (Type I UCM). Overlapping lineaments indicate the elution range of configurational isomers of a given carbon number. Yellow,
black, red, pink, brown, and orange lineaments mark the positions of mono-, bi-, tri-, tetra-, penta-, and hexacyclic compounds, respectively, that contain
between 15 and 27 carbon atoms.
852
G.T. Ventura et al./Organic Geochemistry 39 (2008) 846–867

Page 8

bicyclic ion (decalin) from the C6alkyl substituent of the
molecular ion (m/z 220). The presence of a smaller frag-
ment ion m/z 136 indicates a hydrogen rearrangement,
formed when an alkyl substituent is lost via the cleavage
of the C–C bond a to the ring system. The base peak m/z
81 is formed by the cleavage of the cyclohexane rings of
the decalin ion a to the two tertiary carbon atoms of the
bicyclic system (e.g. Golovkina et al., 1984). Thus, it can
be hypothesized that the compounds forming the first
group are isomers of hexyldecalins. However, the position
of the alkyl substituent (1 or 2) cannot be determined and
it is possible that both isomers are present.
The second group (C in Fig. 6A), includes four peaks
with nearly identical mass spectra that are tentatively
identified as alkyl methyldecalins (Fig. 6C). Cleavage of a
C5alkyl substituent from the molecular ion results in the
formation of a C11bicyclic ion represented by the fragment
ion m/z 151. The base peak m/z 95 indicates fragmentation
a to the tertiary carbon atoms of the methyl-substituted
cyclohexyl ring. The position of both methyl and pentyl
substituents cannot be determined on the sole basis of
the mass spectral data.
The third group (D) includes three peaks with nearly
identical mass spectra (Fig. 6A). The mass spectra of these
compounds have an m/z 81 base peak likely formed by the
cleavage of a cyclohexane ring from a bicyclic system.
However, the mass spectral data is unspecific and no
structure can be proposed at this point.
3.1.3. Tricyclic compounds
Tricyclic compounds volumetrically contribute less to
type I UCMs than monocyclic and bicyclic compounds. Five
Fig. 4. (A) and (B) GC?GC–MS summed mass chromatogram of the fragment ion m/z 224 + 96 + 110 of the Krist Formation sample DMSD. n-Alkanes are
labeled with black dots, alkylcyclopentanes and alkylcyclohexanes are marked by red triangles and squares, respectively. Tentatively identified homologous
series of alkylethylcyclohexanes, alkyldimethylcyclopentanes, alkylmethylcyclopentanes, alkylmethylcyclohexanes, and n-alkycycloheptanes are marked
by yellow ellipses, black rectangles, red circles, larger white ellipses, small black boxes, respectively.
G.T. Ventura et al./Organic Geochemistry 39 (2008) 846–867
853

Page 9

groups of compounds are identified on the mass chromato-
gram of the molecular ion (m/z 220) of C16tricyclic alkanes
(Fig. 7A). All of the peaks within each group have nearly
identical mass spectra. The mass spectra of these com-
pounds (Fig. 7) do not match those of tricyclic terpanes
(e.g. they do not have a fragment ion m/z 123). Interpreta-
tion of the mass spectrometric data was facilitated by Kis-
elev et al’s. (1984) study of perhydrophenanthrene and
perhydroanthracene fragmentation. The mass spectrum
of peak E (Fig. 7E) has a fragment ion M+?29 (m/z 191)
of intensity equal to that of the molecular ion, suggesting
the loss of an ethyl substituent from one of the terminal
rings of the perhydroanthracene or perhydrophenantrene.
Otherwise, the mass spectrum of peak E is similar to those
published by Kiselev et al. (1984). Detail of the fragmenta-
tion pattern is shown in Fig. 7E. The fragment ion m/z 150
correspondstoa perhydro-cyclopropa[a]naphthalene.
Thus, peak E is tentatively identified as an ethyl-perhydro-
anthracene or ethyl-perhydrophenanthrene, though the
location of the ethyl substituent remains unclear.
The compounds of the first group (B) of Fig. 7A are
tentatively identified as isomers of dimethyl-substituted
perhydroanthracene or perhydrophenanthrene on the ba-
sis of mass spectral data (Fig. 7B). The fragment ions
M+?15 and M+?29 correspond to the loss of one and two
methyl substituents, respectively. The fragment ions m/z
149 and m/z 164 correspond to an octahydro-methyln-
aphathaleneandperhydro-cyclopropa[a]naphthalene,
respectively. The position of the methyl substituents, lo-
cated on terminal rings, could not be determined. The
group of peak C and peak D (Fig. 7A) are tentatively inter-
preted to be mono- or bi-substituted dodecahydrofluorene
(Fig. 7C, and D) on the basis of mass spectral fragmenta-
tion. No tentative structure for the group of peak F is
proposed.
3.1.4. Tetracyclic alkanes
Tetracyclic alkanes are also part of type I UCM (Fig. 8).
The mass chromatogram of the molecular ion of C16
tetracyclic compounds (m/z 218) displays one peak, tenta-
tively identified as hexadecahydropyrene by comparison of
its mass spectra with that of published reports (McLafferty
and Stauffer, 1989). The mass chromatogram for the m/z
232 molecular ion of C17tetracyclic compounds displays
five major peaks (Fig. 8A). Peaks B–D are tentatively iden-
tified as methyl-hexadecahydro-pyrenes on the basis of
Fig. 5. Mass spectra of tentatively identified C16(A) alkylcycloheptane, (B) 1-methylalkylcyclohexane, (C) the trans-2-alkylmethylcyclohexane (D) alkyl-
dimethylcyclopentane, and (E) and (F) unidentified monocyclic labeled unknown A and unknown B, respectively.
854
G.T. Ventura et al./Organic Geochemistry 39 (2008) 846–867

Page 10

mass spectral fragmentation patterns (Fig. 8B–D). These
peaks all have M+?15 fragment ions (m/z 217) suggesting
the presence of a methyl substituent. For peak B and C,
methyl substitution at C-1 or C-2 favors cleavage of the
C4H8(isobutylene) or C4H10(isobutyl) from the molecular
ion, resulting in the fragment ions m/z 175 and m/z 177,
respectively (Fig. 8B and C). The fragment ions m/z 189
and m/z 191 are formed by cleavage of the C3H8or C3H6
(propylene) from the ring opposed to that carrying the
methyl substituent. Thus, peaks B and C are tentatively
interpreted to be 1- or 2-methyl-hexadecahydro-pyrenes.
For peak D, methyl substitution at C-4 explains the absence
of the fragment ions m/z 175 and m/z 177 (no possible
cleavage of C4units) and the presences of the fragment
ions m/z 189 and m/z 191, via cleavage of a C3H8or C3H6
unit from the molecular ion (Fig. 8D). Thus peak D is tenta-
tively identified as a 4-methyl-hexadecahydro-pyrene. The
mass spectral fragmentation for peak E (Fig. 8E) cannot be
rationalized and no tentative structure is proposed.
There is no fragment ion M+?15 in the mass spectrum
of peak F (Fig. 8F), suggesting that this compound is not
methyl substituted. Thus, considering the abundance of
the fragment ions m/z 175 and m/z 191, we tentatively
identified peak F as hexadecahydro-benzo[de]anthracene,
the only structure possible.
3.1.5. Pentacyclic alkanes
Pentacyclic alkanes were tentatively identified in type I
UCM where they are in low abundance compared to mono-,
bi-, tri-, and tetracyclic alkanes. The distribution of C19and
C20pentacyclic compounds is monitored on the summed
mass chromatogram m/z 258 + 272 (Fig. 9A). Peaks B–D
are C19compounds (M+258). Peaks E–H are C20compounds
(M+272).
Octadecahydro-benzo[cd]pyrene is the only possible
structural isomer for an hexacyclic compound comprised
solely of six-membered rings having a mass of 258 amu
(Fig. 9B). The mass spectrum of peak B (Fig. 9B) contains
the fragment ions m/z 217 (M+?41) and m/z 215
(M+?43). These fragment ions correspond to tetradecahyd-
ropyrene and dodecahydropyrene ions formed by cleavage
of propylene and propane, respectively, from the fragment
ion (M+?258). Fragment ions m/z 173 and m/z 175 corre-
spond to octahydrophenalene and decahydrophenalene
ions formed by cleavage of saturated and unsaturated C3
units from dodecahydropyrene ions. Thus, on the basis of
mass spectral data, peak B is tentatively interpreted as
octadecahydro-benzo[cd]pyrene.
Peaks C and D must be pentacyclic compounds with at
least one cyclopentane ring. The mass spectrum of peak C
is weak and difficult to interpret and most probably
includes fragment ions derived from more than one
compound. However, the very high intensity of fragment
ion m/z 215 (M+?43; 100%) suggests presence of a methyl
substituent on the pentacyclic ring and cleavage of the
bond between the two adjacent tertiary carbon atoms
(Fig. 9C). Thus, very tentatively, peak C may be octadecahy-
dro-methylcyclopenta[cd]pyrene. The mass spectra of peak
D do not display an M+?15 fragment ion (Fig. 9D). The
abundance of the fragment ion m/z 173, corresponding to
Fig. 6. (A) GC?GC–MS partial mass chromatogram of the fragment ion m/z 222 of the Krist Formation sample DMSD, (B), (C), and (D) are mass spectra of
peaks B, C, and D, respectively, in the mass chromatogram. (C) and (D) are representative mass spectra of peaks tentatively identified as alkylmethyldecalin
and alkyldecalins.
G.T. Ventura et al./Organic Geochemistry 39 (2008) 846–867
855

Page 11

an octahydrophenalene ion, suggests the cleavage of two
rings and loss of two C3units. Thus, an octadedahydro-
cyclopenta[e]pyrene structure is proposed for peak D.
Peaks E and F, both C20pentacyclic hydrocarbons, have
very similar mass spectra (Fig. 9E and F). The fragment ion
m/z 257 (M+?15) suggests the presence of a methyl substi-
tuent on the pentacyclic ring system. The tetradecahydro-
pyrene ion (m/z 215) indicate the cleavage of a C4unit, sug-
gesting that the methyl substituent is located in position 3,
4 or 5 of an octadecahydro-benzo[cd]pyrene. Thus, peak E
andFaretentatively identified
methylbenzo[cd]pyrene isomers.
Peak G and H are both C20pentacyclic hydrocarbon. The
mass spectra of both these peaks do not display an M+?15
as octadecahydro-
Fig. 7. (A) GC?GC–MS partial mass chromatogram of the fragment ion m/z 220 of the Krist Formation sample DMSD showing the distribution of C14
tricyclic compounds. Letters (B), (C), (D), and (F) are mass spectra for peaks inscribed by white (B), red (C) and yellow (F) ellipses and peaks D and E,
respectively of the m/z 220 mass chromatogram. Possible fragmentation patterns are provided for (B) and (E) mass spectra. (For interpretation of the
references to color in this figure legend, the reader is referred to the web version of this article.)
856
G.T. Ventura et al./Organic Geochemistry 39 (2008) 846–867

Page 12

fragment ion. The mass spectral fragmentation patterns for
these compounds are typical of a condensed ring system
but are not informative enough to hypothesize a structure.
3.1.6. Hexacyclic alkanes
Hexacyclic alkanes were also detected in type I UCM.
Mass chromatography of the molecular ion for C22
hexacyclic compounds (m/z 298; Fig. 10A) has three peaks
with mass spectra including m/z 298 as a molecular ion. All
these mass spectra include the fragment ion m/z 257
(M+?41) corresponding to the typical loss of a propylene
fragment from a condensed ring system, as observed for
tetracyclic and pentacyclic alkanes. None of the spectra
display fragment ions with m/z M+?15 nor M+?29
(Fig. 10B) suggesting that these compounds are not methyl
substituted and, thus, do not include five-membered rings.
Only three structural isomers are possible for hexa-
cyclic C22compounds made of cyclohexane rings having
a mass 298 amu: icosahydro-dibenzo[cd,mn]pyrene,
icosahydro-dibenzo[def,mno]chrysene, and docosahydro-
benzo[ghi]perylene (Fig. 10C). However, it is not possible
to determine which of these three structures corresponds
Fig. 8. GC?GC–MS partial mass chromatogram of the fragment ion m/z 232 of the s/u HC fraction of the Krist Formation sample DMSD showing the
distribution of C17tetracyclic compounds. Letters marking the order of mass spectra correspond to the peaks labels of the mass chromatogram.
G.T. Ventura et al./Organic Geochemistry 39 (2008) 846–867
857

Page 13

to each peak in the mass chromatogram m/z 298, the mass
spectral data for peaks C and D (not shown) being too weak.
3.2. Composition of type II UCMs
GC?GC–MS enables identification of compounds form-
ing the bulk of type II UCM. These UCMs are composed of
acyclic and cyclic terpenoids with 1–5 rings, which form
bands separated in the second dimension of the GC?GC
chromatogram (Fig. 11). Each band is composed of one or
more series of compounds. For example, the tetracyclic
band include both steroids and secohopanes, which elute
in the second dimension after a pseudohomologous series
of tricyclic terpanes (cheilathanes) and before a later
eluting band pentacyclic triterpanes (hopanoids; Fig. 11).
The dominant compounds forming type II UCMs are
C36–C41acyclic, mono-, bi-, and tricyclic irregular (head-
to-head) isoprenoids. Acyclic 3,7,11,15,18,22,26,30-octa-
methyldotriacontane (C40H82; biphytane) and C40mono-,
bi-, and tricyclic archaeal lipids were identified on the
basis of mass spectral fragmentation and coelution with
an archaeal lipid standard (Fig. 12). Lower carbon number
derivatives of the acyclic and cyclic biphytanes were
identified on the basis of mass spectral data, following
the rational of the fragmentation of acylic and cyclic biphy-
tane (Figs. 12 and 13). GC?GC–MS analysis of sample OC-
114m enabled the resolution of additional acyclic irregular
isoprenoid with more than 40 carbon atoms (Figs. 11 and
12). No cyclic isoprenoids with more than 40 carbons were
detected and neither tetracyclic biphytane nor derivatives
Fig. 9. GC?GC–MS partial, summed mass chromatogram of the fragment ion m/z 258 + 272 of the Krist Formation sample DMSD showing the distribution
of C19and C20pentacyclic compounds. Letters marking the order of mass spectra correspond to the peaks labels of the mass chromatogram.
858
G.T. Ventura et al./Organic Geochemistry 39 (2008) 846–867

Page 14

of tetracyclic biphytanes were observed. The C40mono-,
bi-, and tricyclic biphytane derivatives and their lower
C36–C39homologs are each, respectively, represented by
doublets, triplets, and at least quadruplets of peaks with
exactly the same mass spectra (Fig. 10). Within each group,
the peaks are clearly separated in the first dimension, but
have nearly the same second-dimension retention time,
suggesting differences in volatility but similar polarity.
3.3. Composition of type III UCMs
The s/u HC fraction of all samples collected from the
Vipond Formation have a type III UCM topology (Fig. 1).
GC?GC–MS showed that type III UCMs are made of nearly
the same compounds as those observed for type I and II
UCMs (Fig. 14). GC?GC chromatograms are dominated
by overlapping bands of alkyl-substituted mono-, bi-,
and tricyclic compounds at early- to mid-retention times.
Although tetracyclic alkanes were not detected, minor
contributions of penta- and hexacyclic compounds were
also present. The compounds forming the mass of the late
eluting part of the UCM are dominated by hopanoids, ste-
roids, tricycic terpanes, secohopanes, and acyclic, mono-,
bi-, and tricyclic irregular isoprenoids (archaeal lipids)
with carbon number ranges similar to those observed
for compounds forming the type II UCM. Thus, the mas-
sive type III UCM appears to be a hybrid of type I and II
UCMs.
4. Discussion
4.1. Origin of PGC UCMs
The UCMs observed in the PGC samples are unlikely
to have arisen by selective preservation of bioresistant
compounds during biodegradation as ascribed to UCMs
of biodegraded oils (e.g. Killops and All-Juboori, 1990).
Easily biodegradable, low
n-alkanes and LMW acyclic isoprenoids, such as 2,6,10-
trimethyltridecane or 2,6,10-trimethylpentadecane (Peters
and Moldowan, 1993), are abundant in all samples. This is
concordant with the observed low concentrations of
biodegraded C-25 norhopanes (Ventura, 2006). Further-
more, biodegradation can explain neither the presence of
highly isomerized archaeal lipids observed in type II UCMs,
nor the abundance of mono-, bi-, tri-, tetra-, penta-, and
hexacyclic alkanes in type I UCMs.
molecularweight (LMW)
4.1.1. Origin of type I UCMs
Type I UCM forming compounds are not likely contribu-
tions from laboratory or field related contamination. Most
of these compounds do not have a carbon number prefer-
ence. Exceptions are n-alkylcyclopentanes, n-alkylcyclo-
hexanes,
n-alkylcycloheptanes
Vipond and Krist Formation, which have concentrations
that are positively correlated with BAQCs. These com-
pounds likely reflect contamination by products of com-
mercial polyethylene products (Grosjean and Logan,
2007; Brocks et al., 2008). Monocyclic alkanes are devoid
of a carbon number preference in samples that do not con-
tain BAQCs (Ventura et al., 2007). Type I UCM forming
compounds are also present in samples that had the sur-
face of the core ground away prior to extraction (Ventura
et al., 2007) and thus are unlikely to derive from surface
contamination of the sample. It is unclear, however,
whether these compounds share a common origin with
one another or with those forming type II UCMs.
Very little information is currently available about the
occurrence or formation of the tentatively identified cyclic
compounds. Payzant et al. (1979, 1980) showed that the
UCMs of extractable hydrocarbons from the Mannville oils,
Lloydminster heavy oil, Cold Lake Bitumen, Athabasca
bitumen, and Athabasca asphaltene pyrolysate are domi-
nated by mono- to hexacyclic hydrocarbons and identified
the dominant compounds as substituted and unsubstitut-
ed monocyclic hydrocarbon and decalin, as observed in
type I UCMs. Payzant et al. (1980) proposed that the
tricyclic component of the UCM of Cold Lake bitumen
is composed of alkyl perhydrophenanthrenes. No struc-
ture was proposed for the tetracyclic, pentacyclic and
of samples fromthe
Fig. 10. (A) GC?GC–MS partial mass chromatogram of the fragment ion
m/z 298 of the Krist Formation sample DMSD showing the distribution of
C22 hexacyclic compounds. (B) Mass spectra of peak B. (C) Tentative
structures of hexacyclic compound corresponding to peak C, B and D.
G.T. Ventura et al./Organic Geochemistry 39 (2008) 846–867
859

Page 15

hexacyclic component of UCMs in the Manville oils
(Payzant et al., 1980).
If the compounds forming type I UCM share a common
origin, several hypothetical scenarios may explain their
formation. These compounds appear unrelated to cyclic
terpenoids and are not known biosynthetic products. Their
presence may thus represent hydrogenation of polycyclic
aromatic hydrocarbons (PAHs). Although many reductive
techniques are known, they typically involve support from
noble metal catalysis under high hydrogen pressure and/or
elevated temperatures, either in the gas phase or in the
presence of organic solvents (Nelkenbaum et al., 2007).
Saturation typically leads to the formation of additional
isomers as observed by Yuan and Marshall (2005) who
used a mixture of supercritical CO2with molecular hydro-
gen to form six tetradecahydroantracene isomers from the
reduction of anthracene.
Alternatively, anaerobic degradation of PAHs has been
demonstrated to yield hydrogenated derivatives of PAHs
(Annweiler et al., 2002; Meckenstock et al., 2004). Denitri-
fication and sulfate reducing conditions can lead to the
transformation of naphthalene and phenanthrene to naph-
thoic and phenanthroic acids (Meckenstock et al., 2004).
Aitken et al. (2004) proposed a reductive pathway involv-
ing carboxylation of naphthalene to 2-naphthoic acid, fol-
lowed by several hydrogenation
account for the presence of decahydro-2-nahthtoic acid
in many biodegraded oils. Additional steps to the reductive
2-naphthoic acid pathway include cleavage of a C2-frag-
ment via beta-oxidation with another metabolite to yield
steps, whichmay
Fig. 11. (A) GC–MS TIC chromatogram of the bitumen from sample OC-114m (of the Hoyle Formation) containing a type I and II UCM. (B) GC?GC–MS TIC
chromatogram of the same sample with the labeled n-alkanes (black circles), mono-, bi-, tri-, tetra- (steranes), and pentacyclic (hopanes). Dotted line
indicates irregular isoprenoids. Boxes encapsulate monomethylalkanes. Open ovals enclose monoethylalkanes.
860
G.T. Ventura et al./Organic Geochemistry 39 (2008) 846–867