Joshua Stern Final

Sulfur-bound biomarkers of a Monterey shale
and a Greenland lake sediment
Joshua Gallant Stern
Submitted in partial fulfillment of the requirements for the degree of

Bachelor of Science in Geology-Biology at
Brown University.
May, 2009
Advisor: Prof. Yongsong Huang
i
Acknowledgements
I am especially grateful to Dr. Yongsong Huang for advising me on this senior
thesis. Thank you to Dr. James Russell and Dr. Timothy Herbert for reading this thesis
and providing helpful feedback. Thank you to Dr. Jan Tullis for helpful discussions and
writing advice.
I am indebted to Dr. Marcelo Alexandre for helpful discussions and for teaching
me laboratory techniques. Thank you to Mr. Rafael Tarozo for help with GC-MS analysis.
Dr. Alex L. Sessions and Dr. William J. D’Andrea generously supplied the rock
and sediment samples, as well as helpful advice. Thank you to Dr. Stefan Schouten for
helpful suggestions.
Thank you to Ms. Li Gao, Mr. Jonathan Nichols, and Ms. Jaime Toney for helpful
discussions and laboratory instruction. Thank you to Mr. Jeffrey Salacup and Mr. Daniel
Scheinerman for helpful advice.
Thank you to my Dad, Mom, Sister, Uncle, and friends for their love, support, and
draft reading.
I am grateful to NASA’s Rhode Island Space Grant Consortium for financial sup-
port during the summer of 2008.
ii
Abstract
Seeking to reconstruct the biogeochemical processes that produced organic
sulfur compounds in two unique depositional environments, we used the nickel boride
desulfurization reaction to release hydrocarbons from sulfur-bound macromolecules
not otherwise amenable to chromatographic analysis. We desulfurized two geochemi-
cal extracts: one sample is a Monterey shale of late Miocene age, and the other sample
is a surface sediment from the Greenland lake Brayasø. Both samples contained organic
sulfur compounds, but the Monterey shale was biologically and thermally modified after
deposition. A comparison of the free and sulfur-bound hydrocarbons from each sample
revealed a precursor-product relationship between tocopherol and pristane, for Monterey.
Greenland’s composition may indicate that photochemical sulfurization occurs in the
Brayasø oxic zone. We found that sulfurization may proceed at different rates for differ-
ent compound families; for example, we did not see any sulfurized alkenones in Brayasø,
but we found an abundance of sulfurized isoprenoids. Greenland’s relatively high overall
desulfurization yield suggests that sulfurization in Brayasø occurs in under 40 years. Our
Greenland findings suggest that photochemical sulfurization may be more widespread
than previously thought, and that sulfurization might not interfere with alkenone pale-
otemperature reconstructions.
iii
Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Section 1 endnotes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2. Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.2 Testing synthetic standards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.3 Organics extraction and fractionation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.4 Desulfurization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.5 GCFID/ GCMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.6 Yield quantification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.7 Experimental control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
3. Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.2 Standards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.3 Monterey desulfurization fraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
3.4 Monterey nonpolar fraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
3.5 Greenland desulfurization fraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
3.6 Greenland nonpolar fraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
3.7 Minor-nonpolar fractions and experimental control . . . . . . . . . . . . . . . . . . . . . 38
4. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
4.1 Standard yields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
4.2 Sample yields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
4.3 Sample composition overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
4.4 Monterey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
4.4.1 Post-depositional modification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
4.4.2 Paleobiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
4.4.3 Precursor-product relationships . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
4.5 Greenland . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
4.5.1 Biomarkers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
4.5.2 Precursor-product relationships . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
4.6 Sulfurization potential . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
5. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
iv
List of figures
Figure 1. Biomarker examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

Figure 2. Base-catalyzed nucleophilic addition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Figure 3. Light-induced free radical addition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
Figure 4. Desulfurization of 1-octadecanethiol and S-benzylthioglycolic acid . . . . . 5
Figure 5. Water column zonation for each depositional environment . . . . . . . . . . . . . 6
Figure 6. Hopanoid and isoprenoid fragmentation . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Figure 7. Overall workflow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Figure 8. Detailed workflow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Figure 9. HMB calibration curves for samples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
Figure 10. Additional Monterey and blank fractions . . . . . . . . . . . . . . . . . . . . . . . . . 20
Figure 11. HMB calibration curves for standards . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Figure 12. GC-FID chromatograms for standard desulfurization yields . . . . . . . . . . 26
Figure 13. Monterey desulfurization fraction, GC-FID trace . . . . . . . . . . . . . . . . . . . 27
Figure 14. Monterey desulfurization compound structures and names . . . . . . . . . . . 29
Figure 15. Monterey nonpolar fraction, GC-FID trace . . . . . . . . . . . . . . . . . . . . . . . . 30
Figure 16. Monterey nonpolar compound structures and names . . . . . . . . . . . . . . . . 32
Figure 17. Greenland desulfurization fraction, GC-FID trace . . . . . . . . . . . . . . . . . . 33
Figure 18. Greenland desulfurization compound structures and names . . . . . . . . . . . 35
Figure 19. Greenland nonpolar fraction, GC-FID trace . . . . . . . . . . . . . . . . . . . . . . . 36
Figure 20. Greenland nonpolar compound structures and names . . . . . . . . . . . . . . . . 37
Figure 21. Greenland and Monterey chromatographic data reconsidered . . . . . . . . . 40
Figure 22. Blank DS-0, GC-FID trace . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
Figure 23. Ion chromatograms showing an unconvoluted component . . . . . . . . . . . . 41
Figure 24. Ion chromatograms showing convoluted components . . . . . . . . . . . . . . . . 42
Figure 25. Comparison of compound-characteristic ions from Monterey DS-3 . . . . . 42
Figure 26. Baseline comparison between Monterey-0 and Monterey fractions . . . . . 43
Figure 27. Biodegradation of Monterey nonpolar fraction . . . . . . . . . . . . . . . . . . . . . 44
Figure 28. Greenland nonpolar and desulfurization GC-FID trace comparison . . . . . 44
Figure 29. Yields of standard desulfurizations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
Figure 30. Desulfurization yield of each sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
Figure 31. Model for organic sulfur compound formation in Brayasø . . . . . . . . . . . . 62
v
List of tables
Table 1. Desulfurizations of synthetic standards and reaction yields . . . . . . . . . . . . 24

Table 2. Mean sulfurization potentials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
Table 3. Normalizing the desulfurization yields . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
Table 4. Raw GC-FID measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
Table 5. Corrected GC-FID measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
Table 6. Mass calibrated . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
Table 7. Procedural blank subtracted . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
vi
The notation “E-n” refers to a section endnote, where n is the number of the endnote.
vii
viii
1. Introduction
We aim to reconstruct biogeochemical processes that produced sedimentary
organic sulfur compounds sampled from two unique depositional environments. Micro-
bial, weathering, and photochemical processes each have roles in converting sulfur
between its many forms, which range from the most oxidized sulfate to the most reduced
sulfide. The oxidation state of an organic sulfur compound (OSC) would fall somewhere
between the opposite states of sulfate and sulfide. OSCs are biomarkers that have been
geochemically sequestered using sulfide linkages (E-1). Biomarkers are the molecular
remains of algae and other organisms, which help reconstruct the history of Earth’s
ecology and climate. Sulfur-bound biomarkers can provide a fuller inventory of the
precursor biochemicals and their sources than free (non-sequestered) biomarkers provide
by themselves. This section introduces the diagenetic process of sulfurization, and several
of the mechanisms by which it operates. We also
introduce the analytical technique of desulfuriza-
tion. Later, we describe our experimental work on
Monterey shale and sediment from the Greenland
lake Brayasø, and we offer paleoenvironmental Steranes

R = H : cholestane; R = Me : ergostane;
R = Et : stigmasterane
and mechanistic explanations for the free and
sulfur-bound compounds in these samples.
Organic carbon can comprise as much as
17% of a sedimentary rock’s mass; however, this Hopanoid (Pentakishomohopane)
Figure 1. Some biomarkers whose precur-

material is 1% of rock mass on average (Katz sors composed lipid membranes or photo-
synthetic pigment. Structures reproduced
and Royle, 2001). Organic compounds are the from Brocks & Summons, 2004.
1
remains of once-living systems. Carbon is by far the most abundant constituent of this
material by mass; hydrogen and other bioelements are also present. Steranes, phytanes,
and hopanes exemplify the compounds found in sedimentary organic matter (Fig. 1).
As photosynthetic organisms exit the productive surface water and move down through
the water column to the sediment, most of them are intercepted and metabolized by
heterotrophs (E-2). Biomolecules vary in their ability to withstand decomposition. Mem-
brane lipids such as cholesterol and long-chain fatty acids tend to be the most durable.
On the other hand, nucleic acids and proteins are poorly preserved (E-3). The degree to
which molecular “fossils” escape remineralization (that is, microbial or thermal decom-
position) determines the ensuing rock’s fraction of total organic carbon (TOC). Particles
accumulating at the sediment-water interface gradually compact the underlying sediment
into rock, creating a net downward movement of the sediment with respect to the plane of
deposition. Sediment burial can help the deposited organics escape respiration by benthic
life.
Another preservative process that occurs during diagenesis is sulfurization. In
this geochemical reaction, reduced sulfur species attack the reactive functionalities on
biomolecules, yielding a compound in which one or more sulfur atoms comprise an intra-
or inter-molecular bridge (Fig. 2). The sulfides originate from sulfate-reducing bacteria,
which respire using sulfate as a terminal electron acceptor instead of oxygen (Werne et
al., 2004). The sulfate ion is abundant in the ocean and some lakes, but it is not reactive
under the mild conditions of the surface sediment (Aizenshtat et al., 1995). The avail-
ability of sulfides depends on the extent of bacterial sulfate reduction, the degree to which
pyrite formation competes with this process for the available sulfate, the flux of iron
2
1 R
H S Sx S
OH
S S
2 S β 3 Sx Sx
S S
Sx R α O
R O R O
S
HO H
4 S 5 O R
Sx (as above) S
O R
S Sx
R O S
R O
Figure 2. Base-catalyzed nucleophilic addition. OH- is the base, which deprotonates the polysulfide nucleophile (1). The
polysulfide reacts with the activated bond (2) to form a carbanion whose lone pair is delocalized (3). The carbanion inter-
mediate deprotonates a water molecule, yielding a polysulfidic organic compound, which can react with another phytenal
(4, 5). Based on Aizenshtat et al. (1995).
oxides to the water column, and the concentration of highly functionalized organic matter
(Werne et al., 2004; Russell and Werne, 2009). Sulfate-reducing bacteria are typically
anaerobic, so organic sulfur compounds are often thought to indicate anoxia. However,
some workers have found evidence of aerobic sulfate-reducing species (Amrani & Aizen-
shtat, 2004).
The abiotic chemical mechanisms by which sulfur incorporates into organic
matter is an area of active research (Werne et al., 2008). As we will argue, there are
two chief mechanisms that pertain to our locations of study. The first mechanism, base-
catalyzed nucleophilic addition, is well-established (Aizenshtat et al., 1995). The second
mechanism, light-induced free-radical addition, seems to be gaining popularity (Amrani
& Aizenshtat, 2004). The base-catalyzed mechanism is thought to occur in marine sedi-
ments, because the waters in these areas tend to be mildly basic (Aizenshtat et al., 1995).
The light-induced mechanism has been argued to occur in anoxic photic zones (Adam et
3
al., 1998). Figure 2 shows the nu-
hν R
1 H2S v=280nm
H SH
cleophilic addition mechanism, and

2 3 H SH 4 SH
figure 3 shows the mechanism for
free-radical addition. Several other

R R R
SH SH
SH
mechanisms not shown in these fig-
Figure 3. Light-induced free-radical addition, based on Vaughan and
Rust (1942). Light abstracts a sulfide radical (1), initiating a radical
ures may also be important to abiotic chain reaction that yields thiophytane (2-4). Adam et al. (1998)
suggest that a similar mechanism forms polysulfide radicals and
intermolecular sulfur.
sulfurization. Adam et al. (1998)
found that ketones catalyze the photochemical sulfurization reaction, probably because
they radicalize at a longer wavelength of light (Vaughan and Rust, 1942). Schneckenburg-
er et al. (1998) found evidence that a radical mechanism facilitates the reaction of sulfides
and ketones to form thioketones, which may be intermediates in OSC synthesis pathways.
Adam et al. (1998) found that photochemical sulfurization occurred in less than one day,
under laboratory conditions.
Natural sulfurization can occur rapidly, as shown by OSCs found in surface
sediments of Ace Lake, Antarctica, (Kok et al., 2000) and Lake Cadagno, Switzerland
(Putschew et al., 1995). Werne et al. (2004) point to evidence of both rapid sulfurization
(days), and less rapid sulfurization (thousands of years). Either way, sulfurization occurs
early in diagenesis, and multiple reactions of different rates probably happen simultane-
ously (Ibid.)
Sulfurized biomarkers can be important to paleoreconstructions, as Werne et al.
(2004) explain. Sulfurization can affect the distribution of free compounds. Despite the
rapid pace of sulfurization, not all biomolecules are sulfurized quantitatively, so their
descendents can occur as OSCs, as defunctionalized hydrocarbons, as both, or as neither.

4
Reconstructing the history of life at a particular depositional environment often involves
interpreting trends in the relative abundance of certain biomarkers over time. Similar
trends suggest similar sources. Therefore, failing to consider the sulfurized biomarkers
can produce an erroneous trend, or overlook a compound altogether. In some sediments,
sulfurized biomarkers compose the majority of the soluble organic matter (Schaeffer et
al., 1995). Sulfurization can preserve carbohydrates (Werne et al., 2004), a class of com-
pounds that would be metabolized rapidly outside of a macromolecular network.
Decoding OSCs requires instruments that isolate, identify and measure them. We
used two analytical instruments: the Gas Chromatograph- Mass Spectrometer (GC-MS)
and the Gas Chromatograph- Flame Ionization Detector (GC-FID) (E-4). Identifying and
measuring the abundance of macromolecular, polar organic sulfur compounds (OSCs)
with the GC-MS and -FID is difficult. OSCs are challenging to analyze because the S
atom is electronegative, and sulfurization often forms macromolecular, sulfide-linked
networks of multiple biomolecules that can contain other electronegative substituents.
Because OSCs tend to be heavy (of a large size) and polar, they often elute slowly or
not at all on a polar GC column. As a workaround, organic geochemists developed a
degradative chemical reaction that replaces sulfide bonds with hydrogen atoms (Schouten
et al., 1993). This “desulfurization” reaction allows us to release, measure and identify
potential biomarkers previously trapped as OSCs (Fig. 4).

MeOH/THF
I SH NiCl2 + NaBH4
50 °C, 1 hr, N2
OH OH
II S as above
O O
Figure 4. I. Desulfurization of 1-octadecanethiol yields octadecane. II. Desulfurization of S-ben-

zylthioglycolic acid yields toluene and acetic acid (acetic acid unconfirmed; see Results).
5
We desulfurized and analyzed synthetic standards and geochemical extracts
with two goals. We sought to optimize this technique’s experimental and analytical
methodology to facilitate future work in our laboratory. In addition, we aim to infer
features of our samples’ depositional environment. The first sample is from an outcrop
on the eastern end of Sulfur Mountain in Monterey, California. The outcrop is part of the
upper member of the Monterey Formation, which deposited somewhere between 6.7 and
7.8 million years ago, about 100 km off the late-Miocene coast. At that time, the area
of deposition was under 1000-1500 m of seawater (Fig. 5) (Isaacs, 2001). Our second
sample is surface sediment from the permanently stratified, oligosaline Greenland lake
Brayasø (Fig. 5) (0-1 cm below the sediment-water interface, at a water depth of 57 feet).
pH, oxygen (mg/l), temperature (°C) and

conductivity (µS cm-1/200)
0 5 10 15 20
0 Mixed layer
~50
~100 Photic zone
5 Mixed layer
10
Water Photic zone Water

depth (m) depth (m)
15
Anoxic zone
Sub-oxic zone
20
~1500 Sediment
pH Sediment
DO
Temp
25 SpCond
A) Greenland, Brayasø B) Monterey

Lacustrine Hemipelagic
Figure 5. (A) Water column zonation for the Greenland sediment’s depositional environment, inferred
by measurements of the lake taken in August 1997 (reproduced from Anderson et al., 1999). (B) Water
column zonation of the Monterey shale’s likely depositional environment, inferred from Isaacs, 2001.
6
The lake sediment is less than 910 years old, and probably less than 40 years old (E-5).
Our blank sample is sand baked overnight at 500 °C.
Section 1 endnotes
1. Not all organic sulfur compounds are formed through a geochemical reaction.
Some OSCs are formed through “assimilatory sulfate reduction”, that is, cellular sulfate
uptake and biosynthetic utilization. Measurements of the 34S depletion in sedimentary
OSCs indicate that 20-25% of marine sedimentary OSC mass is biogenic (Werne et al.,
2004).
2. Most marine life is too small and too buoyant to fall from the surface ocean to
the abyss. However, ocean circulation exports this material (dissolved organic matter and
small particulate organic matter) to depth. Additionally, calcite and opal shells aggregate
buoyant organic matter and drag it to the ocean floor (Sarmiento and Gruber, 2006).
3. Barring favorable circumstances such as a deposition environment that has been
cold for an unusually long time, DNA should degrade about 10,000 years after deposition
(D’Andrea et al., 2006). Nucleic acids and proteins are made with phosphodiester and
amide bonds, respectively. Phosphorus and nitrogen are biolimiting nutrients; their
terrestrial paucity makes them highly coveted, so any biopolymer that they are part of is
vulnerable to microbial attack. Moreover, phosphodiester and amide bonds are vulnerable
to hydrolytic decomposition, so even if microbes do not eat them, these compounds tend
7
to decompose rapidly on geologic timescales (Bada, 1991). Carbohydrates are another
readily metabolized group of compounds not expected to preserve well (Werne et al,
2004).
4. Chromatography is a laboratory technique that takes advantage of a com-
pound’s unique chemical properties in order to physically isolate it. Compound isolation
is crucial if we wish to understand what makes up a mixture of hundreds or thousands of
unidentified compounds from a geochemical extract. Gas Chromatography (GC) sepa-
rates compounds based on size and polarity. The machine injects an aliquot of sample
at the beginning of a column, and, over the course of about 40 minutes, increases the
temperature from about 40°C to about 315°C. Small, nonpolar molecules such as hexane
travel through the column the fastest, eluting at the beginning of the run. Large, polar
molecules such as functionalized cholesterol travel through the column the slowest, elut-
ing at the end of the run. The column is a coiled tube ~60 m in length and ~0.5 mm in
diameter. The column’s inner wall is coated with a polar stationary phase, which attracts
polar compounds from the sample and acts in concert with the gradually rising tempera-
ture to ensure that polar compounds have longer retention (elution) times.
GC is mostly used for fine separation of individual compounds (e.g., nonane,
C9H20, from decane, C10H22), whereas liquid chromatography is used for both the gross
fractionation of broad groups of compounds (e.g., polar versus nonpolar), and fine
separation. Liquid chromatography (LC) and thin layer chromatography (TLC), like
gas chromatography, exploit the fact that different compounds have different polarities.
However, LC and TLC do not use a temperature gradient, and the mobile phase is liquid
8
as opposed to gas. For this report we used gas
chromatography to analyze, identify and measure
individual compounds obtained from sedimentary
extracts that we prepared by separating the extracts
into polar and nonpolar fractions with liquid
chromatography. Figure 6. Top, hopanoid fragmentation.

Bottom, regular isoprenoid fragmenta-
tion. Drawings reproduced from Peters
We can couple a Gas Chromatograph to a et al. (2007).
Mass Spectrometer (MS), so that once a compound
elutes from the GC column, it passes into the MS for further identification. The MS
outputs a spectrum several times per second showing the mass distribution (ion intensity
versus ion mass) for whatever compounds are entering the MS. The mass distribution
helps identify the compound. In our particular setup, the mass spectra are patterns of ion
fragmentation. The instrument breaks the molecule into several pieces, and since each
compound has a unique and predictable fragmentation pattern, we deduce the molecular
structure by comparing the MS detector’s output to a library of fragmentation patterns
for known compounds. (Other types of mass spectrometers do not fragment the ions and
instead provide exact mass data.)
Ion chromatograms are another way to infer the identity of an unknown
compound. In contrast to a mass spectrum, which shows the distribution of ionized
masses at a certain time, an ion chromatogram shows the abundance of a single ion over
the course of the run. The total ion chromatogram (TIC) shows how the sum of all the
ion signals changes over the course of the run. Chromatograms of one or a few ions are
9
useful for detecting common biomarker ion fragments, such as m/z 191, for hopanes, or
m/z 183, for regular isoprenoids (Fig. 6).
To obtain accurate abundance data for hydrocarbon compounds, we use a Flame
Ionization Detector (FID), which burns compounds as they elute off of a GC column and
detects the resulting CO2 gas. Because different compounds have different ionization
efficiencies, using a GC-MS to find a compound’s total ion intensity is an inaccurate way
to compare the abundances of unidentified compounds from diverse samples. Moreover,
the FID is better equipped to analyze large numbers of samples and samples with high
compound abundances.
5. The 40 year estimate is an interpolation based on the radiocarbon date of 910
years for Brayasø sediment between 22.4 cm and 22.5 cm depth (Anderson and Leng,
2004). We assume that the lake’s sedimentation rate has been constant over the past
~910 years, and that the sediment-water interface was unperturbed during this time.
Assuming no perturbation is reasonable since the sediments are laminated and the lake is
meromictic (Anderson et al., 1999).
10
2. Method
2.1. Overview
We used the nickel boride desulfurization reaction to release sulfur-bound
biomarkers from the organic extracts of sedimentary rock samples. We prepared the
organic extract for desulfurization using an Accelerated Solvent Extractor (ASE) and
liquid column chromatography. After the desulfurization reaction, we isolated the
nonpolar yield with another column fractionation. We measured and identified the most
abundant compounds in the nonpolar yield using GC-FID and GC-MS. To optimize the
efficiency of the desulfurization reaction, we performed desulfurizations of synthetic
standards under a variety of conditions.
2.2 Testing synthetic standards
In order to measure the yield of the desulfurization reaction and confirm
its efficacy, we desulfurized the synthetic standards 1-octadecanethiol and
S-Benzylthioglycolic acid (Sigma-Aldrich). For each experiment, we dissolved 1-20 mg
of standard in a 2-4 mL solution of methanol:tetrahydrofuran 1:1, and followed one of
our four distinct desulfurization procedures, which we explain below.
2.3 Organics Extraction and Fractionation
We obtained a Total Organic Extract (TOE) for the Monterey and blank samples
by using an Accelerated Solvent Extractor (ASE) (E-1). We obtained the Greenland TOE
by extracting 1.56 g loose sediment in 15 mL DCM:MeOH 9:1 under ultrasonication at
11
Fractionation #1 Fractionation #2 Fractionation #3
Monterey Desulfurization 1
}
(DS-1)
DS-1
Greenland DS-1
Blank DS-1
Monterey Total Organic Extract
}
(TOE) (45 mg) Monterey Most-Polar Monterey DS-2
Greenland Most-Polar DS-2 Greenland DS-2
Greenland TOE (41.7 mg)
Blank Most-Polar
Blank DS-2
Blank TOE (0 mg)
Monterey Nonpolar Monterey DS-3

DS-3
Greenland Nonpolar Greenland DS-3
Blank Nonpolar Blank DS-3
Monterey Minor-Nonpolar
Greenland Minor-Nonpolar
Blank Minor-Nonpolar
Figure 7. Overall workflow. Red circles indicate chromatographic separations. We do not show the “Polar” fractions
resulting from Fractionation #1. For more details on the chromatographic separations (fractionations), see Figure 7. See E-7
for additional Monterey and Blank fractions not shown in this schematic.
~30°C for 30 minutes (E-2). The Extractable Organic Matter (EOM), as a percentage of
the rock mass, was 1.3% for Monterey and 2.7% for Greenland (E-3).
Monterey Desulfurization-0
}
In order to isolate the organic sulfur compounds
Monterey Most-Polar-0 andDS-0
remove free hydrocarbons,
}
Monterey Total Organic Extract (DS-0)

(TOE) (7.5 mg) Blank Most-Polar-0
Blank DS-0
Blank TOE (0 mg)
we fractionated each total organic extract (TOE) into a nonpolar and polar fraction on
Monterey Nonpolar 0
an alumina gel (Al2O3)Blank

column, using hexane:dichloromethane 9:1 (Hex:DCM 9:1) for
Nonpolar 0
Figure 10 Monterey Minor-Nonpolar 0

the nonpolar fraction and dichloromethane:methanol
Blank Minor-Nonpolar 0 1:1 (DCM:MeOH 1:1) for the polar
fraction. (Fractionation #1, indicated on Fig. 7 and Fig. 8).
Unfamiliar with the practice of asphaltene precipitation, we did not perform it
before the first column fractionation. Omitting this step led us to modify our column
fractionation procedure. This modification had the unintended consequence of
introducing nonpolar cross-contaminants into polar fractions. We controlled for these
cross-contaminants by performing a second fractionation on the polar fraction. Our
second fractionation resulted in a “Most-Polar” fraction and a “Minor-Nonpolar” fraction
(Fractionation #2, indicated on Fig. 7 and Fig. 8). For a detailed explanation of asphaltene
12
Total Organic Extract
Hex/DCM 9/1 I II
Al O Nonpolar GC-FID GC-MS
Fractionation #1 2
DCM/MeOH 1/1
3
Fractionation #2
Polar
Fractionation #3 III IV
Al O
2 3
Hex/DCM 9/1 Minor-Nonpolar GC-FID GC-MS
DCM/MeOH 1/1
Most-Polar
Figure 8. Detailed workflow.
Roman numerals refer to a V
re-dissolution or injection param-
eter, which varied depending on Desulfurization (DS) VI
the experiment. See the table in
E-13 for the parameters we used Hex/DCM 9/1 VII VIII
for each experiment. In this figure
Al O
2 3 DS Nonpolar GC-FID GC-MS
we do not show divisions of the
Most-Polar fraction into separate
aliquots. DS Polar
Discard
precipitation, our modification of the column fractionation procedure, and the resulting
non-polar cross-contamination, see Endnote 4.
2.4 Desulfurization
We had three different TOEs (Monterey, Greenland, and Blank). For each TOE
aliquot we produced one Most-Polar fraction (Figs. 7, 8). For the majority of our experi-
ments, we split the Most-Polar fraction into three aliquots and desulfurized each aliquot
(Fig. 7). For a few of our experiments, we desulfurized the entire Most-Polar fraction at
once (E-7). After the desulfurization reaction, we isolated the nonpolar yield with another
column fractionation (Fractionation #3, indicated on Fig. 7 and Fig. 8).
We produced each desulfurized fraction using one of four distinct desulfurization
procedures. We designed these procedures based on the descriptions in Schaeffer et al.
(1995), Schouten et al. (1993), and Schouten (personal communication, 2008).
13
The four methods were very similar to one another. To prevent water from degrad-
ing the reagent sodium borohydride (NaBH4), we prepared the reaction solvents, metha-
nol (MeOH) and tetrahydrofuran (THF), by drying with Na2SO4 for 15 min or overnight
with 5A molecular sieves. We dissolved 1-20 mg of sample or standard with 2-4 mL
THF:MeOH 1:1 in a test tube with a magnetic stir bar. To the dissolved sample we added
10-100 mg nickel chloride (NiCl2), and (slowly) a roughly equal amount of NaBH4. To
hasten the desulfurization reaction, we heated the mixture to 50-70°C. We used a gentle
nitrogen stream to isolate the reaction from water vapor. After 1 hour, we used Al2O3
isolation of the nonpolar fraction to extract from the reaction mixture any hydrocarbons
released by desulfurization (Fractionation #3, indicated on Fig. 7 and Fig. 8) (E-5).
Three of the desulfurization methods we used are called DS-1, DS-2, and DS-3;
we performed each of these methods on separate aliquots of all three extracts (Fig. 7).
For DS-1, we attempted to reflux the reaction mixture at ~70° C in a long test tube (E-6),
and used MeOH dried with Na2SO4 for 15 min. For DS-2, we used a THF:MeOH mixture
that we had dried overnight with 5A mole sieves. For DS-3, we used the 5A mole dried
solvents and also added ~100 mg additional sodium borohydride 30 minutes after the re-
action start. The fourth desulfurization method we call DS-0, and we performed it only on
aliquots of the Monterey TOE and the Blank TOE (E-7). In DS-0, we heated the reaction
to 50 °C, and used MeOH dried with Na2SO4 for 15 min.
We used some of our blank fractions to test standard desulfurization efficiency, in
addition to using them as procedural controls (E-10).
14
2.5. GC-FID/ GC-MS
In order to determine each sample’s abundance of volatile compounds, we evapo-
rated the nonpolar fractions, redissolved them, and transferred them to 2 mL vials for
GC-FID analysis. We optimized the volume of solvent in the GC vials, as well as the
injector split mode, to obtain peak heights between 20 and 500 pA (E-8). To identify the
compounds in these samples, we performed GC-MS analysis on most of the samples.
Endnotes 11 and 12 describe the GC-FID and GC-MS temperature programs (labelled
with Greek letters in other parts of this report). We determined the compounds present in
each fraction by searching the NIST 2005 mass spectra library with the spectra from our
data.
2.6. Yield Quantification

0.3
To quantify the absolute
desulfurization yield (the mass of

0.25
volatile organic carbon isolated from 0.2
the reaction mixture), we measured 0.15
a hexamethylbenzene (HMB) stan- Back

Front
0.1
Assumed Front
dard at various known concentra- -5

y = 5.8548 x 10 x
0.05 -5
tions with the GC-FID. We observe y = 4.2594 x 10 x
a linear relationship between HMB

0
0 1000 2000 3000 4000 5000 6000 7000
pA*s
Figure 9. Hexamethylbenzene (HMB) calibration curves. Our
mass on column and peak area (Fig- GC-FID has two columns and two detectors (called “Back” and
“Front”), so we collected standard data for each column-detec-
ure 9). For standard concentrations tor setup. The “Assumed Front” point is in lieu of missing data,
and we base it on the typical ratio for these two detectors, for a
particular mass loading, which we observed over the course of
higher than the ones that we mea- our experiments.
15
sured, Krupcík et al. (2004) report a linear FID response to increasing mass on column.
We used the figure 9 curves to quantify the sample desulfurization yield, and we used
similar curves to quantify the standard desulfurization yield (E-9).
2.7. Experimental Control
How will we know that the compounds we detect after desulfurization were
actually released through desulfurization? We performed the desulfurization on the
polar fraction of the total organic extract (TOE) because OSCs are most abundant in
the polar fraction. We expected the desulfurization of this polar fraction to release
nonpolar hydrocarbons. Isolating the nonpolar fraction from the reaction mixture
after desulfurization should therefore isolate only nonpolar hydrocarbons released by
desulfurization. However, we were concerned that the polar fraction contained nonpolar
cross-contaminants before the desulfurization. If that were the case, we would have
difficulty distinguishing contaminant compounds from desulfurized compounds.
Our sample preparation procedure used alumina-gel column fractionation
to separate nonpolar free compounds from polar sulfur-bound compounds. Initially,
we expected this separation to be complete. However, preliminary GC-FID traces of
desulfurized fractions bore an uncanny resemblance to the GC-FID traces of the nonpolar
fractions. This puzzling observation led us to modify our original procedure so that we
could test the hypothesis that our fractionations were incomplete. We introduced a second
alumina-gel column fractionation, shown in Figures 7 and 8, which produced “Minor-
Nonpolar” fractions.
16
If we were to observe that the Minor-Nonpolar fraction contained similar
compounds to those in the Nonpolar fraction, then we would conclude that the first
fractionation was incomplete. We would also suspect that the second fractionation was
incomplete, since it followed the same method as the first fractionation. Even after two
purifications of the polar fraction, we would expect to find a small remainder of nonpolar
cross-contaminants in the Most-Polar fraction. If these contaminants were present in
the desulfurization fraction, they would be difficult to distinguish from desulfurized
hydrocarbons - unless we could deduce a released compound’s authenticity using its
retention time, abundance, or mass spectrum.
Section 2 Endnotes
1. The ASE pumps solvent (DCM:Methanol 9:1) into a vessel containing the
crushed rock sample (10.24 grams). The vessel is next heated to ~100°C and pressurized
to several atmospheres (double-check). After about 1 hour, the vessel depressurizes and
the solvent flows out of the vessel, through a glass wool filter, and into a collection vial.
2. After removing the supernatant, we performed an additional series of 3x 3mL
extractions in DCM:MeOH 9:1.
3. We determined the % EOM by weighing the dried rock powder in a tared
sample bag, and weighing the dried total organic extract in a tared 4 mL vial.
17
4. The purpose of asphaltene precipitation is to separate compounds that are
not amenable to GC analysis (asphaltenes) from compounds that are more likely to be
amenable to GC analysis (maltenes). Asphaltenes have high molecular weights, are
highly functionalized, and are heavily sulfide-linked. Maltenes are less funtionalized,
lightly sulfide-linked, and often have lower molecular weights (Kohnen et al., 1991).
Maltenes are soluble in a light hydrocarbon solvent such as heptane or hexane, in addition
to more powerful (and more polar) solvents such as dichloromethane (DCM) or methanol
(MeOH). On the other hand, asphaltenes are only soluble in DCM or MeOH. Asphaltenes
are nonvolatile (Sessions, pers. comm., 2009). Injecting nonvolatile compounds into a
Gas Chromatography (GC) system will lead to a residual buildup on the inlet liner at the
beginning of the column, eventually interfering with data quality.
Since maltenes and asphaltenes alike contain organic sulfur compounds (OSCs)
(Sinninghe Damsté et al., 1988), we did not attempt to separate them before our first
fractionation. We were indifferent to the possibility of asphaltenes eluting with our polar
fraction, and thought that they could provide more material to desulfurize. Since we
would only analyze the nonpolar yield of the desulfurization, we would not run the risk of
dirtying the equipment with asphaltenes.
Modifying the usual alumina-gel column fractionation procedure, we loaded the
column twice, not once, as follows: we washed the column with 4 mL DCM:MeOH 1:1
followed by 4 mL Hexane:DCM 9:1. We did a 3x 200 ul extraction in Hex:DCM 9:1 of
the dry TOE, loaded the extract on to the column, and eluted it with 3.5 mL Hex:DCM
9:1 to obtain the nonpolar fraction. Then, rather than immediately eluting the polar
fraction with DCM:MeOH, we performed a second 3x 200 ul extraction of the remaining

18
TOE, using DCM:MeOH 1:1. We loaded the extract on to the column, and then eluted it
with 3.5 mL DCM:MeOH 1:1 to obtain the polar fraction (Figs. 7, 8).
We loaded the column a second time because the first extractions with Hex/DCM
9/1 did not completely dissolve the TOE, and they left a lot of material in the sample vial.
We assumed that this residual material contained polar organic sulfur compounds, which
we wanted to desulfurize. We dissolved this (mostly asphaltene) residuum in DCM/
MeOH and loaded it on to the column.
During these experiments, we did not realize that the asphaltene-rich residue
probably contained a small amount of nonpolar (maltene) compounds. Since the first
extractions for column loading did not dissolve the entire dry TOE, they probably left
behind a small amount of nonpolar (maltene) compounds in the asphaltene-rich residue.
When we performed the second set of extractions using DCM:MeOH 1:1, we loaded a
solution rich in polar asphaltenes and tinged with nonpolar maltenes on to a column that
already contained polar maltenes. The second extraction unintentionally loaded a small
amount of the nonpolar compounds (cross-contaminants), which eluted with the polar
fraction.
5. We changed our procedure for isolating the reaction yield because someone
discarded our centrifuge. For the Monterey-0, Blank-0, and standard desulfurizations
A-D, we centrifuged the reaction test tube, transferred the supernatant to a 4 mL vial,
and then performed 2x 2 mL DCM:MeOH 1:1 extractions on the solids remaining in
the test tube, centrifuging as needed. We obtained the nonpolar yield fraction from
this extract by Al2O3 fractionation. For all other desulfurizations (after the centrifuge
19
became unavailable), we waited ~10 minutes for the nickel boride particles to settle, and
transferred the supernatant to a 4 mL vial using a Pasteur pipet. We performed 2x 2 mL
DCM extractions on the solids remaining in the test tube, allowing time for the particles
to settle before supernatant transfer. We dried the yield extract in the 4 mL vial, and then
loaded its nonpolar fraction on to a washed Al2O3 column with 3x 300 ul Hex:DCM 9:1
extractions. We eluted this

Fractionation #1 fraction with 3 mL
Fractionation #2 Hex:DCM 9:1.Fractionation #3
Monterey Desulfurization 1
}
(DS-1)
DS-1
Greenland DS-1
6. However, the solvent dried out within 15 minutes because of either too high Blank DS-1
Monterey Total Organic Extract

a temperature or too strong a nitrogen flow. Monterey
We added additional solvent, lowered the
}
(TOE) (45 mg) Most-Polar Monterey DS-2

Greenland Most-Polar DS-2 Greenland DS-2
Greenland TOE (41.7 mg)
Blank Most-Polar
Blank DS-2
Blank TOE (0 mg)
temperature to 60°C (below reflux), and reduced the N2 stream.
Monterey Nonpolar Monterey DS-3
DS-3
Greenland Nonpolar Greenland DS-3
Blank Nonpolar Blank DS-3
7. We show in Fig. 10 additional Monterey and Blank fractions (not shown in

Monterey Minor-Nonpolar
Greenland Minor-Nonpolar
Figure 7). We desulfurized the most-polar fractions of these aliquots using the procedure
Blank Minor-Nonpolar
Figure 7. Overall workflow. Red circles indicate chromatographic separations. We do not show the “Polar” fractions
resulting from Fractionation #1. For more details on the chromatographic separations (fractionations), see Figure 7. See E-7
DS-0
for (described
additional in the
Monterey and Blanktext).
fractions not shown in this schematic.
Monterey Desulfurization-0
}
}
Monterey Total Organic Extract (DS-0)

Monterey Most-Polar-0 DS-0 Blank DS-0
(TOE) (7.5 mg) Blank Most-Polar-0
Blank TOE (0 mg)
Monterey Nonpolar 0
Blank Nonpolar 0
Figure 10 Monterey Minor-Nonpolar 0

Blank Minor-Nonpolar 0
8. Samples with FID peaks lower than 20 pA may be too dilute for the GC-MS’
sensitivity, while peaks taller than 500 pA can accumulate on the MS source and eventu-
20
ally interfere with the instrument’s 1
function. For fractions with FID

0.8
chromatograms showing heights
outside this range, we concentrated 0.6
or diluted the sample as appropri-
ate.
0.4
y = 0.00045852x
Concentration in 1mL solution (mg/mL) y = 0.00038572x
0.2 Back
Front
9. We made two different
Assumed Front
sets of calibration curves, and 0

0 500 1000 1500
pA*s
2000 2500
Figure 11. Calibration curves for the HMB standard,

applied each to a different part of used to quantify the standard desulfurizations. “Assumed
front” point as explained in figure 8.
our data. We used the calibration
curves in figure 9 to quantify the desulfurization yield for our geochemical extracts,
because we performed all of the GC-FID analyses for these data (samples and serial
HMB dilutions) within the same 24 hours. We used the calibration curves shown in figure
11 to quantify the desulfurization yield for our standards, because we performed our
standard reactions over the course of ~2 months, and we also measured HMB at different
concentrations during this time period (6/12/08 to 8/24/08). The HMB measurements for
the standards were made in 10:1 split mode, and the measurements for the samples were
made in splitless mode.
10. Our blank extracts served both as a procedural control and as a test of standard
desulfurization efficiency. Before dividing the Blank Most-Polar fraction into three
aliquots, we added 3 mg 1-octadecanethiol and 3 mg S-Benzylthioglycolic acid to the 4.5

21
mL solution. On the other hand, our blank-0 extracts served only as procedural controls
(E-7), since we did not add any synthetic standard to these fractions.
11. GC-FID Methods

Program a Program b Program g- Splitless;
10:1 Split 10:1 Split Program d- 10:1 Split
time, min temp °C time, min temp time, min temp
0 40 0 40 0 40
1 40 1 40 1 40
19.33 315 25 100 28.5 315
24.33 315 46.5 315 38.5 315
56.5 315
12. GC-MS Methods

Program e Program z Program h- 10:1 Split
Splitless Splitless Program q- 20:1 Split
time, min temp time, min temp time, min temp
0 40 0 40 0 40
1 40 1 40 1 40
28.5 315 37.67 315 28.5 315
38.5 315 47.67 315 38.5 315
22
13. Reference table for desulfurization and GC injection parameters. Much of the infor-
mation in this table is explained throughout the text as necessary. Greek letters refer to GC-FID
or GC-MS temperature programs, described in E-11 and E-12.
Sample Step of Workflow
I II III IV V VI VII VIII
Nonpol FID Nonpol MS Min-Nonpol FID Min-Non- Division of Most-Polar Desulf reaction Desulf FID Desulf MS
pol MS conditions
Monterey Redissolve in (No GC-MS Redissolve in (No Take 1.5 mL aliquot Heat reaction to Redissolve in 50 Redissolve in
DS-1 50 uL Hex/ data) 50 uL Hex/ GC-MS (One-third of the ini- 60-70 °C; use uL Hex/DCM 10 uL Hex/
DCM 9/1; DCM 9/1; data) tial 4.5 mL Most-Polar MeOH dried 9/1; Inject 1uL DCM 9/1;
Inject 1uL Inject 1uL solution) with Na2SO4. splitless (SL) (g) Manualinject
splitless (SL) splitless (SL) 1uL SL (e)
(g) (g)
Green- Redissolve Inject 1uL Redissolve (No Take 1.5 mL aliquot Heat reaction to Redissolve in 1 Redissolve in
land in 1 mL Hex/ 10:1 split in 1 mL Hex/ GC-MS (One-third of the ini- 60-70 °C; use mL Hex/DCM 100 uL Hex/
DS-1 DCM 9/1; (h) DCM 9/1; data) tial 4.5 mL Most-Polar MeOH dried 9/1; Inject 1uL DCM 9/1;
Inject 1uL Inject 1uL solution) with Na2SO4. SL (g) Inject 1 uL
SL (g) SL (g) SL (e)
Blank Redissolve in (No GC-MS Redissolve in (No Take 1.5 mL aliquot Heat reaction to Redissolve in 1 (No GC-MS
DS-1 50 uL Hex/ data) 50 uL Hex/ GC-MS (One-third of the ini- 60-70 °C; use mL Hex/DCM data)
DCM 9/1; DCM 9/1; data) tial 4.5 mL Most-Polar MeOH dried 9/1; Inject 1uL
Inject 1uL Inject 1uL solution) (E-10) with Na2SO4. 10:1 split (d)
SL (g) SL (g)
Monterey (uses same (No GC-MS (same Minor- (No Take 1.5 mL aliquot ~55 °C; Redissolve in 50 Redissolve in
DS-2 Nonpolar data) Nonpolar GC-MS (One-third of the ini- solvents dried uL Hex/DCM 10 uL Hex/
FID data as FID data as data) tial 4.5 mL Most-Polar overnight with 9/1; Inject 1uL DCM 9/1;
Monterey Monterey solution) 5A molecular SL (g) Manualinject
DS-1) DS-1) sieves. 1uL SL (e)
Green- (uses same (uses same (same Minor- (No Take 1.5 mL aliquot ~55 °C; Redissolve in 1 Inject 1 uL
land Nonpolar Nonpolar Nonpolar GC-MS (One-third of the ini- solvents dried mL Hex/DCM SL (e)
DS-2 FID data as MS data as FID data as data) tial 4.5 mL Most-Polar overnight with 9/1; Inject 1uL
GreDS-1) GreDS-1) GreDS-1) solution) 5A molecular SL (g)
sieves.
Blank (uses same (No GC-MS (same Minor- (No Take 1.5 mL aliquot ~55 °C; Redissolve in 1 Inject 1 uL
DS-2 Nonpolar data) Nonpolar FID GC-MS (One-third of the ini- solvents dried mL Hex/DCM 20:1 split (q)
FID data as data as Blank data) tial 4.5 mL Most-Polar overnight with 9/1; Inject 1uL
Blank DS-1) DS-1) solution) (E-10) 5A molecular 10:1 split (d)
sieves.
Monterey (uses same (No GC-MS (same Minor- (No Take 1.5 mL aliquot Same as DS-2 Redissolve in 50 Redissolve in
DS-3 Nonpolar data) Nonpolar GC-MS (One-third of the ini- but ~100 mg ad- uL Hex/DCM 10 uL Hex/
FID data as FID data as data) tial 4.5 mL Most-Polar ditional NaBH4 9/1; Inject 1uL DCM 9/1;
Monterey Monterey solution) added after 30 SL (g) Manualinject
DS-1) DS-1) mins 1uL SL (e)
Green- (uses same (uses same (same Minor- (No Take 1.5 mL aliquot Same as DS-2 Redissolve in 1 (No GC-MS
land Nonpolar Nonpolar Nonpolar GC-MS (One-third of the ini- but ~100 mg ad- mL Hex/DCM data)
DS-3 FID data as MS data as FID data as data) tial 4.5 mL Most-Polar ditional NaBH4 9/1; Inject 1uL
GreDS-1) GreDS-1) GreDS-1) solution) added after 30 SL (g)
mins
Blank (uses same (No GC-MS (same Minor- (No Take 1.5 mL aliquot Same as DS-2 Redissolve in 1 (No GC-MS
DS-3 Nonpolar data) Nonpolar FID GC-MS (One-third of the ini- but ~100 mg ad- mL Hex/DCM data)
FID data as data as Blank data) tial 4.5 mL Most-Polar ditional NaBH4 9/1; Inject 1uL
Blank DS-1) DS-1) solution) (E-10) added after 30 10:1 split (d)
mins
Monterey Redissolve in Redissolve Redissolve in Inject 1 Entire fraction 50 °C, MeOH Redissolve in 50 Redissolve in
DS-0 50 uL Hex/ in 250 uL 50 uL Hex/ uL SL dried with uL Hex/DCM 10 uL Hex/
DCM 9/1; Hex/DCM DCM 9/1; (e,z) Na2SO4 9/1; Inject 1uL DCM 9/1;
Inject 1uL 9/1; Inject 1 Inject 1uL SL (g) Manualinject
SL (g) uL SL (z) SL (g) 1uL SL (e)
Blank Redissolve in Redissolve Redissolve in Inject 1 Entire fraction 50 °C, MeOH Redissolve in 50 Redissolve in
DS-0 50 uL Hex/ in 250 uL 50 uL Hex/ uL SL dried with uL Hex/DCM 10 uL Hex/
DCM 9/1; Hex/DCM DCM 9/1; (e) Na2SO4 9/1; Inject 1uL DCM 9/1;
Inject 1uL 9/1; Inject 1 Inject 1uL SL (g) Manualinject
SL (g) uL SL (z) SL (g) 1uL SL (e)
23
3. Results
3.1. Overview
Here we report on the desulfurization of synthetic standards and sedimentary ex-
tracts. We show the standard reactions in Table 1 and figure 12. We show the desulfurized
compounds for Monterey in figures 13 and 14, and for Greenland in figures 17 and 18.
We also report on the compounds found in the nonpolar fractions of these extracts, and
compare the free nonpolar hydrocarbons to the nonpolar hydrocarbons released by desul-
furization.
3.2. Standards
We desulfurized the standards 1-octadecanethiol and S-Benzylthioglycolic acid.
(See fig. 4 for the structures of these compounds.) These reactions gave yields between
16% and 92% (Table 1).
Table 1. Desulfurizations of synthetic standards and reaction yields.

Standards and Re- Procedure (cf. sec- Peak Area Com- % Yield (Moles
Trace
Figure 12,
agents tion 2.4) (pA*s) pound yield/ moles

(b)ack column Mass standard)*100
(f)ront column (mg)
16.5 mg 1-octade- DS-0; 4 mL solvent (no GC-FID n/a n/a

canethiol volume data)
106.4 mg NiCl2
104.9 mg NaBH4
2.19 mg 1-octade- DS-0; 2 mL solvent A 156.41 (b) 0.717 36.9%
canethiol volume (using a 10% aliquot)
18 mg NiCl2
14.1 mg NaBH4
2.1 mg 1-octadecane- DS-0; 2 mL solvent B 663.2 (b) 0.304 16.3%
thiol volume
41.1 mg NiCl2
38.7 mg NaBH4
24
Standards and Re- Procedure (cf. sec- Peak Area Com- % Yield (Moles
Trace
Figure 12,
agents tion 2.4) (pA*s) pound yield/ moles
(b)ack column Mass standard)*100
(f)ront column (mg)
2.1 mg 1-octadecane- DS-0; 2 mL solvent C 818.2 (f) 0.316 16.9%

thiol volume
43.8 mg NiCl2
39.1 mg NaBH4
2.1 mg 1-octadecane- Temp=25°C; D 252.5 (f) 0.487 26.1%
thiol solvents= (2 mL (octadecane)
46.8 mg NiCl2 MeOH + 2 mL (using a 20% aliquot)
49.9 mg NaBH4 Hex); MeOH dried
w. Na2SO4
2.16 mg S-Benzylth- Temp=25°C; D 217.4 (f) 0.419 38.4%
ioglycolic acid solvents= (2 mL (toluene)
46.8 mg NiCl2 MeOH + 2 mL (using a 20% aliquot)
49.9 mg NaBH4 Hex); MeOH dried
w. Na2SO4
1 mg 1-octadecane- DS-1; 4 mL solvent E 1980.52 (f) 0.764 86.0%
thiol volume
108 mg NiCl2
111 mg NaBH4
1 mg 1-octadecane- DS-2; 4 mL solvent F 2112.50 (f) 0.815 91.8%
thiol volume
103 mg NiCl2
103 mg NaBH4
1 mg 1-octadecane- DS-3; 4 mL solvent G 1490.29 (f) 0.575 64.7%
thiol volume
90 mg NiCl2
~104 mg NaBH4
+ 130 mg NaBH4 after
30 mins
25
5600
4800 Octadecane G
4000
Octadecane F
3200
2400
Octadecane E
Toluene Octadecane D
1600
Octadecane C
800
Octadecane B
Hexamethylbenzene Octadecane
0
A
6 10 14 18 22 26 30 34 38
time (min)
Figure 12. GC-FID chromatograms for standard desulfurizations. The desulfurized standard
elutes at different times because we used different temperature programs for some runs (Methods,
Endnote 12). Trace A used program a. Traces B-D used program b. Traces E-G used program d.
The octadecane peak elutes slightly sooner in trace B than it does in trace C because these two
runs were each on a different column.
26
3.3. Monterey desulfurization fraction
G15 DS C FID
B
9 B
B 17 12
25 6 B B B
7 B B 13
5 8 B
11 B B
16 B 10 15
14
Intensity (pA)
18
20
1 2 4
3
15
16 20 24 28
Time (minutes)
Figure 13. Monterey Desulfurization Fraction, GC-FID Trace. Numbered peaks refer to figure
Gre DS A (back)
14. (B) indicates the compound is also present in the blank desulfurization fraction at a similar
abundance
100 (E-1). This chromatogram is from analysis of the Monterey DS-3 sample, which typi-
fies the other Monterey samples. 18
80
The Monterey desulfurization fraction contains a straight-chain alkane and a

60 16
Intensity (pA)
11
branched alkane (compounds 6 and 9). Cholestane and other steroids are present (11,
17
14
1 3
13-15). Other isoprenoids are beta-Tocopherol (12), delta-Tocopherol (18), and Lyco-
40 19
12
20
9
2 6 15*
pene (17). Lycopene may also be present in
20 7 the nonpolar fraction.
13* We see heterocyclic
4 5 10*
8*
compounds (3, 5, 10, 16) and nitrogenous compounds

Time (minutes)
(1, 8). Compounds 12 and 15 may
20 25 30
be misidentified terpenoids; their peaks are relatively rich with an ion (m/z 191) that is
characteristic for this compound class (Forster et al., 2004; Schouten et al., 2001). Convo-
luted mass spectra reduce our confidence in the compound assignments for the following
peaks: 5, 10, and 16 (E-2).
Many peaks in the Monterey desulfurization are unidentifiable because a high
baseline obscures them in both the GC-MS and the GC-FID chromatograms. For exam-
ple, unsaturated steroid hydrocarbons may be under-reported here because their charac-
teristic ions (e.g. m/z 257) (Forster et al., 2004) are overwhelmed by the ions of coeluting
27
compounds (e.g. m/z 57) (E-3). The high baseline indicates that this fraction contains an
unresolved complex mixture (UCM) of coeluting compounds. Sutton et al. (2005) esti-
mated that a UCM could contain 250,000 unique compounds.
Some of the compounds in Monterey DS-3 were not found in other Monterey
desulfurized fractions; the DS-0 fraction was particularly disagreeable. We found dif-
ferences between the baseline shape of the Monterey-0 fractions and the other Monterey
fractions (E-4). Such baseline shifts may have affected the signal-noise difference calcu-
lation used to extract a mass spectrum for each peak, thereby leading to disagreements
over peak identities.
28
1 2 3 4
5 6
7 8 9
10 11 12 13
14 15 16
17 18
1 4-(cis-2,3,4,trans-6-Tetramethyl-3-cyclohexenyl)butan-2-one 2,4-dinitrophenylhydrazone
2 2,6-Bis(1,1-dimethylethyl)-4-(1-oxopropyl)phenol
30 4-Acetyl-1,2,3,4-tetrahydro-2-oxoquinoline
40
Gibberellic acid
5 2,10-Dimethyl-2,3,4,5,6,7-hexahydro-1H-2-benzazonine
6 Octadecane
7 8-Heptadecanol
802
[1,1’-Biphenyl]-4,4’-diamine, 3,3’-dimethyl
9 Heptadecane, 9-hexyl-
10 0
8-Methyl-7-phenyl-1,3,8-triazaspiro[4.5]decan-2,4-dione
11 Cholestane
12 beta-Tocopherol
13 Propanoic acid, 2-(3-acetoxy-4,4,14-trimethylandrost-8-en-17-yl)-
14 0
Chol-8-en-24-al, 3-(acetyloxy)-4,4,14-trimethyl-, (3.beta.,5.alpha.)-
15012 Chol-8-en-24-al, 3-(acetyloxy)-4,4,14-trimethyl-, (3.beta.,5.alpha.)-
16 1-(5,5-Dimethyl-1,3-dioxocyclohexan-2-yli den)-2-(N-ethylbenzthiazol-2-yliden)-ethan
17 psi.,.psi.-Carotene, 7,7’,8,8’,11,11’,12,12’,15,15’-decahydro-
18 delta-Tocopherol
Figure 14. Monterey desulfurization compound structures and names.

0
Not found in the Monterey DS-0 fraction; 1Not found in the Monterey DS-1 frac-
tion; 2Not found in the Monterey DS-2 fraction.
29
3.4. Monterey nonpolar fraction
(1/3) Mont2 nonpol (FID)
2000
14
1500
11
Intensity (pA)
1000
2 9 13
4 12 15
35 7 8 16
10 17 18 27
6 21
1 24
500 22
19 20 30
23 25
29 31
26 28 32
0
15 20 Time (min) 25 30
Figure 15. Monterey Nonpolar Fraction, GC-FID Trace. Numbered peaks refer to figure 16.
1/3 Gre Nonpol FID
300
23
20
250
The Monterey nonpolar fraction contains phytane (14), pristane (11),

24
a trimethyl
200
alkane (7), and a dimethyl alkane (2). We find a series of methyl- and isopropyl-substitut-
Intensity (pA)
26
27
ed polycyclic aromatic hydrocarbons (1, 3-5, 6, 8, 9). We see a benzothiophene (13), and
150
other heterocyclic compounds (10, 15, 26, 31). We see a phenolic alcohol (22), a phe-
100 19
22
15* 25
16
21
nolic ester (23),1 3and4 a substituted biphenyl compound (12). We find a series of steroids,
50 14 18*
29
9 12 13 28
5 6 7 17
2 8 11 30
10
whose hydrocarbon skeletons are cholestaneTime

0
16 (19-21,
(min) 25), stigmasterane (28), and unusual
20 24 28 32 36
(16, 24, 30). We see non-steroid polycyclic terpenoids (17, 18 27, 29, 32). Although not
detectable by GC-FID, the GC-MS data show that cyclic octatomic sulfur is present in
the Monterey nonpolar fraction. Convoluted mass spectra reduce our confidence in the
compound assignments for the following peaks: 1, 6, 7, 10, 12, 13, 22, 23, 25, 29, 30, and
32. As is the case for the Monterey desulfurization fraction, the Monterey nonpolar frac-
tion has a UCM with many co-eluting compounds. The m/z 57 ion is very depleted in the
retention range of C15-C35 n-alkanes, although pristane and phytane are abundant (E-5).
30
The depletion of long-chain n-alkanes indicates that this sample’s biodegradation level is
“moderate,” using Wenger and Isaksen’s (2002) scale.
Figure 15 is a composite of information from two nonpolar fractions. The trace is
from GC-FID analysis of the Monterey Nonpolar fraction, and the numbered peaks were
selected using GC-MS data for the Monterey Nonpolar-0 fraction (Methods endnote 7).
The two fractions contain the same compounds, although compound ratios may differ
(Results endnote 4).
31
1 2 3 4 5
6 7 8 9
10 11 12 13
14 15 16 17
18 19 20 21 22
23 24 25 26 27
28 29 30 31 32
1 Naphthalene, 2,6-dimethyl- 17 D-Homopregnane, (5.alpha.)-

2 Dodecane, 4,6-dimethyl- 18 15-Isobutyl-(13.alpha.H)-isocopalane
3 Naphthalene, 1,6,7-trimethyl- 19 Cholest-14-ene, (5.alpha.)-
4 Naphthalene, 1,6,7-trimethyl- 20 Cholest-14-ene, (5.alpha.)-
5 Naphthalene, 2,3,6-trimethyl- 21 Coprostane
6 Naphthalene, 2-(1-methylethyl)- 22 Phenol, 2-(1,1-dimethylethyl)-4-(1-methyl-1-phenylethyl)-
7 Pentadecane, 2,6,10-trimethyl- 23 Phthalic acid, 3,5-dimethylphenyl 4-formylphenyl ester
8 9H-Fluorene, 9-methyl- 24 5.alpha.-Cholest-8-en-3-one, 14-methyl-
9 Naphtho[2,1-b]furan, 1,2-dimethyl- 25 Cholestane
10 1H-Indene-4-carboxylic acid, 2,3-dihydro-1,1- 26 3,5,7-Triazatricyclo[6.3.0.0(3,7)]undec-11-ene-4,6-dione,
dimethyl-, methyl ester 2,2-diphenyl-5-methyl-
11 Pentadecane, 2,6,10,14-tetramethyl- (Pristane) 27 28-Nor-17.beta.(H)-hopane
12 1,1’-Biphenyl, 3,3’,4,4’-tetramethyl- 28 Stigmastane
13 Dibenzothiophene, 4-methyl- 29 Olean-13(18)-ene
14 Hexadecane, 2,6,10,14-tetramethyl- 30 Androst-5-en-17-one, 3-hydroxy-16-(phenylmethylene)-
15 Benzothiazole-2-thiol, 5-dimethylamino- 31 11H-Indeno[1,2-b]quinoline, 2,6-dimethyl-
16 Allopregnane 32 Baccharane
Figure 16. Monterey Nonpolar fraction. Compound structures and names.
32
1
3
I
15
16 20 24 28
Time (minutes)
3.5. Greenland Desulfurization fraction

Gre DS A (back)
100
18
80
60 16
Intensity (pA)
11
17
14
40 1 3 19
12
20
9
2 6 15*
20 7 13*
4 5 10*
8*
20 25 30
Time (minutes)
Figure 17. Greenland Desulfurization Fraction, GC-FID Trace. Numbered peaks refer to figure
18. (*) indicates the compound may be present in the Minor-Nonpolar fraction, at a lower abun-
dance. This chromatogram is from analysis of the Greenland DS-1 fraction, which typifies the
other Greenland DS fractions.
In the Greenland desulfurization fraction, we find C20 regularly branched
isoprenoids (1-3). We find a suite of C17-C27 alkyl methyl esters (4, 6, 8, 10, 12), each with
an odd number of carbons. We find a suite of alkanols (5, 7, 9, 11). Alkanol abundance
increases with chain length in all three desulfurization fractions, and the alkyl methyl
esters do not show a correlation between molecule size and compound abundance. The
steroids in this fraction are functionalized and, usually, unsaturated. Their hydrocarbon
skeletons are of cholestane (13, 14), ergostane (15, 17), and stigmasterane (16, 18). We
observe a C35 alkene (19) and a hopanoid (20). All of the compounds found in the DS-1
fraction are also found in the other Greenland desulfurization fractions, except for one
of the alkyl methyl esters (4). A convoluted mass spectrum reduces our confidence in the
compound assignment for peak 7.
GC-FID retention times suggest that several of the compounds in the Greenland
desulfurization fraction may also present in the nonpolar fractions (8, 10, 13, 15).
33
A detailed retention time comparison supports the idea of similar but not identical
compounds in the two fractions (E-6). For an immature sediment such as Greenland,
we expect to see similar compounds in the desulfurized and nonpolar fractions because
diagenetic processes may have had little opportunity to modify the free hydrocarbons.
Every compound we identify in the Greenland desulfurized fraction is more
abundant than its counterpart in the minor-nonpolar fraction. This relationship suggests
that every labelled compound in the above figure 17 was part of an organic sulfur
compound in the most-polar fraction (before desulfurization). As we will explain below,
the minor-nonpolar fractions provide a “ceiling” for the abundance of nonpolar cross-
contamination in the desulfurization fractions.
34
1 2
3 4
5 6
7 8
9 10
11 12
13 14 15 16
17 19
18 20
1 Hexadecane, 2,6,10,14-tetramethyl-
Figure 18. Greenland desulfurization 2 2-Hexadecene, 3,7,11,15-tetramethyl-, [R-[R*,R*-(E)]]-
compound structures and names. 3 2-Hexadecene, 3,7,11,15-tetramethyl-, [R-[R*,R*-(E)]]-
†
Not found in the other Greenland DS 4†
Hexadecanoic acid, methyl ester
fractions. 5 1-Eicosanol
*May be present in the Greenland 6 Octadecanoic acid, methyl ester
Minor-Nonpolar fraction, at a lower
7 1-Eicosanol
abundance.
8* Docosanoic acid, methyl ester
9 1-Eicosanol
10* Tetracosanoic acid, methyl ester
11 1-Tetracosanol
12 Hexacosanoic acid, methyl ester
13* (3.alpha.,5.beta.)-Cholestan-3-ol Compare to Np 15
14 Cholesterol
15* Campesterol Compare to Np 18
16 Stigmasterol
17 .alpha.-Ergostenol
18 .gamma.-Sitosterol
19 17-Pentatriacontene
20 4,4,6a,6b,8a,11,11,14b-Octamethyl-docosahydropicen-3-ol
35
3 5 7 8 16
10 17 18 27
6 21
1 24
500 22
19 20 30
23 25
29 31
26 28 32
0
15 20 Time (min) 25 30
3.6. Greenland Nonpolar fraction

1/3 Gre Nonpol FID
300
23
20
250
24
200
Intensity (pA)
26
27
150
100 19
22
15* 25
16
21
50 1 14 18*
29
3 4 9 12 13 28
5 6 7 17
2 8 11 30
10
0
16 20 24 Time (min) 28 32 36
Figure 19. Greenland Nonpolar Fraction, GC-FID Trace. Numbered peaks refer to figure 20.
(*) indicates the compound may be present in the Greenland desulfurization fraction.
The Greenland nonpolar fraction contains phytol isomers (1, 3, 4) and a phytene
(2). We observe straight-chain alkanes (8, 11), a branched alkane (7), and a highly
branched alkane (13). We see long-chain alkenes (10, 22) and an ethyl ester dialkene (19).
We find a series of C27-C29 steroids, whose hydrocarbon skeletons are of cholestane (12,
15-17), ergostane (18), and stigmasterane (14, 20, 21). We see a heterocyclic compound
(9). We find a suite of di-, tri-, and tetra- unsaturated C37-C39 methyl and ethyl ketones.
Phthalates are present (5, 6). The GC-MS data show that cyclic octatomic sulfur is pres-
ent in the Greenland nonpolar fraction.
Convoluted mass spectra reduce our confidence in the compound assignments for
the following peaks: 7, 10, 13-15, 17-19, 21, and 22.
36
1 2 5
3 4
6 7 8
9 10 12
11
13 14 15* 16
17 18* 19
20 21 22
1 3,7,11,15-Tetramethyl-2-hexadecen-1-ol 16 Cholestan-3-one, (5.beta.)-

2 2-Hexadecene, 3,7,11,15-tetramethyl-, [R-[R*,R*-(E)]]- 17 Cholestane, 2,3-epoxy-, (2.alpha.,3.alpha.,5.alpha.)-
3 3,7,11,15-Tetramethyl-2-hexadecen-1-ol 18* Ergost-8(14)-en-3-ol, (3.beta.)- Compare to Ds-15
4 3,7,11,15-Tetramethyl-2-hexadecen-1-ol 19 Z,Z-4,15-Octadecadien-1-ol acetate
5 Phthalic acid, butyl 2-pentyl ester 20 Stigmasterol, 22,23-dihydro-
6 Benzyl butyl phthalate 21 .beta.-Sitosterol
7 Heptadecane, 9-hexyl- 22 17-Pentatriacontene
8 Octacosane 23 C37:4Me ketone
9 2-Furanmethanol, tetrahydro-.alpha.,.alph-[2.alpha.,5.beta.(R*)]]- 24 C37:3Me ketone
a.,5-trimethyl-5-(4-methyl-3-cyclohexen-1-yl)-, [2S
10 17-Pentatriacontene 25 C37:2Me ketone
11 Octacosane 26 C38:4Et ketone and C38:4Me ketone
12 Cholesta-3,5-diene 27 C38:3Et ketone and C38:3Me ketone
13 Octadecane, 3-ethyl-5-(2-ethylbutyl)- 28 C39:4Et ketone
14 Stigmasta-5,22-dien-3-ol, acetate, (3.beta.)- 29 C39:3Et ketone

15* Cholesta-5,22-dien-3-ol, (3.beta.)- Compare to Ds-13 30 C39:2Et ketone
Figure 20. Greenland nonpolar fraction: compound structures and names. Assignments for #23-30
are based on relative retention times and peak shapes from D’Andrea and Huang (2005).
*May be present in the Greenland Desulfurization fraction
37
3.7. Minor-Nonpolar fractions and experimental control
The first fractionation (Figs. 7, 8) did not isolate all of the nonpolar compounds
(compare the Nonpolar fractions with the Minor-Nonpolar fractions, in figure 21). Con-
sequently, nonpolar cross-contaminants may be present in the Desulfurization fractions.
Since our fractionation procedure is inherently compromised (see endnote 4 of Method
section, and “Experimental control,” Method section), the second fractionation may have
been just as incomplete as the first. If the second fractionation were incomplete, then we
would assume that a third nonpolar fraction would contain 4-19% of the total volatile
organic carbon (VOC) found in the Minor-Nonpolar fraction. We base this assumption
on the total amount of VOC found in the Minor-Nonpolar fraction compared to the total
amount of VOC found in the Nonpolar fraction (bar graphs, figure 21). The Desulfuriza-
tion fraction is also the third nonpolar fraction (fig. 7). Some compounds are easier to
fractionate than others, as shown by the alkenone (32-35 min) to steroid (27-30 min) ratio
difference between Greenland nonpolar and Greenland minor-nonpolar. Owing to this
compound-specific uncertainty, a compromised isolation procedure, and an abundance of
caution, we conclude: any nonpolar compound that carried through to the minor-nonpolar
fraction must also have carried through to the desulfurized fraction. However, a cross-
contaminant in the desulfurized fraction will be less abundant than its counterpart in the
minor-nonpolar fraction. Therefore, any compound in the desulfurized fraction that is
more abundant than its minor-nonpolar counterpart must, to some extent, result from the
desulfurization reaction.
These cross-contaminants undermine our confidence that desulfurization explains
every compound in the desulfurization fractions. On the other hand, total desulfurization
38
yields are greater than 19% of the minor-nonpolar yields (bar graph, fig. 21), suggesting
that the reaction released many compounds from sulfur linkages. Moreover, several peaks
that are present in the desulfurization fractions are not present in the minor-nonpolar frac-
tions, and several peaks in the desulfurization fractions have smaller counterparts in the
minor-nonpolar fractions. These new or enlarged peaks in the desulfurization fractions,
most noticeable for the Greenland extract, suggest the release of sulfur-bound hydrocar-
bons.
39
Greenland Monterey
300 2000
Greenland Nonpolar Monterey Nonpolar
250
1500
200
150 1000
100
500
50
0 0
20 25 30 35 10 15 20 25 30
150 60
Greenland Minor-Nonpolar Monterey Minor-Nonpolar

Greenland Desulfurization Monterey Desulfurization
50
100 40
30
Intensity (pA)
50 20
10
0 0
20 25 30 35 10 15 20 25 30
Time (min)
Greenland Monterey
Nonpolar 100% 100%
Nonpolar
Greenland Monterey
18.1% 4.06%
Minor-Nonpolar Minor-Nonpolar
Greenland Monterey
Desulfurization 12.1% Desulfurization 2.41%
1 10 100 1 10 100
Total Abundance (% of Greenland Nonpolar fraction) Total Abundance (% of Monterey Nonpolar fraction)
Figure 21: Greenland and Monterey chromatographic data reconsidered. Nonpolar and Desulfurization chro-
matograms (shown previously in figures 13, 15, 17, and 19) are reprinted here with the Minor-Nonpolar chromato-
grams. This figure allows a direct comparison of the volatile hydrocarbon abundance between the fractions of each
extract. For the Greenland extract, we dissolved each fraction in 1 mL solvent, and injected 1 uL splitless. We
dissolved each Monterey fraction in 50 uL of solvent, and injected 1 uL splitless. See E-7 for more on instrument
conditions and the formatting of this figure.
40
Section 3 Endnotes
1. Figure 22 shows a GC-FID trace of Blank DS-0.

40
36
32
28
24
20
16
12
Figure 22
8
16 20 24 28
2. When more than one compound elutes at the same time, the mass spectrom-
eter fragments them simultaneously and they appear together as convoluted mass spectra
(Colby, 1992). It is possible to identify coeluting compounds because their major ions
often have different peak shapes or slight peak offsets, which are visible by extracting the
ion chromatogram. For example, figure 23 shows a peak for which we could not find evi-
dence for coelution. On the other hand, figure 24 shows three peaks from the Monterey
DS-3 data for which we did find evidence of coelution. At least two compounds compose
peak 5; one has major fragments with m/z

Abundance [3.32%] [30411]
100
TIC
= 189, 146, and 160; the other has major
Peak #11
Monterey Nonpolar-0
Figure 23
75.4
fragments with m/z = 83 and 174. Fragment
m/z = 203 may be from a third compound.
At least two compounds compose peak 16;

50.7
141
one has major fragments with m/z = 186
26.1
169
183 and 201; the other has the major fragment
127
113
1.5
19.542 19.563 19.583 19.603 19.623
149, and the m/z = 175 ion may be from a
41
Abundance [3.62%] [3557] Abundance [5.06%] [3484] Abundance [3.34%] [4165]
100
Peak #5 TIC Peak #16 TIC Peak #10 TIC
75
203
50
146
160
83 175 146
25
189 201 128
174 149 155
186 287
0
16.443 16.483 17.677 17.717 17.757 22.411 22.451 22.491 22.531
Figure 24
third compound. The same reasoning applies to peak 10. By extracting major ion chro-
matograms, we searched for evidence of coelution in every peak identified in the Results
section.
3. Figure 25 shows how small the m/z 257 ion abundance is compared to other
ions from coeluting compounds, in GC-MS data for the Monterey DS-3 fraction. In
typical mass spectra 100
for unsaturated steroid
hydrocarbons, the m/z

75
257 ion abundance TIC

(Intensity * 5606)
dwarfs the m/z 57 ion

50
abundance (e.g., this is
the case for Monterey

25
nonpolar compounds 19 57
m/z 57, m/z 257
and 20). (Intensity * 5606 * 6.20%)
257
0
22.71 22.97 23.23 23.49 23.76 24.02 24.28 24.54
Time (min)
Figure 25
42
1000 80
Monterey Nonpolar-0 Monterey Minor-Nonpolar-0 50 Monterey Desulfurization-0 (DS-0)
70
800
60
40
600 50
30
40
400
Intensity (pA) Intensity
30 (pA) Intensity
20 (pA)
20
200
10
10
0 0 0
10 15 20 time 25 30 10 15 20 time 25 30 10 15 20 time 25 30
2000 64 20
Monterey Nonpolar Monterey Minor-Nonpolar Monterey DS-3
56
1500 48 15
40
1000 32 10
Intensity (pA) Intensity

24 (pA) Intensity (pA)
500 16 5
0 0 0
10 15 20 time 25 30 10 15 20 time 25 30 10 15 20 time 25 30
Figure 26
4. Figure 26 shows that light compounds have a higher relative abundance in the
Monterey nonpolar fractions than they do in the Monterey-0 nonpolar fractions. We show
the Monterey-0 fractions in black and the Monterey fractions in red. The DS-3 and DS-0
profiles are similar to one another, although DS-0 gives a higher yield than DS-3, and
DS-3 has a higher relative abundance of heavy compounds than DS-0. DS-3 has a very
similar profile to DS-1 and DS-2 (not shown). By the term “light”, we mean compounds
with relatively short retention times, and by “heavy” we mean compounds with relatively
long retention times.
One possible explanation for the systematic baseline difference between the
nonpolar fractions is that the light compounds are vulnerable to evaporative loss. We
began each of these two fractionation procedures with an aliquot of Monterey TOE. We
produced the Monterey-0 fractions on 8/15/08, 8/18/08, and 8/19/08. One month later, we
produced the Monterey fractions (on 9/18/08 and 9/22/08). We analyzed all six fractions
on the same day (12/16/08) using GC-FID, with the same column, using the same instru-
ment method. In the evaporation scenario, the fractions that we produced earlier lost more
of their light compounds because they had more time to do so.

43
5. Figure 27 shows the extracted m/z 57 ion chromatogram for the Monterey-0
Nonpolar fraction.
Abundance [7.68%][96866] TIC
100
75
50
25

0 57

9.15 11.63 14.11 16.58 19.06 21.54 24.01 26.48 28.96 31.44 33.92 36.39 38.86 41.34 43.82 46.29
Figure 27

6. Figure 28 compares the Greenland GC-FID traces for the Nonpolar and Des-
ulfurization fractions. Peak Desulf-3 is the same compound as peak Nonpol-2, based on
nearly identical retention times and GC-MS data. The desulfurization peaks 13 and 15
appear to have nonpolar counterparts with similar enough retention times that we assume
they are identical. On the other hand, the desulfurization peaks 14, 16, 18, 19, and 20 do
not have counterparts with similar retention times.

280
18 20
Nonpolar Desulfurization
80
240
200
Greenland Desulfurization Intensity (pA)
60
Greenland Nonpolar Intensity (pA)
160
14 40
120
3 19 19
80 22
16 16
20
21
40 18
15
15
17
2 13 20
17
0 0
27 27.5 28 28.5 29 29.5 30 30.5
17.5 17.75 18 time
Figure 28
44
7. The six chromatograms shown in figure 21 were obtained from GC-FID analy-
sis of six fractions. To allow comparison of volatile compound abundances between
different fractions of the same extract (e.g., the Greenland desulfurization fraction and the
Greenland minor-nonpolar fraction), the instrument conditions were constant for each ex-
tract. For the Greenland extract, we dissolved each fraction in 1 mL of Hexane:DCM 9:1
and injected 1 uL using program g (splitless). For the Monterey extract, fractions in 50 uL
of Hex:DCM 9:1 were injected 1 uL splitless, using program g. The Greenland traces are
data for fractions Greenland nonpolar, Greenland minor-nonpolar, and Greenland DS-1.
The Monterey traces are data for fractions Monterey nonpolar, Monterey minor-nonpolar,
and Monterey DS-3. Each desulfurization started with 1/3 of the most-polar fraction;
therefore, to allow a direct abundance comparison between different fractions of the same
extract, we scaled the nonpolar and minor-nonpolar intensity measurements to 1/3 of their
original size. For visual clarity we offset the Greenland desulfurization trace by 10 pA
with respect to the Greenland minor-nonpolar trace. We did not offset the Monterey traces
with respect to one another.
45
4. Discussion
4.1. Standard yields
We sought to optimize the desulfurization efficiency by performing several exper-
iments on synthetic standards. For most of our standard experiments, we changed several
variables at once. We sought to add NaBH4 at a molar excess to its reactant NiCl2, be-
cause the product of this reaction (Ni2B) decomposes NaBH4 (Back et al., 1992). Higher
rates of Ni2B formation per unit volume should make product formation more energeti-
cally favorable. Using a graphing program, we determined how well each variable (mass
of reagents, solvent volume, reagents/standard, etc.) correlated to the percent yield. Our
analysis did not find a distinctive variable or combination of variables that could explain
the variance in % yield (Fig. 29) better than other variables or combinations of variables.
However one ratio that has a robust correlation to the % yield is (mass reagents/ mass
standard). To achieve high desul-

DS-0, 2mL (2, 18, 14)
DS-0, 2mL (2, 41, 39)

furization yields on geochemical
DS-0, 2mL (2, 44, 39)
25 °C, 2mL MeOH + 2mL Hex samples, we suggest: that the ratio
(2, 47, 50)
25 °C, 2mL MeOH + 2mL Hex
of reagents to OSCs should be high

(toluene)
(2, 47, 50)
DS-1, 4mL (1, 108, 111)
DS-2, 4mL (1, 103, 103) (~100 mg each reagent, <10 mg

DS-3, 4mL (1, 90, 104+130)
0 20 40 60 80 100 sample), the reaction concentra-
% Yield
Figure 29. Yields of standard desulfurizations. The
tion should be high (4 mL solvent),
experiment labelled “toluene” was a desulfurization
of S-benzylthioglycolic acid; all the other experiments
desulfurized 1-octadecanethiol. Beside each yield bar, and the molar excess of NaBH4 to
we report: Procedure description, solvent volume (mg
standard, mg NiCl2, mg NaBH4). We used the solvents NiCl2 should be about 3:1 (cf. Back
MeOH/THF 1/1, unless otherwise noted.
et al., 1992).

46
4.2. Sample yields
We assume that the desulfurized fractions contain volatile cross-contaminants
with abundances between 4% and 19% of the minor-nonpolar fraction (Results, Minor-
Nonpolar section). These best and worst-case contamination assumptions allow us to ap-
proximate the actual desulfurization yield and compare it to the yields achieved by other
workers (Fig. 30). Our yields are on the same order of those reported elsewhere, except
Monterey, Shell Beach

11.1 Ma
(1) Monterey, Naples Beach From this report
6.7-7.8 Ma
(2) Monterey, Naples Beach From literature
6.7-7.8 Ma
Monterey DS-0
Monterey DS-1
Monterey DS-2
Monterey DS-3
Greenland DS-1
Greenland DS-2
Greenland DS-3
Lake Cadagno, 0-6 Ya

Lake Cadagno, 50-56 Ya
Vena del Gesso basin
Upper Miocene shale
0.01 0.1 1 10 100
Desulfurization Yield
(mg/g TOC)
Figure 30. Desulfurization yield of each sample, normalized to the total organic carbon (TOC).
Darker bars indicate the minimum yield, and lighter bars indicate the maximum yield. See E-3
for the procedure we used to calculate these yields.
for Schouten et al.’s (1993) desulfurization of a shale from the Vena del Gesso basin,
and we suggest that this discrepancy relates to the Messinian salinity crisis. Intra-sample
variations within our own data are probably due to procedural inconsistencies. Our
Greenland sample tends to give higher desulfurization yields than our Monterey sample
gives, indicating the rapidity of sulfurization in Brayasø. Organic matter concentration is
relatively high for both Greenland and Monterey, consistent with each sample’s oxygen-
poor depositional environment.

47
The Vena del Gesso shale gives a desulfurization yield an order of magnitude
higher than every other sample shown in Fig. 30. Such a large enrichment of released
hydrocarbons cannot easily be explained by experimental variability; this sample’s diage-
netic conditions must have been significantly different from those of the other samples.
Since this sample is from the Gessoso-solfifera formation outcropping in a northern Italy
evaporitic basin (Kohnen et al., 1991; Vai and Ricci Lucchi, 1977), its age is between 5.5
and 6 Mya (Roveri et al., 2003). This time coincides with the Messinian salinity crisis,
when the Mediterranean sea became isolated from the Atlantic ocean. This isolation
contributed to evaporitic conditions throughout the Mediterranean basin (Krijgsman et
al., 1999), including in the Vena del Gesso area (Roveri et al., 2003). Evaporitic condi-
tions, coupled with anoxic bottom waters (Vai and Ricci Lucchi, 1977) may have been an
ideal setting for prolific bacterial sulfate reduction and an anomalously high sulfurization
efficiency in the Vena del Gesso basin. Another Messinian evaporitic sample, desulfur-
ized by Schaeffer et al. (1995) also seems to record a very high sulfurization efficiency.
These authors found that a desulfurized TOE yielded alkanes at an abundance 20-30
times greater than the abundance of free saturated hydrocarbons. On the other hand, the
Monterey samples deposited ~1 Myr before the Messinian salinity crisis, and outside of
the Mediterranean basin. In our Monterey samples, we did not observe any desulfurized
compounds that were more abundant than any of the free hydrocarbons.
Among our Monterey desulfurizations, DS-0 seems to have been significantly
more effective than DS 1-3. The increased yield may be due to a difference between the
method we used to isolate the desulfurization yield for DS-0 and the method we used for
the six other sample desulfurizations (Method endnote 5) (E-1). Among our Monterey
48
and Greenland desulfurizations, the DS-1 yields tend to be higher than those of DS-2
and DS-3. We suspect that the DS-1 yields are higher than those of DS-2 and 3 because
the DS-1 reactions began with slightly more sample than the other two reactions (E-2).
Greenland desulfurization yields are 1.5x-3.5x higher than Monterey yields, when the
maximum or minimum yields are averaged and Monterey DS-0 is excluded from the
comparison. Since the Greenland sample is very young, sulfurization occurred rapidly in
its depositional environment (the lake Brayasø).
Desulfurization yield can vary significantly even between samples from the same
stratigraphic unit (Fig. 30, Monterey Naples Beach samples). Sedimentary compounds
amenable to desulfurization have a heterogeneous distribution.
Our Greenland extractable organic matter (EOM), relative to the rock mass, is
higher than our Monterey EOM (2.7% versus 1.3%). The Greenland sediment has a Total
Organic Carbon abundance (TOC) of 11.7% (D’Andrea, pers. comm., 2009), which is
higher than the TOC for Monterey (probably 2-5%; cf. Katz and Royale, 2001). This
relationship is consistent with the observation that, generally, lake rocks have a higher
TOC than ocean rocks (Peters et al., 2007, pp. 15-16). The same authors reason that lakes
have relatively high organics preservation because higher lacustrine sedimentation rates
provide less opportunity for aerobic remineralization than the lower marine sedimenta-
tion rates provide. Although our Greenland lake probably does have a higher sedimenta-
tion rate than our hemipelagic Miocene shale, we hesitate to suggest that this difference
caused Greenland to have a higher TOC than Monterey. The depositional environment
of Monterey was suboxic, and dissolved oxygen was not an important influence on
organic matter preservation in the Monterey sequence (Isaacs, 2001). The Greenland
49
lake’s bottom waters are anoxic (Anderson et al., 1999). Aerobic remineralization was
not the dominant threat to organic matter preservation in the benthic waters and surface
sediments of these two environments. Perhaps anaerobic remineralization (e.g., bacterial
sulfate reduction) occurs in our two environments, which might explain the TOC relation-
ship. Since our Greenland sample is surface sediment, such a process may not have had
time to reduce this sample’s TOC.
4.3. Sample composition overview
The diagenesis of sediment into sedimentary rock involves thermochemical
processes that add layers of complexity to the already difficult puzzle of geochemical
sulfurization. When a mature sample has been uplifted into a terrestrial environment, it
may be subject to further modification by microbes (Sec. 3.4). The lacustrine Greenland
compounds are fewer in number and easier to identify than the marine Monterey com-
pounds (Secs. 3.3-3.6). This difference may reflect a more diverse paleobiota during the
Monterey sample’s deposition (Sec. 4.4.2) than has recently been present at the Green-
land lake. The Monterey sample also shows more signs of biological and thermal post-
depositional modification (Sec. 4.4.1) than does the Greenland sample.
The Greenland sample (surface sediment from the oligosaline, meromictic lake
Brayasø) is much younger than the ~7 million year old Monterey shale (Sec. 1). Nor-
matively, such a young sediment provides a window into the sulfurization process that
relatively mature sediments may not. Greenland’s nonpolar fraction should contain OSC
precursors and its desulfurized fraction should contain OSC degradative products. A
50
surface sediment’s organic material should have undergone much less alteration than the
organics in a relatively mature rock.
4.4. Monterey
The chromatographic and mass-spectral data for the Monterey sample indicate
that this rock underwent biological and thermal modification, after and possibly during
organic sulfur formation. Bacterial, eukaryotic, and terrestrial sources contributed to the
free and bound organic matter from Monterey. Although the paleobiotic information re-
vealed by desulfurization is more limited than we had initially expected, this fraction may
uncover a precursor-product relationship not previously reported for Monterey bitumens.
4.4.1. Post-depositional modification
Both the desulfurization fraction and the nonpolar fraction give a high baseline.
For the nonpolar fraction, this baseline reflects biodegradation, as we discussed in section
3.4. Heterotrophs, probably aerobic microbes, metabolized the C15-C35 straight-chain
alkanes in the Monterey nonpolar fraction. The nonpolar fraction contains a regular iso-
prenoid that seems to be missing a methyl group (Nonpolar-7) (Np-7), possibly suggest-
ing biodegradation. However, Np-7 also has a convoluted mass spectrum.
The desulfurization fraction is more difficult to interpret. Organic sulfur com-
pounds are thought to be more resistant to biodegradation than free hydrocarbons
(Schouten et al., 2001), so we are hesitant to interpret the high baseline as indicating
microbial metabolism of organic sulfur compounds. Many of the organic sulfur com-
pounds in this Monterey sample may not be macromolecules. Rather, these OSCs may be
51
polar sulfoxides, which Schouten et al. (2001) report may comprise 40% of polar OSCs
in Monterey. Perhaps polar sulfoxides are less resistant to microbial attack than sulfide
macromolecules are. Another possibility is that the biodegraded free hydrocarbons car-
ried through as cross-contaminants to the desulfurized fraction. However, reasonable
estimates of the amount of cross contamination (section 3.7) imply that the latter scenario
is unlikely.
The nonpolar fraction containes cholestoids and stigmasteroids, which are also
reported by Schouten et al. (2001). Each of these compounds (Nonpolar 16, 19-21, 24,
25, 28, 30) seems to have undergone some modification. The C5-C6 unsaturation is not
present in any of these steroids, except for in Nonpolar-30 (Np-30). The structure of Np-
30, however, shows an aromatic substituent on the D-ring, which is unusual for biogenic
steroids. Aromatized steroids are considered indicative of diagenetic or catagenetic ther-
mal alteration (Brocks and Summons, 2004), although usually the aromatic ring is part of
the sterane skeleton. (Np-30’s identification is based on a convoluted mass spectrum, so
this compound’s diagnostic power is questionable.) The steroid Np-16 is missing much
of its alkyl side chain, which may indicate biodegradation or thermally-induced cracking.
The cholestenone Np-24 may be the product of anaerobic biohydrogenation, a pathway
known to produce keto-steroids (Kok et al., 2000), and Np-24 would be a good candidate
for sulfurization due to the reactivity of the ketone group.
Monterey’s nonpolar fraction also may contain a substituted dibenzothiophene
(Np-13). In contrast to Schouten at al. (2001), we did not find thiophenes with long-chain
alkyl groups. However, we only sought to identify the most abundant peaks, so we do
not suggest that these compounds are absent from our nonpolar fraction. The relatively
52
abundant dibenzothiophene may originate from the catagenesis of less-mature macromo-
lecular OSCs (Aizenshtat et al., 1995). Since the upper member of the Monterey forma-
tion is thought to be thermally immature (Sessions, personal communication, 2008), the
possibility of intense thermal modification to our sample is surprising. Np-13’s identity is
obfuscated by a convoluted mass spectrum.
Aizenshtat et al. (1995) suggested that, while sulfide (H2S) and polysulfides (Sx2-)
are reactive with functionalized organics at low temperatures, elemental sulfur (e.g. S8)
is only reactive at high temperatures. The authors reasoned that elemental sulfur required
a homolytic cleavage (220-250 °C) to become reactive. Although we found elemental
sulfur (S8) in the Monterey nonpolar fraction, it does not necessarily indicate thermal ma-
turity for our sample, because we also found S8 in the Greenland nonpolar fraction, which
is (extremely immature) surface sediment. Although our samples are not both thermally
mature, their diagenetic environments may have both activated S8 through a homolytic
pathway. In the case of Monterey, this pathway may have been thermolytic, and in the
case of Greenland, this pathway may have been photolytic.
Phenolic alcohols and esters (Np-22, Np-23) present in the nonpolar fraction
would suggest an immature sediment, because alcohol groups and C-O bonds tend to
be thermally unstable. We suggest that these compounds are misidentified, because they
have convoluted mass spectra. Alternatively, the Monterey sample contains non-endog-
enous compounds, a possibility that would confound our reconstruction of Monterey’s
diagenetic environment.
53
4.4.2. Paleobiology
Bacterial, eukaryotic, terrestrial, and photosynthetic organisms contributed to
the organic compounds observed in the Monterey sample. Bacteria are the likely parent
organism(s) of 28-norhopane (Np-27), which is commonly found in Monterey sediments
(Yamamoto et al., 2005). The biochemical precursors to hopanoids are poorly character-
ized, however, hopanoids are rarely found in eukaryotes and widely found in bacteria
(Brocks and Summons, 2004). Bacterial hopanoids are thought to regulate the fluidity of
cell membranes in a similar manner to eukaryotic steroids (Ibid.). Hopanoids have anaer-
obic synthetic pathways, although they have so far been observed mostly in aerobic bac-
teria (Ibid.). A similar compound to Np-27 is 28,30-dinorhopane, which Schouten et al.
(2001), identified as a free hydrocarbon in Monterey. This compound is associated with
euxinic environments and sulfidic water columns, but its parent organism is unknown
(Brocks and Summons, 2004). We suspect that anaerobic sulfate-reducing bacteria are the
parent organisms of 28-norhopane and 28,30-dinorhopane in the Monterey sample. Ste-
roids, which we found in both the desulfurized and nonpolar Monterey fractions, as did
Schouten et al. (2001), are diagnostic for eukaryotic organisms (Brocks and Summons,
2004).
The nonpolar fraction also containes a series of substituted PAHs (Np 1, 3-5, 6, 8,
9). We suspect that these PAHs are of biogenic, terrestrial origin. Since they are substi-
tuted, they probably do not result from combustion (Jiang et al., 1998). The same authors
identified cadalene, simonellite, and retene as plant-derived biomarkers. Although we do
not detect these PAHs, the ones we identified are structurally similar. While we interpret
these PAHs as indicators of terrestrial input, exogenous contamination cannot be ruled

54
out, since this sample is from an exposed outcrop. Moreover, oil-rich formations are
known to have organics diffusion across stratigraphic layers.
We observe phytane in the nonpolar fraction, as does Schouten et al. (2001). This
compound is abundant as a side-chain to chlorophyll, which is the photosynthetic pig-
ment used by algae and plants (Brocks and Summons, 2004).
4.4.3. Precursor-product relationships
Pristane, which we found in the Monterey nonpolar fraction (Np-11), has two
biological precursors: the phytyl side-chain of chlorophylls (Brocks and Summons,
2004), and tocopherols (Goossens et al., 1984). Tocopherols occur as ether-linked lipids
in the algae Botryococcus braunii, and these compounds may have a role as anti-oxi-
dants (Metzger and Rager, 2002). We observed beta-tocopherol and delta-tocopherol in
Monterey’s desulfurized fraction (Ds-12, Ds-18; Ds-12’s mass spectrum is convoluted),
but we did not find these compounds in the nonpolar fraction. We suggest that tocoph-
erols are a precursor to pristane in this part of the Monterey formation. The precursor
to sulfurized tocopherol must have had an additional functionality, since geochemical
sulfurization operates by replacing reactive functionalities with C-S bonds. Such a func-
tionality on tocopherol could have occurred as conjugated double bonds on the branched
side-chain, or as an alcohol or aldehyde group on an alkyl substituent of the compound’s
aromatic moiety.
Schouten et al. (2001) found lycopane as a free hydrocarbon in Monterey extracts,
and they suggested that this compund had a bacterial or algal source. Lycopane may be
a diagenetic product of lycopene (Brocks and Summons, 2004), a compound that we

55
observed in the desulfurized fraction (Ds-17). Lycopene is a biosynthetic precursor to
carotenoids, which are common across several taxa (Brocks and Summons, 2004), nota-
bly as photosynthetic pigment in sulfur bacteria. Since Monterey’s depositional environ-
ment was probably sulfidic, phototrophic, anaerobic sulfide-oxidizing bacteria could be
a source for the lycopene. However, such an organism would imply photic zone euxinia,
and we are unaware of evidence for large-scale die-offs in the marine photic zone during
the late Miocene. Lycopene has many double bonds, and an activating group near any
of them could have promoted sulfurization via the base-catalyzed nucleophilic addition
mechanism. In the minor-nonpolar fraction at Ds-17’s retention time, we found a mass
spectrum similar to that of Ds-17. Therefore, desulfurized lycopene may be cross-con-
tamination from the nonpolar fraction. In any case, we suggest a diagenetic relationship
between lycopene and lycopane.
Steroids can reveal changes in sulfurization efficiency in the Monterey sediments,
because they occur as mono- and di-unsaturated forms in both nonpolar and desulfur-
ized fractions (Schouten et al., 2001). For example, the abundance of a free diasterene
in some sample may be lower than its abundance in a slightly older sample. Perhaps this
compound’s sulfurized counterpart (if it could be found) is more abundant in the younger
sample than it is in the older sample. Such a relationship would be consistent with sulfu-
rization efficiency increasing with time. Unfortunately, practical constraints do not allow
us to make such an interpretation for our Monterey sample. We only have data for the
extract of one Monterey rock, and the steroids that we do identify have problematic mass
spectra. We find cholestane in the desulfurized and nonpolar fractions (Ds-11 and Np-
21), consistent with Schouten et al. (2001). These authors found cholestanes, ergostanes,
56
and stigmasteranes in the desulfurized fraction. Besides cholestane, we identified two
unusual steroids (Ds-14, Ds-15). We are unaware of sedimentary steroids with a dimethyl
substitution at carbon #4. Ds-15 has a convoluted mass spectrum. It is also relatively en-
riched in the m/z 191 ion, suggesting the presence of a terpenoid, or possibly a hopanoid.
Schouten et al. (2001) did identify hopanoids in their desulfurized fraction, suggesting a
bacterial contribution to the sulfurized compounds.
The nonpolar fraction contains a C23 tricyclic terpenoid (Np-18), whose struc-
ture is similar to a compound that Schouten et al. (2001) found only in their desulfurized
fraction. They argued that this terpenoid was a sulfur-bound moiety whose precursor was
functionalized, since they did not observe this compound as a free hydrocarbon. How-
ever, we suggest that our compound, nonpolar-18, is identical to Schouten et al.’s (2001)
desulfurized compound. Therefore, these terpenoids were either biosynthesized as such or
are products of a degradative reaction. This interpretation does not preclude the terpenoid
from having a functionalized precursor that was incorporated into an organic sulfur com-
pound. If our sample had a more intense thermal history than the samples that Schouten
et al. (2001) analyzed, the terpenoid may have historically been part of a sulfur-linked
macromolecule that underwent catagenic degradation (cf. Aizenshtat et al., 1995). In
this scenario, our terpenoid underwent geochemical desulfurization, and their terpenoid
underwent experimental desulfurization.
4.5. Greenland
Biomarkers in the Greenland sample suggest that at least two different species of
phototrophic algae, as well as two different species of sulfur bacteria, inhabit Brayasø.
57
The composition of the desulfurized fraction suggests that the rapid sulfurization in
Brayasø may employ a photochemical mechanism.
4.5.1. Biomarkers
We found a pentacyclic triterpenoid at low abundance in the desulfurized fraction
(Ds-20), but absent from the nonpolar fraction. This compound has a structure identical
to β-amyrin, if β-amyrin did not have a double bond. β-amyrin is a terpenoid diagnostic
for land plants (Brocks and Summons, 2004), so we interpret Ds-20 as indicative of low-
level terrestrial input to Braysø.
The C20 regularly branched isoprenoids found in Greenland’s nonpolar and
desulfurized fractions (Np 1-4; Ds 1-3) indicate photosynthetic parent organisms, such as
algae. Steroids, which are also present in Greenland’s desulfurized and nonpolar frac-
tions, indicate eukaryotic parent organisms (Brocks and Summons, 2004). Alkenones
present in the nonpolar fraction at high abundance are likely the remains of a specific
clade of prymnesiophytic algae (D’Andrea et al., 2006). D’Andrea and Huang (2005)
argued that the steroids and some other compounds in Brayasø are produced mostly by
non-prymnesiophytic phototrophs, because these compounds are 13C enriched relative to
the alkenones.
Purple sulfur bacteria inhabit Brayasø’s photic zone (McGowan et al., 2008).
These organisms are the likely source for the S8 found in Greenland, since they can
produce elemental sulfur by oxidizing H2S (Proctor, 1997). We strongly suspect that
sulphate-reducing bacteria inhabit Brayasø, since biological sulfate reduction is the most
important pathway to sedimentary sulfide formation (Werne et al., 2004).

58
4.5.2. Precursor-product relationships
We assume that the phytane (Ds-1) and phytenes (Ds 2, 3) in the desulfurized
fraction were part of an OSC prior to desulfurization. The precursors to these OSCs must
have had a reactive functionality additional to any functionalities that the desulfurized
compounds have. The phytene in the nonpolar fraction (Np-2) is a plausible precursor to
the desulfurized phytane (Ds-1), if photochemical sulfurization occurred. The lake has
pH ~10, so the nucleophilic base-catalyzed addition reaction would initially seem appli-
cable. However, this reaction is only effective on activated unsaturated bonds, such as α,
β-unsaturated carbonyl compounds (Fig. 2) (Aizenshtat et al., 1995). Instead, we invoke a
photochemical mechanism for sulfurization (Fig. 3), as proposed by Adam et al. (1998).
Absent a photochemical mechanisms, the desulfurized phytenes (Ds 2,3) would
present a similar problem to that of Ds-1: we do not observe phytadienes or phytenals in
the nonpolar fraction. Without activated double bonds, nucleophilic addition is unlikely.
It is possible that phytadienes or phytenals are present in the nonpolar fraction, but we
missed them because they have a low abundance; we do not rule out the base-catalyzed
nucleophilic mechanism. Another possibility is that the “desulfurized” phytenes are actu-
ally products of a reductive ester cleavage (E-4) rather than desulfurization. However, a
photochemical mechanism would suggest that Np-1, Np-3, and Np-4 are the precursors to
the desulfurized phytenes. Allylic alcohols such as phytol can become photochemically
oxidized to activated aldehydes (Amrani and Aizenshtat, 2004). Such compounds would
then be amenable to the sulfide nucleophilic addition reaction. Alternatively, the phytols
may enter a bio-hydrogenation pathway (cf. Kok et al., 2000), exit as phytanals (with no
59
C=C double bonds), and then undergo photochemical sulfurization via a thioketone inter-
mediate (Schneckenburger et al., 1998).
The desulfurized steroids all have alcohol substituents at the third carbon, indi-
cating that the sulfurization did not occur at this functional group. While many steroids
have more than one C=C double bond, we are not aware of any biogenic steroids with
activated double bonds (e.g., α, β-unsaturated carbonyls or conjugated dienes). Thus we
suggest that the precursors to the desulfurized steroids are di- and tri- unsaturated sterols
that underwent photochemical sulfurization. The absence of activated double bonds, or
activated carbonyls, in the free steroids leads us to rule out the nucleophilic mechanism
for these compounds. The desulfurized steroids have the same carbon skeletons as the
steroids in the nonpolar fraction, although we are less sure of how steroid functionalities
compare between the two fractions. This uncertainty stems from convoluted mass spectra
for several of the nonpolar-fraction steroids. We do not interpret these convolutions as
indicative of a high baseline (E-5).
The pentacyclic triterpanoid (Ds-20), similar to β-amyrin, lacks a double bond
between C12 and C13. We suggest that β-amyrin is the biological precursor to Ds-20, and
that photochemical sulfurization added a sulfide moiety to β-amyrin’s C12=C13 double
bond.
We do not find any likely nonpolar-fraction precursor compounds to the medium-
chain desulfurized alcohols (Ds-5,7,9,11). However, D’Andrea and Huang (2005) ob-
served free medium-chain monoacids (alkanoic acids) in several oligosaline Greenland
lakes, including Brayasø. Trace abundance of monoacids would be consistent with rapid,
efficient sulfurization via the photochemical mechanism. Such monoacids would origi-
60
nate from non-prymnesiophytes, as those observed by D’Andrea and Huang (2005) are
13C-enriched relative to alkenones.
For the alkanoates in the desulfurized fraction (Ds 6, 8, 10, 12) we do not find
any obvious nonpolar-fraction precursors. Alkenoates have been observed as free com-
pounds in other samples from this Greenland lake (D’Andrea, personal communication,
2009), but we do not observe them here, perhaps because they are efficiently sulfurized.
We suggest that alkenoates are the precursors to our desulfurized alkanoates. Depending
on the position of the unsaturation, alkenoates could have sulfurized by either the nucleo-
philic or the photochemical mechanism. Marlowe et al. (1984) reported on haptophytes
(the taxon in which prymnesiophytes reside) that produce C37-C39 alkenoates, and the
alkanoates that we identified have shorter chains (C19-C27).
Consistent with Kok et al.’s (2000) desulfurization of Ace lake sediments, we
do not observe long-chain alkenones in our desulfurized fraction. As those authors sug-
gest, the sulfurization reaction could differ drastically in efficiency between classes of
compounds. Alternatively, the sulfurized alkenones are part of macromolecular networks
too large to elute in the most-polar fraction; they may be present only as asphaltenes or
kerogen. In the latter scenario, we did not encounter alkenones in the desulfurized frac-
tion because we did not desulfurize the asphaltenes or the kerogen.
Figure 31 presents a model for sulfurization in the Greenland lake Brayasø, which
integrates our results with those of Anderson et al. (1999) (Fig. 5). Adam et al. (1998)
suggested that photochemical sulfurization would be limited to anoxic photic zones. If
that suggestion were correct, then we would be hard-pressed to explain the sterols re-
leased by our desulfurization. Brayasø’s photic zone is relatively deep (D’Andrea and
61
Huang, 2005), but measurements by Anderson et al. (1999) indicate that the entire zone
is oxic. We suggest that photochemical sulfurization occurs in the chemocline, where

2-
Sulfide concentration SO4
there are steep gradients of light Minimum Maximum
0
penetration, dissolved oxygen, and
presumably, dissolved sulfides,

5 Mixed layer
which would diffuse upward from

A) CO2 + H2O CH2O
the lake’s bottom. On the other B) CH2O + H2S CH2OS
10
Photosynthetic
hand, sunlight is not the only way Water zone
depth (m)
to explain rapid sulfurization, even

15
if this reaction involves a radical

2- Anoxic zone
C) SO4 H2S
mechanism (E-7). Sulfurization 20 D) CH O + H S
2 2 CH2OS
in Brayasø might occur via some Sediment
CH2O, CH2OS
mechanism that we have not re- 25
viewed (cf. Filley et al., 2002 and Figure 31. Model for organic sulfur compound formation in
the Greenland lake Brayasø. We show four major reactions
(A-D). A is the carbon fixation reaction. B is the photochemi-
Schneckenburger et al., 1998). cal sulfurization of organic matter. C is bacterial sulfate
reduction. D is the base-catalyzed sulfurization of organic
matter. We suggest that the dissolved sulfide has an inverse
abundance relationship to the dissolved oxygen, because these
two species are reactive with one another.
4.6. Sulfurization potential
We estimated the “sulfurization potential” for the major compounds in each of our
samples (Table 2; see endnote 8). These values correspond to the amount of a compound
that is sulfur-bound relative to the total amount of the compound. Table 2 implies that
regular isoprenoids are more susceptible to sulfurization than steroids, that the measured
compounds in Greenland are more susceptible to sulfurization than their counterparts in
Monterey, and that alkenones are not sulfurized in either environment. (“N/a” indicates
62
Table 2. Mean sulfurization Monterey Greenland
potential for the major com- Mean S.E. Mean S.E.
pound families of Monterey
and Greenland. Regular isoprenoids 1.3% 0.5% 23.5% 5.5%
Steroids 0.9% 0.3% 7.8% 6.8%
Alkenones n/a n/a -3.2% 2.9%
S.E. = one standard error of the mean
that we did not identify alkenones in any of the Monterey fractions, whereas the negative
value for Greenland indicates that the area in the alkenone retention time was larger in
the blank desulfurization than it was in the Greenland desulfurizations.) These sulfuriza-
tion potentials reflect desulfurization of maltene-dominated fractions; they do not reflect
desulfurization of the asphaltene or kerogen fractions, which may contain organic sulfur
compounds beyond the scope of this study.
Section 4 endnotes
1. When we did not centrifuge the reaction mixture, we allowed ~10 mins for the
Ni2B particles to settle and transferred the supernatant to a clean vial. However, the super-
natant remained cloudy. After drying down the supernatant, a residual of Ni2B remained
in the vial. To obtain the nonpolar yield, we extracted this vial with hexane. The Ni2B
may have impeded the hexane’s dissolution of the nonpolar yield.
2. We intentionally used different reaction conditions between DS-1, 2, and 3 to
test their effect on the yield. However, we introduced a confounding variable by inac-
curately measuring the volume of solvent used for the Most-Polar sample vials. We
assumed each most-polar vial contained 4.5 mL of solvent, but they actually contained
about 4 mL. We performed all of the DS-1 reactions together, followed by the DS-2
63
reactions, followed by the DS-3 reactions. We did not realize our error until after we had
taken the first 1.5 mL aliquots.
3. We normalized the desulfurization yield to the Total Organic Carbon (Table 3),
using data from several sources, which we explain below Table 3. All of the desulfuriza-
tion yields from the literature were obtained using a nickel boride reaction on the polar
fraction of an asphaltene-precipitated sedimentary extract. For some samples, we report
the TOC as a range, in lieu of a measurement of the actual rock whose extract was desul-
furized. This range is our estimate based on data from stratigraphically similar samples.
For our samples, we report the desulfurization yield as a range (min released/TOE and
max released/TOE), which we explain in E-6. A sample that uses one or more measure-
ment ranges requires us to express its TOC-normalized desulfurization yield as a range.
Table 3 Min Max min

(released (released TOE/ TOC/
max
TOC/
min
TOE/
max
TOE/
Min Max
released/ released/
Description / TOE)a / TOE)b rockc rockd rocke TOCf TOCg TOCh TOCi
Monterey, Shell Beach,
11.1 Ma 0.00265 0.059 1.6E-04
Monterey, Naples Beach,
6.7-7.8 Ma (KG-7) 0.0005 0.002 0.037 0.067 3.4E-05
Monterey, Naples Beach,
6.7-7.8 Ma (KG-8) 0.0035 0.008 0.058 0.131 4.6E-04
Monterey DS-0 0.0028 0.0034 0.013 0.020 0.050 0.260 0.650 7.3E-04 2.21E-03
Monterey DS-1 0.0009 0.0013 0.013 0.020 0.050 0.260 0.650 2.3E-04 8.45E-04
Monterey DS-2 0.0005 0.0009 0.013 0.020 0.050 0.260 0.650 1.3E-04 5.85E-04
Monterey DS-3 0.0007 0.0011 0.013 0.020 0.050 0.260 0.650 1.8E-04 7.15E-04
Greenland DS-1 0.0056 0.0074 0.027 0.117 0.230 0.230 1.3E-03 1.70E-03
Greenland DS-2 0.0019 0.0037 0.027 0.117 0.230 0.230 4.4E-04 8.51E-04
Greenland DS-3 0.0011 0.003 0.027 0.117 0.230 0.230 2.5E-04 6.90E-04
Lake Cadagno, 0-6 Ya 0.0177 0.092 1.6E-03
Lake Cadagno, 50-56 Ya 0.00276 0.063 1.7E-04
Vena del Gesso basin,

Upper Miocene shale 0.117 0.002 0.002 0.013 0.154 1.000 1.8E-02 1.17E-01
64
Table 3 footnotes
a. Monterey, Shell Beach measurement from Schouten et al. (1997): sample
“SB-18”, from their Figure 7. Monterey, Naples Beach data from Schouten et al. (2001):
samples “KG-7” and “KG-8”, from their Figure 9.10. Monterey DS and Greenland DS
measurements are explained in endnote 6. Lake Cadagno measurements from Putschew
et al. (1995): their samples “0-2 [cm]” and “18-20 [cm]”, from their Table III. Vena del
Gesso data from Schouten et al. (1993): their sample “VDG polar fraction,” from their
Table 1.
b. Monterey DS and Greenland DS measurements are explained in endnote 6.
c. Monterey, Naples Beach data from Katz and Royale (2001), their Table 6.5,
samples KG-7 and KG-8. Monterey DS and Greenland DS EOM measurements are
explained in section 2.3 of this report. Vena del Gesso datum from Kohnen et al. (1991),
“Extraction and Fractionation” section.
d. Monterey, Naples Beach data from Katz and Royale (2001), their Table 6.1,
samples KG-7 and KG-8. Monterey DS TOC measurements are from Katz and Royale
(2001), their figure 6.2. Greenland DS TOC measurements are from D’Andrea (pers.
comm., 2009). Vena del Gesso datum from Lugli et al. (2007), their Table 2.
e. Monterey DS TOC measurements are from Katz and Royale (2001), their figure
6.2. Vena del Gesso datum from Lugli et al. (2007), their Table 2.
f. The Monterey Shell Beach value is from Schouten et al. (1997), their Table
2, sample SB-18. For Monterey Naples Beach, we divided the (TOE/rock) values by
the (min TOC/rock) values. For Monterey DS, Greenland DS, and Vena del Gesso, we
65
divided the (TOE/rock) values by the (max TOC/rock) values. Lake Cadagno values are
from Putschew et al. (1995), their samples “0-2 [cm]” and “18-20 [cm]”, from their Table
I.
g. For all samples in this column, we divided the (TOE/rock) values by the (min
TOC/rock) values. We forced the quotient for Vena del Gesso to 1, since TOE cannot be
greater than TOC.
h. For all samples in this column, we multiplied the “Min (released/TOE)” value
by the “Min TOE/TOC” value.
i. For all samples in this column, we multiplied the “Max (released/TOE)” value
by the “Max TOE/TOC” value.
4. Nickel boride is considered a gentle desulfurization reaction in comparison
to alternatives such as raney nickel (Back et al., 1992; Schouten et al., 1993). However,
Putschew et al. (1996) found that nickel boride desulfurization can reductively cleave
ester bonds in chlorophyll a to produce phytenes. Those authors suggest the nickel boride
reaction be subjected to further systematic study.
5. The Greenland nonpolar fraction’s convoluted mass spectra are likely an in-
strument artifact. The GC-MS data for this fraction show much broader peaks after ~22
minutes than the GC-FID data show, and we are unsure of the reason.
6. The four tables (4-7) shown below are the four steps we used to calculate our
desulfurization yields. Table 4 shows the raw data (total area under each GC-FID trace,
66
omitting the solvent peaks). Beneath the areas are letters indicating a scaling correction
(explained below). Table 5 shows the data with the corrections applied to each value.
Table 6 shows the mass equivalent for each area in milligrams, which we found using
the HMB calibration curves. Table 7 shows the mass of each fraction minus the mass
in the procedural blank. Table 7 also shows best- and worst-case estimates of the mass
of nonpolar cross-contaminants in the desulfurized fraction (4% VOC residue and 19%
VOC residue). Table 7 finds the minimum desulfurization yield by subtracting the “19%
VOC residue” column from the “Desulf” column. We find the maximum desulfurization
yield by subtracting the 4% VOC column from the Desulf column. We find the % yields
by dividing the yield columns by the TOE weight column.
67
Table 4. Raw GC-FID measurements (pA*s)
Detector &
Nonpolar Minor-Nonpolar Desulf column
Blank-0 234.472 171.888 850.718 front Corrections
correction B B B,C A Divide by 3 for triplicate design
B Divide by 20 for 20x concentration
Monterey-0 187307.826 16227.104 12330.353 front C Multiply by 10/9 for missing BSTFA aliquot
correction B B B,C
Calibration curves
Monterey 1 1314132.125 53396.427 10542.730 front front detector y= 4.26E-05 x
Monterey 2 7869.352 back detector y= 5.85E-05 x
Monterey 3 9454.325
correction A,B A,B B
Explanation of scaling corrections:
Greenland 1 47662.066 8608.460 1916.234 back We divided some of the non-
polar areas by 3 (correction A), because
Greenland 2 1033.257
Greenland 3 849.907
correction A A our workflow divided their most-polar
fractions into 3 aliquots (Fig. 7).
Table 5. Corrected GC-FID Measurements (pA*s)
Detector & We divided many of the areas
Nonpolar Minor-Nonpolar Desulf column by 20 (B), because for these FID runs,
the volume in the sample vial was 50
Blank-0 11.724 8.594 47.262 front
Monterey-0 9365.391 811.355 685.020 front uL, twenty times as small as the volume
used for the HMB calibrations (1 mL).
Monterey 1 21902.202 889.940 527.137 front
Monterey 2 21902.202 889.940 393.468
Monterey 3 21902.202 889.940 472.716 We multiplied some of the
runs by 10/9 (C), because these frac-
Greenland 1
Greenland 2
15887.355
15887.355
2869.487
2869.487
1916.234
1033.257
back
tions were missing 10% of their yield
Greenland 3 15887.355 2869.487 849.907 (removed for a different experiment that
we do not report on).
Table 6. Mass calibrated (mg)
Nonpolar Minor-Nonpolar Desulf
Blank-0 0.000 0.000 0.002
Monterey-0 0.399 0.035 0.029
Monterey 1 0.933 0.038 0.022

Monterey 2 0.933 0.038 0.017
Monterey 3 0.933 0.038 0.020
Greenland 1 0.930 0.168 0.112

Greenland 2 0.930 0.168 0.060
Greenland 3 0.930 0.168 0.050
Table 7. Procedural blank subtracted (mg)

TOE
4% VOC 19% VOC min desuf max desulf weight min % max %
Nonpolar Minor-Nonpolar Desulf residue residue yield yield (mg) yield yield
Monterey-0 0.398 0.034 0.027 0.0014 0.0065 0.0207 0.0258 7.5 0.28% 0.34%
Monterey 1 0.932 0.038 0.020 0.0015 0.0071 0.0133 0.0189 15 0.09% 0.13%
Monterey 2 0.932 0.038 0.015 0.0015 0.0071 0.0076 0.0132 15 0.05% 0.09%
Monterey 3 0.932 0.038 0.018 0.0015 0.0071 0.0110 0.0166 15 0.07% 0.11%
Greenland 1 0.930 0.168 0.110 0.0067 0.0319 0.0783 0.1035 13.9 0.56% 0.74%
Greenland 2 0.930 0.168 0.058 0.0067 0.0319 0.0266 0.0518 13.9 0.19% 0.37%
Greenland 3 0.930 0.168 0.048 0.0067 0.0319 0.0159 0.0410 13.9 0.11% 0.30%
68
7. The mitochondrial electron transport chain constantly produces free rad-
icals (superoxides) (Nelson and Cox, 2005). Perhaps these species could abstract
sulfur radicals.
8. We calculated the sulfurization potentials for each of the four Monterey

desulfurized fractions: SM,j , and SM 0 .
minimum desulfurization yieldM,j

SM,j = ,
minimum desulfurization yieldM,j + 13 nonpolar yieldM
where M denotes the M onterey measurement set, and 1 ≤ j ≤ 3. j corre-

sponds to three of the four Monterey desulfurized fractions (Sec. 2.3; Fig. 7).
minimum desulfurization yieldM 0

SM 0 = ,
minimum desulfurization yieldM 0 + nonpolar yieldM 0
where M 0 denotes the M onterey DS-0 fraction, and the M onterey-0 nonpolar
fraction (Sec. 2, endnote 7).
The mean sulfurization potential for the Monterey sample, SM , is
 3
1
SM = (SM 0 + SM,j ).
4 j=1
The standard error of SM is
  
 3
S.D.M 1 
S.E.M = √ , where S.D.M = (SM − SM 0 )2 + (SM − SM,j )2
4 3 j=1
69
We calculated Si,j , the sulfurization potentials for each of the three Greenland
desulfurized fractions, for each of the two Greenland measurement sets.
minimum desulfurization yieldi,j

Si,j = ,
minimum desulfurization yieldi,j + 13 nonpolar yieldi
where i = GA or GB, and 1 ≤ j ≤ 3.
i denotes one of two measurement sets: Greenland-A, or Greenland-B (see
below).
j denotes one of the three desulfurized fractions in each measurement set.
The mean sulfurization potential for the Greenland sample, SG , is
1 1
SG = SGA  + SGB , where
2 2
3 3
1 1
SGA  = SGA,j , and SGB  = SGB,j .
3 j=1 3 j=1
We performed the two sets of Greenland measurements, GA and GB, as GC-

FID runs on September 30, 2008, and December 14, 2008, respectively. The GA
data show regular isoprenoids with a high relative abundance and steroids with a
low relative abundance; the GB data show regular isoprenoids with a low relative
abundance and steroids with a high relative abundance. We suspect that the following
variations in experimental conditions contributed to this discrepancy: diﬀerent split
modes between GA (10:1 split) and GB (splitless); isoprenoid evaporation between
September and December; and column replacement.
The standard error of SG is

S.D.G
S.E.G = √ , where S.D.G = (SG − SGA )2 + (SG − SGB )2 .
2
Our calculations of desulfurization potential each corrected for the yield of a

procedural blank.
70
5. Conclusion
Sulfurization occurs within 40 years of an organism’s death in the Greenland lake
Brayasø. The kinetics of this process seem to differ between major compound families.
Alkenones in the Greenland setting are not sulfurized, although we base this conclusion
on a limited sampling. Alkenone sulfurization could potentially confound paleotem-
perature reconstructions, which use free alkenone abundance ratios (e.g., D’Andrea and
Huang, 2005). Since alkenones compose the majority of Greenland’s nonpolar fraction,
and the concentration of these compounds in Brayasø is the highest yet reported for
lacustrine surface sediments (D’Andrea and Huang, 2005), we would be surprised if sul-
furized alkenones were absent from the asphaltene or kerogen fractions.
We suspect that sulfurization progresses in the Brayasø chemocline, where the
availability of sunlight, oxygen, and sulfides changes rapidly. If photochemical sulfuriza-
tion can occur in oxic water, then its operation is more widespread than previously recog-
nized. Studies on the organic matter dissolved throughout Brayasø’s water column would
bear on our argument for photochemical sulfurization. Future study of Brayasø could
also desulfurize multiple sub-surface samples, which may reveal trends in the relative
abundances of free and bound biomarkers. Sub-surface desulfurizations would also test
for alkenone sulfurization late in diagenesis, which this study’s surface desulfurizations
could not have revealed.
Stable-isotope measurements often show differences between free compounds
and their sulfur-bound counterparts. In view of the isotopic differences already observed
between alkenones and other Brayasø lipids, such measurements on the desulfurization
yield would be interesting. Desulfurizing the asphaltene or kerogen fractions of Brayasø

71
samples may reveal sulfur-bound alkenones. Brayasø kerogen may also be a useful test
for the feasibility of desulfurizing meteoritic insoluble organic matter.
The ~7 million year old Monterey shale that we desulfurized contains steroids,
hopanoids, regular isoprenoids, and substituted PAHs. Respectively, each of these com-
pound classes suggests input from eukaryotic, bacterial, photosynthetic, and terrestrial
organisms. Sulfur-bound tocopherols may be precursors to free pristane. We observe
some indicators of thermal modification to this rock, such as aromatic moieties and a
dibenzothiophene. Microbial biodegradation explains the near absence of n-alkanes as
free hydrocarbons in Monterey, and this process may also have modified the organic sul-
fur compounds.
Biomarker sulfurization efficiency may depend on sulfate availability and the ac-
tivity of sulfate-reducing bacteria, as suggested by a comparison of the Monterey desulfu-
rization yields with those from Vena del Gesso. Future work on samples from Monterey
may require asphaltene precipitations, and carefully planned chromatographic separations
of the desulfurized fraction (cf. Schouten et al., 2001). We suggest that future experi-
ments proceed with an understanding of the diversity and complexity that can be present
in a geochemical extract.
72
References
Adam, P.; Philippe, E.; Albrecht, P. Photochemical sulfurization of sedimentary organic

matter: A widespread process occurring at early diagenesis in natural environments? Geo-
chimica et Cosmochimica Acta. 1998, 62, 265-271.
Aizenshtat, Z.; Krein, E. B.; Vairavamurthy, M. A.; Goldstein, T. P. Role of sulfur in the
transformations of sedimentary organic matter: a mechanistic overview. In: Geochemical
transformations of sedimentary sulfur; Vairavamurthy, M. A., Schoonen, M. A. A., Eds.;
American Chemical Society: Washington, DC, 1995; 16-37.
Amrani, A.; Aizenshtat, Z. Photosensitized oxidation of naturally occurring isoprenoid

allyl alcohols as a possible pathway for their transformtion to thiophenes in sulfur rich
depositional environments. Organic Geochemistry. 2004, 35, 693-712.
Anderson, N. J.; Leng, M. J. Increased aridity during the early Holocene in West Green-
land inferred from stable isotopes in laminated-lake sediments. Quaternary Science
Reviews. 2004, 23, 841–849.
Anderson, N. J.; Bennike, O.; Christoffersen, K.; Jeppesen, E.; Markager, S.; Miller,
G.; Renberg, I. Limnological and palaeolimnological studies of lakes in south-western
Greenland. Geology of Greenland Survey Bulletin. 1999, 183, 68–74.
Bada, J. L. Amino acid cosmogeochemistry. Phil. Trans. R. Soc. Lond. B. 1991, 333, 349-
358.
Back, T. G.; Yang, K.; Krouse, H. R. Desulfurization of Benzo- and Dibenzothiophenes

with Nickel boride. J. Org. Chem. 1992, 57, 1986-1990.
Brocks, J. J.; Summons, R. E. Sedimentary hydrocarbons, biomarkers for early life. In

Biogeochemistry 8, Treatise on Geochemistry; Schlesinger, W. H., Ed.; Elsevier: Oxford,
2004; 63–116.
Colby, B. N. Spectral Deconvolution for Overlapping GC/MS components. J. Am. Soc.

Mass. Spectrom. 1992, 3, 558-562.
Conte, M. H.; Thompson, A. T.; Lesley, D.; Harris, R. P. Genetic and physiological influ-
ences on the alkenone/alkenoate versus growth temperature relationship in Emiliania
huxleyi and Gephyrocapsa oceanica. Geochimica et Cosmochimica Acta. 1998, 62, 51-
68.
D’Andrea, W. J.; Lage, M.; Martiny, J. B. H.; Laatsch, A. D.; Amaral-Zettler, L. A.;
Sogin, M. L.; Huang, Y. Alkenone producers inferred from well-preserved 18S rDNA in
Greenland lake sediments. Journal of Geophysical Research. 2006, 111, G03013.
73
D’Andrea, W. J. Personal communication, 11 March 2009.
D’Andrea, W. J. Personal communication, 16 April 2009.
D’Andrea, W. J.; Huang, Y. Long chain alkenones in Greenland lake sediments: Low
del13C values and exceptional abundance. Organic Geochemistry. 2005, 36, 1234–1241.
Filley, T. R.; Freeman, K. H.; Wilkin, R. T.; Hatcher, P. G. Biogeochemical controls on re-
action of sedimentary organic matter and aqueous sulfides in holocene sediments of Mud
Lake, Florida. Geochim. Cosmochim. Acta. 2002, 66, 937-954.
Forster, A.; Sturt, H.; Meyers, P. A.; the Leg 207 Shipboard Scientific Party. Molecular
Biogeochemistry of Cretaceous Black shales from the Demerara Rise: preliminary ship-
board results from sites 1257 and 1258, Leg 207. In Proceedings of the Ocean Drilling
Program, Initial reports. Volume 207. Erbacher, J., Mosher, D. C., Malone, M. J., et al.,
2004.
Goossens, H.; de Leeuw, J. W.; Schenck, P. A.; Brassel, S. C. Tocopherols as likely pre-
cursors of pristane in ancient sediments and crude oils. Nature. 1984, 312, 440-442.
Isaacs, C. M. Depositional framework of the Monterey formation, California. In The

Monterey Formation: from rocks to molecules; Isaacs, C. M., Rullkötter, J., Eds.; Colum-
bia University Press: New York, Chichester, West Sussex, 2001, 1-30.
Jiang, C.; Alexander, R.; Kagi, R. I.; Murray, A. P. Polycyclic aromatic hydrocarbons
in ancient sediments and their relationships to palaeoclimate. Org. Geochem. 1998, 29,
1721-1735.
Katz, B. J.; Royle, R. A. Variability of source rock attributes in the Monterey Formation,
California. In The Monterey Formation: from rocks to molecules; Isaacs, C. M., Rul-
lkötter, J., Eds.; Columbia University Press: New York, Chichester, West Sussex, 2001,
107-129.
Kohnen, M. E. L.; Sinninghe Damsté, J. S.; Dalen, A. C. K.; de Leeuw, J. W. Di- or

polysulphide-bound biomarkers in sulphur-rich geomacromolecules as revealed by selec-
tive chemolysis. Geochimica et Cosmochimica Acta. 1991, 55, 1375-1394.
Kok, M. D.; Rijpstra, W. I. C.; Robertson, L.; Volkman, J. K.; Sinninghe Damsté, J. S.
Early steroid sulfurisation in surface sediments of a permanently stratified lake (Ace
Lake, Antarctica). Geochimica et Cosmochimica Acta. 2000, 64, 1425–1436.
Krijgsman, W.; Hilgen, F. J.; Raffi, I.; Sierro, F. J.; Wilson, D. S. Chronology, causes and
progression of the Messinian salinity crisis. Nature. 12 August 1999, 400, 652.
74
Krupcík, J.; Oswald, P.; Oktavec, D.; Armstrong, D. W. Calibration of GC-FID and IR
Spectrometric methods for determination of high boiling petroleum hydrocarbons in envi-
ronmental samples. Water, Air, and Soil Pollution. 2004, 153, 329–341.
Lugli, S.; Bassetti, M. A.; Manzi, V.; Barbieri, M.; Longinelli, A.; Roveri, M. The
Messinian ‘Vena del Gesso’ evaporites revisited: characterization of isotopic composition
and organic matter. Geological Society, London, Special Publications. 2007, 285, 179-
190.
Marlowe, I. T.; Brassell, S. C.; Eglinton, G.; Green, J. C. Long chain unsaturated ketones
and esters in living algae and marine sediments. Org. Geochem. 1984, 6, 135-141.
McGowan, S.; Juhler, R. K.; Anderson, N. J. Autotrophic response to lake age, conductiv-
ity and temperature in two West Greenland lakes. J. Paleolimnol. 2008, 39, 301-317.
Metzger, P.; Rager, M. Lycopanerols H, two high molecular weight ether lipids from
Botryococcus braunii comprising an alpha-tocopherol unit. Tetrahedron Letters. 2002, 43
2377–2380.
Nelson, D. L.; Cox, M. M. Lehninger Principles of Biochemistry, Fourth Edition; W. H.

Freeman and Company: New York, 2005, 722.
Peters, K. E.; Walters, C. C.; Moldowan, J. M. The Biomarker Guide, Second Edition;
Cambridge University Press: Cambridge, UK, 2007.
Putschew, A.; Scholz-Böttcher, B. M.; Rullkötter, J. Organic geochemistry of sulfur-rich

surface sediments of meromictic Lake Cadagno, Swiss Alps; In Geochemical transforma-
tions of sedimentary sulfur; Vairavamurthy, M. A., Schoonen, M. A. A., Eds.; American
Chemical Society: Washington, DC, 1995; 59-79.
Putschew, A.; Scholz-Böttcher, B. M.; Rullkötter; J. Early diagenesis of organic matter

and related sulphur incorporation in surface sediments of meromictic Lake Cadagno in
the Swiss Alps. Org. Geochem. 1996, 25, 379 –390.
Proctor, L. M. Nitrogen-fixing, photosynthetic, anaerobic bacteria associated with pelagic

copepods. Aquat. Microb. Ecol. 1997, 12, 105-113.
Roveri, M.; Manzi, V.; Ricci Lucchi, F.; Rogledi, S. Sedimentary and tectonic evolution
of the Vena del Gesso basin (Northern Apennines, Italy): Implications for the onset of the
Messinian salinity crisis. GSA Bulletin. April 2003, 115, 387–405.
Russell, J. M.; Werne, J. P. Climate change and productivity variations recorded by

sedimentary sulfur in Lake Edward, Uganda/D. R. Congo. Chemical Geology. 2009,
doi:10.1016/j.chemgeo.2009.03.020
75
Sarmiento, J. L.; Gruber, N. Ocean Biogeochemical Dynamics; Princeton University
Press: Princeton and Oxford, 2006.
Schaeffer, P.; Reiss, C.; Albrecht, P. Geochemical study of macromolecular organic mat-
ter from sulfur-rich sediments of evaporitic origin (Messinian of Sicily) by chemical
degradations. Org. Geochem. 1995, 23, 567-581.
Schneckenburger, P.; Adam, P.; Albrecht, P. Thioketones as key intermediates in the

reduction of ketones to thiols by HS- in natural environments. Tetrahedron Letters. 1998,
39, 447-450.
Schouten, S.; Pavlovic, D.; Sinninghe Damsté, J. S.; de Leeuw, J. W. Nickel boride: an
improved desulphurizing agent for sulphur-rich geomacromolecules in polar and as-
phaltene fractions. Org. Geochem. 1993, 20, 901-909.
Schouten, S.; Schoell, M.; Rijpstra, W. I. C.; Sinninghe Damsté, J. S.; de Leeuw, J. W. A
molecular stable carbon isotope study of organic matter in immature Miocene Monterey
sediments, Pismo basin. Geochimica et Cosmochimica Acta. 1997, 61, 2065-2082.
Schouten, S.; De Loureiro, M. R. B.; Sinninghe Damsté, J. S.; de Leeuw, J. W. Molecular

biogeochemistry of Monterey sediments, Naples Beach, California, I: Distributions of
Hydrocarbons and organic sulfur compounds. In The Monterey Formation: from rocks
to molecules; Isaacs, C. M., Rullkötter, J., Eds.; Columbia University Press: New York,
Chichester, West Sussex, 2001, 150-173.
Schouten, S. Personal communication, 27 August 2008.
Sessions, A. L. Personal communication, 15 March 2009.
Sessions, A. L. Personal communication, 16 August 2008.
Sinninghe Damsté, J. S.; Rijpstra, W. I. C.; de Leeuw, J. W.; Schench, P. A. Origin of

organic sulfur compounds and sulfur-containing high molecular weight substances in
sediments and immature crude oils. Org. Geochem. 1988, 13, 593-606.
Sutton, P. A.; Lewis, C. A.; Rowland, S. J. Isolation of individual hydrocarbons from the
unresolved complex hydrocarbon mixture of a biodegraded crude oil using preparative
capillary gas chromatography. Organic Geochemistry. 2005, 36, 963–970.
Vai, G. B.; Ricci Lucchi, F. Algal crusts, autochthonous and clastic gypsum in a cannibal-
istic evaporite basin: a case history from the Messinian of Northern Apennines. Sedimen-
tology. 1977, 24, 211-244.
Vaughan, W. E.; Rust, F. F. The photo-addition of hydrogen sulfide to olefinic bonds. J.

Org. Chem. 1942, 07 (6), 472-476.
76
Wenger, L. M.; Isaksen, G. H. Control of hydrocarbon seepage intensity on level of bio-
degradation in sea bottom sediments. Organic Geochemistry. 2002, 33, 1277–1292.
Werne, J. P.; Hollander, D. J.; Lyons, T. W.; Sinninghe Damsté, J. S. Organic sulfur bio-
geochemistry: recent advances and future research directions. In Sulfur biogeochemistry -
Past and present; Amend, J. P., Edwards, K. J., Lyons, T. W., Eds.; Geological Society of
America: Boulder, Colorado, 2004; Special Paper 379, 135-150.
Werne, J. P.; Lyons, T. W.; Hollander, D. J.; Schouten, S.; Hopmans, E. C.; Sinninghe
Damsté, J. S. Investigating pathways of diagenetic organic matter sulfurization using
compound-specific sulfur isotope analysis. Geochimica et Cosmochimica Acta. 2008, 72,
3489-3502.
Yamamoto, M.; Naraoka, H.; Ishiwatari, R.; Ogihara, S. Carbon isotope signatures of
bacterial 28-norhopanoic acids in Miocene-Pliocene diatomaceous and phosphatic sedi-
ments. Chemical Geology. 2005, 218, 117–133.
77

Joshua Stern Final

Uploaded by

Document Informationclick to expand document information

Copyright:

Available Formats

Joshua Stern Final

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Joshua Stern Final

Uploaded by

Copyright:

Available Formats

Sulfur-bound biomarkers of a Monterey shale

and a Greenland lake sediment

Joshua Gallant Stern

Submitted in partial fulfillment of the requirements for the degree of

Advisor: Prof. Yongsong Huang

I am especially grateful to Dr. Yongsong Huang for advising me on this senior

Scheinerman for helpful advice.

port during the summer of 2008.

Seeking to reconstruct the biogeochemical processes that produced organic

desulfurization reaction to release hydrocarbons from sulfur-bound macromolecules

not otherwise amenable to chromatographic analysis. We desulfurized two geochemi-

revealed a precursor-product relationship between tocopherol and pristane, for Monterey.

Greenland’s composition may indicate that photochemical sulfurization occurs in the

but we found an abundance of sulfurized isoprenoids. Greenland’s relatively high overall

Greenland findings suggest that photochemical sulfurization may be more widespread

Figure 1. Biomarker examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

Table 1. Desulfurizations of synthetic standards and reaction yields . . . . . . . . . . . . 24

We aim to reconstruct biogeochemical processes that produced sedimentary

of the mechanisms by which it operates. We also

introduce the analytical technique of desulfuriza-

tion. Later, we describe our experimental work on

Monterey shale and sediment from the Greenland

lake Brayasø, and we offer paleoenvironmental Steranes

sulfur-bound compounds in these samples.

Organic carbon can comprise as much as

17% of a sedimentary rock’s mass; however, this Hopanoid (Pentakishomohopane)

Figure 1. Some biomarkers whose precur-

heterotrophs (E-2). Biomolecules vary in their ability to withstand decomposition. Mem-

accumulating at the sediment-water interface gradually compact the underlying sediment

Another preservative process that occurs during diagenesis is sulfurization. In

The abiotic chemical mechanisms by which sulfur incorporates into organic

catalyzed nucleophilic addition, is well-established (Aizenshtat et al., 1995). The second

mechanism, light-induced free-radical addition, seems to be gaining popularity (Amrani

cleophilic addition mechanism, and

free-radical addition. Several other

under laboratory conditions.

Natural sulfurization can occur rapidly, as shown by OSCs found in surface

Sulfurized biomarkers can be important to paleoreconstructions, as Werne et al.

descendents can occur as OSCs, as defunctionalized hydrocarbons, as both, or as neither.

can produce an erroneous trend, or overlook a compound altogether. In some sediments,

pounds that would be metabolized rapidly outside of a macromolecular network.

measuring the abundance of macromolecular, polar organic sulfur compounds (OSCs)

atom is electronegative, and sulfurization often forms macromolecular, sulfide-linked

networks of multiple biomolecules that can contain other electronegative substituents.

not at all on a polar GC column. As a workaround, organic geochemists developed a

potential biomarkers previously trapped as OSCs (Fig. 4).

Figure 4. I. Desulfurization of 1-octadecanethiol yields octadecane. II. Desulfurization of S-ben-

methodology to facilitate future work in our laboratory. In addition, we aim to infer

pH, oxygen (mg/l), temperature (°C) and

Water Photic zone Water

A) Greenland, Brayasø B) Monterey

Our blank sample is sand baked overnight at 500 °C.

uptake and biosynthetic utilization. Measurements of the 34S depletion in sedimentary

3. Barring favorable circumstances such as a deposition environment that has been

4. Chromatography is a laboratory technique that takes advantage of a com-

is crucial if we wish to understand what makes up a mixture of hundreds or thousands of

unidentified compounds from a geochemical extract. Gas Chromatography (GC) sepa-

GC is mostly used for fine separation of individual compounds (e.g., nonane,

chromatography to analyze, identify and measure