A review of current applications of mass spectrometry for biomarker/molecular tracer elucidations

Authors

  • Bernd R.T. Simoneit

    Corresponding author
    1. Environmental and Petroleum Research Group, College of Oceanic and Atmospheric Sciences and Department of Chemistry, Oregon State University, Corvallis, Oregon 97331-5503
    • College of Oceanic and Atmospheric Sciences, COAS Administration Building 104, Oregon State University, Corvallis, OR 97331-5503.
    Search for more papers by this author

Abstract

Mass spectrometry, especially coupled with gas chromatography or tandem, is the analytical method of choice in elucidations of biomarker compounds present in organic mixtures extracted from geological, environmental, or biological samples. This review describes the biomarker concept; i.e., the precursor natural products to the geological/environmental derivatives, and their application as tracers in the geosphere and ambient environment. The mass spectrometric methods currently utilized for such analyses are reviewed, and typical examples of applications are described with a general key to the literature. © 2004 Wiley Periodicals, Inc., Mass Spec Rev 24:719–765, 2005

I. INTRODUCTION

This review is written to give the reader an insight and a literature that are key into the potential of using biomarker determinations by mass spectrometrists in research on global and environmental problems. It is not an exhaustive literature survey about every application of mass spectrometry to biomarker research. Furthermore, mass spectrometry applications will be discussed that apply to molecular biomarkers and not microbial (i.e., ecosystem species) biomarkers (McCarthy & Shugart, 1990).

A. Biomarker Definitions

Biomarkers, in general, are organic indicator compounds that can be used as tracers for geological and environmental processes. The carbon skeletons of the natural product precursor compounds were synthesized by biota, and are found either as such or as altered directly or indirectly by diagenetic changes to derivative products. The term “biomarker” evolved from early product–precursor relationships proposed in the 1960s (Streibl & Herout, 1969). The relevance of those naturally-derived saturated and aromatic hydrocarbons (and sometimes oxygenated analogs) to environmental geochemistry became evident when polynuclear aromatic hydrocarbons (PAHs) from anthropogenic or geogenic origins were found in recent and contemporary sediments (Simoneit, 1977a; LaFlamme & Hites, 1979; Wakeham, Schaffner, & Giger, 1980a; Tan & Heit, 1981). In the geological record, the biomarkers were originally called chemical fossils (Eglinton & Calvin, 1967; Mackenzie et al., 1982; Johns, 1986), and the concept has been extended to environmental chemistry (Simoneit et al., 1993a; Eganhouse, 1997).

1. Higher-Plant Biomarkers

Higher-plant biomarkers are derived from biosynthesis in higher-order flora (i.e., vascular plants = angiosperms, gymnosperms, ferns), and were elucidated initially in brown coals, peats, and resins (amber) (Johns, 1986). Homologous aliphatic (lipids) and cyclic (terpenoids) compounds are both utilized as biomarkers. The homologous aliphatic compounds have strong carbon-number predominances and are derived from the epicuticular waxes and related lipids of higher plants. The compounds are primarily n-alkanes (C27, C29, C31, C33), n-alkanols, and n-alkanoic acids (both homologous series have dominant C24, C26, C28, or C30), with lesser amounts of other oxygenated homologous species.

a. Lipids

Lipids are important biomarkers because they carry a strong carbon-number predominance that is inherited from their biosynthesis (e.g., buildup from acetate), and their homolog distribution can reflect biogenic origin (e.g., marine vs. terrestrial vegetation; Simoneit, 1977a,b). Mass spectrometry is an important method to identify homologous lipids such as alkanes, alkenes, fatty acids, fatty alcohols, alkanones, esters, wax esters, amides, etc. (Simoneit & Mazurek, 1982; Evershed, 1992; Murphy, 1993; White et al., 1997; Simoneit, 1999).

b. Terpenoids

The terpenoid biomarkers from higher plants are the natural products and derivative products from the reductive and oxidative alteration of the precursors. That alteration can occur during transport, by diagenesis in sedimentary environments, or by thermal transformation processes (Currie & Johns, 1989; Poinsot et al., 1995; Versteegh et al., 2004). Reductive alteration generally yields the parent compound skeleton with various isomerizations of chiral centers and, in some cases, loss of carbon due to decarboxylation and other reactions. Oxidative alteration occurs mainly by successive ring aromatizations that usually commence from a ring that has a functional group (e.g., OH, C[DOUBLE BOND]C, C[DOUBLE BOND]O; typically on ring-A) by direct dehydrogenation, dehydration, ring rearrangement, or ring opening and subsequent loss.

Various processes have been proposed for aromatization, including disproportionation, microbial hydrogen abstraction, sulfurization (i.e., reaction with SOmath image to yield H2S), and pyrolytic (anthropogenic) or catagenetic (heating during progressively deeper burial in sedimentary basins) thermal processes. Geological aromatization commences in recent (geologically) and contemporary sedimentary environments and progresses during early diagenesis of organic matter (i.e., organic matter remineralization and stabilization under low temperature and euxinic conditions). However, the early diagenetic reactions that yield aromatic biomarkers are not well understood. Thus, aromatic biomarker formation by thermal processes is illustrated here.

Biomass burning is of great environmental concern. Many organic compounds are not destroyed completely during burning, but are altered to the same biomarkers as elucidated in the geological record (Simoneit, Cardoso, & Robinson, 1990a, 1993a; Abas et al., 1995; Simoneit, 1998, 2002a). Thus, the natural product precursors in the biomass fuel and their transfer as such or altered to the smoke can be studied (Simoneit, Cardoso, & Robinson, 1990a, 1993a). The geologic and pyrolysis biomarker record (oxidation) is illustrated by the progressive aromatization of terpenoids (Figs. 1–4), which yields aliphatic, functionalized, or aromatic derivatives. The utility of aromatic derivatives in organic geochemistry and environmental chemistry is to complement the corresponding saturated alicyclic terpanes as proxies for tracers of sources and processes (Johns, 1986; Bouloubassi & Saliot, 1993a).

Figure 1.

Scheme for the alteration of sesquiterpenoid precursors to saturated and aromatic hydrocarbons (natural product examples are boxed).

Figure 2.

Scheme for the alteration of diterpenoid precursors to saturated and aromatic derivatives (natural product examples are boxed, stereochemistry is indicated where known).

Figure 3.

Scheme for the alteration of triterpenoid precursors from higher plants to saturated and aromatic derivatives (natural product examples are boxed, stereochemistry is indicated where known).

Figure 4.

Scheme for the alteration of sterols to saturated and aromatic hydrocarbons (natural product examples are boxed).

Examples of reductive and oxidative alterations of sesqui- and diterpenoids are shown in Figures 1 and 2. Mono-, sesqui-, and some diterpenoids are found in marine and terrestrial flora. They are, therefore, not always unambiguous tracers for higher-plant sources. However, diterpenoids with the abietane and pimarane, and less common phyllocladane and kaurane, skeletons are predominant constituents in resins and supportive tissue of coniferous vegetation (Coniferae), which evolved in the late Paleozoic (200–300 million years ago). Diterpenoid biomarkers have been characterized in ambers, coals, sediments, contemporary environments, and anthropogenic emissions (Streibl & Herout, 1969; Johns, 1986; Wang & Simoneit, 1990; Simoneit, 1998). The product–precursor relationship for diterpenoids has been presented by many authors (Fig. 2; Simoneit, 1977a; LaFlamme & Hites, 1979; Wakeham, Schaffner, & Giger, 1980a; Simoneit & Mazurek, 1982; Johns, 1986). Reductive preservation retains the C20 skeletons, which are the major diterpane biomarkers in the geological record. Decarboxylation of resin acids with subsequent reduction yields biomarkers with structures ≤C19. Oxidative alteration of diterpenoids to aromatic derivatives is discussed in more detail in a later section.

Triterpenoids are a major group of natural products in higher plants. The tetracyclic triterpenoids, based on the lanostane, euphane, onocerane, dammarane, limonin, and cucurbitacin skeletons, are found mostly in vascular plants, and the first four have been reported in sedimentary rocks (cf., Fig. 3; Kimble et al., 1974; Johns, 1986). The steroids (I, key chemical structures cited in the text are given in Appendix I), natural product derivatives from triterpenoids, are ubiquitous in the geosphere. The phytosterols (mainly C29, minor C28 skeletons) have been used as indicators for higher-plant sources, although many marine algae also produce C28 and C29 sterols with the same or different alkyl substitutions on the side chain (Johns, 1986).

The reductive and oxidative alteration of steroids is illustrated in Figure 4 and shows the geosteranes, aromatic steroid hydrocarbons, and thermal cracking derivatives. Reductive processes, mainly geological, produce steranes and diasteranes (geosteranes) with typically the parent skeletons and various epimerizations of chiral centers (Fig. 4). Products with dealkylation of the side chain are also encountered. Steroid aromatization reactions (Fig. 4) have also been elucidated mainly for geological samples (Mackenzie et al., 1982). The products found are primarily from the regular aromatization of ring-C to ring-A, with some loss of the side chain and subsequent cracking to alkylphenanthrenes (Fig. 4).

Many pentacyclic triterpenoids [e.g., the oleanane, ursane, taraxerane, lupane, friedelane, serratane, or bauerane skeletons (Fig. 3)] are characteristic biomarkers for an origin from terrestrial higher plants. They occur as functionalized (e.g., alcohols, acids, ketones, esters) precursors but are not necessarily specific to individual classes of biota. The most frequently encountered compounds in sedimentary or contemporary environments are those derived from α- and β-amyrins (II, III, ursanes and oleananes, respectively) (Johns, 1986; Wang & Simoneit, 1990). The reductive and oxidative alteration of the amyrins is illustrated in Figure 3. It should be noted that the presence of a functionality at C-3 in those triterpenoids makes them more susceptible to microbial or photochemical degradation to yield ring-A opened products and ultimately compounds without ring-A (e.g., des-A-oleanane).

Reductive alteration of triterpenoids yields mainly the parent skeleton with epimerization of key chiral centers. Aromatic triterpenoids were first isolated and characterized from brown-coal extracts, and were inferred to be derived from triterpenoids based on structural similarity and the origin of the coal (Ruhemann & Raud, 1932; Münch, 1934; Streibl & Herout, 1969). The impetus for the extensive characterization of the aromatic triterpenoid hydrocarbons was the synthesis of authentic standards (Spyckerelle, 1975). Aromatization of higher-plant triterpenoids (terrestrial source) generally commences in ring-A, and progresses to ring-E either directly or after loss of ring-A due to initial photo- or thermo-chemical reactions (Fig. 3). That alteration scheme has been presented by various authors (Spyckerelle, 1975; LaFlamme & Hites, 1979; Wakeham, Schaffner, & Giger, 1980a; Johns, 1986), and the numbers of known compounds and alteration pathways continually increases. The amyrins (α and β), lupeol, friedelanol, taraxerol, bauerenol, and others are readily oxidized in the environment to the 3-oxo derivatives or dehydrated by heat to the Δ2-olefin (Fig. 3). The 3-oxo compounds can be photolyzed in sunlight to the seco acids (e.g., dihydroroburic acid from α-amyrone; note that α-amyrin photolyzes to roburic acid), which have been found in tropical environments. The seco acids upon aromatization lose the remaining carbons of ring-A and yield the hydrochrysene series (Fig. 3). The Δ2-olefins or 3-oxotriterpenoids undergo successive ring aromatization to yield the hydropicene series (Fig. 3).

Tetraterpenoids and polyterpenoids are minor components of higher-plants, and are generally overwhelmed by the input of those compounds from microbial biomass in marine and lacustrine environments or sedimentary rocks. The natural cyclic tetraterpenoids have a maximum of two alicyclic rings, and thus the saturated and aromatic derivatives are limited. The known parent compounds are lycopane (IV), carotane (V), 1-(2′,2′,6′-trimethylcyclohexyl)-3,7,12,16,20,24-hexamethylpentacosane (VI), and biphytane (VII).

Higher-plant biomarkers provide a geochemical tool to identify terrestrial organic detritus in global compartments (e.g., atmosphere, sediments, etc.). They have been used to study the export of carbon by rivers to the ocean, and have revealed a surprisingly large pool of refractory land-derived organic matter in marine sediments (Prahl et al., 1994). Other applications include the identification of the sources of petroleum and the correlation of oils with their source rocks (Ekweozor & Telnaes, 1990; ten Haven, Peakman, & Rullkötter, 1992; Peters & Moldowan, 1993). Higher-plant biomarkers are useful in paleoenvironmental studies, although some caution is needed because those compounds appear to be more resistant to degradation than many marine biomarkers (Conte & Weber, 2002a,b). They are also utilized in environmental studies; for example, in source apportionment of organic matter in the atmosphere (Simoneit, 1977a, 1989; Peltzer & Gagosian, 1989; Schauer et al., 1996).

The major application for aromatic biomarkers is in studies on diagenesis of sedimentary organic matter to supplement the aliphatic biomarkers and in studies of environmental contamination by fossil fuels (coal and oil), by their products, and by recycled (eroded) sediments that contain those tracers (Wakeham, Schaffner, & Giger, 1980a,b; Bouloubassi & Saliot, 1993b). Also, their natural product precursors can be altered to the same aromatic biomarkers by combustive/thermal processes such as burning of biomass, cooking, rendering, and other waste-disposal methods. As such, they are proving to be useful atmospheric tracers (Simoneit et al., 1993a, 1999, 2000; Rogge et al., 1998; Elias et al., 2001; Nolte et al., 2001; Simoneit, 2002a).

2. Microbial Biomarkers

The initial biomarkers attributable to bacteria are the hopane series (Ensminger et al., 1974; Ourisson, Albrecht, & Rohmer, 1979, 1982; Kajukova et al., 1981). The 17α(H),21β(H)-hopanes (VIII) that range from C27 to C35 (no C28) were encountered in numerous ancient sediments and petroleums, and diagenesis and maturation of the microbial precursors (e.g., bacteriohopanepolyol, IX, and diploptene, X, Fig. 5) were elucidated (Ourisson, Albrecht, & Rohmer, 1982; Rohmer, 1993). The diagenesis of diploptene in contemporary sediments proceeds by double bond migration from Δ22,29 via Δ21,22 to Δ17,21 and possibly to neohopene (Fig. 5). Oxidation of bacteriohopanetetrol yields mainly the homohopanoic acids (XI), ranging from C31 to C34, and minor homohopanols (XII).

Figure 5.

Scheme for the alteration of hopenes to aromatic hopanoids and hopane (natural product examples are boxed).

The aromatization of hopanoids, when derived from bacterial detritus in sediments, proceeds with the diagenetic alteration or binding of the precursors (e.g., diploptene to neohopene, bacteriohopanepolyol) and dehydrogenation from ring-D to ring-A (Fig. 5) (Spyckerelle, 1975). However, the process by which the C-30 methyl group is lost from the side chain has still not been clarified. The precursor (e.g., diploptene, hopan-22-ol) may have been chemically bonded to the kerogen and released by cracking because those aromatics are encountered mainly in geological samples (black shales).

Isoprenoidyl di- and tetra-ether lipids are major biomarkers for Archaea (Kates, 1997). Caldarchaeol (XIII) is found mainly in methanogens and thermophiles, and archaeol (XIV) in the previous as well as in halophiles (DeRosa et al., 1977; Tornabene & Langworthy, 1979; Kates, 1997; Hinrichs et al., 2000). A significant proportion of the isoprenoid hydrocarbons in sediments and petroleums can be derived from tetraether lipid cores (Moldowan & Seifert, 1979; Chappe, Albrecht, & Michaelis, 1982; Albaiges, Borbón, & Walker, 1985; Vorobjeva et al., 1987). A predominant tetraterpenoid aromatic product that has been identified is octadecahydroisorenieratene (XV, also called isorenieratane). It is derived from the isorenieratenes found in photosynthetic green sulfur bacteria with other minor diarylisoprenoids (Schaeflé et al., 1977; Hartgers et al., 1994, 1996; Koopmans et al., 1996).

A group of C20 (XVI), C25 (XVII), and C30 (XVIII) highly branched isoprenoid (HBI) hydrocarbons has been identified in numerous sediments (Rowland & Robson, 1990). Those compounds occur as mono- to tri-unsaturated, and mono- and bi-cyclic analogs, as well as sulfurized (thiophene) derivatives in contemporary and recent sediments. The saturated parent compounds (XVIXVIII) are found in ancient and also in some contemporary sedimentary environments (Barrick, Hedges, & Peterson, 1980; Yon, Ryback, & Maxwell, 1982; Requejo & Quinn, 1983; Vorobjeva, Zemskova, & Petrov, 1986; Rowland & Robson, 1990; Hird, Evens, & Rowland, 1992; Jaffé et al., 2001). The inferred origin of these compounds is from algal lipids (plankton) in blooms and periphyton (Gearing et al., 1976; Rowland & Robson, 1990; Jaffé et al., 2001). They have been isolated from freshwater and marine diatoms and photosynthetic algae (Rowland et al., 1985; Volkman, Barret, & Dunstan, 1994; Sinninghe Damsté et al., 1999; Wraige et al., 1999). Tetrahymanol (XIX) is the precursor of gammacerane (XX), which occurs in hypersaline sediments as an indicator of protozoa such as Tetrahymena (ten Haven et al., 1989).

B. Anthropogenic Tracer Compounds

There are more than 70,000 organic compounds in commercial production; approximately 1,000 are added annually. Most of the mass spectra of those synthetic (anthropogenic) compounds are in the MS libraries and the standards are commercially available. The numerous uses of synthetic compounds can result in the release, directly or indirectly, to the surrounding environment. Many of those compounds are of environmental concern due to their production quantities, toxicity, persistence, and tendency to bioaccumulate (Schwarzenbach, Gschwend, & Imboden, 1993, 2003; DeCaprio, 1997; Eganhouse, 1997; Takada et al., 1997; Aboul-Kassim & Simoneit, 2001). They can enter the hydrosphere (Hites, 1973, 1977; Lopez-Avila & Hites, 1980; Poiger et al., 1996; Leeming et al., 1997; Suter et al., 1997, 1999; Simonich et al., 2000), atmosphere (Atlas et al., 1993; Baker & Hites, 1997; Hillery, Basu, & Hites, 1997; Strandberg et al., 2001; Whiteaker & Prather, 2003), geosphere into soils and sediments (Hom et al., 1974; Eglinton, Simoneit, & Zoro, 1975; Beller & Simoneit, 1986, 1988; Kumata, Takada, & Ogura, 1997; Reiser, Toljander, & Giger, 1997; Stoll et al., 1997; Rogge, Medeiros, & Simoneit, 2004a–c), and up the food chain in biota (Risebrough et al., 1967; Bowes et al., 1973; Goldberg et al., 1978; Spitzer et al., 1978; Jarman et al., 1992, 1993; Simonich & Hites, 1995a). These applications of MS will keep increasing as more problems become evident, and as the sensitivities of the instrumental methods increase to ever-improved detection limits.

II. MASS SPECTROMETRY OF BIOMARKERS

The mass spectrometry of biomarker precursors was carried out by natural product chemists during the 1950–1970s. The mass spectra were generally determined with low- and high-resolution instruments for the underivatized natural product or its acetate ester (also methyl ethers). Sample introduction to the ion source was by direct-insertion probe. Those mass spectra are in the NIST and Wiley MS libraries.

The overall field of mass spectrometry has been extensively described, and the early references are still useful, especially for interpretation of mass spectra (Biemann, 1962; McLafferty, 1963, 1966; Budzikiewicz, Djerassi, & Williams, 1967; Burlingame & Schnoes, 1969; Burlingame, 1970; Porter & Baldas, 1971; Aczel, 1972; McFadden, 1973; Zaretskii, 1976; Philp, 1985; Davis & Frearson, 1987; McLafferty & Tureček, 1993; Murphy, 1993). Current books also provide excellent introductions and applications (De Hoffmann & Stroobant, 2001; Hübschmann, 2001; Ardrey, 2003). The advances in mass spectrometry have been presented periodically at meeting proceedings, and are reviewed biennially by Burlingame et al. since 1974 (Burlingame, Boyd, & Gaskell, 1996, 1998; Gelpi, 2002; Richardson, 2002, and references therein).

The coupling of gas chromatography with mass spectrometry, i.e., GC-MS, was the major breakthrough in developing biomarker geochemistry initially for petroleum exploration research and then organic geochemistry in general. Current instrumentation and derivatization methods (especially persilylation of polar functional groups) extend the biomarker concept into other disciplines and applications where hydrocarbons and polar, as well as synthetic (anthropogenic), compounds are encountered. Other mass spectrometric techniques such as chemical ionization (CI) MS, GC-high-resolution MS, field-ionization and -desorption MS, tandem mass spectrometry (MS-MS), pyrolysis GC-MS, and high-pressure liquid chromatography (HPLC)-MS, all with the associated online computers and processors, are being applied to biomarker fingerprinting. Natural product elucidation continues and is fostered by the pharmaceutical and food industries, and will provide novel biomarkers of geochemical interest. However, the classical natural product database and standards are still of greatest utility for environmental biomarker research.

A. Common Mass Spectrometry Methods

The commonly used mass spectrometry methods discussed here do not include MS as used in biochemical and biomedical applications for highly polar and labile large molecular weight compounds. The latter are reviewed elsewhere (Green & Lebrilla, 1997; Beck et al., 2001; Cristoni & Bernardi, 2003). Those methods have not yet been used in geochemical and environmental research. The MS methods used for biomarker research are summarized in Table 1.

Table 1. Biomarkers and molecular tracers utilized in geochemical and environmental researchThumbnail image of
  • GC, gas chromatography; Py, pyrolysis (flash or hydrous heating); MS, mass spectrometry; IRMS, isotope-ratio mass spectrometry; HPLC, high-pressure liquid chromatography; EC, electron capture; CI, chemical ionization.

  • aLabile markers are stable for brief geological periods after diagenetic preservation. Group 2 compounds are stable for longer geologic periods.

  • bSlash indicates that a distinction is possible.

  • 1. Gas Chromatography—Mass Spectrometry

    Gas chromatography-mass spectrometry is a coupled analytical technique, where the GC is the compound-separation instrument and the MS analyzes the ions from compounds according to their mass-to-charge ratios (m/z). It has been the predominant analytical method in the development of organic geochemistry and environmental chemistry, especially for biomarkers. GC-MS typically involves three stages of choice: (1) sample introduction into the GC, (2) GC parameters, and (3) MS parameters (Message, 1984; Davis & Frearson, 1987; Evershed, 1993). Current separation technology uses fused silica capillary columns for GC to achieve high resolution. The detection limits of GC-MS systems range from ng to fg for compounds, using the base peak as the key ion.

    2. Gas Chromatography—Isotope Ratio Mass Spectrometry

    A recent development for stable carbon isotope analysis is the introduction of GC-combustion-isotope MS (Hayes et al., 1990; Schoell et al., 1992). Here, the GC effluent passes through a combustion tube with CuO at 800°C and the CO2 generated is analyzed in real time for its carbon isotope composition by monitoring m/z 44 and 45 (expressed as δ13C). That method is also referred to as compound-specific isotope analysis (CSIA) and is finding rapid application in biomarker analysis (Schoell, Hwang, & Simoneit, 1990, 1994a,b; Simoneit et al., 1993b, 1995; Simoneit & Schoell, 1995; Albrecht, 2003). CSIA is also available for stable hydrogen and stable nitrogen isotope analyses (i.e., δD, δ15N; Xie et al., 2003). The sample requirements and GC analytical conditions for CSIA of carbon isotopes are the same as for GC-MS, and the detection limits are also in the same range.

    Compound-specific isotope analysis is illustrated with examples of a hydrothermally formed petroleum (Simoneit & Schoell, 1995) and a brown coal (Schoell, Simoneit, & Wang, 1994b). The mass chromatogram (m/z 44) of an alkane fraction from the petroleum is shown in Figure 6. The peak resolution is the same as for TIC under electron impact MS or FID with GC only. The results, calculated as δ13C versus the PDB standard, for homologous biomarker compounds (e.g., n-alkanes) are usually plotted versus carbon number to assess sources and similarities among samples (Fig. 7). The carbon isotope compositions of the alkanes and polycyclic aromatic hydrocarbons have δ13C means of −23.1 and −21.2‰, respectively, which reflect their marine source and confirm their generation from the sedimentary kerogen (Simoneit & Schoell, 1995). The CSIA results for more complex mixtures of compounds (e.g., brown coal) are presented in decreasing δ13C values to reveal a clustering of compounds from the same sources (e.g., Fig. 8). For example, terrestrial higher plants (diterpenoids, triterpenoids, alkanes) and bacteria (hopanes) have different carbon isotope compositions of their lipids, and thus the derivative biomarkers are distinguishable.

    Figure 6.

    Representative mass chromatogram (m/z 44) for compound-specific isotope analysis (CSIA) of an alkane fraction (CO2 = reference standard, numbers refer to carbon chain length) (Simoneit & Schoell, 1995; reproduced with permission, Elsevier Science Publishers, copyright 1995).

    Figure 7.

    Plots of δ13C versus carbon number for the n-alkanes in hydrothermal petroleums from Guaymas Basin, Gulf of California (Simoneit & Schoell, 1995; reproduced with permission, Elsevier Science Publishers, copyright 1995).

    Figure 8.

    Compounds in baise coal extract arranged according to their carbon isotopic composition. Biomarkers cluster according to their chemical structures. Sesqui- and diterpenoids from resins tend to be enriched in 13C compared to the bulk coal and oleanane-ursane-lupane derivatives (cuticular compounds) are slightly depleted in 13C compared to the coal. (Compounds denoted by asterisks are plotted in the isotopic trend and not with their assumed source.) All land plant derived compounds are within ±2‰ of the bulk coal isotopic value (Schoell, Simoneit, & Wang, 1994b; redrawn with permission, Elsevier Science Publishers, copyright 1994).

    3. Tandem Mass Spectrometry

    Tandem or triple-stage quadrupole mass spectrometers coupled with GC (GC-MS-MS) have a major advantage over GC-MS systems because they can resolve individual biomarkers or biomarker groups in complex mixtures such as petroleum or coal extracts. A typical MS-MS consists of three quadrupoles linked in series, where the first is the precursor quadrupole, the middle the collision-cell quadrupole, and the third the product quadrupole. GC-MS-MS can be operated in three modes, i.e., precursor, product, and neutral (Philp, Oung, & Lewis, 1988). Other hybrid mass spectrometer combinations are in use, especially for natural product, environmental, and biomedical applications (Schroeder, 1996; Reilly et al., 1997; Pyle, Marcus, & Robertson, 1998; White et al., 1998; van Pelt, Haggarty, & Brenna, 1998; Demirev, Ramirez, & Fenselau, 2001; Poulsen et al., 2003). GC-MS-MS is used in petroleum geochemistry for rapid biomarker screening of mainly the sterane and hopane series (Wang et al., 1988; Johnson, Yost, & Wong, 1989; Lin et al., 1990; Gallegos & Moldowan, 1992; Philp & Oung, 1992). The methods and procedures of biomarker analyses by mass spectrometry, including MS-MS, in fossil fuel research have been described in detail with typical examples (Peters & Moldowan, 1993).

    B. Other Mass Spectrometry Methods

    1. High-Temperature Gas Chromatography—Mass Spectrometry

    High-temperature gas chromatography-MS (HTGC-MS) is being applied to direct analyses of lipid and bitumen mixtures that contain high-molecular weight (HMW) compounds. The GC is usually equipped with an on-column injector and a metal-clad capillary column, and the oven is operated from ambient temperature to 400°C with a rapid heating ramp. HTGC-MS is used to fingerprint HMW paraffins (to C100) in petroleums (Philp, 1994; Philp et al., 1995). It is being used in archeological chemistry to identify HMW lipid and steroid derivatives (Evershed et al., 1987, 1990), and is of utility to characterize biomarkers in environmental and geological samples such as wax esters, terpenoidyl esters, phytyl esters, and tocopheryl esters and intact lipids such as glyceryl esters and glyceryl ethers (Elias et al., 1997, 1999; Pereira et al., 2002; Siqueira et al., 2003).

    High-temperature gas chromatography-MS is illustrated with the following examples from fingerprinting the tracers in smoke particulate matter from biomass burning (Elias et al., 1999). Many samples have HMW compounds in the range of 350–900 daltons (Da).

    The total extract of smoke filtrate from the burning of Vismia (Vismia guineensis) contained a series of compounds that eluted above 350°C as shown in the HTGC-MS trace (Fig. 9a). The major compounds consist of levoglucosan (XXI, 1,6-anhydro-β-glucopyranose) and other monosaccharide derivatives from the thermal breakdown of cellulose (Simoneit et al., 1999). The second most abundant class of compounds comprises various methoxyphenols, which were produced by the thermal degradation of lignin (Hawthorne et al., 1988; Simoneit et al., 1993a). Typically, those two classes of components encompass approximately 85% of the total extractable organic compounds in smoke samples. Compounds of relatively HMW (e.g., triterpenoids, phytosterols, and fatty acids) that elute earlier than 310°C (i.e., the temperature limit of conventional GC columns) are also present in the total extracts but at lower abundances than the lower molecular weight compounds mentioned above. The compounds that eluted >350°C comprised wax esters, triterpenoidyl esters, triglycerides, and other lipid esters as illustrated next.

    Figure 9.

    Representative TIC traces from high-temperature GC-MS analyses of smoke particles from: (a) the total extract (methylated and silyated) from fine particles of smoke from the burning of Vismia, (b) TIC trace of the fraction that contained wax esters in smoke extract from the burning of Cupuaçú (numbers refer to the total carbon chain length of the wax esters, ni = normal alkane carbon chain length), and (c) the fraction that contained triglycerides in smoke extract from the burning of Andiroba. (Numbers refer to the triglycerides: 1 = dipalmitoylolein, 2 = 1-palmitoyl-2-oleoyl-3-stearin, 3 = triolein, · = unknown lipids) (Elias et al., 1999; reproduced with permission, American Chemical Society, copyright 1999).

    Long-chain wax esters (LCWE) are present in smoke from the burning of many plant species. For example, smoke aerosol from the burning of Cupuaçú (Theobroma grandiflorum) contained LCWE from 38 up to 58 total carbon numbers, with a typical strong even-carbon number predominance as reported for plant wax esters (Fig. 9b). That LCWE series is comprised of mainly palmitic acid esterified with the fatty alcohols, and ranged from C22 to C34 with minor amounts of stearic and arachidic acids esterified to C32 and C34 alcohols. Some ester fractions separated from extracts of smoke particulate matter from the burning of different species of plants contained triglycerides (Fig. 9c, Elias et al., 1999). The major triglyceride compounds in Andiroba (Carapa guineensis) smoke are dipalmitoylolein (C53H100O6, MW 832), 1-palmitoyl-2-oleoyl-3-stearin (C55H104O6, MW 860), and triolein (C57H104O6, MW 884). It should be noted that those common energy-storage lipids are not present in plant epicuticular waxes (Kolattukudy, 1976), but are internal constituents of the plant that were burned. Triglycerides have a very low-vapor pressure, and their presence in smoke, as well as other HMW lipids, is further evidence of their direct volatilization as HMW natural products during the burning of plant matter.

    2. Gas Chromatography—High-Resolution Mass Spectrometry

    Gas chromatography coupled to high-resolution mass spectrometry (GC-HRMS) has had limited application in biomarker research. The impediment seems to be the intimidating data volume; for example, 500–1,000 high-resolution mass spectra per mixture analyzed by GC-HRMS. The method has a potential utility to elucidate key-ion searches for mixtures, and to provide accurate mass compositions of unknown biomarkers.

    An example of a biomarker elucidation by GC-HRMS follows (Simoneit et al., 1981). The analysis of hydrothermal bitumen for phenolic compounds was carried out with a Hewlett-Packard 5890 Series II gas chromatograph coupled to a JEOL SX-102 double-focusing high-resolution mass spectrometer. The mass spectrometer was operated at 70 eV over the 100–500 Da range, resolution of 10,000, and a cycle time of 2.0 s/decade. Perfluorokerosene was used for the calibration of the accurate-mass scale. The GC oven temperature was programmed at isothermal at 60°C for 5 min, 10°C/min to 120°C, 5°C/min to 310°C, and isothermal for 20 min; the injector was in the splitless mode at 300°C, and helium was the carrier gas. The MS data were acquired and processed with an on-line JEOL MS-MP7000 computer data system.

    The compounds of interest in those mixtures were in peaks A and B in the GC-MS data (Fig. 10). Compounds A and B have mass spectra with ions at m/z 107, 122 (base peak), 135, 161 and 388 (molecular ion, M·+), and at m/z 121, 136 (base peak), 149, and 402 (M·+), respectively (Fig. 11a,b). Treating that polar fraction with a silylating agent produced the trimethylsilyl ether derivatives of compounds A and B (labeled A′ and B′) with characteristic ions at m/z 179, 194 (base), 207 and 460 (M·+), and at m/z 193, 208 (base), 221, and 474 (M·+), respectively (Fig. 11c,d). Analyses by GC-HRMS provided the accurate masses for the dominant ion peaks, and two accurate-mass chromatograms are illustrated in Figure 12. The GC-HRMS data confirmed the presence of oxygen in the molecules and indicated two isomers of compound B.

    Figure 10.

    GC-MS TIC traces of the polar (NSO) fraction of hydrothermal petroleum (a) underivatized and (b) treated with silylating reagent (BSTFA). Numbers refer to carbon chain length of the n-alkan-2-ones (dots) and alkylphenols (triangles), A = isoprenoidyl phenol (MW = 388), B = isoprenoidyl phenol (MW = 402), A′ = isoprenoidyl phenol-TMS (460), B′ = isoprenoidyl phenol-TMS (MW = 474), UCM = unresolved complex mixture (Simoneit, Leif, & Ishiwatari, 1996; reproduced with permission, Elsevier Science Publishers, copyright 1996).

    Figure 11.

    Mass spectra of the isoprenoidyl phenols present in the polar fraction of hydrothermal petroleum (free phenols a, b, silyl ethers c, d): (a) compound A, C27H48O; (b) compound B, C28H50O (peaks A and B in figure); (c) compound A′, C30H56OSi; and (d) compound B′, C31H58OSi (Simoneit, Leif, & Ishiwatari, 1996; reproduced with permission, Elsevier Science Publishers, copyright 1996).

    Figure 12.

    Examples of accurate-mass chromatograms from the GC-HRMS analysis of the polar fraction of hydrothermal petroleum: (a) m/z 387.9340–389.0540, molecular ion of compound A; (b) m/z 136.0593–136.1490, base peak of compound B (two isomers) (measured accurate mass and millidalton difference from calculated mass are shown in the figure, Simoneit, Leif, & Ishiwatari, 1996; reproduced with permission, Elsevier Science Publishers, copyright 1996).

    Free and derivatized phenols A and B both undergo β-cleavage to yield the respective base peaks that indicated methylalkyl-, dimethylalkyl-, and trimethylalkylphenyl substitution. The mass spectra of chromans and tocopherols (standards and geological compounds) similarly lack characteristic alkyl-branch fragments from the isoprenoid side chain (Elliott & Waller, 1972; Scheppele et al., 1972; Sinninghe Damsté et al., 1987). Thus, by analogy and lack of authentic standards, the alkyl side chain of compounds A and B was assigned to be isoprenoidal. The structures of those compounds were tentatively interpreted to be 3,7,11,15-tetramethylhexadecyl-methyl phenol (structure A, Fig. 12) for compound A, and 3,7,11,15-tetramethylhexadecyl-dimethyl phenol (structure B, Fig. 12) for compounds B1 and B2 (Simoneit, Leif, & Ishiwatari, 1996).

    3. Pyrolysis Gas Chromatography—Mass Spectrometry

    Pyrolysis is the non-oxidative thermal degradation of macromolecular organic matter, such as biopolymers (Larter & Douglas, 1982; Lewan, 1985). Thermal degradation fragments the polymer by scission at the weakest bonds, and the low-molecular weight products are thus transferred into the GC-MS. More detailed descriptions of pyrolysis (i.e., pyGC-MS) have been published (van de Meent et al., 1980; Horsfield, 1984). There are essentially two types of pyrolysis. The first type is dry, in which the sample is heated rapidly to a specified temperature (e.g., 610°C) in the absence of water-usually by inductively heating an alloy to its Curie point. On-line dry pyrolysis tends to produce unsaturated hydrocarbons as a result of bond scission from the kerogen (Simoneit et al., 1978, 1981; van de Meent et al., 1980). The second type is hydrous pyrolysis, where the sample is heated off-line in a bomb with water (350°C for 3 days) (Lewan, 1983, 1997; Leif, Simoneit, & Kvenvolden, 1992; Simoneit, 1992a). The resultant product (the pyrolysate) is compositionally more saturated than in dry pyrolysis, and the technique approximates petroleum generation from a source rock. Reduction can be further enhanced by the addition of oxalic or formic acids, which decompose to H2, CO2, and CO under the pyrolysis conditions (McCollom, Ritter, & Simoneit, 1999a,b; Rushdi & Simoneit, 2001). That off-line reductive hydrous pyrolysis method is of particular utility to produce biomarkers that retain the full precursor structure from macromolecular organic matter and biomembranes.

    4. High-Pressure Liquid Chromatography—Mass Spectrometry

    High-pressure liquid chromatography (HPLC) is a powerful separation technique when coupled to MS for the analysis of polar, non-volatile, or thermally unstable compounds that are otherwise not separable by conventional GC-MS methods. HPLC-MS evolved from the prior combinations of liquid chromatrographic (LC) systems with mass spectrometers that also dealt with polar and highly functionalized compounds. Reviews of LC/MS systems, interface applications, and designs were presented (Vestal, 1984a,b; Linscheid, 1992; Creaser & Stygall, 1993; Gelpi, 2003). Various experimental systems have been described and used to evaluate, for example, coal-liquefaction products in terms of the number of condensed rings separated by the HPLC into fractions that were further separated and analyzed by LC/MS (Dark, McFadden, & Bradford, 1977; McFadden et al., 1977; McFadden, 1982). Currently, HPLC-MS is used to elucidate the structures of polar and highly functionalized biomarkers and natural products (McFadden et al., 1979; Voyksner, 1994; Thomson, 1998; Harada & Fujii, 2000) and organic compounds in environmental samples (Schilling et al., 1996; Suter, Riediker, & Giger, 1999).

    5. Time-of-Flight Mass Spectrometry

    Time-of-flight mass spectrometry (TOFMS) is beginning to be applied in atmospheric chemistry to elucidate organic compounds including biomarkers, on single aerosol particles (Alvarez, Vezmar, & Whetten, 1998; Noble & Prather, 2000; Angelino, Suess, & Prather, 2001; Whiteaker & Prather, 2003). The organic matter of individual aerosol particles is vaporized by laser and ionized in the TOFMS. Total low-resolution mass spectra are recorded during discrete successive time intervals, and the ion signatures are interpreted in terms of precursor compounds that are assignable to sources (Silva & Prather, 2000).

    III. APPLICATIONS OF MASS SPECTROMETRY IN BIOMARKER RESEARCH

    The following sections cover the applications of MS in biomarker, or molecular-tracer, research. The biomarker classes or compound groups with their sources and the common MS methods applied for analysis are summarized in Table 1.

    A. Geochemistry

    1. Biomarkers in the Fossil Record

    Most of the research on biomarkers in the fossil record has dealt with the derivative hydrocarbons in petroleum, coals, and sedimentary rocks. Reports on the biomarker signatures in discrete fossils compared to the host rocks are sparse because (1) previous studies focused on the highly degraded “geoterpenoids,” i.e., saturated and aromatic hydrocarbons (Brassell, Eglinton, & Maxwell, 1983; Chaffee et al., 1986; Simoneit, 1986; Peters & Moldowan, 1993), and (2) the preservation potential of polar compounds (“bioterpenoids”) was believed to be low (Streibl & Herout, 1969). However, recent investigations of Eocene and Miocene conifer fossils showed that unaltered bioterpenoids can be preserved in resin material (Otto & Simoneit, 2001; Otto, White, & Simoneit, 2002, 2003).

    Most of the terpenoids reported from plant fossils are degradation products from the natural precursor compounds synthesized by the living plant. Despite various biological and physicochemical degradation processes during the decay and diagenesis of the plant material, the geoterpenoids generally retain their basic natural product skeleton and can be assigned to their respective structural classes (Simoneit, 1986; Peters & Moldowan, 1993; Otto & Wilde, 2001). The composition of geoterpenoids in fossil plants can thus be compared directly to the distribution of certain bioterpenoid classes in extant plants. However, it should be emphasized that the degradation of some functionalized natural product skeleton types may yield the same series of geoterpenoids. Therefore, the bioterpenoids are more specific biomarkers than the geoterpenoids, because their occurrence in extant plants is more specific than the distribution of structural classes of terpenoids (Otto & Wilde, 2001). Terpenoids are preserved in fossil plants, and may thus be used as chemical markers for (paleo)chemosystematic studies, especially in conifers (Chaloner & Allen, 1969; Thomas, 1970, 1986). The analyses of terpenoids and lipids from fossil plant remains started early with fossil wood and resins (Sterling & Bogert, 1939; Langenheim, 1969; Gough & Mills, 1972), and with sediments rich in fossil plants (Knoche & Ourisson, 1967; Knoche, Albrecht, & Ourisson, 1968; Anderson and Crelling, 1995). Analytical techniques have greatly improved since then. Whereas Knoche, Albrecht, & Ourisson (1968) had to extract several kilograms of fossiliferous sediment to obtain enough material, it is now sufficient to analyze a few μg of organic matter from individual plant macrofossils. The extracts of the fossil materials are analyzed with the same analytical methods as applied for extant plant analyses, i.e., GC-MS. Numerous terpenoids and aliphatic lipids have been reported from fossil conifer shoots and cones (Stout, 1992; Staccioli, Mellerio, & Alberti, 1993; Vávra & Walther, 1993; Anderson & LePage, 1995; Otto, Walther, & Püttmann, 1997, 1999, 2001, 2002; Otto & Simoneit, 2001), fossil angiosperm leaves (Huang et al., 1995; Logan, Smiley, & Eglinton, 1995; Lockheart, van Bergen, & Evershed, 2000), and fossil resins (Grantham & Douglas, 1980; Czechowski et al., 1996).

    This procedure is illustrated here with the analysis of the extractable organic matter from the seed cones of two fossil Cupressaceae from the Miocene Clarkia Flora of Idaho and of a related extant species to evaluate the preservation of characteristic biomarkers in the fossils (Fig. 13, Otto, Simoneit, & Rember, 2003). The major compounds in the fossil and extant samples are oxygenated terpenoids. They are dehydroabietane (XXII), 6,7-dehydroferruginol (XXIII), ferruginol (XXIV), taxodione acetate (XXV), 3-oxo-12-hydroxysimonellite (XXVI), 7-acetoxy-6,7-dehydroroyleanone (XXVII), sugiol (XXVIII), chamaecydin (XXIX), isochamaecydin (XXX), 18- or 19-hydroxyferruginol (XXXI), and trans-communic acid (XXXII), and the minor geoterpenoids diaromatic totarane, simonellite (XXXIII), and 12-hydroxysimonellite (XXXIV) for the fossil cones. The extant Taxodium distichum contained additional bioterpenoids, such as royleanone (XXXV), taxoquinone (XXXVI), 7-hydroxytaxodione acetate (XXXVII), 6-hydroxytaxoquinone (XXXVIII), and taxodone (XXXIX); some were described previously (Kupchan, Karim, & Marcks, 1969).

    Figure 13.

    Annotated GC-MS (TIC) traces of the total extracts (TMS derivatives) of seed cones of: (a) Glyptostrobus oregonensis (Miocene), (b) Taxodium dubium (Miocene), and (c) Taxodium distichum (extant) (Otto, Simoneit, & Rember, 2003; reproduced with permission, Elsevier Science Publishers, copyright 2003).

    Such results show that the preservation potential of bioterpenoids, as the alcohols and acids, is higher than expected. The compositions of bioterpenoids in fossil materials indicate only minor degradation, due to a rapid burial in anaerobic sediments and/or a protection of the terpenoids in the resinous material with only minor exposure to microbial or abiogenic alteration. The in situ preservation of bioterpenoids and their diagenetic products in plant fossils allows chemotaxonomic classification, and yields unique information on the possible precursors of geoterpenoids.

    2. Petroleum Biomarkers

    Biomarkers have had extensive applications in petroleum geochemistry as organic matter maturity, source, and alteration indicators (Tissot & Welte, 1984; Peters & Moldowan, 1993; Hunt, 1996). However, the saturated and aromatic hydrocarbon biomarker derivatives (geomarkers), rather than the natural product precursors, are mainly utilized for petroleum geochemistry. The full spectrum of biomarkers, i.e., natural product precursors to derivative products, is employed in organic geochemistry and environmental chemistry. The major biomarker series of interest in petroleum geochemistry are the hopanes, tricyclic terpanes, steranes, and specific or unique biomarkers (i.e., other less-common terpanes; e.g., oleananes).

    Biomarkers derived from pigment natural products (e.g., carotenoids, chlorophylls, bacteriochlorophylls) are of utility in some applications, and the carotanes, renieratanes, and porphyrins have been analyzed by MS and GC-MS (Boylan & Calvin, 1967; Boylan, 1970; Eglinton, Evershed, & Gill, 1984; Gill et al., 1985; Repeta & Gagosian, 1987; Kruge et al., 1990; Sinninghe Damsté et al., 1993; Hartgers et al., 1994, 1996; Goericke, Shankle, & Repeta, 1999; Ocampo & Repeta, 1999, 2002). Carotanes (V) are used as indicators of a saline (marine) paleo-environment, and the renieratanes and isorenieratanes (XV) are biomarkers from photosynthetic sulfur bacteria. Porphyrins are generally derivatives from chlorophylls and bacteriochlorophylls, and are used as maturity biomarkers.

    Hydrothermal petroleums (bitumens), which are products of rapid diagenesis/catagenesis/metagenesis, have alkane and biomarker distributions analogous to those of conventional crude oils (Simoneit & Lonsdale, 1982; Simoneit, 1985, 1990a,b, 1992a,b, 1994). The carbon number distributions, biomarker compositions, and other geochemical parameters of marine hydrothermal petroleums generally reflect the source organic matter and the degree of thermal alteration or maturity (Simoneit & Lonsdale, 1982; Simoneit, 1985, 1990a,b; Kvenvolden et al., 1986; Kawka & Simoneit, 1987, 1990). Thus, the examples of biomarker patterns for hydrothermal petroleums are chosen here, because the full alteration range from immature biogenic precursors to fully mature biomarkers is found.

    a. Immature precursors

    Immature biomarkers occur in low-temperature regions of sedimentary hydrothermal systems, and maturation is observed with an increasing temperature and depth below the seafloor (sub-bottom depth) (Simoneit et al., 1979, 1984; Brault, Simoneit, & Novelli, 1988; Simoneit, 1994; Rushdi & Simoneit, 2002a,b). Sterols (I) are major components in unaltered sedimentary sections, and range from C27 to C29, with cholesterol (I, R = H) dominant or equal to β-sitosterol (I, R = C2H5). Diagenetic alteration of sterols accelerated by thermal stress yields stenones (XL) and stanones (XLI) with the same range from C27 to C29.

    Terrestrial triterpenoid markers are present in immature sediments, but their mature derivatives are only found as traces due to the high concentrations of hopanes also generated during hydrothermal maturation. The dominant biological precursors identified are α-amyrin (II, R = OH) and β-amyrin (III, R = OH) from higher plants (Brassell, Eglinton, & Maxwell, 1983; Simoneit, 1986). Those compounds are altered primarily to α-amyrone (urs-12-en-3-one, II, R = O) and β-amyrone (olean-12-en-3-one, III, R = O), and to lesser amounts of olean-12-ene (XLII) and urs-12-ene.

    Various triterpenoid precursors for the microbial hopanes are found in immature sediments. Diploptene (X) occurs in shallow sections and is altered to 17β(H)-hop-21-ene (XLIII) and subsequently to hop-17(21)-ene (XLIV).

    b. Triterpane maturation

    The hopanes undergo maturation from immature precursors, including the 17β(H),21β(H)-hopanes (XLV) and moretanes [17β(H),21α(H)-hopanes, XLVI], to the 17α(H),21β(H)-hopanes (VIII) over geological time periods of millions of years. The isomer configuration of the biological precursors is 17β(H),21β(H) for this series, and as the C-22 R epimer for the extended homologs >C31 (Ensminger et al., 1974, 1977). Maturation converts the precursors to the thermodynamically most-stable isomer configuration of 17α(H),21β(H), with the S- and R-epimers at C-22 for the extended homologs >C31 at an equilibrium ratio [S/(S + R)] of approximately 0.6 (Ensminger et al., 1974, 1977; Seifert & Moldowan, 1978). An example for this series is shown in Figure 14 as defined by the m/z 191 key ion fragmentogram (Simoneit, 1994).

    Figure 14.

    Examples of representative mass fragmentograms of m/z 189 and 191 for the triterpenoid hydrocarbons in the bitumen of sediments from the Ocean Drilling Program (ODP) Leg 139 in the NE Pacific hydrothermal systems. m/z 189 (a) sample 858D-1H-1, m/z 191 (b) sample 858D-1H-1, (c) sample 858B-1H-1, (d) sample 858C-3H-1, (e) sample 858A-1H-1. (1 = 17β(H)-hop-21-ene; 2 = diploptene; 3 = 17β(H)-22,29,30-trisnorhopan-21-one; 4 = hop-17(21)-ene; 5 = 17β(H),21β(H)-homohopan-31-ol; O = oleananes, G = gammacerane; iα = 17α(H),21β(H)-hopane series, with S and R entantiomers at C-22; iβα = 17β(H),21α(H)-hopane series, i.e., moretanes.)

    Some deltaic petroleums contain additional minor biomarkers from terrestrial sources, such as oleananes derived from the amyrins. The 18α(H)- and 18β(H)-oleananes (XLVII) elute just prior to 17α(H),21β(H)-hopane by GC or GC-MS (Ekweozor & Udo, 1988), and their epimerization has been used as a maturity parameter for terrestrial organic matter.

    c. Sterane maturation

    Sterane hydrocarbons, useful for oil-source rock and maturity comparisons (Seifert & Moldowan, 1978, 1979; Mackenzie et al., 1982), also undergo maturation in hydrothermal systems (Fig. 15, Simoneit, 1994). The steranes have C27 or C29 dominant, C28 intermediate, and C30 as a minor component. The C27(20R)/C29(20R) isomer ratios range from 0.22 to 3.2, where values <1 indicate a stronger influx of C29 terrestrial marker steroids. This ratio assumes a constant influx of marine-derived C29 steroid residues (Volkman, 1986). Immature samples contain significant concentrations of C27 to C30 steranes (XLVIII), primarily with the 5α,14α,17α-20R and smaller amounts of the thermally less-stable 5β,14α,17α-20R configurations (Fig. 15a). Mature samples exhibit the onset of epimerization at C-20, which occurs during thermal maturation (Fig. 15b–d; Seifert & Moldowan, 1978; Mackenzie et al., 1980; Simoneit, 1994). The epimerization ratio at C-20, [S/(S + R)], of C29 isomers increases to 0.45 at equilibrium (full maturity). Analyses of sedimentary rocks showed that this change in the epimerization ratio is due to the preferential thermal degradation of the C-20R isomers (Marzi & Rullkötter, 1992). As the maturity increases, additional isomerization is evident as the 5α,14β,17β-20R and 20S-steranes (XLIX), and the diasteranes (L,10α,13β,17α-20S/R) increase in relative concentration to the regular steranes (Fig. 15a vs. d). The diasterane epimerization parameter at C-20, [S/(S + R)] for C27, varies from 0.38 to 0.65 for those samples.

    Figure 15.

    Examples of representative mass fragmentograms (m/z 217, 218, and 259) for the steranes in the bitumen extracts of ODP Leg 139 sediments. (a, m/z 217) sample 858C-2H-CC, (b, m/z 217; c, m/z 218; d, m/z 259) sample 858A-1H-1 (ααα or α = 5α(H),14α(H),17α(H)-steranes, βαα or β = 5β(H),14α(H),17α(H) configuration, αββ = 5α(H),14β(H),17β(H) configuration, S and R are diastereomers at C-20, D = diasteranes, numbers refer to the carbon skeleton).

    3. Aromatic (Coal) Biomarkers

    Aromatic biomarkers are defined here as alkylated aromatic hydrocarbons derived from natural products (primarily terpenoids) by diagenetic or catagenetic alteration (i.e., dehydrogenation and dealkylation). Aromatic biomarkers have been elucidated in detail by organic geochemists for application as chemical fossils (tracers) in the geological record, especially for coals, and complement the aliphatic biomarkers, which were described above. The relevance of PAHs and naturally-derived aromatic hydrocarbons (and sometimes oxygenated analogs) to ore deposits and hydrothermal mineralization became evident after the initial key publications that reported that those compounds were found in bitumen concretions of hydrothermal mercury deposits (Geissman, Sim, & Murdoch, 1967; Blumer, 1975; Wise et al., 1986).

    As already mentioned, the alkylated aromatic compounds are derived from terpenoids, which are ubiquitous lipid components of vascular plants (including ferns) and microbiota (Simoneit, 1998). The typical precursors are sesqui-, di-, tri-, and tetraterpenoids (Figs. 1–4). The aromatization processes, discussed earlier, proceed by dehydrogenation and dealkylation accelerated by thermal or catalytic interactions. Aromatization starts in recent (geological time) and contemporary sedimentary environments during early diagenesis of organic detritus. Aromatic biomarker formation will be illustrated here with thermal alteration processes as observed in smoke from burning of biomass, which yields the same biomarkers as elucidated in the geological record (Simoneit et al., 1993a; Simoneit, 1998). The natural product precursors in the biomass fuel, and their transfer as such or altered to the smoke (Figs. 1–4), are utilized in such studies (Simoneit, 1990a, 1993a; Abas et al., 1995).

    The product–precursor relationship for sesquiterpenoids yields mainly cadalene, and for diterpenoids commences by dehydrogenation of abietic acid (LI) to dehydroabietic acid (LII), with subsequent decarboxylation to dehydroabietin (LIII) and full aromatization to retene (LIV; Fig. 2; Simoneit, 1977a, 1986; LaFlamme & Hites, 1979; Wakeham, Schaffner, & Giger, 1980a; Simoneit & Mazurek, 1982). Dehydroabietane (XXII), which is found as the hydrocarbon in many resins, would dehydrogenate to simonellite (XXXIII) and proceed to retene. The various pimaric acids have been shown to rearrange and also dehydrogenate to dehydroabietic acid, or to dealkylate and dehydrogenate to 16,17-bisnordehydroabietic acid (LV), which aromatizes to pimanthrene (LVI). The tetracyclic kaurane, phyllocladane, and beyerane skeletons also rearrange under appropriate conditions to tricyclic products. Numerous other aromatic diterpenoid intermediates have been identified in smoke from the burning of biomass (i.e., conifer wood; Standley & Simoneit, 1994; Simoneit, Oros, & Elias, 1998; Simoneit, 2002a).

    The impetus for the extensive characterization of the aromatic triterpenoid hydrocarbons was the synthesis of authentic standards (Spyckerelle, 1975; Spyckerelle et al., 1977a,b), and it continues to be the first step necessary for progress. Aromatization of higher-plant triterpenoids (terrestrial source) generally commences in ring-A, and progresses to ring-E either directly or after the loss of ring-A due to initial photo- or thermo-chemical reaction (Fig. 3; Spyckerelle, 1975; LaFlamme & Hites, 1979; Wakeham, Schaffner, & Giger, 1980a; Simoneit, 1986). The precursor triterpenoids are ubiquitous in vascular plant waxes and gums, usually with an oxygen functionality at C-3 of ring-A. The numbers of compounds and alteration pathways continue to increase. The amyrins (α and β), lupeol, friedelanol, taraxerol, bauerenol, and others are readily oxidized in the environment to the 3-oxo derivatives, or are dehydrated by heat to the Δ2-olefin. The photolytic seco acids lose the remaining carbons of ring-A upon aromatization to yield the hydrochrysene series, and the Δ2-olefins or 3-oxotriterpenoids undergo successive ring aromatization to the hydropicene series (Fig. 3). The aromatization of hopanoids, when derived from bacterial detritus in sediments, proceeds with the diagenetic alteration of the precursors and dehydrogenation from ring-D to ring-A (Fig. 5) (Spyckerelle, 1975).

    The sterols that are present in higher-order biota also undergo aromatization reactions (Fig. 4), which have mainly been elucidated for geological samples (Mackenzie et al., 1982; Brassell, Eglinton, & Maxwell, 1983; Riolo, Ludwig, & Albrecht, 1985; Moldowan et al., 1992). The found products are primarily from the regular aromatization of ring-C to ring-A, and at higher temperatures the side chain is lost with subsequent cracking to mainly 3,4-dimethylphenanthrene (Fig. 4). Additional minor aromatic steroid derivatives have been identified as, for example, aromatic seco-, abeo-, dia-, spiro-, and nor-steranes (Hoffman, 1984).

    4. Sulfurized Biomarkers

    Sulfurized biomarkers and lipid compounds were first described for immature crude oils and recent sediments (Valisolalao et al., 1984; Sinninghe Damsté et al., 1987, 1988a,b, 1989). Those biomarkers consist of alkylated thiophenes, thiolanes, thianes, and benzothiophenes, and are derived from various lipid and terpenoid precursors during early diagenesis by reactions with HSmath image and H2S (Adam et al., 2000; Werne et al., 2000; Hebting, Adam, & Albrecht, 2003). Sulfurization also occurs under high thermal stress in deep sedimentary sequences to produce sulfur-rich kerogen and other sulfurized compounds such as 2-thia-adamantanes (Hanin et al., 2002).

    5. Novel Biomarkers

    Novel biomarkers are usually encountered in the geological or environmental records typically as hydrocarbons. Detection and determination are by interpretation of mass spectra in GC-MS analyses. The proofs of chemical structures are based on the proposed interpretation of the MS data, separation and purification of the unknown compound, exact structure determination by NMR methods or X-ray crystallography (if the compound is a solid that can be crystallized), and finally, comparison with a synthetic standard (Smith, Fowell, & Melsom, 1970; Hussler et al., 1984; Chaffee & Fookes, 1988; Moldowan et al., 1991; Hauke et al., 1992a,b). The next question concerns the biological source of the biomarker precursor compound. Many biomarkers still have no proven precursors nor known biological sources (e.g., perylene, tricyclic terpanes, Aquino Neto et al., 1982; Venkatesan, 1988; Simoneit et al., 1990b).

    The characterization of a novel series of biomarkers is illustrated with the gem-dialkylalkanes in bitumen from a hydrothermal system on the Mid-Atlantic Ridge (Simoneit et al., 2004). The total bitumen consists of hydrocarbons, a major UCM (unresolved complex mixture of branched and cyclic compounds) and mature biomarkers (e.g., hopanes) (Fig. 16a,b). The bitumen contains a series of cyclopentylalkanes (CnH2n) that range from n = 14 to 34, with only even-chained pseudohomologs and a maximum at n = 18. That source is biogenic, based on the presence of only even-carbon number homologs, but the precursors are unknown. The series is defined by the mass spectra and the comparison of one compound with an authentic standard. Fragmentation is simple, similar to an alk-l-ene, and consists of a molecular ion, M-C2H4, a cyclopentyl ion with H-transfer as the base peak (m/z 68), and typical alkane cleavage.

    Figure 16.

    Salient features of the GC-MS data for the saturated fraction of bitumen sample M2 from the Mid-Atlantic Ridge: (a) total ion current trace, (b) m/z 191 fragmentogram (key ion for tricyclic terpanes and hopanes), (c) m/z 85 fragmentogram (key ion for 3-ethyl-3-methylalkanes and n-alkanes), (d) m/z 99, key ion for 3,3-diethylalkanes with n-alkanes, (e) m/z 127, key ion for 5,5-diethylalkanes with n-alkanes, and (f) m/z 141, key ion for 6,6-diethylalkanes. (Numbers refer to total carbon number, dots over peaks are n-alkanes; Simoneit, Lein, Peresypkin, Osipov, 2004; reproduced with permission, Elsevier Science Publishers, copyright 2004.)

    Four other homologous series are gem-diethyl substituted n-alkanes. The gem prefix designates geminal-substituted compounds; i.e., two substituents on the same atom of a disubstituted compound. Several research groups have reported the presence of 5,5-diethylalkanes, as well as lesser amounts of other gem-dialkylalkanes, in sedimentary rocks back to the Precambrian and in hydrothermal fluids (Mycke, Michaelis, & Degens, 1988; Logan et al., 2001; Simons et al., 2002; Greenwood et al., 2004; Kenig et al., 2004). The correct structural assignments of the 3,3-diethylalkanes (LVII) and the 5,5-diethylalkanes (LVIII) by comparison with synthetic standards were described recently (Kenig et al., 2002, 2004). The other gem-dialkylalkane series were tentatively identified based on an interpretation of their characteristic mass-spectrometric fragmentation patterns and gas-chromatographic retention factors (Kenig et al., 2002; Greenwood et al., 2004).

    The 3-ethyl-3-methylalkanes (i.e., 2,2-diethyl substitution) range from C14 to C34, with only even-chained pseudohomologs detectable and a Cmax at 18 (Fig. 16c). The structures are interpreted from the mass spectra (Fig. 17a), which consist of a cleavage of C2–C3 to yield the base peak (C6H13, m/z 85), M-C2H5, minor alkane cleavage, and no molecular ion. The fragmentation patterns match that of the library mass spectrum for 3-ethyl-3-methyldecane, but no authentic coinjection standard is available. Their GC retention times are earlier than the normal alkanes (Kováts, KF, and relative retention factors, RRF, are calculated; KF/RRF C18 = 1,752/−0.48, C26 = 2,558/−0.42; Simoneit et al., 2004).

    Figure 17.

    Representative mass spectra of gem-dialkylalkanes in bitumen of sample M2: (a) 3-ethyl-3-methylpentacosane, C28H58; (b) 3,3-diethyltricosane, C27H56; (c) 5,5-diethyltricosane, C27H56; and (d) 6,6-diethyltetracosane, C28H58 (Simoneit, Lein, Peresypkin, Osipov, 2004; reproduced with permission, Elsevier Science Publishers, copyright 2004).

    The 3,3-diethylalkanes range (LVII) from C15 to C39, with a Cmax at 27 and only odd-carbon numbered pseudohomologs (Fig. 16d). The structures are based on an interpretation of the mass spectrometric fragmentation patterns (Fig. 17b) and coinjection of authentic 3,3-diethylpentadecane. The mass spectra have a base peak at m/z 57, an intense ion at m/z 99 (C7H15, key ion) from C3–C4 cleavage, M-29 (C2H5), minor M-57, typical alkane fragments, and no molecular ion. Their GC retention times precede the n-alkanes (KF/RRF, C23 = 2,262/−0.38, C29 = 2,872/−0.28; Simoneit et al., 2004).

    The 5,5-diethylalkanes (LVIII) also range from C15 to C39, with odd-carbon numbered pseudohomologs and a Cmax at 29 (Fig. 16e). The structures are based on interpretation of the mass-spectrometric fragmentation patterns, and on prior reports of the occurrence of those compounds in sedimentary sulfides and rocks (Mycke, Michaelis, & Degens, 1988; Logan et al., 2001; Kenig et al., 2002, 2004; Greenwood et al., 2004). The mass spectra generally exhibit a base peak at m/z 57 (C4H9), an intense key ion at m/z 127 (C9H19), M-29 (C2H5), M-57 (C4H9), other general alkane fragments, and no or a low-intensity molecular ion (Fig. 17c). Their GC retention times further precede those of the n-alkanes (KF/RRT, C25 = 2,403/−0.97, C33 = 3,211/−0.89).

    The 6,6-diethylalkanes range from C16 to C38, with a Cmax at 30 and only even-carbon numbered pseudohomologs (Fig. 16f). The structures are based on the analogous interpretation of the mass spectra (Fig. 17d), which have a base peak at m/z 57 (C4H9), an intense key ion at m/z 141 (C10H21), M-29 (C2H5), M-71 (C5H11), general alkane fragments, and no molecular ion. Their GC retention times further precede those of the n-alkanes (KF/RRF, C24 = 2,290/−1.10, C30 = 2,891/−1.09).

    Those branched gem-alkane series are biomarkers because of the locale of their occurrence and the presence of only alternate pseudohomologs (even- or odd-carbon numbers only). Their inferred origin is from probable microbial precursors of unknown species, where methylation and ethylation, diethylation, and butylation and ethylation occurred during biosynthesis at the C2, C3, C5, or C6 positions of odd- or even-carbon chained substrates (i.e., C11–C33). Those compound series occur in samples of relict or weathered hydrothermal talus at the base of active vent systems, consistent with a lower input of detritus from archea and with an enrichment of lipid residues from microbial mats and/or sulfide-oxidizing bacteria as proposed for the ancient examples (Mycke, Michaelis, & Degens, 1988; Logan et al., 2001; Simons et al., 2002; Greenwood et al., 2004) and hydrothermal fluids on the Juan de Fuca ridge flank (Kenig et al., 2002, 2004).

    B. Environmental Chemistry

    Environmental chemistry is the study of the sources, reactions, transport, effects, and fates of chemical species, organic and inorganic, in water, soil, and air environments of Earth. The topic has considerable overlap with environmental biogeochemistry—the study of effects from environmental chemical species on life—and with toxicological chemistry—the study of toxic substance effects on living organisms. It is important in those fields to be able to distinguish natural and geological biomarker compounds from the superimposed synthetic (anthropogenic) compounds. This distinction can be accomplished by mass spectrometric techniques. Numerous texts and handbooks are available on those topics (Hutzinger, 1988; Manahan, 1989, 1991; Schwarzenbach, Gschwend, & Imboden, 1993, 2003). Mass spectrometry is a major analytical tool for research and for monitoring studies in environmental chemistry (Lee, Novotny, & Bartle, 1981; Hites, 1985; Evershed, 1992, 1993; Eganhouse, 1997; Stoll et al., 1997; Takada et al., 1997; White et al., 1997; Poster, Sander, & Wise, 1998). The basic literature of environmental chemistry has been extensively reviewed elsewhere (Hites, 1977, 1992, 1998; Atlas et al., 1993; Ong & Hites, 1994; Suter et al., 1997). Initial research involved the assessment of PAH and related toxic compounds (Giger & Blumer, 1974; Wakeham, Schaffner, & Giger, 1980a,b), and branched into fate and accumulation of synthetic compounds such as pesticides/herbicides in the food chain (Simonich & Hites, 1995a,b; Baker & Hites, 1997; Hillery, Basu, & Hites, 1997; Buehler, Basu, & Hites, 2001). Currently, the concerns are with risk assessment, human health effects, and bioactive compounds as endocrine disruptors (DeCaprio, 1997; Hanselman, Graetz, & Wilkie, 2003; McArdell et al., 2003; Sexton, Needham, & Pirkle, 2004).

    Mass spectrometry, especially GC-MS, HPLC-MS, HRMS, negative-ion MS, and chemical-ionization MS, are the major analytical methods for compound characterization and for monitoring in environmental chemistry. HPLC-MS is used for polar trace compounds (Poiger et al., 1996; Suter, Riediker, & Giger, 1999; McArdell et al., 2003). The application of the chemical ionization (CIMS) technique is of utility for compounds with low-intensity molecular ions or low abundances of initial fragments from M·+. CI occurs via ion-molecule reactions instead of ionizing with an energetic electron beam. The ion, often referred to as a reagent ion, reacts with a sample molecule by transferring a proton or by abstracting an H- or an electron; those processes impart a +1 charge to the sample molecule. CI is used in environmental analyses to increase the MS sensitivity of pesticide/herbicide compounds (Koester, Simonich, & Esser, 2003).

    1. Natural Product and Fossil Fuel Biomarkers

    The assessments of ecosystems and environmental pollution are connected in that biomarkers derived from natural products are utilized as proxies. With the present threat of rapidly induced climate change, numerous ecosystem studies are underway. For example, tracking past mangrove ecosystems along continental coasts is a sensitive dynamic indicator for sea-level changes and river-transported sediment erosion. Mangrove pollen and taraxerol (LIX), the dominant internal biomarker of mangroves, have been proposed as tracers for past erosional events, which transported sediments into the southeast Atlantic (Versteegh et al., 2004). GC-MS was used to screen and quantitate the biomarkers in cores.

    Trace level pollution with fossil fuel residues in a pristine natural environment is illustrated for the lake of Crater Lake National Park (Oros, Collier, & Simoneit, 2004). The lake is one of the clearest bodies of water in the world, but is impacted by tourism and other anthropogenic inputs. Tour boats operate on the lake during summer and introduce unknown quantities of hydrocarbons into the water. The GC-MS data for typical samples of water surface slick and lake-bottom sediments are shown in Figure 18. Aliphatic and higher molecular weight hydrocarbons (C10–C30) are hydrophobic and therefore concentrate at the water–air interface, forming a slick (film) with the natural lipids, which also accumulate there. Those surface films (also termed surface microlayer, upper 100 μm) are important for concentrating lipophilic higher molecular weight compounds such as petroleum hydrocarbons and natural lipids (Morris & Culkin, 1975). The more water-soluble and volatile petroleum components (e.g., BTEX) partition into the water column or evaporate, and are consequently depleted in the water–air interface slick. Therefore, sampling of surface slicks is a way to analyze an enriched upper limit in concentration of hydrophobic petroleum hydrocarbons superimposed on the natural background lipids (Fig. 18a). The GC-MS data indicate n-alkanes from C18 to C25, with a Cmax at 22 and UCM, typical of motor exhaust. The petroleum hydrocarbons in lake-bed sediments are mixed with natural lipids. The sediment extract from the boat dock area exhibits an UCM and the n-alkanes, as discerned in the m/z 85 mass fragmentogram, range from C15 to C33, with a Cmax at 24 (Fig. 18b,c). A minor amount of higher-plant wax is also superimposed. The presence of petroleum products is confirmed by the tricyclic terpane and 17α(H)-hopane series (Fig. 18d) (Peters & Moldowan, 1993; Rogge et al., 1993; Bieger, Hellou, & Abrajano, 1996). A sediment extract from a deep part of the lake (590 m depth, North Basin) had an extremely low-hydrocarbon content (Fig. 18e,f). The total extract is comprised of n-alkanoic acids, n-docosanol, and elemental sulfur. The n-alkanes (m/z 85 key ion trace) are derived from minor petroleum components, plant wax, and plankton lipids.

    Figure 18.

    GC-MS data for extracts of surface slick and sediments from Crater Lake in C.L. National Park: (a) TIC trace for surface slick reflecting boat exhaust, (b) TIC trace for sediment, 5 m water depth near boat docks, (c) m/z 85 key ion for n-alkanes, (d) m/z 191 key ion for tricyclic terpanes and 17α(H)-hopanes from fuel use (petroleum products), (e) TIC trace for sediment, 590 m water depth, and (f) m/z 85 key ion for n-alkanes. Numbers refer to carbon-chain length of n-alkanes in (a, c, f) and n-alkanoic acids in (e), or to carbon number of biomarkers in (d), UCM = unresolved complex mixture, S8 = sulfur, OH = alkanol.

    The total hydrocarbons attributable to petroleum in the deep sediment sample amount to 0.02 mg/kg. Significant petroleum residues (0.3–1.4 mg/kg) from engine exhaust are detectable in the sediments of the boat dock cove, but the values are low compared to other areas (e.g., Rhone River Estuary, Mediterranean, 167 mg/kg sediment; Bouloubassi & Saliot, 1993b). It should be pointed out that the guideline cutoff for non-polluted (by oil and grease, assumed equivalent to petroleum residues) harbor sediment is <1,000 mg/kg and moderate pollution is from 1,000–2,000 mg/kg (EPA (Environmental Protection Agency), 1977).

    2. Anthropogenic Compounds

    The application of mass spectrometry to determine anthropogenic compounds in environmental, as well as forensic chemistry is illustrated with the following example. The question was to determine whether the GC-MS data obtained by a commercial laboratory showed any organic compounds due to the burning of material other than firewood in a wood stove. A perusal of the data indicates that it is excellent, and that numerous compounds can be unambiguously identified (Fig. 19). The major peaks in the TIC plot of a sample are the surrogate standards, internal standards, coinjection standards, and compounds in the blank (probably solvent impurities) (compounds am). The compounds in the sample extract (139) can be categorized into two groups: (1) those from wood smoke and (2) those from non-wood fuels.

    Figure 19.

    GC-MS TIC trace for the extract from a soot sample taken in the interior firebox of a wood stove to show the annotations of standards (am) and identified compounds (139).

    Burning is an incomplete combustion process that is analogous to laboratory hydrous pyrolysis, and thus yields smoke that consists of particles and vapors (gases) (Simoneit et al., 1993a, 1999; Simoneit, 1999, 2002a). Wood and other organic fuels are dominantly water in addition organic matter, and upon burning the water steam strips organic compounds into the vapor. Those compounds condense onto and into the particles upon cooling in a firebox or chimney or upon mixing in the atmosphere. Therefore, most of the tracers of the fuel are emitted in the smoke. Furthermore, the condensation/deposition of soot (analogous to activated carbon) entraps and protects a portion of those organic tracers even though some tracers are quite volatile and thermally unstable. The compounds from wood-smoke deposition (soot) are listed in Table 2 in normal print, and those from non-wood fuels in bold type.

    Table 2. Organic compounds in a fireplace soot residue*Thumbnail image of
  • *All concentrations are normalized to the highest GC response.

  • aCompounds in bold type are not found in wood smoke.

  • The compounds that did not originate from wood smoke nor from the experimental and analytical procedures can be attributed to non-wood fuels burned in the stove over extended times. Those compounds are, in part, unaltered chemicals and some are specific derivatives from the burning of their precursors. They consist of plasticizers, wetting agents, perfumants (fragrances; Simonich et al., 2000), antioxidants, solvents, and greases/lubricants. Actual pyrolysis products of plastics (polyethylene, polystyrene, etc.) are not detectable, because if formed, then they would be in the vapor phase due to their extreme volatility (ethylene, vinylbenzene).

    The plasticizers are mainly the phthalate esters with the associated antioxidant, 2,6-di-t-butyl-4-hydoxy-4-methyl-2,5-cyclohexadien-1-one (Table 2). The phthalates are of special interest because they are found as the esters and as derivatives (Fig. 20). The esters and antioxidant are vaporized directly from plastics, and the derivatives are generated by thermal alteration. Di-(2-ethylhexyl)phthalate is the least volatile of the phthalates, and thus most susceptible to thermal cracking, to yield 2-ethylcaproic acid and 1,2-benzenedicarboxylic (o-phthalic) acid. The latter compound occurs at trace levels, and the presence of phthalimide and N-methylphthalimide indicate combustion with a major nitrogen source such as NH3 (ammonia) from nitrogen-rich organic matter (Fig. 20). Wood, paper, and plastics are nitrogen-poor. Thus, possible nitrogen sources are urea, organic waste, etc. The cyanobenzene and dicyanobenzene (Fig. 20) are inferred high-temperature alteration products from 1,2-benzenedicarboxylic acid diamide, derived from the amination of the diacid during burning. The wetting agents, emulsifiers, and fragrances are obvious in the table. Therefore, those tracers from non-wood fuels probably derive from burning of hospital/nursing home wastes.

    Figure 20.

    Incomplete combustion products from dioctyl phthalate as detected in wood-stove soot.

    C. Archeological and Forensic Chemistry

    Archeological and forensic chemistry are related because both are similar inquiries about the sources and alteration of organic matter (Beck, 1974). The various techniques of mass spectrometry are used extensively to characterize organic mixtures and to identify compounds.

    1. Biomarkers in Archeology

    Archeological chemistry addresses a wide spectrum of topics, and also branches into art restoration and preservation (Nazaroff & Cass, 1991; Grosjean, Salmon, & Cass, 1992; Ye, Salmon, & Cass, 2000). For example, it can document past agricultural practices, decipher embalming agents, assess food intake in coprolites, or determine contents of pottery vessels (Dudd & Evershed, 1998; Evershed et al., 1999, 2002; Buckley & Evershed, 2001; Copley et al., 2001; Bull et al., 2002).

    2. Use of Biomarkers in Forensics

    Forensic chemistry and the related medical jurisprudence and toxicology sciences also utilize mass spectrometry and occasionally encounter the biomarkers described in this review. The same situation is the case for food and wine chemistry (Webb, 1974; Flament, 2002; Bell et al., 2003). Those topics are not discussed further, except to mention that adulteration of products is one aspect in which the new method of CSIA by GC-isotope MS is of utility (Woodbury et al., 1995).

    D. Organic Cosmochemistry

    The NASA-Exobiology Program initially supported research on chemical evolution in the early 1960s, and vigorously developed the contemporary state-of-the-science expertise and instrumentation (especially GC-MS) for the analyses of organic matter and provided the impetus for the advancement of organic geochemistry.

    Carbonaceous chondrites and ancient sediments (1–3 × 109 years) were analyzed for biogenic organic residues (molecular fossils), such as lipid, amino acid, and pigment compounds. That new research field (chemical evolution as applied to exobiology) was eloquently summarized for example, by Calvin (1969) and in a broader context by Mason (1992). It had the goal to ultimately examine lunar samples for biogenic/prebiotic organic matter, but none was found (Eglinton, Maxwell, & Pillinger, 1972; Gibson & Chang, 1992). Also, organic analyses were performed to test for contemporary life on the Viking Mission to Mars (Biemann et al., 1977; Levin & Straat, 1977; Kieffer et al., 1992). The results were inconclusive. Over the past 25 years, organic geochemistry has expanded into a myriad of research ideas and has built a vast database on biomarkers that are indicative of biochemistry. That database can now be applied to the search for tracers of prebiotic organic chemistry and of extinct and extant life on Mars, elsewhere in the solar system, and even in the cosmos (Schwartz & Mancinelli, 1989; Carle, Schwartz, & Huntington, 1992; Bock & Goode, 1996; Dobrijevic et al., 2001).

    1. Biomarkers in Extraterrestrial Samples

    The utility of biomarkers as indicators of biogenic, paleoenvironmental, and geochemical processes on Earth has been widely accepted (Didyk et al., 1978; Mackenzie et al., 1982; Johns, 1986; Simoneit et al., 1986; Brassell, 1992; Imbus & McKirdy, 1993; Mitterer, 1993; Simoneit, 1998, and references therein). The detailed characterization of biomarker mixtures in terms of sources and degrees of alteration permits the assessment of: (a) extant life and major contributing species, and (b) extinct life, major contributing species and geo/hydrothermal alteration.

    The Moon has been shown to be sterile with no evidence of past life (Henderson et al., 1971; Oyama et al., 1971; Eglinton, Maxwell, & Pillinger, 1972). Most of the carbon is inorganic, is implanted by the solar wind, or is residual in the igneous rocks. In the case of Mars, the biomarkers should be based on carbon biochemistry, and those derived from lipids and biopolymers would be expected to be preserved best if life evolved there during its early history (3–4 × 109 years ago) (Bock & Goode, 1996; Committee on Planetary and Lunar Exploration, 2002). Oxidized and reduced products would both be expected. The most versatile and specific methods available now for biomarker characterization are based on mass spectrometry (Simoneit, 2002b).

    2. Biomarkers of Past Microbial Ecosystems

    The biomarker concept can be applied to bacterial lipids, which represent the development and evolution of oxygenic photosynthesis and aerobiosis, because those processes are the basis of the modern carbon cycle and sustain complex life (Summons, Jahnke, & Simoneit, 1996). Thus, detection of life on Mars is initially concentrated on microorganisms, especially bacteria (Committee on Planetary and Lunar Exploration, 2002). Microbiota, i.e., bacteria and archaea, are also the major contributors to the biomarker record of the early Earth (Brocks et al., 1999; Summons et al., 1999). The isoprenoids and hopanoids are the most extensively documented alkyl and cyclic biomarkers for bacteria (Ourisson, Albrecht, & Rohmer, 1979; Summons & Jahnke, 1990; Jahnke et al., 1992, 1999; Ourisson & Rohmer, 1992; Rohmer, 1993; Kates, 1997). It is important to note that a single natural product precursor (phytol in Fig. 1 or diploptene in Fig. 2 of the report by Simoneit, 2004a) can be altered to a large number of derivative biomarkers whose structures are still correlatable to the precursor. Thus, a sequence or set of biomarkers must be identified to assess organic matter source.

    3. Biomarkers of Extant Microbial Ecosystems

    Contemporary microbial ecosystems on Earth are found mainly in extreme environments (hot springs, desert) and most others (estuaries, lakes) are complicated by the admixture of higher-plant detritus (White et al., 1997; Winfree et al., 1997). However, such ecosystem biomarkers should be illustrated with an example in case this type of extant organic matter is encountered in extraterrestrial locales (e.g., Mars). The GC-MS data of a total extract (derivatized) from an anoxic estuarine sediment is shown in Figure 21. The dominant compounds are mostly from microbial (including algal) sources, and comprise lipid and steroid natural products with early diagenetic alteration derivatives. The biomarkers indicative of anoxic alteration of microbial lipids are phytadienes, isoprenoid thiophenes, methylpentadecyl thiophene, phytol, and dihydrosqualene. The steroids, i.e., cholesterol, campesterol, and β-sitosterol, are indicators for algae and higher plants. The n-alkanes, n-alkanols, and n-alkanoic acids are ubiquitous biogenic lipids, where the long-chain (>C20) homologs are more concentrated in higher plants. α-Glycerophosphoric acid is a labile cell component of all biota derived from the active phospholipid pool, and is short-lived in the environment. Thus, samples of contemporary environments exhibit biomarker compositions indicative of extant life with early diagenetic alteration derivatives and reflect the gross species input.

    Figure 21.

    GC-MS TIC trace for the total extract of an anoxic estuarine sediment, analyzed as the trimethylsilyl (TMS) derivatives. The homologous lipids are indicated by the key, and other biomarkers are labeled (Simoneit, 2004; reproduced with permission, Elsevier Science Publishers, copyright 2004).

    4. Abiogenic Organic Compounds

    Organic compounds, which are obviously synthesized from inorganic precursors (e.g., CO) by hydrothermal activity or other non-biogenic processes (e.g., volcanism), must also be considered in the search for biomarkers indicative of life (Holm, 1992). Such de novo products could be overwhelmed by an excess of products formed from alteration of contemporary natural organic precursors (e.g., hydrocarbons, alkanoic acids, etc.) or occur solely as prebiotic organic compounds. Thus, such organic syntheses are being demonstrated in the laboratory to be able to distinguish abiogenic compounds from those compounds derived from organic matter alteration in geologic environments (Simoneit, 1992b, 2004a,b). Organic synthesis under hydrothermal conditions is theoretically possible (Shock, 1990) and is well-established in industry (e.g., synthetic fuels) (Anderson, 1984; Simoneit, 1995).

    Strecker-type synthesis experiments have, for example, been carried out under hydrothermal conditions (Hennet, Holm, & Engel, 1992; Marshall, 1994). Those generated amino acids such as glycine, DL-alanine, DL-aspartic acid, and DL-glutamic acid. The formation of lipid compounds during an aqueous organic synthesis reaction (Fischer–Tropsch-type) was studied with solutions of oxalic acid (also formic acid) as the carbon and hydrogen source (McCollom, Ritter, & Simoneit, 1999a; Rushdi & Simoneit, 2001). The reactions were conducted in stainless-steel vessels by heating aqueous oxalic acid solutions at discrete temperatures from 100 to 400°C at increments of 50°C for 2 days each. The maximum lipid yield, especially for oxygenated compounds, is produced in the temperature range of 150–250°C. The lipid components ranged from C6 to >C33 and included n-alkanols (fatty alcohols), n-alkanoic acids (fatty acids), n-alkyl formates (esters), n-alkanals (aldehydes), n-alkanones (ketones), n-alkanes, and n-alkenes (both hydrocarbons), all with no carbon number preferences. Biomarkers were not detectable. An example of the GC-MS data for the homologous polar components from the aqueous thermosynthesis using oxalic acid at 200°C is shown in Figure 22 (McCollom, Ritter, & Simoneit, 1999a). The n-alkanols (C7–C23) and n-alkanoic acids (C6–C16) (both as the trimethylsilyl derivatives) are the major compounds in the total extract. At temperatures >300°C, synthesis competed with cracking and reforming (also rearrangement) reactions, and >400°C resulted in major cracking and an extensive formation of PAHs and their alkylated homologs.

    Figure 22.

    GC-MS data for the polar compound mixture synthesized from aqueous oxalic acid at 200°C (2 days), analyzed as the (TMS) derivatives: (a) m/z 103 fragmentogram (key ion for TMS ethers of n-alkanols), and (b) m/z 117 fragmentogram (key ion for TMS esters of n-alkanoic acids). Numbers refer to carbon chain length, dots are acids (Rushdi & Simoneit, 2001; Simoneit, 2004; reproduced with permission, Elsevier Science Publishers).

    Such abiotic amino acids, lipids, and other possible compounds are not biomarkers per se because they do not necessarily originate from biosynthesis. Rather, they should be considered as a distinctly separate group, and termed as either prebiotic or abiotic organic compounds.

    5. Non-Biomarkers

    High-temperature thermal alteration of organic matter, incomplete combustion of methane, and de novo synthesis from free-radical recombination produces polycyclic aromatic compounds, especially the ubiquitous PAHs. The unsubstituted parent PAHs are the compounds derived mainly from those various processes. They are commonly analyzed by GC-MS and other methods for the compound range from naphthalene (MW 128) and phenanthrene/anthracene (MW 178) to coronene (MW 300) (Clar, 1964; Lee, Novotny, & Bartle, 1981). Free radicals, such as acetylyl radical (HC[TRIPLE BOND]C·), recombine successively to form the unsubstituted PAHs. Once formed, the simple low-molecular weight PAHs (e.g., naphthalene) undergo further synthesis buildup by the “zig-zag addition process” to the thermodynamically most stable PAH series (Stein, 1986; Simoneit & Fetzer, 1996). PAHs are generally not biomarkers as inferred by various authors (Bernstein, Sandford, & Allamandola, 1999; Ehrenfreund, 1999; Ehrenfreund & Charnley, 2000; Blake & Jenniskens, 2001), because they can form from abiological carbonaceous matter by free-radical buildup (Simoneit & Fetzer, 1996) or by aromatization of biomarker natural products (Simoneit, 1998).

    The abiogenic PAHs occur primarily as the parent compounds without alkyl or alicyclic substituents as is the case for the aromatic biomarkers. Thus, phenanthrene is not a biomarker, but 1-methyl-7-(1-methylethyl)phenanthrene (i.e., retene, LIV) is, because the latter is generated from numerous natural product precursors by dehydrogenation. Retene has the specific arrangement of four carbons on phenanthrene as 1-methyl and 7-isopropyl, which was inherited from the sequential alteration products and natural precursors (e.g., 1,1-dimethyl-7-isopropyl-1,2,3,4-tetrahydrophenanthrene, abieta-8,11,13-triene, XXII, etc., in Fig. 2; also compare aromatization of the hopanes, Fig. 5), thus making it a biomarker. On the other hand, random alkyl substitution on aromatic compounds in the geological record is mainly with methyl groups, and such aromatic compounds are not biomarkers, termed herein non-biomarkers. Thus, the C4-phenanthrenes that occur, for example, in petroleum have methyl groups at all isomeric positions, and retene occurs at extremely low concentrations.

    Polynuclear aromatic hydrocarbons, but not aromatic biomarkers, have been detected in carbonaceous chondrites, in dust and graphite grains of interstellar clouds, interplanetary dust, and cometary ice (Pering & Ponnamperuma, 1971; Clemett, Maechling, & Zare, 1992, 1993, 1995; Ehrenfreund & Charnley, 2000). However, the rapid contamination of newly fallen meteoritic material (e.g., Allende meteorite fall in 1968) by terrestrial organic compounds (Han et al., 1969) suggests caution in interpreting results, and requires rigorous background control and blank analyses. Lunar samples contained no endogenous PAHs (Burlingame et al., 1971; Henderson et al., 1971; Holland et al., 1972). The report of PAHs and possible fossil microbial cells in a Martian meteorite collected in Antarctica with its implications for the origin of life was intriguing (McKay et al., 1996). If the PAHs were not contaminants or artifacts, then the exciting hypothesis that life evolved during the early history of Mars had merit. However, it has been shown that those PAHs and the related organic matter were of a terrestrial and partially recent in origin, i.e., pervasive contamination (Becker, Glavin, & Bada, 1997; Jull et al., 1998). Such PAHs generally have low-molecular weights and typically range to pyrene, with naphthalene and phenanthrene predominant, and are characteristic of terrestrial background from fossil fuel utilization. Thus, the carbon and deuterium isotope compositions (i.e., CSIA) of individual PAHs should be used in conjunction with the presence of alkylated and alicyclic analogs to assign a biogenic or abiogenic origin.

    IV. PROGNOSIS

    This review has presented an insight into the literature base about the applications of mass spectrometry for the elucidation of biomarkers, and the applications of biomarkers as molecular tracers for numerous processes and in various scientific disciplines. Biomarker analysis and elucidation has been illustrated with various examples to clarify the concepts and procedures. The number of novel compounds as well as the applications of biomarker tracers is expected to keep increasing. The utility of mass spectrometry in the contemporary interdisciplinary sciences will also continue to expand, especially as the analytical instrument capabilities develop further with increased sensitivities, ease of operation, lower maintenance, better separation technology, improved ionization efficiencies, etc. Because many applications of mass spectrometry are interdisciplinary, it would be of great utility to find a cost-effective means of incorporating the missing mass spectra of the numerous natural products, synthetic compounds, geological biomarkers, environmental biomarkers, and their various derivatives (TMS, Ac), documented by organic geochemists and environmental chemists, into the commercial MS libraries (Wiley, NIST). This would greatly enhance the quality and reliability of future literature data.

    Most MS analytical methods for current biomarker applications usually deal with compound concentrations in the pg to μg/g range, small sample sizes (5–1,000 μL solution), and as complex mixtures. At those low concentrations the contamination by extraneous organic matter during sample acquisition, preparation, extraction, and analysis can be a major problem. Analysts need to be aware of this. The help of an organic mass spectrometrist/chemist should be sought when mass spectra of unknown compounds are encountered. Preliminary mass spectrum interpretation may lead to characterization of novel biomarkers and their ultimate structure proof by use of other chemical methods (NMR, XRF). This is an expanding field and knowledgeable organic mass spectrometrists are and will continue to be needed.

    Acknowledgements

    I thank the National Aeronautics and Space Administration (Grant NAS5-13502) for partial support during the preparation of this review, Professor D.M. Desiderio for the invitation to prepare this work and his editorial assistance, and two anonymous reviewers for their thorough and helpful comments.

    APPENDIX

    Key Chemical Structures Cited in the Text

    Thumbnail image of Thumbnail image of Thumbnail image of

    Biographical Information

    original image

    Dr. Bernd R.T. Simoneit received a B.S. in Chemistry from the University of Rhode Island in 1960 and a Ph.D. in Organic Geochemistry under the supervision of Prof. Geoffrey Eglinton from Bristol University, England in 1975. He has been involved in academic research since 1966, initially at the University of California in Berkeley and then in Los Angeles. He joined the faculty at Oregon State University in 1981, where he is currently a professor emeritus in the College of Oceanic and Atmospheric Sciences and the Department of Chemistry. His research has dealt with the distribution and fate of organic matter in the geo- and biospheres (organic and petroleum geochemistry, biogeochemistry, and analytical chemistry applied to pollution and ecology), the geochemistry of organogenic elements and compounds in the lithosphere (hydrothermal systems, fluid inclusions, sediment subduction and accretion, organic metamorphism in supercritical fluids, prebiotic organic synthesis), and the origin and evolution of life on Earth and the search for evidence of extraterrestrial life. In all the research topics, he has extensively applied mass spectrometry as the analytical tool.

    Ancillary