- © 2014 by the Mineralogical Society of America
The unambiguous identification of graphitic carbons as remains of life in ancient rocks is challenging because fossilized biogenic molecules are inevitably altered and degraded during diagenesis and metamorphism of the host rocks. Yet, recent studies have highlighted the possible preservation of biosignatures carried by some of the oldest graphitic carbons. Laboratory simulations are increasingly being used to better constrain the transformations of organic molecules into graphitic carbons induced by sedimentation and burial processes. These recent research advances justify a reevaluation of the putative biogenicity of numerous ancient graphitic carbons, including the presumed oldest traces of life on Earth.
- experimental fossilization
- abiotic processes
- Archean/early life
The search for evidence of early life in the geological record has often involved graphitic carbons and has been a subject of intense debates and divisive controversies (e.g. Mojzsis et al. 1996; Rosing 1999; Schopf et al. 2002; Brasier et al. 2002; Van Zuilen et al. 2003). Numerous challenges exist, such as contamination and abiotic pathways that might lead to the formation of graphitic carbons similar to fossilized biological remains (Pasteris and Wopenka 2003; McCollom and Seewald 2006). The main complication resides in the poor preservation of biosignatures in the oldest metasedimentary rocks. Bioalteration processes, as well as the increase of temperature and pressure conditions during burial, inevitably alter the original biochemical signatures of organic molecules. As a result, the general paleobiological perception has long been that burial processes are detrimental to the preservation of signatures of life by graphitic carbons (Schiffbauer et al. 2007). However, although numerous challenges still remain, newly available microanalytical techniques create opportunities for the assessment of the biogenicity of graphitic carbons.
This contribution begins with an overview of the processes involved in the conversion of biogenic organic molecules into graphitic carbons. We describe remarkable examples of preservation of biosignatures by graphitic carbons in metasedimentary rocks. We then explore possible solutions and new approaches to evaluate the biogenicity of graphitic carbons. Finally, we discuss the question of the putative biogenicity of some Eoarchean and Paleoarchean graphitic carbons in light of some representative examples from the recent literature.
CONVERSION OF ORGANIC BIOLOGICAL REMAINS INTO GRAPHITIC CARBONS
When organisms die, their organic molecules are usually rapidly biodegraded. Most biomolecules released to the environment are either oxidized or assimilated by living microbes. Yet, some organic macromolecules resistant to biodegradation may escape the biological cycle and be incorporated in sediments by selective preservation. For instance, some labile organic molecules may bypass biodegradation through degradation/recondensation reactions (e.g. Vandenbroucke and Largeau 2007). Adsorption on inorganic surfaces and encapsulation within mineral microcavities or larger molecules are also processes permitting the incorporation of organic molecules into sediments (e.g. Vandenbroucke and Largeau 2007). The respective contribution of each of these processes remains debated. Of note, organic molecules incorporated into sediments through selective preservation, adsorption, or encapsulation have likely been only slightly degraded and thus may still be identified. In contrast, labile molecules incorporated into sediments through degradation/recondensation reactions become geopolymers that may exhibit chemical signatures distinct from those of the initial biomolecules. All these processes lead to the formation of insoluble, polymeric, sedimentary, organic macromolecules, collectively called kerogen (Vandenbroucke and Largeau 2007).
Subsequent diagenetic processes may further complicate the recognition of biosignatures (Fig. 1). Partially preserved biosignatures carried by kerogen may suffer from compaction and fluid migration. As fluid migration is limited within low-permeability sediments, they tend to favor the preservation of isotopic, elemental, and molecular biosignatures, whereas high-permeability sediments are unlikely to preserve indigenous biosignatures (e.g. Vandenbroucke and Largeau 2007). During burial, reducing sediment may minimize redox reactions with biogenic organic matter and improve its preservation potential, through, for example, kerogen sulfurization (e.g. Vandenbroucke and Largeau 2007).
When burial depth increases, sedimentary rocks are metamorphosed under increasing temperature and pressure conditions. Burial-induced biomolecule degradation processes (carbonization/graphitization processes) may alter the morphological and chemical signatures of life remains in rocks (Fig. 1).
With increasing temperature, weaker organic bonds are thermally broken and dehydrogenation reactions become predominant. Metamorphic degradation of kerogen is associated with an increase in aromaticity and the release of heteroatoms such as O, N, and S. By promoting structural reorganization, thermal metamorphism may ultimately lead to pure graphite that may have completely lost its original biosignatures. Finally, graphitic carbons remain susceptible to late oxidation and hydrogenation, as well as to alteration and contamination by younger biological carbon from the subsurface (particularly if carried by groundwater) during exhumation processes.
Biogenic organic molecules may thus experience multiple stages of degradation induced by a combination of biological, chemical, and physical factors during the geological history of their host rocks. Yet, it has been demonstrated that products released by kerogen degradation (i.e. thermally matured bitumen, molecular fossils, and organic biomarkers) may sometimes preserve structural information diagnostic of biological and environmental provenance (e.g. Vandenbroucke and Largeau 2007). This can also be the case for graphitic carbons, as detailed in the following section.
PRESERVATION OF BIOSIGNATURES BY NATURAL GRAPHITIC CARBONS
Graphitic fossils have been reported from rocks of various ages and metamorphic facies, including the greenschist facies (e.g. Butterfield et al. 2007), the amphibolite facies (e.g. Schiffbauer et al. 2007), and the blueschist facies (e.g. Galvez et al. 2012).
Some of the most illustrative examples include the “soft-bodied” graphitic fossils found in the Middle Cambrian Burgess Shale in the Rocky Mountains of southern British Columbia, Canada (Fig. 2). This formation has experienced greenschist facies metamorphism at burial depths of around 10 kilometers (Powell 2003). Yet, the Burgess fossil assemblage constitutes one of the most diverse records of Cambrian animal life. Graphitic fossils that are preserved as films or compressions are interpreted as the remains of recalcitrant organic exoskeletons, whereas more labile cellular tissues have been preserved in three dimensions via early diagenetic mineralization (Butterfield et al. 2007). Although the mechanisms that allowed their exceptional preservation remain uncertain, recent investigations have documented the preservation of chitin-protein molecular signatures in some of these fossils (Cody et al. 2011; Ehrlich et al. 2013).
Also remarkable is the morphological preservation of graphitic plant macrofossils in Mesozoic blueschist metamorphic rocks from the Marybank Formation in New Zealand (Galvez et al. 2012). Although these rocks have been buried to a minimum depth of 15 km, they contain fossil leaves and stems that can be identified unambiguously. Graphitic polygonal structures coated with micas may represent original cellular structures. Yet, despite the excellent morphological preservation, the organic biomarkers and molecular functional groups of these graphitic carbons have been completely lost during metamorphism.
Other spectacular, morphologically preserved graphitic fossils are lycophyte megaspores from Triassic metasedimentary carbonate concretions from Vanoise in the western Alps, France (Fig. 3; Bernard et al. 2007). These partially mineralized microfossils experienced blueschist facies metamorphism (~360 °C, ~14 kbars, i.e. a burial depth of about 40 km) during the late Mesozoic and Cenozoic Alpine orogeny. While mainly composed of graphitic carbons, these microfossils exhibit chemical and structural heterogeneities, which have been interpreted as remnants of original biochemical signals. In particular, the chemical signature of degraded sporopollenin, the resistant biopolymer composing the walls of modern spores and pollen grains, has been identified within the fossilized megaspore walls using synchrotron-based scanning transmission X-ray microscopy (STXM) and X-ray absorption near edge structure (XANES) spectroscopy at the carbon absorption edge (Bernard et al. 2007).
As original biosignatures carried by organic molecules are degraded at least partially, if not completely, during bioalteration and burial processes, unambiguously identifying fossils of microorganisms in ancient rocks remains a challenging goal. As detailed in the following section, the next step in interpreting the ancient organic fossil record lies in quantitatively constraining the impact of bioalteration and burial processes on organic molecule degradation.
Although laboratory simulations do not perfectly simulate natural diagenesis, experimental investigations are a promising route towards quantitative analyses of the transformations of organic molecules induced by burial. Most recent attempts at experimental fossilization have been aimed at reproducing and better constraining early-diagenetic processes that contribute to the morphological preservation of organic structures (e.g. Li et al. 2013). Numerous studies now also explore the molecular evolution of organic molecules during artificial maturation (e.g. Li et al. 2013).
The use of isotope compositions to evaluate the biogenicity of graphitic carbons appears necessary but insufficient. Abiotic pathways, such as the decarbonation of siderite and Fischer-Tropsch Type (FTT) synthesis, produce organic molecules with isotopic signatures similar to those of biological carbons, that is, depleted in 13C (McCollom and Seewald 2006). Recent hydrous pyrolysis experiments on kerogen in isotopically labeled aqueous solutions have shown that temperature increase promotes isotopic exchange between macromolecular organic matter and reactive inorganic compounds (Schimmelman and Lis 2010). Kinetic extrapolation from long experiments performed at 100 °C suggests that isotopic exchange may occur in natural settings even at temperatures as low as 40–50 °C. If not taken into account, such modifications of isotopic signatures may lead to erroneous interpretations regarding the search for evidence of early life in the geological record.
Structural and molecular biosignatures appear a bit more robust. For instance, using heating experiments to simulate the thermal alteration of organic matter, Schopf et al. (2005) showed that Raman spectroscopy may provide sensitive indicators of the low-temperature alteration of graphitic carbons. More recently, Schiffbauer et al. (2012) exposed natural fossilized acritarchs from the Mesoproterozoic Ruyang Group to experimental heating at approximately 500 °C for up to 250 days, under both oxic and anoxic conditions. Notably, anoxic replicates retain biological morphologies despite an increasing degree of carbonization with continuous heating, whereas oxic replicates experienced more aggressive degradation. Following a similar approach, Li et al. (2013) investigated the structural and chemical transformations of bacterial cells during thermal-aging experiments performed under anoxic conditions. Combined microscopy and spectroscopy data demonstrate partial morphological and chemical preservation of bacterial cells completely encrusted by iron phosphates despite heat treatments at hundreds of degrees for several hours under an argon atmosphere. This study illustrates that inorganic phases may favor morphological and chemical preservation during carbonization processes.
Altogether, the trends in structural, elemental, molecular, and isotopic changes reported in experimental fossilization studies provide new milestones towards a generalized mechanistic model of organic molecule degradation processes during diagenesis and metamorphism. In addition to simulating biodegradation and taphonomic processes in the laboratory, future studies should investigate the impact of key parameters, such as the geochemical nature of the fluid and the mineral matrix, on the extent of biomolecule degradation during burial in natural settings. Artificial maturation experiments performed in open or closed systems under well-constrained physical and chemical conditions have been used for decades to better understand kerogen degradation processes (e.g. Vandenbroucke and Largeau 2007). In addition, the recent development of advanced characterization techniques allows the evolution of chemical, structural, and isotopic signatures of graphitic carbons within natural maturation series to be precisely documented (Bernard and Horsfield 2014).
CONTROVERSIES OVER THE EARLIEST TRACES OF LIFE
Although many claims for Eoarchean and Paleoarchean life have been made on the basis of graphitic carbons, many controversies have arisen regarding their origin (e.g. Mojzsis et al. 1996; Rosing 1999; Brasier et al. 2002; Schopf et al. 2002; Van Zuilen et al. 2003). Because of the inevitable alteration of organic molecules during burial, biogenic graphitic carbons can be difficult to differentiate from carbons coming from contamination. Contamination in geological specimens may have several origins: (1) an experimental origin, where samples are contaminated during sampling, handling, and/or preparation; and (2) a natural origin, where samples are contaminated by natural abiotic, prebiotic, or younger biogenic carbons during fluid migration or biodegradation. Because of these poorly understood difficulties, the distinction of biogenic from nonbiogenic graphitic carbons in Archean rocks remains an issue, as illustrated by a number of controversial studies.
For instance, 13C-depleted diamonds and graphitic carbons included in Hadean zircons from the Jack Hills in Western Australia have been interpreted as possible evidence of the earliest carbon cycle and/or early life (Menneken et al. 2007). Yet, nanoscale TEM observations performed on foils microfabricated by focused ion beam (FIB) techniques have recently revealed that the observed graphite–diamond mixtures included in the Hadean zircons are embedded in epoxy, leading to the conclusion that they are contaminants derived from polishing compounds during sample preparation (Dobrzhinetskaya et al. 2014).
Among other controversial findings is the isotopically light graphite (with δ13C values between −21 and −49‰) from ca 3.83 Ga granulite facies quartz–pyroxene rock on the island of Akilia, southwest Greenland. This isotopic signature has led researchers to interpret this graphite as the oldest remains of life (Mojzsis et al. 1996). Yet, it has been demonstrated in the laboratory that abiotic hydrothermal pathways, such as siderite decarbonation and FTT synthesis, may lead to the formation of carbonaceous compounds similarly depleted in 13C (McCollom and Seewald 2006). Further, the graphite from this rock has been shown by two independent analytical methods to not be as fractionated as previously reported, with δ13C values between −4.1 and −23.6‰ (Papineau et al. 2010b). Mojzsis et al. (1996) argued that, because graphite is completely included within apatite crystals, isotope exchanges with fluids during long crustal residence times did not occur. While recent nanoscale investigations of graphite associated with apatite in the Akilia quartz–pyroxene rock indicate the presence of crystalline domains consistent with the granulite facies metamorphic grade of this rock, detailed petrography has revealed clear evidence for fluid-deposited graphite in the rock, which blurs interpretations (Fig. 4a). So far, these investigations have reported graphite as coatings on apatite grains in these rocks, but not as inclusions (Fig. 4b; Papineau et al. 2010a). Similar graphite coatings on apatite grains within banded iron formations from the ca 3.75 Ga old Nuvvuagittuq Supracrustal Belt, northern Québec, are poorly crystalline and inconsistent with amphibolite facies metamorphism, and have thus been interpreted as deposited by fluids late in the history of the rock (Papineau et al. 2011). NanoSIMS analyses of H, O, N, S, and P heteroatoms in the Akilia graphite indicate their presence, but nothing is known regarding how these heteroatoms are bonded to carbon. While all these data are not inconsistent with a biogenic origin for the carbon in the Akilia quartz–pyroxene rock, these examples illustrate the importance of unambiguously demonstrating indigenicity before discussing the age and biogenicity of graphitic carbons.
Graphite particles in the ca 3.77 Ga turbiditic schists from the Isua Supracrustal Belt (western Greenland) have also been interpreted as biogenic remains based on their abundance and isotopic signatures (Rosing 1999). Yet, additional measurements showing less 13C-depleted graphite particles in Isua metacarbonate rocks have led to the interpretation that these particles are the result of the abiotic thermal decarbonation of siderite (Van Zuilen et al. 2003). Since then, turbiditic schists from Isua have been reported to contain graphitic layers with a crystallinity consistent with amphibolite facies metamorphism (Ohtomo et al. 2014). However, these recent interpretations are debated, notably because they rely on samples from a single formation.
Perhaps the most famous report of graphitic microfossil evidence for earliest life is the discovery of microbe-like structures composed of graphitic carbon from the ca 3.46 Ga Apex Chert in the Warrawoona Group, western Australia (Schopf et al. 2002). While these microscopic objects were originally interpreted as cyanobacterial microfossils and subsequently became the benchmark for bona fide Paleoarchean microfossils, their biogenic origin has come into question after reexamination of their morphology and their geological and geochemical contexts (Brasier et al. 2002). The origin of this graphitic carbon is still debated. Although the presence of heteroatoms reported by De Gregorio et al. (2009) suggested a biological origin, FTT synthesis (Brasier et al. 2011) and contamination by younger organic matter (Olcott Marshall et al. 2012) have also been proposed.
A target of several investigations is the 3.35 Ga Strelley Pool Formation, Pilbara Craton, Western Australia, as it hosts different types of graphitic carbons (Fig. 4c) as well as stromatolites (Fig. 4d). Wacey et al. (2012) recently concluded on the biogenicity of microfossil-like structures containing spheroidal nanograins of silica embedded in graphitic carbon from this formation. Their conclusions rely mostly on the morphological similarities that these objects share with Gunflint microfossils as observed using FIB-SEM and TEM.
REMAINING QUESTIONS AND CONCLUDING REMARKS
The controversies detailed above highlight the need for gathering the multiple lines of evidence necessary for unambiguously identifying ancient graphitic carbons as biological remains. To prevent misinterpretations, we propose a number of criteria, listed in Table 1, to be considered when investigating the possible biological origin of graphitic carbons. Still, several fundamental questions still need to be addressed.
For instance, almost nothing is known regarding the possibility that abiotic carbons accumulate as aromatic compounds and polymerize within sediments or sedimentary rocks. FTT synthesis generally proposed as the source of controversial graphitic carbons dominantly produces short aliphatic hydrocarbons (McCollom and Seewald 2006). The fate of such abiotic organic molecules in sedimentary rocks over geological timescales remains unknown. Under what conditions can these hydrocarbon chains become aromatic macromolecules, which can be difficult to distinguish from graphitic carbons resulting from the degradation of organic molecules of biogenic origin? What geological mechanisms and conditions would have to work in concert to yield accumulations of abiotic, poorly crystalline graphitized carbon from such aliphatic precursors? What type of FTT synthesis or diagenetic pathways could produce graphitic microbe-like features?
Galvez et al. (2013) reported graphite formation from abiotic calcite reduction at low temperature in an exhumed serpentinite–sediment contact in Alpine Corsica (France). Based on isotopic and spectroscopic measurements, these authors argued that this graphite formed from an abiotic source of carbon and could be distinguished from the biogenic graphitic carbons observed in the same rocks. Yet, such reports of robust geochemical evidence for geological deposits of abiotic graphitic carbons remain scarce, in contrast to what should be expected based on the null hypothesis. In the present context, the null hypothesis stipulates that all possible abiotic pathways need to be disproven before a biogenic origin can be accepted. This cautionary approach may be valuable for the oldest sedimentary rocks and for rocks in an exobiological context. However the null hypothesis could likewise be used in the opposite way from a uniformitarian perspective, depending on the context and as far as terrestrial organic matter is concerned. Ultimately, such debates are healthy for this field of science as they stimulate additional research, and they can only be solved by using new analytical approaches and sample suites.
All these questions can be summarized as follows: “How can ancient graphitic carbon be unambiguously assigned an origin?” This apparently simple question is full of hidden fundamental and methodological issues. A great deal of caution is required to prepare samples for microanalyses and to gather the necessary multiple lines of evidence (Table 1). In addition, we suggest that the identification of biogenic graphitic carbons in ancient rocks should not be based on results obtained from a single locality. Most importantly, the determination of syngenicity and indigenicity of putative microfossils and/or graphitic carbons within metasedimentary rocks requires an accurate assessment of their thermal history, as well as quantitative constraints on the biosignatures that may be preserved by these graphitic carbons in light of their thermal history. Experimental fossilization should allow substantial progress in that regard in the future.
As described here, the assessment of the origin of graphitic carbons in ancient rocks remains challenging and debatable because the behaviors of biogenic and abiotic carbons during diagenesis and metamorphism have not yet been accurately constrained. Likewise, while there remain uncertainties about the possible biogenic origin of graphitic carbons in Eoarchean and Paleoarchean metasedimentary rocks, significant advances have recently been made with the introduction of new microanalytical techniques. Such tools have helped to expand databases and generate reproducible observations of samples from ancient sedimentary rocks. Such progress opens up new subfields in Precambrian micropaleontology and nanobiogeochemistry and holds great promise to contribute to the current challenges of addressing the origin of graphitic carbons.
SB acknowledges support from the ERC (project Paleo-NanoLife, PI: F. Robert) and DP acknowledges support from the University College London and the NASA Astrobiology Institute (grant numbers NNA04CC09A, NNA09DA81A, and NNX12AG14G). STXM-based C-XANES data shown in Figure 3 were acquired at beamlines 220.127.116.11 and 11.0.2 at the Advanced Light Source (ALS), Berkeley, CA, USA, which is supported by the Director of the Office of Science, Department of Energy, under Contract No. DE-AC02-05CH11231. Special thanks go to David Kilcoyne and Tolek Tyliszczak for their expert support on STXM at the ALS.