Hydrogen Isotope Biogeochemistry

Octopus Spring in Yellowstone National Park. The orange and yellow colors are bacterial mats, one of the environments where we have been studying the D/H ratios of lipids.

One of the major focuses for our research is an effort to use (and understand) the distribution of stable hydrogen isotopes (2H, known as 'deuterium', and 1H, known as 'protium') in individual organic compounds. While measurements of hydrogen isotopes go all the way back to the initial work of Harold Urey and Harmon Craig, an important distinction is that those time-tested methods measured the abundance of isotopes in bulk materials, such as water, minerals, and total organic matter. We have been conducting studies which look at hydrogen isotopes in individual compounds; the greater specificity of the measurements gives us a more detailed understanding of relevant processes, but of course is more difficult. The measurements are achieved by connecting a standard gas chromatograph (GC) to a standard isotope-ratio mass spectrometer (IRMS) via a high-temperature pyrolysis (P) oven. This hyphenation yields the acronym GC/P/IRMS. To learn more about this technique, have a look at these papers: 1, 23 (papers 1 and 2 are research articles; 3 is a more recent review). Because all of our analytes must be readily volatilized (we are using a gas chromatograph), it is mainly applicable to lipids, hydrocarbons, and other nonpolar moleculars.

At the moment we are working on three general research questions that utilize these techniques:

1) What is the distribution of hydrogen isotopes in natural organic materials from a wide range of environments? One of the first things to do with any new analytical technique is to go explore the world with your eyes wide open. How much variability exists? Are there systematic patterns? What areas, compounds, or sample types show the most promising signals? To answer these questions, we have been surveying lipids extracted from various natural environments. Completed work includes marine algae and grasses from Woods Hole, MA (4, 5), crude oils from Australia (6), suspended marine particulate matter from near Santa Barbara, CA (7), and microbial mats in hot springs from Yellowstone National Park (8). It is hard to summarize the results of these diverse studies succinctly, but one of our most important results has been to show that in many places there is a lot of variability, and the greatest D/H variability seems to occur in environments where microbial processes (as opposed to plants, or algae) are important. For example, in the anoxic sediments at the bottom of Santa Barbara Basin, we find lipid D/H ratios that vary by as much as 30% (300 permil in the conventional dD notation). In a world where differences of 0.5% are significant, this is a very big signal.

2) What are the biochemical controls on hydrogen isotopes in organic matter? While its nice to document large differences in natural environments, this becomes much more useful when we understand why those differences exist. The second part of our research in this field has thus focused on working with pure cultures of bacteria in the laboratory. These organisms grow rapidly, can be readily manipulated, and can grow under a variety of conditions and on different organic substrates. In essence, we grow an organism under defined conditions, measure the H inputs to the system (water and organic H), then look at the D/H ratios of produced lipids. Then change the conditions, see how the lipids respond, and repeat ad nauseum. Early results from this approach revealed that organisms able to grow on H2 produce lipids that are very strongly depleted in D (9,10), while those growing on CH4 produce lipids with moderate D enrichment (11). More recently, we have been working with several facultative heterotrophs (which means they are able to grow on a wide range of organic substrates). We have found that D/H fractionation between lipids and growth water varies dramatically (by up to 50%, or 500 permil, in a single organism) depending on what they are fed. We believe that this effect represents differences between central biochemical pathways that are used to metabolize the different substrates. Combined with the results of topic 1, described above, this research has lead us to the hypothesis that D/H ratios of lipids in environmental samples can be used to understand the metabolic capabilities of the organisms that produced them.

3) How well are these biotic signals preserved over geologic timescales? The patterns described above will certainly be useful to understanding modern environments, but how far back in time might these records persist? The molecules we study - lipids and their hydrocarbon derivatives - are extremely stable, and under appropriate conditions can be preserved for millions (and sometimes billions) of years. However, the fact that the molecules are stable does not necessarily guarantee that the isotopic information they contain will be preserved. This is because H bound to carbon atoms in organic molecules can undergo exchange: in essence, one H atom pops off and another (different) H atom quickly takes its place in the molecule. Thus the molecular structure is preserved, but the D/H ratio can change. To understand this process, we first need to know what the endpoint of that exchange process is. In other words, what is the D/H ratio of a fully exchanged molecule? It might seem like a simple question, but turns out to be very hard to answer because the molecules are so stable. We could do an experiment in the lab, but we would have to wait around for millions of years for it to finish! We have thus been working to combine selected lab experiments (on molecules that exchange quite rapidly, such as the alpha-carbonyl position in ketones) with theoretical calculations (using the tools of molecular modeling) to predict these equilibrium effects. When we compare our results (in review... be patient) to D/H ratios measured in old rocks, we can show that most rocks that have been heated enough to turn organic matter into petroleum (roughly 75-100C) have completely lost their original D/H fingerprints, and become completely equilibrated. However, in some exceptional cases where rocks have not been significantly heated (e.g., the ~50 million-year-old Green River Formation in Wyoming), we believe that original biotic and environmental information can be preserved.

Recent papers on this subject:

Zhang, Xinning, Gillespie, Aimee, Sessions, Alex L. (2009). Large D/H variations in bacterial lipids reflect central metabolic pathways. Proceedings of the National Academy of Science, USA (Aug 4 issue).

Li, Chao, Sessions, Alex L., Kinnamen, Frank, Valentine, David L. (2009). Hydrogen-isotopic variability in lipids from Santa Barbara Basin sediments. Geochimica et Cosmochimica Acta 73, 4803-4823.

Campbell, Brian J., Li, Chao, Sessions, Alex L., and Valentine, David L. (2009) Hydrogen isotopic fractionation in lipid biosynthesis by H2-consuming Desulfobacterium autotrophicum. Geochimica et Cosmochimica Acta 73, 2744-2757.

    Jones, Ashley, Sessions, Alex L., Campbell, Brian J., Li, Chao, Valentine, David L. (2008). D/H ratios of fatty acids from marine particulate organic matter in the California Borderland BasinsOrganic Geochemistry 39, 485-500.

   Schimmelmann A., Sessions A.L., and Mastalerz M. (2006). Hydrogen isotopic (D/H) composition of organic matter during diagenesis and thermal maturationAnnual Reviews of Earth and Planetary Sciences 34, 501-533.

   Sessions A.L. (2006) Seasonal changes in lipid D/H fractionation by Spartina alternifloraGeochimica et Cosmochimica Acta70, 2153-2162.

    Sessions A. L. and Hayes J.M. (2005) Calculation of hydrogen isotopic fractionations in biogeochemical systemsGeochimica et Cosmochimica Acta 69, 593-597.

    Valentine D.L., Sessions A.L., Tyler S.C., and Chidthaisong, A. (2004) Hydrogen isotope fractionation during H2/CO2 acetogenesis: hydrogenase efficiency and the origin of lipid-bound hydrogenGeobiology 2, 179-188.

    Schimmelmann A., Sessions A.L., Boreham C.J., Edwards D.S., Logan G.A., and Summons R.E. (2004) D/H ratios in terrestrially sourced petroleum systemsOrganic Geochemistry 35, 1160-1195.

    Sessions A.L., Sylva S.P., Summons R.E., and Hayes J.M. (2004) Isotopic exchange of carbon-bound hydrogen over geologic timescalesGeochimica et Cosmochimica Acta 68,1545-1559.

    Sessions A. L., Jahnke L. L., Schimmelmann A., and Hayes J. M. (2002) Hydrogen isotope fractionation in lipids of the methane-oxidizing bacterium Methylococcus capsulatusGeochimica et Cosmochimica Acta 66, 3955-3969.

Copyright 2011 by Alex Sessions. All rights reserved.