In 1999 the late Thomas Gold, cosmologist and quite a lot more, annoyed the geoscience community with publication of his book The Deep Hot Biosphere: The Myth of Fossil Fuels (Springer-Verlag: New York). In that book Gold reached the acme of his lone campaign for recognition that oil, gas and even coal formed from carbon and hydrogen feedstock that had been residing in the mantle since the Earth’s accretion. He suggested that it was mediated by a hidden yet teeming biosphere at much deeper levels than suspected at the time. I did my share of carpet gnawing, but was sorry to learn of the death in 2004 of such a supreme scientific provocateur. Although without mentioning Gold, a recent paper hints at a possibility that he may have been on to something (Proskurowski, G. et al. 2008. Abiogenic hydrocarbon production at Lost City hydrothermal field. Science, v. 319, p. 604-607).

Hydrocarbons are often found as blobs in fluid inclusions within gangue minerals of a variety of ore bodies. The US-Swiss research team examined hydrocarbons within the fluids that gush from a hydrothermal vent at 30˚N on the Mid-Atlantic Ridge; i.e. where there is no older sediment that might host biologically generated hydrocarbons, but where heat-loving microbial life could play a role. Molecular structure and carbon-isotope composition of the hydrocarbons point strongly to their formation by reduction of CO₂ to methane and low molecular weight hydrocarbons by the catalytic action of mineral surfaces in the presence of a great deal of hydrogen. This is known as a Fischer-Tropsch reaction, the basis for making oil from coal, as in Nazi Germany and South Africa when under economic blockade.

The CO₂ could have come from two possible sources: seawater or the mantle beneath the Lost City vents. Hydrogen can form abundantly when the olivine in peridotite beaks down to serpentinite as seawater is convected through the oceanic mantle. The vents have created towers made partly of carbonates, in whose pores there are microbes whose metabolism is based on use of hydrogen. However, the key finding is that the hydrocarbons contain no radioactive ¹⁴C, which forms by cosmic-ray interaction with nitrogen atoms in the atmosphere and is easily detectable in seawater. This rules out a seawater source for the CO₂, but supports a mantle origin.