The impact of the upper crust composition on the production of natural hydrogen during serpentinization in continental settings

Natural hydrogen, or white/gold hydrogen, can potentially be a key factor in the energy transition to mitigate climate change and many exploration efforts are underway in various part of the globe. Yet, in continental settings where its exploitation would be easiest, the hydrogen cycle is poorly constrained. Most of the hydrogen produced naturally is through the oxidation of Fe in silicates due to the hydration of iron-rich rocks at depth, either ultramafic rocks or banded iron formation (BIF). Nevertheless, the fluid responsible for the alteration of such rocks is usually assumed to be pure water, which is an oversimplification of fluid-rock interaction processes. Here I present the results of reaction path thermodynamic modeling of successive interactions of an initial meteoritic fluid with different rocks that would compose the upper crust in a continental setting, i.e., granites, quartzites, carbonates and evaporites, which is then reacted with ultramafic rocks.

Thermodynamic modeling was computed at 300 °C and 5 kbar using the DEW model and the EQ3/6 package. Meteoritic fluid (initial concentrations set at 10^-6 molal (mol/kg) of water for all dissolved elements, except Cl at 0.1 molal and 5 molal for evaporites) after the first interaction had higher dissolved Na, K, Fe, Al and Si after interacting with granites than the other three lithologies, although Si was also high after interaction with quartzites. Dissolved C, Ca and Mg were higher after interaction with carbonates and evaporites. These fluids were then reacted with 861 compositions of ultramafics spanning the entire range of Ol-Opx-Cpx compositions. Maximum serpentinization degree was always achieved for peridotite composition intermediate in Ol-Opx, reaching about 45-50 vol% for fluids that interacted with granites or quartzites, while only reaching 20-25 vol% for fluids having interacted with carbonates and almost no serpentinization for fluids having interacted with evaporites. Magnetite and H₂ content were coupled for all settings but were decoupled from serpentinization degree, with highest contents of H₂ produced for low dunitic initial peridotite compositions. The maximum amount of H₂ produced was 0.1 molal for fluids interacted with granites, quartzites and carbonates, an order of magnitude higher than for fluids that interacted with evaporites.

The above results highlight the vital importance of taking into account the actual chemistry of fluids that are responsible for the serpentinization of ultramafic bodies in continental settings. Especially, the presence of evaporites in sedimentary sequences, for example in the widely studied Pyrenees could hamper the natural production of hydrogen. Thus, further exploration of areas of economic importance for natural hydrogen production should carefully map the lithologies in contact with doing fault that are believed to be the carrier of surface fluid toward peridotite bodies at depth.