Mapping Subsurface Natural Hydrogen Accumulation Windows: A Hybrid Quantum-Chemical and Kinetic Modelling Framework
PosterProspectivity to Decision-ready Assessment
2026-11-03 | 08:30 AM - 06:00 PM
Abstract
Natural hydrogen accumulation in subsurface ophiolitic settings is emerging as a low-carbon energy source which is governed by coupled mineral redox processes. However, the exploration of such resources requires understanding and accurate prediction of depth-temperature window in which hydrogen-producing serpentinization overtakes hydrogen consuming hematite reduction. The understanding of mechanistic origin of temperature-dependent hydrogen release remains limited. We present a modelling and benchmarking framework which links the experimental observation through active-space quantum chemistry to a depth target in Andaman Ophiolitic setting.
Temperature-resolved XRD reveals progressive serpentinization, magnetite formation peaking at 80°C, and hematite content falling below detection limit at 120°C. Five Fe coordination environments identified by XRD (olivine (Fo80), clinopyroxene, lizardite, vermiculite, and hematite) are modelled as finite molecular clusters and treated with hierarchical quantum chemistry. The approach included density functional theory (DFT/PBE0), active-space full configuration interaction (CASCI/FCI), CASSCF+NEVPT2 with Newton-Raphson and automated CASCI fallback and simulated near-term variational quantum algorithms (UCCSD-VQE, ADAPT-VQE) on 8-qubit circuits. Bronsted-Evans-Polanyi (BEP) scaling of the active-space-corrected molecular oxidation energy onto periodic DFT+U study barriers yield adjusted activation energies of 0.62 (R1, olivine serpentinisation) and 0.28 eV (R5, hematite reduction). The kinetic-competition model predicts a crossover at 115+/-12°C when combined with the XRD-derived hematite-availability function. This prediction is consistent within 5°C of the experimental hematite-depletion observation. Considering the geothermal gradient of Andaman as 40°C/km, this corresponds to a 2.0-2.3 km model-prioritized interval for future hydrogen exploration activities. The hydrostatic-pressure correction may shift the modelled depth by 70m, and a depth-stratified scenario analysis gives the range of 2.06-2.13 km.
The quantum-algorithm benchmarks indicate that UCCSD-VQE and ADAPT-VQE reach chemical accuracy (<43 meV vs. FCI) for two Fe environments dominated by a single electronic determinant. They also exhibit failure for the three multi-reference cases by 62-1537 meV. Within the tested minimal active spaces and operator pools, this maps the practical accuracy frontier of UCCSD-class near-term quantum algorithms on first-row transition-metal redox. This limitation suggests that the next-generation hardware-efficient Ansatze needs further development for implementation of quantum-accelerated chemistry for regional-scale subsurface modelling. This study is based on single outcrop sample, cluster-model approximations and BEP scaling propagation to an overall +/-10-12°C per 300 m modelling uncertainty.
Authors: Raj Kiran¹, Rakesh Behera¹, Parth Shroff¹
¹IIT (ISM) Dhanbad, Dhanbad, Jharkhand, India-826004
