Searching beyond the lamppost: atomistic simulation of mineral-water interfaces

by Dr Ardalan Hayatifar (University of Saskatchewan)

Wednesday 4 Nov 2026, 15:00 → 16:00 Europe/Zurich

OVGA/200

The fate of metals and metalloids in natural waters is controlled by their interactions at mineral-water interfaces. These interfaces are difficult to probe experimentally, because the interfacial signal is obscured by the bulk. Surface complexation models reproduce macroscopic adsorption data, but they rely on postulated surface species and fitted, often non-unique constants that must be validated independently. Atomistic modeling faces its own limits: density functional theory (DFT) is restricted in length and time scale, particularly for the geochemically important Fe-bearing phases, while classical force fields lack transferability and cannot describe reactivity.

In this talk, I describe my journey from synthetic chemistry to molecular simulation. I first summarize my doctoral work on germanium chemistry, including germylium-mediated C-F bond activation and amide bond formation. I then discuss how combining ab initio molecular dynamics (AIMD) with synchrotron X-ray scattering resolves the uptake of vanadium by schwertmannite, a nanocrystalline mineral common in acid mine drainage environments. Next, I examine the interaction of V with ferrihydrite, another abundant Fe mineral, using AIMD snapshots and embedded-cluster time-dependent DFT to calculate V K-edge pre-edge features for comparison with experimental X-ray absorption spectra. I further present AIMD simulations of Li₂SO₄·H₂O under pressure that provide insight into the prograde "sulfate surge" observed in deep hydrothermal fluids. Finally, I discuss the MACE-MP-0 foundation model as a starting point for machine-learning frameworks to simulate these geochemical systems.