Vacancy–Solute Coupling in Ni–Fe–Cr–Al Alloys Active

PhD, Queen's University, 2024–present

DFT and molecular dynamics show that Cr additions systematically raise the equilibrium vacancy concentration in Ni–Cr–Al and promote Al enrichment at grain boundaries — mechanisms with direct implications for oxidation resistance in high-temperature superalloys. A first-author manuscript has been submitted to Scripta Materialia (preprint, SSRN 5759303). This project anchors the broader programme: the vacancy–solute coupling framework developed here directly motivates parallel studies in Zr cladding alloys and informs how vacancy-mediated surface composition controls catalytic selectivity on alloyed Cu. Current work extends these calculations to Fe-containing quaternary compositions and larger supercells for finite-temperature validation with MACE-driven k-ART trajectories — closing the loop between formation energetics from static DFT and the segregation profiles seen in atom-probe tomography of model superalloys.

Defect Kinetics via k-ART Active

PhD, Queen's University, 2025–present

k-ART couples the Activation-Relaxation Technique nouveau (ARTn) with kinetic Monte Carlo to access defect clustering, segregation, and microstructural evolution at seconds-to-hours timescales — providing the temporal bridge between DFT energetics and experimentally observable outcomes. k-ART is now parameterised with MACE and ACE potentials trained on DFT data, expanding the range of alloy compositions and defect configurations accessible while maintaining near-DFT fidelity for saddle-point energetics. Production trajectories in Ni–Cr–Al are revealing vacancy-mediated clustering pathways inaccessible to conventional MD, with barrier-statistics benchmarking against NEB reference calculations validating the ML-potential accuracy for transition-state searches. The same workflow is being adapted for hcp Zr, where the lower symmetry and anisotropic migration barriers stress-test the saddle-search infrastructure.

CO₂ Catalysis on Metal Surfaces Active

PhD, Queen's University, 2025–present

DFT screening of electrochemical CO₂ reduction on transition-metal and alloy surfaces, leveraging the surface-energetics expertise from Ni-alloy grain-boundary work. Current screening covers flat, stepped, and defected facets of alloyed Cu, mapping how vacancy and adatom sites modulate CO₂ binding energies, reaction pathways, and product selectivity toward C₂+ products. The same vacancy-mediated segregation framework used in the Ni–Cr–Al project predicts which alloy surface compositions are thermodynamically accessible under reaction conditions — connecting structural-alloy research to sustainability-driven catalysis. Foundation-model screening (MACE-MP-0) accelerates exploration of the combinatorial composition space before targeted DFT validation.

Radiation Damage in Zirconium Alloys Active

PhD, Queen's University, 2025–present

Point-defect formation and clustering in Zr-based fuel-cladding alloys, combining MD with empirical and ML potentials alongside DFT validation. Focus: how Nb, Sn, and Fe additions affect defect stability and irradiation-induced microstructural evolution in hcp Zr. Hybrid ML/EAM benchmarking against DFT is establishing the accuracy floor needed for reliable k-ART simulations in these systems — a prerequisite for extending the long-timescale kinetic modelling capabilities proven in fcc Ni alloys to the hcp crystal structure of reactor-grade Zr.

Ion-Pair Trapping in Functionalized Organic Systems Completed

MSc, Ontario Tech, 2021–2023

Applied quantum chemistry methods (DFT, Hartree-Fock, MP2) to model noncovalently bound molecular complexes and ion-pair trapping mechanisms in functionalized cyclic hydrocarbons. Explored potential applications in energy storage and light detection.

Synthetic Chemistry Completed

Ontario Tech, 2016–2018

Laboratory experience with organic synthesis, NMR spectroscopy, UV-Vis characterization, and GC-MS analysis supporting research in molecular systems and functionalized materials.

Foundation Models for Atomistic Simulation Emerging

Universal neural-network potentials (MACE-MP-0, CHGNet, M3GNet) are now viable for screening applications, reducing the DFT data generation burden for new alloy compositions. These pre-trained foundation models enter the DFT → ML potential pipeline at two points: upstream, for rapid hypothesis testing across chemical spaces before committing to full active-learning cycles; and downstream, as zero-shot baselines for k-ART saddle-point searches in new chemistries. This directly supports the Zr-alloy project, where hcp training sets remain expensive to build, and the CO₂ catalysis screening, where the combinatorial space of alloy surface compositions exceeds what targeted DFT campaigns can cover. Fine-tuning protocols that adapt foundation models to domain-specific configurations with minimal additional DFT data are an active area of development.

Autonomous Simulation Workflows Emerging

LLM-driven agents integrated with high-throughput DFT and MD workflows to automate convergence testing, error detection, and adaptive sampling. Current applications include automated CO₂ adsorption screening across facet–composition space and self-correcting SLURM pipelines for the Ni–Cr–Al k-ART production runs. Longer-term, these agents will close the loop between k-ART kinetic trajectories and active-learning potential refinement, enabling autonomous exploration of defect-evolution pathways in alloy systems too complex for manual campaign design. This direction reflects a broader shift toward AI-accelerated materials discovery, where simulation workflows become increasingly self-directed.