Research
My work sits where electronic-structure theory, lattice dynamics, and beyond-DFT corrections meet. The unifying question is methodological: I look for systems where standard functionals break — flat bands, strong correlation — and build the machinery (hybrids, SOC, MLIPs, spin-wave theory) needed to recover the right physics. Below are the active threads.
2D materials & heterostructures
First-principles defects and phonons in transition-metal dichalcogenides and their heterostructures, with a focus on 1T-SnSe₂ / 2H-WSe₂. Recent work: substitutional W defects in 1T-SnSe₂ (CRYSTAL23, hybrid functionals), diagnosing and curing a spurious imaginary out-of-plane phonon mode that was a basis-set artifact rather than a true dynamical instability, and full phonopy force-constant workflows on 100+ atom supercells.
Methods: CRYSTAL23 (hybrids, SOC), phonopy/VASP/Quantum ESPRESSO, broken-symmetry DFT.
Chiral materials & spintronics
Chirality and spin transport in low-dimensional tellurium: spin-resolved band structures, Brillouin-zone spin textures, the spin Hall tensor via pseudo-atomic orbitals (PAOFLOW), and chirality analysis of finite Te–H oligomers (L/R enantiomers, Kabsch-RMSD validation) in the context of the chiral-induced spin selectivity (CISS) effect.
Methods: VASP, Quantum ESPRESSO, PAOFLOW, Wannier90, VMD-based analysis pipelines.
Magnetic materials & spin dynamics
Magnetic ground states and excitations in 2D and bulk magnets (CrI₃, CrSBr, Mn₅Si₃). The central project: extracting exchange tensors from CRYSTAL23 hybrid + SOC broken-symmetry calculations and feeding them into a linear spin-wave theory engine to find magnetic ground states where semilocal DFT and DFT+U disagree — plus skyrmion-scale physics (DMI, micromagnetics) via the QE → Wannier90 → TB2J → UppASD chain.
Methods: Quantum ESPRESSO, TB2J, UppASD, LSWT.
Methods: Machine-learned interatomic potentials
Cross-cutting methods work on extending Machine-learned interatomic potentials to the dynamical matrix (phonons), and machine-learned interatomic potentials for accelerated screening.
Methods: Machine-learned interatomic potentials, high-throughput screening.
For the formal write-ups behind some of this — spin-wave theory, the acoustic sum rule, density-driven error — see the THE GRID ↗. A full list of publications is on the publications page.