The discovery of quantum materials with nontrivial electronic band topology and robust surface or edge states has transformed modern condensed matter physics. Our group investigates a broad spectrum of topological phases—ranging from topological insulators to various classes of topological semimetals—with an emphasis on their fundamental electronic structure, symmetry protection, and potential device applications. Topological insulators (TIs) represent a unique state of matter where the bulk behaves as an electrical insulator, yet the surface hosts highly conductive electronic states. These states arise from strong spin–orbit coupling and are protected by time-reversal symmetry, enabling spin-momentum locking and dissipationless transport channels. Such characteristics make TIs promising for next-generation spintronics and low-power electronics. Beyond TIs, the landscape of topological semimetals (TSMs) offers even richer physics, as their bulk bands themselves carry nontrivial topology. Unlike conventional metals, TSMs exhibit symmetry-protected band crossings that give rise to a variety of exotic quasiparticles e.g. (1) Dirac Semimetals (DSMs): Characterized by four-fold degenerate Dirac points, hosting low-energy excitations analogous to relativistic Dirac fermions (2) Weyl Semimetals (WSMs): Feature two-fold degenerate Weyl nodes acting as monopoles of Berry curvature, leading to unusual phenomena such as the chiral anomaly and Fermi-arc surface states. (3) Nodal-Line Semimetals (NLSMs): Exhibit band crossings that form one-dimensional loops or extended lines within the Brillouin zone, often protected by mirror or inversion symmetry (4) Triple-Point Semimetals (TPSMs): Represent an intermediate topological phase where three-fold band degeneracies arise due to rotational or crystalline symmetries.
The dimensionality and topology of these band crossings govern the electronic, magnetic, and transport responses of the materials. Importantly, the crystalline symmetries that protect these degeneracies can be tuned—via strain, alloying, external fields, or pressure—to induce topological phase transitions, enabling control over material functionalities.
Our work involves first-principles calculations, symmetry analysis, and topological invariants to predict, classify, and engineer new topological quantum materials. These studies not only expand the catalogue of novel quasiparticles in solids but also provide pathways toward applications in spintronics, quantum computation, optoelectronics, and energy-efficient devices.
Thermoelectric (TE) materials enable the direct conversion of heat into electricity and vice versa, offering clean and reliable solutions for waste-heat recovery, solid-state refrigeration, and energy-efficient power generation. Their performance is quantified by the dimensionless figure of merit, ZT = S2 σ T/κ, which depends on the Seebeck coefficient (S), electrical conductivity (σ), and total thermal conductivity (κ). Achieving high ZT is challenging because these parameters are inherently coupled—enhancing one often deteriorates another. Thus, designing high-efficiency thermoelectrics requires simultaneously achieving metal-like electronic transport and glass-like phonon transport, a delicate balance rarely found in conventional materials. Our group employs density functional theory (DFT) combined with Boltzmann transport theory to explore and design advanced TE materials. We investigate the intricate interplay between lattice dynamics, electronic structure, carrier scattering, and phonon transport, enabling predictive understanding of TE behavior. We tune thermoelectric performance through a variety of strategies, including band engineering, strain modulation, chemical doping and alloying, nanostructuring, and quantum confinement—each providing a handle to decouple electronic and phononic transport pathways.
A particularly exciting direction in our research is the integration of topological quantum materials into thermoelectric science. Topological systems frequently host coexisting flat and highly dispersive bands, which enable large density of states alongside high carrier mobility—an ideal electronic structure for boosting the power factor without increasing lattice thermal conductivity. Furthermore, topological insulators, Dirac/Weyl semimetals, and nodal-line semimetals often exhibit Berry-curvature–driven transport phenomena, giving rise to unconventional thermoelectric responses such as the anomalous Nernst effect, spin Nernst effect, and planar Nernst effect. Such transverse thermoelectric effects open new avenues for harvesting waste heat in geometries beyond traditional longitudinal devices. In parallel, we also integrate machine learning and high-throughput materials screening to accelerate the discovery of next-generation TE materials. By combining physics-guided descriptors with data-driven models, we rapidly identify promising candidates from large materials databases and optimize them for performance under realistic operating conditions.
Together, these efforts aim to establish topology-driven and AI-accelerated thermoelectrics as a transformative platform for efficient energy conversion technologies of the future.
Photovoltaic and optoelectronic materials power technologies that convert light into electricity and enable light generation, detection, and communication. Solar cells, LEDs, lasers, and photodetectors rely on materials with finely tuned electronic and optical properties for high efficiency and long-term stability.
Our group uses density functional theory (DFT) to design and understand advanced materials for these applications, including lead-halide perovskites, lead-free chalcogenide, oxide, and nitride perovskites, as well as various chalcogenides and 2D materials. We investigate phase stability, band structure, excitonic effects, dielectric behavior, carrier mobility, and defect-related recombination pathways to optimize material performance. We engineer electronic and optical properties through band engineering, strain tuning, alloying, heterostructure design, and dimensional confinement. A growing focus of our work explores how ferroelectricity, multiferroicity, and magnetism can enhance charge separation and enable unconventional photovoltaic responses such as the Bulk/Anomalous Photovoltaic Effect (BPVE/APVE).
We collaborate closely with experimental groups to validate predictions and to develop interface-engineered, high-efficiency devices, including multi-junction and tandem solar cells.
Magnetic materials and spintronics leverage the spin of electrons—along with their charge—to enable ultra-fast, low-power, and non-volatile electronic technologies. Devices such as magnetic tunnel junctions, spin valves, spin-torque memories, and skyrmion-based racetrack devices highlight the transformative potential of spin-based information processing. Our group employs first-principles calculations and spin-transport theory to study diverse magnetic systems, including ferromagnets, ferrimagnets, antiferromagnets, half-metals, Heusler alloys, 2D magnets, and van der Waals heterostructures. We analyze their magnetic ground states, spin dynamics, and switching mechanisms to identify materials suitable for nanoscale and ultrafast devices.
A central part of our research focuses on spin–orbit coupling (SOC) and its associated phenomena—magnetocrystalline anisotropy, Dzyaloshinskii–Moriya interactions, spin Hall and Rashba effects, and topological spin textures—which underpin the stability and control of domain walls, skyrmions, and other functional spin configurations. We also advance antiferromagnetic spintronics, leveraging THz dynamics and zero stray fields for dense device integration.
Our overarching goal is to discover materials with high operating temperatures, strong spin polarization, low damping, and robust magnetic stability, paving the way for next-generation spintronic memory, sensing, spin logic, neuromorphic computing, and quantum information technologies.
Catalysis underpins many reactions essential for clean energy, environmental sustainability, and industrial chemistry. Our work focuses on the first-principles design and mechanistic understanding of catalytic materials for CO₂ reduction, hydrogen production, and electrochemical fuel conversion. Using density functional theory, we examine adsorption energetics, reaction pathways, transition states, and activation barriers to reveal how atomic-scale interactions control catalytic activity and selectivity.
We study transition-metal catalysts, single-atom catalysts, and 2D materials, tuning their performance through defect engineering, strain, functionalization, and controlled metal dispersion. Systems such as biphenylene-based single-atom catalysts and functionalized MXenes allow us to probe how local bonding environments influence CO₂ activation, intermediate stabilization, and product formation. Our analyses show how defect sites enhance adsorption, how isolated metal adatoms remain stable, and how specific coordination motifs promote key intermediates leading to value-added products like formic acid, methanol, and methane. Ultimately, our goal is to identify catalytic platforms that are efficient, selective, stable, and composed of earth-abundant elements suitable for large-scale deployment.
ML model for Lattice Thermal Conductivity prediction (Ref.)
Image Credit : Nature Communications volume 11,3509 (2020)
Machine learning is transforming materials science by enabling rapid prediction, screening, and design of new materials far beyond the reach of conventional computational methods. While high-throughput DFT calculations remain the gold standard for accuracy, they become prohibitively expensive for large or complex systems. Our group develops machine-learning models that learn structure–property relationships from high-quality DFT datasets, allowing us to predict formation energies, band gaps, thermal transport, catalytic activity, and other key properties at a fraction of the computational cost.
A major thrust of our research is building ML frameworks that closely emulate DFT-level accuracy. By training models on reliable data from specific material families and calculation protocols, we effectively replace the most computationally intensive steps of traditional workflows while retaining strong physical interpretability. We use feature-engineering approaches, graph neural networks, and generative models to capture both local atomic environments and global crystal symmetry.
This integrated strategy—combining physics based understanding with fast, data-driven inference—enables rapid exploration of vast chemical spaces, accelerates the discovery of high-performance materials for energy, electronic, and catalytic applications, and establishes a scalable paradigm for next-generation materials design.