Light-Induced Phase Transitions in Semimetals from First Principles

Methods

Approaches developed in my group for modelling ultrafast phenomena combine multiple levels of theory, including density-functional theory and many-body Green’s-function. We have developed new theoretical frameworks and codes to address emerging classes of ultrafast processes, including the microscopic description of coherent phonon excitation and lattice-driven nonequilibrium dynamics. Our developments are released through widely used, open-access simulation packages — including Quantum ESPRESSO and the EPW code — so that the broader community can adopt, validate, and extend these methods. More specifically, we focus on implementing new capabilities for describing the nonequilibrium dynamics of the crystal lattice, with an emphasis on accuracy at the first-principles level. The simulation workflows developed in our group typically rely on solving large-scale linear-algebra problems (in particular, demanding eigenvalue problems) together with time-propagation algorithms for coupled electron–lattice dynamics.

Results

Using the Lichtenberg high-performance computing cluster, we have carried out large-scale simulations of ultrafast and nonequilibrium phenomena in a variety of condensed matter systems. Our computational investigations focused on several classes of quantum materials, including two-dimensional semiconductors, and topological semimetals, where microscopic modeling is essential to capture the interplay of electronic structure, topology, and lattice dynamics. A central focus has been the first-principles description of coherent phonon excitations — collective atomic motion triggered by excitation with ultrashort laser pulses. Additionally, we have investigated polaron dynamics and the buildup and relaxation of nonequilibrium phonon populations, providing insight into how energy and coherence flow between electrons and the crystal lattice after photoexcitation. These efforts have resulted in several publications in peer-reviewed journals and have helped establish new methodological frameworks for ultrafast-dynamics simulations. Importantly, our computational results directly complement experimental investigations by providing material-specific information about the origin of spectroscopic features. Overall, access to HPC resources throughout project p0021280 has been a critical component for establishing and consolidating the research activities of my group during the ramp-up phase of the junior professorship.

Discussion

We identified the emergence of non-thermal phonon populations in the two-dimensional semiconductor MoS. Together with signatures of ultrafast diffuse scattering, these results provide evidence for a light-induced modification of the electron–phonon interaction, highlighting how photoexcitation can transiently reshape scattering pathways and energy flow between electrons and the lattice. We also carried out the first computational investigation of real-time polaron formation on the Lichtenberg cluster. These simulations enabled us to pinpoint key dynamical fingerprints of polaron formation—such as the characteristic buildup of lattice dressing and the associated redistribution of phonon populations—which can guide the identification of polarons in pump–probe experiments. In addition, our work on coherent phonons established a new formalism for treating driven lattice coherence in a fully microscopic framework. This approach is now becoming a reference methodology and contributes to the emerging state of the art in the field. Building on these advances, we will extend our modelling to new classes of nonequilibrium phenomena, including light-induced phase transitions in ferroelectrics and charge-density-wave materials. We will also address more complex composite quasiparticles, such as exciton–polarons, where correlations between electronic and lattice degrees of freedom require a unified treatment. These studies are already underway using the HPC infrastructure of the Lichtenberg clusters, and we anticipate additional peer-reviewed publications on these topics in the coming years.

Additional Project Information

DFG classification: 307-02 Theoretical Condensed Matter Physics
Software: EPW, QuantumEspresso
Cluster: Lichtenberg

Publications

Pan, Y., Caruso, F. (2024): Strain-induced activation of chiral-phonon emission in monolayer WS2. npj 2D Materials and Applications, 8(1), 42.

Emeis, C., Jauernik, S., Dahiya, S., Pan, Y., Jensen, C. E., Hein, P., Caruso, F. (2025).: Coherent phonons and quasiparticle renormalization in semimetals from first principles. Physical Review X, 15(2), 021039.

Pan, Y., Emeis, C., Jauernik, S., Bauer, M., & Caruso, F. (2025). Ab initio theory of coherent phonon damping in semimetals. Physical Review B, 112(24), 245111.

Pan, Y., Hildebrandt, P. N., Zahn, D., Zacharias, M., Windsor, Y. W., Ernstorfer, R., … & Seiler, H. (2025): Momentum-resolved signatures of carrier screening effects on electron–phonon coupling in MoS2. ACS nano, 19(11), 11381-11389.

Garcia-Herrero, V., Emeis, C., Dai, Z., Lafuente-Bartolome, J., Giustino, F., Caruso, F. (2026): Watching Polarons Form in Real Time. arXiv preprint arXiv:2601.21810.

Pan, Y., Fragkos, S., Descamps, D., Petit, S., Caruso, F., Beaulieu, S. (2025): Ultrafast Strongly Anisotropic Valleytronics in SnSe. arXiv preprint arXiv:2512.15400.