The research activity of the Institute of Physics covers five main domains:
- Quantum Science and Technology
- Condensed Matter Physics
- Biophysics and complex systems
- Particle and astrophysics
- Physics for energy
Two-dimensional crystals of semi-metallic van der Waals materials hold much potential for the realization of novel phases, as exemplified by the recent discoveries of a polar metal in few-layer 1T’-WTe2 and of a quantum spin Hall state in monolayers of the same material. Understanding these phases is particularly challenging because little is known from experiment about the momentum space electronic structure of ultrathin crystals. In this talk, I will discuss direct electronic structure measurements of exfoliated mono- bi- and few-layer 1T’-WTe2 by laser-based micro-focus angle resolved photoemission. This is achieved by encapsulating a flake of WTe2 comprising regions of different thickness with monolayer graphene. Our data support the recent identification of a quantum spin Hall state in monolayer 1T'-WTe2 and reveal strong signatures of the broken inversion symmetry in the bilayer. We finally discuss the sensitivity of encapsulated samples to contaminants following exposure to ambient atmosphere.
About the research of the speaker: https://dqmp.unige.ch/baumberger/
Deep learning has been immensely successful at a variety of tasks, ranging from classification to artificial intelligence. Yet why it works is unclear. Learning corresponds to fitting training data, which is implemented by descending a very high-dimensional loss function. Two central questions are (i) since the loss is a priori not convex, why doesn't this descent get stuck in poor minima, leading to bad performance? (ii) Deep learning works in a regime where the number of parameters can be larger, even much larger, than the data to fit. Why does it lead to very predictive models then, instead of overfitting?
Here I will discuss an unexpected analogy between the loss landscape in deep learning and the energy landscape of repulsive ellipses, that supports an explanation for (i). If times permit I will discuss (ii), more specifically the surprising finding that predictive power continuously improves by adding more parameters.
By: Prof. Jan S. Hesthaven, Doyen de la Faculté des Sciences de Base,
Electron-phonon interactions (EPIs) are ubiquitous in condensed matter and materials physics. For example EPIs play a central role in the electrical resistivity of metals, the carrier mobility of semiconductors, the pairing mechanism of conventional superconductors, and the optical properties of indirect-gap materials. More fundamentally, the EPI is the simplest realization of the interaction between fermion and boson fields, arguably one of the pillars of many-particle physics and quantum electrodynamics. The EPI has been studied for almost a century, however only during the last two decades predictive, non-empirical calculations have become possible. In this talk I will outline the theoretical and computational framework underlying modern electron-phonon calculations from first principles, and illustrate recent progress in this area by discussing representative work from our group. In particular I will touch upon our recent investigations of polarons in the angle-resolved photoelectron spectra of transition metal oxides [1,2], the superconducting pairing mechanism in transition metal dichalcogenides , non-adiabatic Kohn anomalies in the inelastic X-ray scattering spectra of doped semiconductors , and the phonon-induced renormalization of carrier effective masses in halide perovskites . I will conclude by discussing opportunities for future work, and the key challenges for advancing theoretical and computational research on electron-phonon physics .
 C. Verdi et al., Nat. Commun. 8, 15769 (2017).
 J. M. Riley et al., Nat. Commun. 9, 2305 (2018).
 C. Heil et al., Phys. Rev. Lett., 119, 087003 (2017).
 F. Caruso et al., Phys. Rev. Lett. 119, 017001 (2017).
 M. Schlipdf et al., Phys. Rev. Lett. 121, 086402 (2018).
About the speaker
Feliciano Giustino is Full Professor of Materials at the University of Oxford, and during AY 2017/18 he was the Mary Shepard B. Upson Visiting Professor in Engineering at Cornell University. He holds an MSc in Nuclear Engineering from Politecnico di Torino and a PhD in Physics from the Ecole Polytechnique Fédérale de Lausanne. Before joining the Department of Materials at Oxford he was a postdoc in the Physics Department of the University of California at Berkeley. He specialises in electronic structure theory and the atomic-scale design of advanced materials for electronics, photonics, and energy. He is author of 120+ research papers and one book on Materials Modelling using Density Functional Theory. He started the open-source software project EPW, which is currently distributed as a core module of the Quantum ESPRESSO materials simulation suite.
17th NCCR MARVEL "Distinguished Lecture"
By: Prof. Feliciano Giustino (University of Oxford, UK)
Particle collisions are the most common experiment in Particle Physics. In real experiments the collision energy is limited by our current technology. Thought experiments are not subject to this limitation and can provide a window into the strongly coupled regime of Quantum Field Theory (QFT) and Quantum Gravity. I will explain how basic physical principles can be used to constrain scattering amplitudes in QFT.