Currently, we have three open projects at the level of “Laurea Magistrale” and in exceptional cases also the “Laurea di Fisica.”
Machine Learning Room Temperature Superconductors
This year the field of superconducting hydrides has witnessed a huge step towards delivering the first near room-temperature superconductors. A couple of months ago two independent groups have measured in LaH10, critical temperatures of superconductivity of ~250 K and ~260 K, respectively. The discovery of these compounds, previously predicted, tremendous proves the progress in theoretical and computational methods. Further progress in computational techniques is highly desirable in order to speed up the discovery of other high-Tc materials. Help could come from advanced methodologies such as machine learning models, fingerprints and neural networks that are robust and well tested in other fields. During the development of this thesis, you will be brought up top-speed into the state of the art material simulation methods: we will build a machine learning model and test its accuracy versus a curated database. Succeeding in this task and you will be able to predict superconducting values of TC. Only a handful subset of materials from hundreds that have been proposed have been measured. Considering a large number of systems that contain hydrogen and the vast space of thermodynamic conditions yet to be explored, there is plenty of room for discovering your new high-Tc material! You will be working in a new team, highly active, young atmosphere and you will have access to the fastest supercomputer in Europe. Basic Linux and basic programming is a plus.
Tags: superconductors, computational modelling, machine learning, and descriptors. Tools: DFT, supercomputer, linux, fortran, python, jupyter.
Complex Phase Transformation in 2D materials
Two-dimensional materials are substances with a thickness of a few nanometres or less. Electrons in these materials are free to move in the two-dimensional plane, but their restricted motion in the third direction is governed by quantum mechanics. Prominent examples include quantum wells and graphene. Other systems of interest are materials that easily exfoliate or are fabricated from 3D counterparts. We will focus on intermetallic compounds that, in their bulk, have been extensively studied for the purpose of thermoelectricity, superconductivity and as hard materials. In this work, we propose to revive and study a subset of these materials and understand the process that leads to a specific transformation from 3D to 2D. This transformation is not simply peeling like in graphite that leads to graphene, but involves re-arranged of bonds and is driven by an energetic reaction. We aim to understand this complex transformation by mapping the lattice vibrations and to derive the results to an experimental observable (spectroscopy, Raman, IR, etc). Succeeding in this task, and you will be able to unmask novel thermoelectric materials.
Tags: 2D intermetallic materials, computational modelling, anharmonic phonons, thermoelectricity. Tools: DFT, supercomputer, linux, bash, fortran.
Computational Evaluation of Prospective High-Performance p-Type Transparent Conductors
Transparent electronic and photonic devices have been a long-coveted technological advancement that would enable the creation of many futuristic applications, allowing us to turn windows into power generators (solar panels). At the centre of this technological advancement are unique materials –transparent to visible light, yet capable of sustaining an electric current. Such materials, known as transparent conductors (TCs), are typically formed from wide bandgap semiconductors, where the mobile charge carriers are introduced by heavy doping. The electrons (n-type) or holes (p-type) are generated through point defect incorporation. Though such point defects, as the name implies, create structural imperfections in the host crystalline lattice, their presence is key to the performance of TC materials. High-throughput studies are gaining increasing popularity as a tool for predicting p-type transparent conductors, however, the lack of appropriate descriptors makes the efficient screening of materials for this purpose challenge. Yet, in order to identify and validate novel descriptors, more robust data on non-oxide p-type transparent conductors is needed. During this project, we will focus on point-defect (native and impurity) calculations in a variety of wide bandgap materials. We suggest an in-depth investigation of the p-type propensity of all recently speculated, but not experimentally verified, p-type TC materials through an exhaustive study of native defects. Succeeding in this goal will open possibilities to advance the field of transparent conductors and bring futuristic applications at hand.
Tags: Semiconductors, p-type conductivity, computational modelling, high-throughput. Tools: DFT, hybrid calculations, supercomputer, linux, python.