Computational Physics Group

First prize for this year’s 2023 IoP CPG Thesis Prize has been awarded to Michael Davies, University College London and University of Cambridge. Michael’s thesis, titled Solving mysteries of ice formation with simulation and data-driven methods, applied molecular dynamics simulations and machine learning techniques to understand ice formation at the molecular level.

At first glance the formation of ice might seem a mundane everyday phenomenon. But its impacts are vast, ranging from glaciers, to cryopreservation, to climate modelling. And its formation is perplexing: in its pure state water must be cooled to around -40 °C for ice to form and a foreign material is almost always required. To understand how materials control ice formation, Michael used high-throughput computational simulations in combination with deep learning. The work uncovered a path to an elusive “cubic ice” polymorph and produced an AI model that beat experts from across the globe in an open head-to-head challenge despite 80 years of human endeavour. Michael also investigated the formation of “amorphous ice”, which is believed to be the most common form of water in the universe. In collaboration with experiment, he discovered a new form of amorphous ice with the same density as liquid water. The discovery raises questions about the very nature of liquid water.

We look forward to reading more about Michael’s work in the next IoP Computational Physics Group Newsletter. In the meantime, Michael’s thesis is available online.

Second prize for this year’s 2023 IoP CPG Thesis Prize has been awarded to Dimitrios Bachtis, Swansea University. Dimitrios’s thesis, titled Quantum field-theoretic machine learning and the renormalization group, explores the derivation of neural networks from quantum field theories and utilizes machine learning techniques to study phase transitions.

An indispensable tool in the study of phase transitions is the renormalization group, which investigates how a system changes when viewed at different scales. The application of a renormalization group transformation can be intuitively understood as the “zooming out” of a map: as the image becomes smaller some of the fine details within the map have disappeared. Dimitrios explored how machine learning algorithms enable an approximate inversion of the renormalization group which, analogously to the previous example, can be understood as the “zooming in” where new fine details are now introduced by the machine learning algorithm. The method opens up the opportunity to conduct high precision computational studies of phase transitions. Dimitrios additionally investigated the derivation of neural networks from quantum field theories via the use of the Hammersley-Clifford theorem, thus establishing a mathematically rigorous connection between quantum field theory, machine learning, and probability theory.  

We look forward to reading more about Dimitrios’ work in the next IoP Computational Physics Group Newsletter. In the meantime, Dimitrios’ thesis is available online.

This year’s PhD Thesis Prize is now accepting nominations. The Committee of the Institute of Physics Computational Group offers an annual prize for the author of the PhD thesis that, in the opinion of the Committee, contributes most strongly to the advancement of computational physics.

The prize submission deadline of 30 April 2023. More information can be found on the CPG page.

Nominations can be made by emailing t.shendruk@ed.ac.uk:

• a four-page (A4) abstract
• a one-page (A4) citation from the PhD supervisor
• a one-page (A4) confidential report from the external thesis examiner

Second prize for this year’s 2022 IoP CPG Thesis Prize has been awarded to Mary Coe, University of Bristol. Mary’s thesis, titled Hydrophobicity Across Length Scales: The Role of Surface Criticality, employed Monte Carlo simulations and density function theory to elucidating the behaviour of water near a hydrophobic solid surface. Despite its ubiquity in everyday life and in many scientific disciplines, the underlying physical mechanism relating hydrophobicity on the microscopic scale to hydrophobicity on macroscopic length scales has remained a difficult problem. Mary studied density depletion and enhanced fluctuations in the vicinity of the drying critical point for several fluid-fluid and fluid-solid interactions near curved surfaces, and so extended her work beyond hydrophobicity to consider more generally solvophobicity. Mary’s results provide strong numerical evidence that the mechanism underlying both hydrophobicity and solvophobicity across microscopic and macroscopic length scales is a drying surface critical point.

We look forward to reading more about Mary’s work in the next IoP Computational Physics Group Newsletter. In the meantime, Mary’s thesis is available online.