skip to content

Winton Programme for the Physics of Sustainability

Department of Physics

Computational design of thermally-functional materials
Sabbatical visit host Professor Jeffrey Neaton, Molecular Foundry and Kavli ENSI

Stephen Elliott spent the month of May 2018 at Berkeley during his sabbatical year as part of the Winton-Kavli, Cambridge-Berkeley exchange scheme.  His host for the visit was Professor Jeff Neaton, Director of the Molecular Foundry at the Lawrence Berkeley National Laboratory (LBNL).  

One of the aims of the visit was to learn more about the work being carried out at Berkeley and LBNL on materials for thermoelectricity generation, the latter being allied to the Winton's mission for sustainability. Another area of interest was work in the Physics Department, in the group of Professor Frances Hellman, on the preparation of amorphous silicon (a-Si) films having extremely low values of mechanical loss. This work is directly relevant to some of Stephen’s very recent computer-simulation work [1], in which he has used a recently-developed linear-scaling (O(N)) interatomic potential for Si, derived using a machine-learning approach, to generate models of a-Si with unprecedently high degrees of structural order, and hence correspondingly low values of mechanical loss. It is hoped that a collaborative effort can be initiated between the two groups to make a-Si in glassy form, directly from the liquid, which should have the highest degree of structural order, and thus the lowest value of mechanical loss, for this material.


[1] J. Phys. Chem. Lett. 9, 2879-2885 (2018)

Latest news

Manipulation of Quantum Entangled Triplet Pairs

7 January 2021

Researchers have uncovered a new technique to create and manipulate pairs of particle-like excitations in organic semiconductors that carry non-classical spin information across space, much like the entangled photon pairs in the famous Einstein-Podolsky-Roden “paradox”.

Machine learning algorithm helps in the search for new drugs

20 March 2019

Researchers have designed a machine learning algorithm for drug discovery which has been shown to be twice as efficient as the industry standard.