Exploiting the European XFEL for a New Generation of High Energy Density and Materials Science

Lead Research Organisation: University of Cambridge
Department Name: Materials Science & Metallurgy

Abstract

Abstracts are not currently available in GtR for all funded research. This is normally because the abstract was not required at the time of proposal submission, but may be because it included sensitive information such as personal details.
 
Description The ability to predict materials structure from first principles has revolutionised materials science, permitting the design of new materials "in the computer". Our previous work has used traditional density functional theory to describe the energy landscape of interacting atoms. As part of this project we have demonstrated that ab initio random structure searching (AIRSS) can be accelerated for many metallic systems through the orbital free density functional theory. That is, a form of density functional theory that does not require the computation of any quantum mechanical wave functions, following instead just the electronic density. As part of this work we used our newly developed SHEAP method for the dimensional reduction and visualisation of energy landscapes. The development of Ephemeral Data Derived Potentials (EDDPs) has the potential to accelerate structure search by several orders of magnitude.
Exploitation Route The AIRSS, PROFESS, EDDP and SHEAP computer codes used in this work have been made available under open source licenses.
Sectors Aerospace, Defence and Marine,Chemicals,Construction,Digital/Communication/Information Technologies (including Software),Electronics,Energy,Manufacturing, including Industrial Biotechology
URL https://www.msm.cam.ac.uk/news-and-events/papers-month-archive/2022-01-cjp
 
Title EDDP 
Description Ephemeral Data Derived Potentials (EDDP): The ddp package contains a suite of tools to construct and test data derived interatomic potentials. They are designed to be used with the airss first principles structure prediction package. Ab initio random structure searching (AIRSS) can be used to generate data, and exploit the generated ddp potentials to potentially accelerate searches. They are referred to as ephemeral data derived potentials as they are designed to be constructed for a particular set of structure searching parameters, discarded and regenerated as those parameters change. The methodology is introduced in Pickard, Ephemeral data derived potentials for random structure search, 2022. 
Type Of Technology Software 
Year Produced 2022 
Open Source License? Yes  
Impact The EDDP package has been downloaded over 140 times since its release - an entirely organic uptake, as the focus has been on exploring its impact locally. Given its compatibility with the CASTEP and AIRSS codes there is scope for significant adoption over time. 
URL https://www.mtg.msm.cam.ac.uk/Codes/EDDP
 
Title Stochastic Hyperspace Embedding and Projection (SHEAP) 
Description Stochastic Hyperspace Embedding And Projection (SHEAP) is a dimensionality reduction method designed for visualising potential energy surfaces. Computational structure prediction can assist the discovery of new materials. One searches for the most stable configurations of a given set of atomic building blocks, which correspond to the deepest regions of an energy landscape-the system's energy as a function of the relative positions of its atoms. To explore these landscapes efficiently, it is important to understand their topologies. However, they exist in spaces with very large numbers of dimensions, making them difficult to visualise. SHEAP uses dimensionality reduction through manifold learning to effectively visualise the distribution of stable structures across a high-dimensional energy landscape. 
Type Of Technology Software 
Year Produced 2021 
Open Source License? Yes  
Impact The SHEAP code is being routinely used in our structure searches to map the energy landscape, and help to steer the searches. 
URL https://www.mtg.msm.cam.ac.uk/Codes/sheap