Crystal Structure Prediction in Practice: Setting Up Searches and Reading Results with Matlantis CSP

Crystal structure prediction is an important approach for exploring unknown stable crystals and for evaluating phase stability. In practice, however, it is not always clear at which phase of a research project it should be used, or how.

Using Matlantis CSP as its subject, this webinar covers how to select a search method, how to design the search conditions, and how to use the results. The speaker is a researcher involved in the development of Matlantis CSP.

Who should attend

• Computational materials scientists evaluating crystal structure prediction for their own workflows
• Experimental and R&D researchers working on compositions with no reported structure
• Teams screening candidate phases ahead of synthesis

Outline

Crystal structure prediction is an important approach for exploring unknown stable crystals and for evaluating phase stability. At the same time, it is a technique whose place in actual research is not always easy to see: at which phase of a project it should be applied, and in what way.

Matlantis CSP runs these searches on Matlantis PFP, a universal machine learning interatomic potential covering 96 elements, so no system-specific model needs to be trained.

Using Matlantis CSP as its subject, this webinar covers the following:

1. How crystal structure prediction can support decision-making in the early phase of research, while the crystal structure is still undetermined
2. How to select a search method, global search or derivative structure search, according to the research objective
3. How to design the search conditions
4. How to post-analyze the search results, and how to connect them to experiments and to subsequent steps

The speaker is involved in the development and research of Matlantis CSP, and will draw on actual use cases. The aim is to present crystal structure prediction not as the task of running calculations, but as a process that supports research decisions.

The session is intended for anyone who wants to clarify where crystal structure prediction fits in their research, from those who are about to use Matlantis CSP for the first time to those already using it who want to get more out of it.

Speaker

Kohei Shinohara

Researcher,
Preferred Networks, Inc.

Kohei received his B.Eng. in Engineering Science from Kyoto University in 2018, and his M.Eng. (2020) and Ph.D. (2023) in Materials Engineering from the Graduate School of Engineering, Kyoto University. He joined Preferred Networks, Inc. as a researcher in 2023. His work centers on computational crystal structure prediction, computational crystallography, and scientific software development.

Webinar Details

Apply here