AIChemy

Generative Machine Learning for Characterising Atomic Structure of Nanomaterials

Project Overview

X-ray scattering is essential for characterization of the atomic structure of materials, but the data analysis required to go from data to structure is a bottleneck in materials discovery and development. There is a huge potential for significantly advancing data analysis methods in X-ray science through data science. Here, building on advancements in deep learning, we proposed a new generative, data-driven approach to identify and determine atomic structures of nanomaterials directly from scattering data, as well as introducing new approaches to material discovery and generative machine learning.

Project Members

  • Collaborators: Kirsten Jensen (PI), Erik Dam (PI)
  • Raghavendra Selvan (Co-Applicant)
  • Ulrik Friis-Jensen (PhD Candidate)
  • Frederik Johansen (PhD Candidate)

Related Publications

  1. deCIFer: Crystal Structure Prediction from Powder Diffraction Data using Autoregressive Language Models
    Frederik Lizak Johansen, Ulrik Friis-Jensen, Erik Bjørnager Dam, Kirsten Marie Ørnsbjerg Jensen, Rocío Mercado, Raghavendra Selvan
    Transactions on Machine Learning Research (TMLR), 2026
  2. Tackling Real-World Crystal Structure Prediction from Powder X-ray Diffraction Data
    Frederik Lizak Johansen, Adam F Sapnik, Erik Bjørnager Dam, Raghavendra Selvan, Kirsten M Ø Jensen
    Digital Discovery, 2026
  3. CHILI: Chemically-Informed Large-scale Inorganic Nanomaterials Dataset for Advancing Graph Machine Learning
    Ulrik Friis-Jensen, Frederik Lizak Johansen, Andy Sode Anker, Erik Bjørnager Dam, Kirsten Marie Ørnsbjerg Jensen, Raghavendra Selvan
    ACM International Conference on Knowledge Discovery and Data Mining (ACM KDD), 2024
  4. A GPU-Accelerated Open-Source Python Package for Calculating Powder Diffraction, Small-Angle-, and Total Scattering with the Debye Scattering Equation
    Frederik Lizak Johansen, Andy Sode Anker, Ulrik Friis-Jensen, Erik Bjørnager Dam, Kirsten Marie Ørnsbjerg Jensen, Raghavendra Selvan
    Journal of Open Source Software (JOSS), 2024
  5. DeepStruc: Towards structure solution from pair distribution function data using deep generative models
    Emil T. S. Kjær, Andy S. Anker, Marcus N. Weng, Simon J. L. Billinge, Raghavendra Selvan, Kirsten M. Ø. Jensen
    Digital Discovery, 2023
  6. Characterising the Atomic Structure of Mono-Metallic Nanoparticles from X-Ray Scattering Data Using Conditional Generative Models
    Andy Sode Anker, Emil TS Kjær, Erik B Dam, Simon JL Billinge, Kirsten MØ Jensen, Raghavendra Selvan
    16th International Workshop on Mining and Learning with Graphs, 2020