Materials Discovery Analytics

Materials Discovery Analytics

X-ray diffraction computed tomography reveals inner structure in batteries (project below)

The goal of our Working Group is to enable Columbia to take a leadership position in the emerging field of Materials Genomics and to exploit group funding opportunities in this space. This will be accomplished by a partnership between data science materials science professionals through the center, development of modeling concepts, tools, techniques and systems to integrate and leverage Data Science methodologies with fundamental physics and chemistry underlying material science to speed up the discovery of novel materials. We will also partner with hardware and software engineers to examine the development and use of computers and software platforms that are appropriate for the tasks at hand.


Simon Billinge, Applied Physics and Applied Mathematics (Chair)
Michael Hill, Chemical Engineering
Sanat Kumar, Chemical Engineering
Venkat Venkatasubramanian, Engineering


Katy Barmak, Applied Physics and Applied Mathematics
Louis Brus, Chemistry
Siu Chan, Applied Physics and Applied Mathematics
Jingguang Chen, Chemical Engineering
Qiang Du, Applied Physics and Applied Mathematics
Jacob Fish, Civil Engineering and Engineering Mechanics
Lauren Hannah, Statistics
Irving Herman, Applied Physics and Applied Mathematics
James Hone, Mechanical Engineering
Barry Honig, Biochemistry & Molecular Biophysics
Daniel Hsu, Computer Science
Chris Marianetti, Applied Physics and Applied Mathematics
Andy Millis, Physics
Colin Nuckolls, Chemistry
Rick Osgood, Electrical Engineering & Applied Physics and Applied Mathematics
Jon Owen, Chemistry
Abhay Pasupathy, Physics
James Teherani, Electrical Engineering
Michael Weinstein, Applied Physics and Applied Mathematics
Chris Wiggins, Applied Physics and Applied Mathematics
Sebastian Will, Assistant Professor of Physics
John Wright, Electrical Engineering
Xiaoyang Zhu, Chemistry


      1. Thin film PDFFunctional thin films are deposited onto a substrate, often silicon, silicon dioxide, or aluminum oxide. This makes it difficult to study the thin film structure because of the small amount of film material and large amount of substrate. To minimize the scattering of x-rays off the substrate, thin film x-ray scattering studies are usually done using grazing incidence (GI) X-ray experiments. However, GI studies are challenging, and are yet to be successfully applied for PDF analysis. Nanostructure analysis of thin films have therefore not been possible – until now.

        Taking advantage of the high X-ray flux at the XPD beamline at NSLS-II, we spent our first beamtime at the new beamline studying thin films. Instead of using the complicated GI setup, we used simple normal incidence measurements, letting the X-rays pass through the thick substrate before hitting the film. 99% of the collected total scattering signal therefore originates from the substrate, but with careful background subtraction using xPDFsuite, we were able to isolate the signal from the film and obtain high quality PDFs for nanostructure analysis of the films.

      1. Inner structure in batteriesCombining X-ray diffraction and PDF with computed tomography allows for spatially resolved structural analysis, which can help unravel structures in heterogeneous materials and map out the phases present in the system. We recently demonstrated this in a paper on complex structures in batteries, where we were able to map out the components in commercial AAA batteries and lab-scale coin cell batteries.

        Read more:

      1. Atomic structure at the nanoscaleThe work of the Complex Modeling team, Pavol Juhas, Kevin Knox, Mike McKerns and Xiaohao Yang, was recently highlighted on the Brookhaven National Laboratory website. The full article can be accessed here.

        Understanding atomic structure at the nanoscale is a massive challenge, but one at the heart of solutions to modern technological problems in energy, health and the environment. Complex modeling is an attempt to obtain more robust structure models for nanomaterials from the large amounts of data coming available at modern x-ray, neutron and electron sources such as the National Synchrotron Light Source-II (NSLS-II) at BNL. The work, presented by Prof. Billinge at the 2014 Supercomputing Conference in New Orleans in November 2014, is described in the BNL news article.

        Figure 1 of a recent group paper published in Journal of Applied Crystallography ( has been selected by the journal’s Editors as the cover art for the April 2014 issue. The paper from Farrow, pioneered a complex modelling approach to study the stuy the structure of CdS nanoparticles by incorporating both small angle X-ray scattering and PDF data for the first time, and found more robust morphological and structural results than using either single technique alone.

      1. dPDF methods used to detect nanoparticles at very dilute concentrationsMany modern material systems contain heterogeneous mixtures of components, and it is important to be able to characterize the structures of the components in such mixtures. Working with low concentrations is often desirable, either to achieve the proper pharmacokinetic properties of a drug molecule or to prevent particle aggregation and settling. To study structure in these systems, difference atomic pair distribution function (dPDF) methods are often used, in which a signal from only the component of interest is extracted by subtracting scattering contributions from other components, background, and environment.

        By studying small nanoparticles of a proprietary active pharmaceutical ingredient (API), we recently showed that dPDF methods can be used to detect nanoparticles at very dilute concentrations, in this experiment as low as 0.25 wt%. This allows for components to be studied in situ rather than in contrived situations with exaggerated concentrations. In this case, the API can be studied at the actual concentration of the marketed drug product.

550 W. 120th St., Northwest Corner 1401, New York, NY 10027    212-854-5660
©2018 Columbia University