An illustration of the inner workings of the novel machine learning algorithm for determining crystal structure: a convolutional neural network computes the probability that the input diffraction pattern belongs to a given class (e.g. Bravais lattice or space group). Image: Vecchio lab/Science.Nanoengineers at the University of California (UC) San Diego have developed a computer-based method that could make it less labor-intensive to determine the crystal structures of various materials and molecules, including alloys, proteins and pharmaceuticals.
Their method uses a machine learning algorithm, similar to the type used in facial recognition and self-driving cars, to independently analyze electron backscatter diffraction (EBSD) patterns produced by a scanning electron microscope (SEM), and do so with at least 95% accuracy. The nanoengineers, led by UC San Diego nanoengineering professor Kenneth Vecchio and his PhD student Kevin Kaufmann, report their work in a paper in Science.
Compared to other electron diffraction techniques, such as those used with transmission electron microscopy (TEM), SEM-based EBSD can be performed on large samples and analyzed at multiple length scales. This provides local sub-micron information mapped to centimeter scales. For example, a modern EBSD system allows the determination of fine-scale grain structures, crystal orientations, relative residual stress or strain, and other information in a single scan of the sample.
The drawback of commercial EBSD systems is the software's inability to determine the atomic structure of the crystalline lattices within the material being analyzed. This means a user of commercial software must select up to five crystal structures presumed to be in the sample. The software then attempts to find probable matches to these candidate structures in the diffraction pattern.
Unfortunately, the complex nature of the diffraction pattern means the software often finds false matches in the user-selected list. As a result, the accuracy of existing software's determination of the lattice type is dependent on the operator's experience and prior knowledge of the sample.
The method that Vecchio's team developed does all this autonomously, as the deep neural network independently analyzes each diffraction pattern to determine the crystal lattice, out of all possible lattice structure types, with a high degree of accuracy (greater than 95%). According to the researchers, a wide range of research areas, including pharmacology, structural biology and geology, could benefit from using similar automated algorithms to reduce the amount of time required for identifying crystal structures.
This story is adapted from material from the University of California, San Diego, with editorial changes made by Materials Today. The views expressed in this article do not necessarily represent those of Elsevier. Link to original source.