Materials Design Workbench: An Open-Access Database for Materials and Molecular Design

Monday, October 17, 2011: 9:30 AM
200 B (Minneapolis Convention Center)
Glen Allen Ferguson1, Jeff Greeley2, Peter Zapol1, Larry A. Curtiss3, Svetlozar Evtimov Nestorov4, Ian T. Foster5, David Landis6, Jens Strabo Hummelshøj7, Karsten W. Jacobsen8, Thomas Bligaard9 and Jens Nørskov10, (1)Material Science Department, Argonne National Laboratory, Argonne, IL, (2)Center for Nanoscale Materials, Argonne, IL, (3)Materials Science Division, Argonne National Laboratory, Argonne, IL, (4)Computer Science, University of Chicago, Chicago, IL, (5)Mathematics and Computer Science and Department of Computer Science, Argonne National Laboratory and University of Chicago, Argonne, IL, (6)Department of Physics CAMD, Technical University of Denmark, Kongens Lyngby, Denmark, (7)SLAC National Accelerator Laboratory, SUNCAT, Menlo Park, CA, (8)Center for Atomic-scale Materials Physics, Dept. of Physics, Technical University of Denmark, Lyngby, Denmark, (9)Center for Atomic-scale Materials Design, Technical University of Denmark, Kgs. Lyngby, Denmark, (10)Chemical Engineering, Stanford University and SUNCAT, Stanford, CA

The use of modern computing resources and efficient computational chemistry software packages has resulted in the production an extraordinary amount of computational data related to materials and catalytic problems. Very few such data, however, are available for public browsing; they tend to be stored locally and are rarely publically disseminated. A significant reason for this is absence of an easy to use, open-access public database for computational materials and molecular data. To address this problem, we are implementing an open-access database, the Materials Design Workbench (MDW), that will store data from computational chemistry software. The data are in a form that can be conveniently searched, visualized, analyzed, and retrieved by users.

            The implementation of the Materials Design Workbench is a collaboration between scientists at Argonne National Laboratory, the University of Chicago, the Center for Sustainable Energy through Catalysis at the Stanford Linear Accelerator Center, and the Technical University of Denmark. The implementation consists of several modules. The central piece is a MySQL database that stores information from computational chemistry software and associated experimental data. To facilitate entering computational data into the database, a stand-alone python interface named the Computational Materials Repository (CMR) has also been developed. One of the most important roles of the CMR is to convert the output from calculations into XML format files. These XML files can then be modified to add additional data and uploaded to the database. Two additional applications are under development for visualization and analysis of data that have been entered into the database. The first is a client-side application that runs in any HTML 5 enabled web browser. The second is a server-side web interface.

            In this talk, we will discuss two important parts of the MDW. The first is how the MySQL database stores and organizes the computational data. The second, which will include a preview, is the server-side application. This application, which is in development at the University of Chicago and Argonne National Laboratory, currently consists of a web interface for viewing and retrieving structures from the database. A user can query the database using any of the stored criteria and can display the results as a single molecule or as a collection of molecules. These collections can be analyzed, saved for later analysis, or downloaded. Users can also attach citations or other any other desirable data to their calculations. It is also possible to restrict access to data sets to particular users and groups or to make the data globally readable.  These features, combined with the powerful visualization and analysis software that is also being developed, make MDW an attractive solution for the storage, retrieval and analysis of materials and molecular data.

Extended Abstract: File Not Uploaded