The results from this tool are estimates based on data-driven analytics on DFT-computed and experimentally-measured
data (for the 39 DFT target properties, the prediction is by definition at 0 K temperature).
Note that these models are trained on the known most stable structure for a given composition, and thus only
take the composition as input. In other words, no structure information is needed to get the property predictions
from this software tool, which on the one hand is advantageous as structure information can be unavailable or
difficult to obtain in many cases, while on the other hand, these models cannot distinguish between structure
polymorphs of a given composition.
All results are provided for informational purposes only, in furtherance of the developers'
educational mission, to complement the knowledge of materials scientists and engineers, and
assist them in their search for new materials with desired properties. The developers may not
be held responsible for any decisions based on this tool.
Welcome to the online materials property predictor. This tool deploys data mining models to predict
the various materials properties of a materials (given below)
based on its chemical composition or physical attribute as described in the paper presented below.
The predictive models deployed here have been built on hundreds
of thousands of Density Functional Theory (DFT) calculations on crystalline materials from the
Open Quantum Mechanical Database (OQMD) and Joint Automated Repository for Various Integrated Simulations (JARVIS),
and run many orders of magnitude faster than DFT.
In order to use this tool, please provide the list of chemical compositions in the text box below,
and click Submit. Please ensure that each chemical formula respects the charge balance condition
with common oxidation states of individual elements. The elements indicated with red color in the
period table below may not be used.
This work was performed under the following financial assistance award 70NANB19H005 from U.S. Department of Commerce, National Institute of Standards and Technology as part of the Center for Hierarchical Materials Design (CHiMaD). Partial support is also acknowledged from NSF Award CMMI-2053929 and DOE Awards DE-SC0014330, DE-SC0019358 and DE-SC0021399, and Northwestern Center for Nanocombinatorics.
-
V Gupta, K Choudhary, F Tavazza, C Campbell, WK Liao, A Choudhary, A Agrawal. Cross-property deep transfer learning framework for enhanced predictive analytics on small materials data. Nature communications. 2021 Nov 15;12(1):6595. [url]
-
V Gupta, K Choudhary, Y Mao, K Wang, F Tavazza, C Campbell, WK Liao, A Choudhary, A Agrawal. Mppredictor: An artificial intelligence-driven web tool for composition-based material property prediction. Journal of Chemical Information and Modeling. 2023 Mar 27;63(7):1865-71. [url]
Center for Ultra-scale Computing and Information Security (CUCIS) , EECS Department, Northwestern University, Evanston, IL 60208, USA