Welcome to Seebeck coefficient predictive component of our toolkit

The calculator predicts Seebeck coefficient of a given compound at 300K, 400K, 700K, and 1000K.
This calculator is trained on nearly 300 thermoelectric compounds. The predictive models are built using the Random Forest technique. For each compound entered by the user, 187 attributes are generated based on the physical and chemical properties of the constituent elements, and fed into the predictive models to obtain predictions for Seebeck coefficient.
The following rules must be remembered when entering the compound information:
This application is written for compounds with two or more elements.
A coefficient should be always entered after each element. Space should be used as separator between elements and their coefficients.
An element can be entered only once in a compound.

Example 1: formula Bi2Te3 must be entered as Bi 2 Te 3
Example 2: formula Ca0.9Bi0.1Mn0.9Nb0.1O3 must be entered as Ca 0.9 Bi 0.1 Mn 0.9 Nb 0.1 O 3
Example 3: formula Ca(Mn2Cu)Mn4O12 must be entered as Ca 1 Mn 6 Cu 1 O 12


  • A. Furmanchuk, J. Saal, J. W. Doak, G. B. Olson, A. Choudhary, A. Agrawal, "Prediction of Seebeck coefficient for compounds without restriction to fixed stoichiometry: A machine learning approach", Journal of Computational Chemistry, 2017, DOI: 10.1002/jcc.25067 [url]