Valence problems for some elements in Structure Predictor

Hi MP,

I can’t find the element Te (tellurium) with the valence of -2 when using the Structure Predictor APP.
Hope you can fix this.

Regards!
Yuchong

I believe this is because there were not sufficient materials containing Te2- (or Te-) in the data used to train the structure predictor. Only Te4+ and Te6+ are currently supported. Was there a specific Te2- material you were interested in?

@Anubhav_Jain reports that, per the original structure predictor paper:

“In this work, we consider a compound to be ionic if it contains one of the following anions: O2-, N3-, S2-, Se2, Cl-, Br-, I-, F-”

https://pubs.acs.org/doi/abs/10.1021/ic102031h

These were the only anions considered. We are due to re-train the structure predictor however so can look at potentially including Te -2 in the future. This will probably take some time, but I’ll update this thread if/when we have a chance to do that.

Hi mkhorton,

There are lots of telluride compounds in thermoelectric semiconductor field, e.g. Bi2Te3, PbTe, GeTe, SnTe, AgSbTe2, etc. All these tellurides are potential candidates for thermoelectric application, which converts temperature gradient into electrical potential and vice versa.

I hope the negative two valence of Te (Te2-) will be soon available here for a better understanding and application of the thermoelectrics.

Regard!
Yuchong