Garnet Neural Network

This application uses a deep neural network to predict the stability of garnet crystals, as described in Ye, W.; Chen, C.; Wang, Z.; Chu, I.-H.; Ong, S. P. Deep neural networks for accurate predictions of crystal stability. Nat. Commun. 2018, 9 (1), 3800 DOI: 10.1038/s41467-018-06322-x.

Instructions

Enter species in the C (24c), A (16a) and D (24d) sites of the prototypical C₃A₂D₃O₁₂ garnet crystal.

• Species allowed on C site:
  Bi³⁺, Hf⁴⁺, Zr⁴⁺, La³⁺, Pr³⁺, Nd³⁺, Sm³⁺, Eu³⁺, Gd³⁺, Tb³⁺, Dy³⁺, Ho³⁺, Er³⁺, Tm³⁺, Yb³⁺, Lu³⁺, Y³⁺ , Cd²⁺, Zn²⁺, Ba²⁺, Sr²⁺, Ca²⁺, Mg²⁺, Na⁺
• Species allowed on A site:
  Rh³⁺, Ru⁴⁺, Cr³⁺, Sb⁵⁺, Ta⁵⁺, Nb⁵⁺, Sn⁴⁺, Ge⁴⁺, Hf⁴⁺, Zr⁴⁺, Ti⁴⁺, In³⁺, Ga³⁺, Al³⁺, Lu³⁺, Yb³⁺, Tm³⁺, Er³⁺, Ho³⁺, Dy³⁺, Y ³⁺ , Sc³⁺, Zn²⁺, Mg²⁺, Li⁺
• Species allowed on D site:
  As⁵⁺, P⁵⁺, Sn⁴⁺, Ge⁴⁺, Si⁴⁺, Ti⁴⁺, Ga³⁺, Al³⁺, Li⁺
• "Ca2+" means only one type of species (Ca²⁺) is present in the site.
• "Ca2+:2 Mg2+:1" means Ca²⁺ and Mg²⁺ are present in the site in the ratio of 2:1.

Only 2:1 mixing is allowed in the C and D sites, and 1:1 mixing is allowed in the A sites.