Figure2

Crystal structure prediction using neural network potential and age-fitness Pareto genetic algorithm

Figure 2. Sample structure prediction by ParetoCSP. Every ground truth structure is followed by the predicted structure. (A- P) show that the structures of MnAl, ZrSe$$ _2 $$, GeMo$$ _3 $$, SrTiO$$ _3 $$, Ta$$ _2 $$N$$ _3 $$O, and GaBN$$ _2 $$ were successfully predicted, while (Q - T) show that ParetoCSP was unable to predict the structures of GaSeCl, and NdNiSnH$$ _2 $$. All the structures were visualized using VESTA. For better visualization, we set the fractional coordinate ranges of all axes to a maximum of $$ 3 $$ for Ta$$ _2 $$N$$ _3 $$O, GaBN$$ _2 $$, and GaSeCl, and we used the space-filling style for Ta$$ _2 $$N$$ _3 $$O, and GaSeCl. Besides these, we set the fractional coordinate ranges of all axes to a maximum of $$ 1 $$ for all structures and used the ball-and-stick style.

Journal of Materials Informatics
ISSN 2770-372X (Online)
Follow Us

Portico

All published articles are preserved here permanently:

https://www.portico.org/publishers/oae/

Portico

All published articles are preserved here permanently:

https://www.portico.org/publishers/oae/