Paper detail

ConfPred: A layered intergrowth structure prediction model based on confinement self-assembly in two-dimensional interlayer space

We constructed a simple but effective model to predict the layered intergrowth structures by combining the self-assembly phenomenon in confined space and the sandwich configuration of layered materials. In this model, a two-dimensional confined space is constructed by two known block layers, such as the Fe$_2$As$_2$ block of iron-based superconductors. Then, the crystal structure prediction is carried out only inside the confined space to search for brand-new block layers. We realized this model on the basis of the USPEX9.4 code. In the test, the already existing iron-based superconductors can be always successfully found, such as Ba$_2$Ti$_2$Fe$_2$As$_4$O, Sr$_3$Sc$_2$Fe$_2$As$_2$O$_5$, Sr$_4$V$_2$Fe$_2$As$_2$O$_6$, and so on. The comparison test suggests that our model has remarkable advantages in searching for intergrowth structures. With this space confinement prediction model, a structure prediction of layered intergrowth materials even with up to six elements can be performed with an acceptable machine time consumption. So far, we have done some multi-composition crystal structure predictions of iron-based superconductor, and found several stable and metastable structures, such as Ba$_2$Fe$_2$As$_3$, Eu$_2$Fe$_2$As$_3$, La$_2$O$_2$ClFeAs, LiOMn$_2$As,Li$_4$OFe$_2$As$_2$.