Fig. 1

high-throughput calculations with multi-precision transmission strategy (a) and the process of establishing a structure-property relationships model based on machine learning techniques (b)"

Fig. 2

Flowchart of high-throughput calculations with multi-precision transmission strategy"

Fig. 3

Accelerated strategy for screening fast ion conductor materials and optimal doping scheme exploration[26]"

Fig. 4

Schematic diagram of the model architectures of several typical ML interatomic potential functions (a) workflow of building DP descriptors[29]; (b) CHGNet network architecture[30]; (c) NequIP network architecture[31]"

Fig. 5

The crystal structure of Li1-2x Mg x BiOS(x=0.1875). The corresponding MSD versus time curves obtained by AIMD simulations (simulation temperature is 525 K)[26]"

Fig. 6

The thermodynamic phase diagrams of NSC (a) and LSC (e). The diagrams of crystalline structures of NSC (b) and LSC (f). The potential energy surface of migration ions of NSC (c), (d) and LSC (g), (h) using BV method; (i) The Arrhenius curves of LSC and NSC by AIMD simulations, respectively[41]"

Fig. 7

(a) The workflow of constructing crystal structures through polyanionic groups. The entire crystal structures of Na3Y2Cl9 are obtained by combining (Y2Cl9)3- and Na+; (b) Three different crystalline of Na3Y2Cl9 designed by polyanionic groups, which space groups are Cc, P63 and R32, respectively; (c) The results of thermodynamic stability of Na3Y2Cl9 constructed by polyanionic groups. The Na3Y2Cl9 with R32 symmetry exhibit the lowest magnitude of energy above hull in virtual structures[44]"

Fig. 8

The atomic structure of amorphous LiSiON (a) and energy barrier of Li+ migration calculated by AIMD (b)[46]"

Fig. 9

(a), (b) The atom position distribution in (a) crystal LiAlCl4 and (b) amorphous LiAlCl2.5O0.75. Green, blue, yellow and red spheres represent Li, Al, Cl and O, respectively. The bright-yellow points signify the movement trajectories of Li+ in LiAlCl4 and LiAlCl2.5O0.75 during 100 ps; (c) The MSD of Li, Al, Cl and O in LiAlCl4 and LiAlCl2.5O0.75 during 100 ps obtained by AIMD; (d) The RDF curve of LiAlCl4 and LiAlCl2.5O0.75[48]"

Fig. 10

The core functions of ESM cloud toolkit"