Zhenbin Wang

Insights into the Performance Limits of the Li7P3S11 Superionic Conductor: A Combined First-Principles and Experimental Study

The Li7P3S11 glass-ceramic is a promising superionic conductor electrolyte (SCE) with an extremely high Li(+) conductivity that exceeds that of even traditional organic electrolytes. In this work, we present a combined computational and experimental investigation of the material performance limitations in terms of its phase and electrochemical stability, and Li(+) conductivity. We find that Li7P3S11 is metastable at 0 K but becomes stable at above 630 K (∼360 °C) when vibrational entropy contributions are accounted for, in agreement with differential scanning calorimetry measurements. Both scanning electron microscopy and the calculated Wulff shape show that Li7P3S11 tends to form relatively isotropic crystals. In terms of electrochemical stability, first-principles calculations predict that, unlike the LiCoO2 cathode, the olivine LiFePO4 and spinel LiMn2O4 cathodes are likely to form stable passivation interfaces with the Li7P3S11 SCE. This finding underscores the importance of considering multicomponent integration in developing an all-solid-state architecture. To probe the fundamental limit of its bulk Li(+) conductivity, a comparison of conventional cold-press sintered versus spark-plasma sintering (SPS) Li7P3S11 was done in conjunction with ab initio molecular dynamics (AIMD) simulations. Though the measured diffusion activation barriers are in excellent agreement, the AIMD-predicted room-temperature Li(+) conductivity of 57 mS cm(-1) is much higher than the experimental values. The optimized SPS sample exhibits a room-temperature Li(+) conductivity of 11.6 mS cm(-1), significantly higher than that of the cold-pressed sample (1.3 mS cm(-1)) due to the reduction of grain boundary resistance by densification. We conclude that grain boundary conductivity is limiting the overall Li(+) conductivity in Li7P3S11, and further optimization of overall conductivities should be possible. Finally, we show that Li(+) motions in this material are highly collective, and the flexing of the P2S7 ditetrahedra facilitates fast Li(+) diffusion.

Deep neural networks for accurate predictions of crystal stability

Predicting the stability of crystals is one of the central problems in materials science. Today, density functional theory (DFT) calculations remain comparatively expensive and scale poorly with system size. Here we show that deep neural networks utilizing just two descriptors—the Pauling electronegativity and ionic radii—can predict the DFT formation energies of C3A2D3O12 garnets and ABO3 perovskites with low mean absolute errors (MAEs) of 7–10 meV atom−1 and 20–34 meV atom−1, respectively, well within the limits of DFT accuracy. Further extension to mixed garnets and perovskites with little loss in accuracy can be achieved using a binary encoding scheme, addressing a critical gap in the extension of machine-learning models from fixed stoichiometry crystals to infinite universe of mixed-species crystals. Finally, we demonstrate the potential of these models to rapidly transverse vast chemical spaces to accurately identify stable compositions, accelerating the discovery of novel materials with potentially superior properties.Crystal stability prediction is of paramount importance for novel material discovery, with theoretical approaches alternative to expensive standard schemes highly desired. Here the authors develop a deep learning approach which, just using two descriptors, provides crystalline formation energies with very high accuracy.

Correction to Insights into the Performance Limits of the Li7P3S11 Superionic Conductor: A Combined First-Principles and Experimental Study

Author(s): Chu, Iek-Heng; Nguyen, Han; Hy, Sunny; Lin, Yuh-Chieh; Wang, Zhenbin; Xu, Zihan; Deng, Zhi; Meng, Ying Shirley; Ong, Shyue Ping