Computational screening and experimental validation of promising Wadsley-Roth Niobates

TL;DR

This study used high-throughput DFT screening to identify new Wadsley-Roth niobates, synthesized a promising candidate, and demonstrated its potential for high-capacity lithium-ion battery applications.

Contribution

The paper expands the known set of WR niobates from less than 30 to over 1300 using computational screening and validates a new compound experimentally with superior battery performance.

Findings

Successfully synthesized and validated a new WR niobate, MoWNb24O66.

Achieved high lithium diffusivity and capacity in the new material.

Provided a large computational dataset to guide future experimental discovery.

Abstract

The growing demand for efficient, high-capacity energy storage systems has driven extensive research into advanced materials for lithium-ion batteries. Among the various candidates, Wadsley-Roth (WR) niobates have emerged as a promising class of materials for fast Li+ storage due to rapid ion diffusion within their ReO3-like blocks in combination with good electronic conductivity along the shear planes. Despite the remarkable features of WR phases, there are presently less than 30 known structures which limits identification of structure-property relationships for improved performance as well as the identification of phases with more earth-abundant elements. In this work, we have dramatically expanded the set of potentially (meta)stable compositions (with $\Delta$ Hd < 22 meV/atom) to 1301 (out of 3283) through high-throughput screening with density functional theory (DFT). This large space of compound was generated through single- and double-site substitution into 10 known WR-niobate prototypes using 48 elements across the periodic table. To confirm the structure predictions, we successfully synthesized and validated with X-ray diffraction a new material, MoWNb24O66. The measured lithium diffusivity in MoWNb24O66 has a peak value of 1.0x10-16 m2/s at 1.45 V vs. Li/Li+ and achieved 225 mAh/g at 5C. Thus a computationally predicted phase was realized experimentally with performance exceeding Nb16W5O55, a recent WR benchmark. Overall, the computational dataset of potentially stable novel compounds and with one realized that has competitive performance provide a valuable guide for experimentalists in discovering new durable battery materials.