First-Principles High-Throughput Screening Pipeline for Nonlinear Optical Materials: Application to Borates

Discovering new nonlinear optical (NLO) materials that require an optimization between multiple properties is a time-consuming and high-cost process. To speed up material development, a first-principles high-throughput screening pipeline for nonlinear optical materials (FHSP-NLO) that combines density functional theory (DFT) codes, linear and nonlinear optical property calculation codes, and data transformation and extraction codes has been developed for searching promising NLO materials from crystals collected in the Inorganic Crystal Scientific Structure Database (ICSD). The tests for a dozen of well-known NLO crystals covering deep-ultraviolet (DUV), UV, vis-NIR, and middle/far-infrared (M-F-IR) wavelength ranges verify the high-accuracy of FHSP-NLO. Subsequently, nearly 300 noncentrosym- metric borates are tested with FHSP-NLO. The screened deep-ultraviolet (DUV) NLO crystals are fully consistent with previously reported results. Besides, five crystals, whose NLO properties have not been reported, i.e., B2S2O9, Al4B6O15, HP-Na2B4O7, KB(SO3CI)(4), and H3BO3-3T, are identified as new promising NLO materials. Two hydrated borates [Ca2B5O9]center dot[H(OH)(2)] and Ca(B8O11(OH)(4)), whose second-harmonic generation (SHG) responses are primarily measured, are screened out and suggested to grow large-size crystals for further evaluation. FHSP-NLO provides a powerful and efficient tool to screen and reduce the total number of experiments necessary for searching NLO materials from reported crystals.