High-Performance Infrared Nonlinear Optical Crystals Discovery Guided by High-Throughput Computation, Machine...

Area of Science:

• Materials Science
• Optoelectronics
• Computational Chemistry

Background:

• Infrared nonlinear optical (NLO) materials are crucial for advanced laser and photonic technologies.
• Current material discovery is hindered by fragmented systems, long development times, and trial-and-error methods.

Purpose of the Study:

• To develop an integrated computational-experimental framework for accelerating the discovery of novel infrared NLO materials.
• To establish a comprehensive figure of merit (CFOM) for quantifying NLO material performance trade-offs.

Main Methods:

• Utilized first-principles high-throughput calculations and machine learning.
• Created a multidimensional dataset of 1807 non-centrosymmetric compounds.
• Developed a crystal graph neural network classifier with high predictive accuracy (AUC = 0.95).

Main Results:

• Uncovered key composition-structure-performance relationships for enhanced NLO properties.
• Identified 12 promising, previously unreported NLO material candidates.
• Experimentally validated HgAl2Q4 (Q=S, Se, Te) exhibiting wide band gaps and strong NLO responses.

Conclusions:

• The integrated framework significantly accelerates the discovery of high-performance optoelectronic materials.
• This approach provides a pathway to overcome limitations in traditional NLO material development.
• Identified novel defect-chalcopyrite compounds with significant potential for photonic applications.