Cited by
- BibTex
- RIS
- TXT
Nonlinear optical (NLO) materials, with their unique wavelength conversion capabilities, play a crucial role in a wide range of scientific and industrial applications. Despite significant progress, the development of novel NLO materials, particularly those in the deep ultraviolet and mid-infrared regions, remains a challenge. Recent advancements in machine learning (ML) technologies have injected new momentum into materials science research. In this work, we present an integrated data platform incorporating advanced ML techniques, designed to drive the discovery and exploration of inorganic NLO materials. The platform currently includes about 1000 entries with their structures and key properties. Users can apply built-in ML models developed in our group for immediate predictions of NLO properties or train their own models based on specific research needs. Additionally, the platform provides access to the results of deep generative models, allowing users to retrieve newly generated virtual crystal structures, thus expanding the chemical space for NLO materials exploration. This platform not only provides reliable data support for researchers but also holds the potential to accelerate the discovery of novel NLO materials.
}, issn = {2617-8575}, doi = {https://doi.org/10.4208/cicc.2025.61.01}, url = {http://global-sci.org/intro/article_detail/cicc/24049.html} }Nonlinear optical (NLO) materials, with their unique wavelength conversion capabilities, play a crucial role in a wide range of scientific and industrial applications. Despite significant progress, the development of novel NLO materials, particularly those in the deep ultraviolet and mid-infrared regions, remains a challenge. Recent advancements in machine learning (ML) technologies have injected new momentum into materials science research. In this work, we present an integrated data platform incorporating advanced ML techniques, designed to drive the discovery and exploration of inorganic NLO materials. The platform currently includes about 1000 entries with their structures and key properties. Users can apply built-in ML models developed in our group for immediate predictions of NLO properties or train their own models based on specific research needs. Additionally, the platform provides access to the results of deep generative models, allowing users to retrieve newly generated virtual crystal structures, thus expanding the chemical space for NLO materials exploration. This platform not only provides reliable data support for researchers but also holds the potential to accelerate the discovery of novel NLO materials.