AI & Materials
Cases where AI was actually used in materials research and scientific experiments.Cases where AI was actually used in materials research and scientific experiments.Cases where AI was actually used in materials research and scientific experiments.Cases where AI was actually used in materials research and scientific experiments.Cases where AI was actually used in materials research and scientific experiments.Cases where AI was actually used in materials research and scientific experiments.Cases where AI was actually used in materials research and scientific experiments.Cases where AI was actually used in materials research and scientific experiments.
매일 00·06·12·18시(한국시간)에 웹에서 새 소식을 찾습니다. AI 활용 방법과 결과가 확인된 새 사례만 게시합니다.
Checking notifications…
A Nature Communications paper reported that combining a machine learning design paradigm with low-temperature thermal-force coupling carbonization controlled the microstructure of hard carbon and improved sodium storage performance. It is meaningful as an example of AI use in battery materials development because it combines a low-energy process with ML-based design.
A Nature Machine Intelligence paper proposed using a large language model as an optimizer for experimental discovery while combining uncertainty calibration to improve the reliability of candidate proposals. Strong performance was reported on catalyst spaces and materials and molecular design tasks, drawing attention to its potential use in materials search and autonomous experimental design.
A study published in npj Computational Materials on 2026-09-03 is a computational materials study that evolves transition-state search using an agentic LLM. It shows a trend in chemistry and materials computation in which AI agents assist search strategies.
A Communications Materials paper addressed the bottleneck of defining the search space in autonomous closed-loop materials discovery by dynamically adjusting it with an LLM. It can be seen as an example of AI managing the search range at the intersection of autonomous experimental workflows and materials informatics.
An MIT research team proposed CrysVCD, which combines valence constraints with a generative materials model to reduce unstable candidates before they are generated. It was reported that about 70% of the computationally generated products passed the lattice dynamical stability assessment.