淡马锡（Temasek）：2025人工智能（AI）在材料发现领域的应用研究报告（英文版）（22页）.pdf

1、12AI is Unlocking the Door to Endless Possibilities in Materials DiscoveryFrom generating novel material candidates to exploring vast chemical spaces,AI is ushering in a transformative leap in material science.all in a matter of daysThrough extraordinary speedand precise calculations:Accelerated Dis

2、coveryMore accurate predictions:Reduced Costand WasteStarting with the desired material properties:Targeted Discovery and DesignCharacterise material properties based on experimental validationGenerate novel material candidatesDesign and propose synthesis and experimental validation pathwaysOptimise

3、 synthesis of desired materialsAI can.3AI MethodsChallengesPossibilitiesAI-driven materials discovery is not without its challenges:Among AI methods,these offer promising potential noting their neural network backbone.Innovation in the materials science field could revolutionise many other industrie

4、s.Novel battery chemistries could vastly improve energy storage,making renewables a stable,24/7 power source.Re-engineered photovoltaic materials may boost solar efficiency,reducing costs and land use.At the frontier,discovering materials that endure extreme conditions in nuclear reactors could enab

5、le safer,cleaner power.AI is accelerating this shift by navigating complex material classes like high-entropy alloys and metal-organic frameworks.Powered by large,high-quality datasets and grounded in a deep understanding of chemistry and physics,AI is becoming an essential tool in materials discove

6、ry.Existing synthesis methods and tools are slow and limitedData is scarce,noisy,and complexMultiple scales of interaction make modelling a complicated processDiffusion ModelsGraph Neural NetworksGenerative Adversarial Networks4 4MATERIALS DISCOVERY 2.0:THE AI REVOLUTION015Material science,in partic