Microstructure–Defect Engineering of Spinels for Self‐Powered and Recoverable Near‐Infrared Mechanoluminescence

Different from traditional stimulus‐responsive luminescence systems, near‐infrared mechanoluminescence (NIR‐ML) materials have unique advantages in biomechanical sensing and deep‐tissue imaging. Their unique mechano‐optical conversion ability makes them promising candidates for emerging smart platforms; however, unclear mechanisms underlying self‐recoverable and self‐powered ML have hindered their further development toward practical applications. Here, we report a series of Cr 3+ ‐activated ML phosphors through spinel microstructure‐defect engineering in centrosymmetric hosts. By systematically regulating defect types, we clarify the distinct energy origins and mechanisms of self‐recoverable and self‐powered ML, both of which are governed by defect‐piezoelectric coupling effects. Here, Self‐recoverable ML mainly originates from intrinsic point defects associated with electron capture and release process during pre‐irradiation, while self‐powered ML arises from anti‐site defects associated with local structural distortion. Furthermore, by combining binary coding with convolutional neural network algorithms, a proof‐of‐concept NIR biomechanical monitoring and sensing model is designed to simulate luminescence during human biting. These findings elucidate the structure‐defect‐piezoelectric‐luminescence relationship and provide rational design principles for high‐performance NIR ML materials toward potential practical applications.