Lab-on-device investigation of phase transition in MoOx semiconductors

Precise tuning of phase transition material properties enables multifunctional devices for information processing and energy conversion, but controlling on-device phase transitions and monitoring microscopic mechanisms remains challenging. Here, we develop a lab-on-device system for molybdenum oxide to probe operando hydrogenation mechanisms through in situ electrical and spectral characterization with density functional theory calculations, revealing threshold-driven proton dynamics that govern the transition between nonvolatile memory operation and catalytic hydrogen evolution. Moderate proton intercalation (flux < 10¹⁷cm^-2) achieves a five-order conductance modulation under ambient conditions via polaron formation and stoichiometric optimization (H/Mo up to 22%, Mo/O approaching ideal ratios), outperforming oxygen vacancy engineering. Beyond this threshold (flux ~10¹⁷ cm^-2), intensive proton intercalation triggers electric-to-chemical energy conversion, directly linking proton history to catalytic activity. Leveraging these principles, we achieve nonvolatile electrochemical memory with linear synaptic and accumulative neuronal functionalities, and demonstrate an all electrochemical random-access memory neural network hardware that executes memory-efficient rank-order coding for sparse signals even under noisy conditions.