Defect-engineered hydrogen-terminated diamond optoelectronic synapses for UV-driven neuromorphic computing
Abstract
Bioinspired neuromorphic computing offers an energy-efficient route to information processing beyond the von Neumann architecture. Wide-bandgap semiconductors such as diamond combine high carrier mobility with outstanding thermal and chemical robustness, yet the potential of hydrogen-terminated diamond (H-diamond) in neuromorphic computing remains largely unexplored. Here, we report an H-diamond-based optoelectronic synaptic transistor realized by engineering bulk defects, hydrogen-induced surface states, and interface traps. Under 365 nm ultraviolet (UV) illumination, the device displays tunable persistent photoconductivity governed by defect-assisted carrier trapping and delayed release, enabling the emulation of essential synaptic behaviors and logic operations. Furthermore, a spiking neural network constructed from the experimentally measured conductance states attains a recognition accuracy of 84.95% on the Fashion-MNIST dataset. This work establishes a foundational benchmark for diamond-based neuromorphic devices and underscores the promise of H-diamond for robust, UV-sensitive information processing systems.
Fields of Study & Subjects
Advanced Memory and Neural Computing
Neural Networks and Reservoir Computing
Quantum-Dot Cellular Automata
Publication details
- Journal / source
- Journal of Colloid and Interface Science
- Publisher
- Elsevier BV
- Volume
- 722
- Pages
- 140876–140876
- Publication date
- 3 June 2026
- Language
- EN