Ultrafast Photodetectors for Secure Communication, Logic Processing, and Machine Learning-Assisted Optical Material Classification

Despite significant advancements in high-speed photodetection, existing ultrafast photodetectors remain constrained by fundamental limitations in responsivity, dynamic range, and signal integrity, particularly for applications requiring secure communication and adaptive processing. An ultrafast photodetector that captures optical transients on nanosecond timescales, far surpassing the ≈µs speed limitations of conventional photosensors is presented. Achieving a 61 ns response time (33 ns halfwidth) via a coplanar Schottky-capacitive design, this device leverages instantaneous photo-induced capacitance modulation to generate transient current spikes, effectively bypassing RC time-constant limitations. The resulting transient detection mode offers a large linear dynamic range (>93 dB) and a 6000% enhanced sensitivity compared to conventional steady-state photocurrent operation. This ultrafast speed and sensitivity are harnessed for secure high-speed data transmission and logic processing via an electro-optical modulation scheme that ensures reliable, tamper-resistant information encoding. Furthermore, the photodetector's nonlinear, bias-tunable photoresponse captures distinct material-dependent optical signatures, allowing machine learning classification of metals, insulators, and semiconductors with over 82% accuracy. By integrating ultrafast optical detection with secure communication and logic processing capabilities, this photodetector platform represents a transformative solution for next-generation robotics, automation, intelligent sensing, and high-security materials characterization.