Photo-Enhanced Output in Memdiodes Based on Hybrid Materials for Artificial Neural Network

We report that memdiodes (MDs) have been created based on metal halide-embedded polymeric hybrids in ambient conditions, presenting unusual photo-enhanced electronic transport behaviors. The hybrid materials show nonconventional optoelectronic properties, providing alternatives to traditional semiconductors such as silicon. The blending of inorganic metal halides and organic polymers combined with an insulating polymer membrane facilitates resistive memory and diode behaviors. Two-terminal memristive devices are made of a hybrid photoactive layer of CuCl₂ blended with poly(ethylene glycol) (PEG) and a layer of poly(methyl methacrylate) (PMMA) deposited on substrates to form a pn-junction memdiode. Thin films were characterized using ATR-FTIR, SEM, UV-Vis-NIR microspectrophotometer, and IV curves. Cyclic voltage sweepings present polarity-related current curves, which manifest the MD characteristics as the current stays near zero in negative voltages (reverse), but significantly increases in positive voltages (forward). Cyclic IV curves also present a moderate hysteresis, featuring a memory device. Surprisingly, the current magnitude increases by 4 times when the light radiation turns on, implicating that the electronic output is significantly enhanced by photoexcitation. The resistive transport mechanism is hypothetically attributed to electron-ion couplings, where both electronic mobility and ionic-hopping contribute to carrier transport as electrons migrate in an ionic “train”, where dielectric dipoles induce IV hysteresis. All materials in the MD devices are bio-compatible, stable, and flexible, potentially applicable to bio-electronic circuits, artificial neuromorphic synapses, and brain-inspired quantum computing.