2-Aminobenzazolium chlorocuprate(ii) complexes for memristive devices: structural studies, synaptic plasticity, and random number generation

Abstract

Hybrid organic–inorganic materials are promising candidates for next-generation memristive devices due to their structural tunability and rich electronic properties. Here, we investigate a series of three 2-aminobenzazolium chlorocuprate(II) complexes, based on benzothiazole (THIA), benzoxazole (OXA), and benzimidazole (IMI), to elucidate the relationship between molecular structure, electronic properties, and memristive performance. Single-crystal X-ray diffraction reveals distinct coordination geometries of the chlorocuprate anions, ranging from square planar to tetrahedral and trigonal bipyramidal, governed by heteroatom substitution (S, O, N) in the organic cations. Comprehensive spectroscopic (UV-vis-NIR, FTIR, XAS), morphological, and density functional theory (DFT) analyses show that these materials are p-type semiconductors with band gaps of ∼2.2–2.4 eV and electronic structures characterised by strong coupling between inorganic and organic components. Thin-film devices fabricated in an ITO|complex|Cu architecture exhibit reproducible bipolar resistive switching with pinched hysteresis loops and stable cycling endurance. Electrical, impedance, and noise spectroscopy collectively indicate that switching is governed by an interfacial, electronically driven mechanism involving charge trapping/detrapping and modulation of the Schottky barrier at the metal/semiconductor interface, rather than filamentary conduction or ion migration. The interfacial, noise-dominated switching dynamics give rise to stochastic conductance fluctuations that can be harnessed for hardware random number generation. The heteroatom-dependent electronic structure critically controls device performance: the sulphur-containing THIA system shows the strongest memristive response and synaptic plasticity due to enhanced hybridisation and higher trap density, whereas OXA exhibits weaker behaviour and IMI intermediate characteristics supported by extended hydrogen-bonding networks. The devices demonstrate key neuromorphic functionalities, including potentiation/depression and spike-timing-dependent plasticity, as well as stochastic conductance fluctuations enabling hardware-based random number generation. These findings establish a clear structure–electronic structure–device property relationship and highlight heteroatom engineering as an effective strategy for tuning interfacial memristive switching in hybrid materials.