Silicon circuitry has dominated the semiconductor industry for decades but is constrained in its power efficiency by the Fermi-Dirac distribution of electron energies. Electron-correlated transition metal oxides exhibiting metal-to-insulator transitions (MITs) are excellent candidates for energy-efficient computation, which can further emulate the spiking behavior of biological neural circuitry. We demonstrate that β′-Cu_xV₂O₅ exhibits a pronounced nonlinear response to applied temperature, voltage, and current, and the response can be modulated as a function of Cu stoichiometry. We show that polaron oscillation, coupled to the real-space shuttling of Cu ions across two adjacent sites, underpins the MIT of this material. These results reveal the interplay between crystal structure distortions and electron correlation in underpinning the metal-insulator transition of a strongly correlated system. The utilization of coupled cation diffusion and polaron oscillation further demonstrates a means of using ionic vectors to obtain highly nonlinear conductance switching as required for neuromorphic computing.

硅电路已经在半导体行业中占据了主导地位，但其功率效率受到电子能量的费米-狄拉克分布的限制。表现出金属到绝缘体转变（MIT）的电子相关过渡金属氧化物是进行节能计算的极佳选择，可以进一步模拟生物神经电路的尖峰行为。我们证明β'铜_X V ₂ Ø ₅对施加的温度，电压和电流表现出明显的非线性响应，并且可以根据Cu化学计量来调节响应。我们表明，极化子振荡与跨两个相邻位点的Cu离子的实际空间穿梭相结合，是这种材料的MIT的基础。这些结果揭示了在支撑强相关系统的金属-绝缘体过渡过程中，晶体结构畸变与电子相关之间的相互作用。利用耦合的阳离子扩散和极化子振荡进一步证明了一种使用离子载体来获得神经形态计算所需的高度非线性电导切换的方法。