Efficient Synaptic Emulation and Ultralow Power Digital-Analog Conversion in Cellulose-Based Neural Devices through Molecular Polarization

Abstract Efficient hardware-cell communication is crucial in understanding cellular states and controlling cells, serving as a crucial pathway in advancing next-generation human-machine interfaces. Here, we propose an energy-efficient neural device based on natural cellulose, addressing limitations in conventional interface communication hardware, particularly concerning material biocompatibility and biological signal matching. The cellulose-based device effectively emulates the plasticity of biological synaptic connection and exhibits learning behavior under continuous pulse stimuli as low as 10 mV. Significantly, it demonstrates exceptional digital-to-analog conversion performance with a minimal power consumption of 0.1 nJ, facilitating efficient interface biological signal matching. Furthermore, a molecular-level model is introduced to elucidate the rotation of intramolecular polar bonds in cellulose induced by electrical stimulation. This rotation alters the material's relative dielectric constant, unveiling the digital-to-analog conversion ability and neuro-like behavior. Moreover, the transparent cellulose thin film serves as both a dielectric layer and mechanical support, enabling the device to maintain functional stability under various curvatures. The flexible and biocompatible cellulose-based neural device in this work not only efficiently emulates synapses but also, with its low-power signal conversion, holds promise for effective biological signal matching in brain-machine interface applications.