Neuromorphic properties of flexible carbon nanotube/polydimethylsiloxane nanocomposites

Neuromorphic materials are promising for fabricating artificial synapses for flexible electronics, but they are usually expensive and lack a good combination of electronic and mechanical properties. In this paper, low-cost flexible carbon nanotube/polydimethylsiloxane (CNT/PDMS) nanocomposites were prepared by solution processing. Their neuromorphic properties were studied as a function of PDMS macromolecular network structure. Specifically, the structural defects of the polymer network originating from intermolecular crosslinking reactions were tuned to tailor the electron transfer between carbon nanotubes, resulting in a recorded low switching power consumption (1.40 × 10^–10 W) and a high working bending radius of curvature (5 mm) compared to other organic, flexible neuromorphic materials. As-fabricated CNT/PDMS composites demonstrate robust performance for 10⁴ operating cycles under mechanical deformation. Emulation of synaptic functions was also presented, showing long-term potentiation (LTP) and long-term depression (LTD) characteristics. These results lay a foundation for networked polymer-based multifunctional nanocomposites for flexible neuromorphic electronic devices.