Orientation-Controlled In-Material Physical Reservoirs Based on Polymer-Wrapped Carbon Nanotube Nanofilms
Semiconducting polymer poly(m-phenylenevinylene-co-2,5-dioctoxy-p-phenylenevinylene) (PmPV)-wrapped single-walled carbon nanotube (SWCNT-PmPV) nanofilms were fabricated using the Langmuir-Blodgett (LB) method to investigate the effects of structural orientation on their electrical and reservoir computing properties. Polarized ultraviolet-visible absorption spectroscopy and atomic force microscopy revealed that repeated compression-expansion cycles during LB deposition produced highly oriented SWCNT nanofilms. The oriented films exhibited an anisotropic electrical response, pronounced hysteretic current-voltage characteristics, and enhanced generation of higher-order harmonics, while water adsorption further enhanced the current and nonlinear electrical response. In waveform generation tasks, the advantage of the highly oriented films was most pronounced for the sawtooth and sin(2ω) targets, which involve more complex waveform features and frequency components. Additional delayed-input reconstruction and nonlinear autoregressive moving average (NARMA) 2 tasks revealed enhanced fading memory and nonlinear temporal processing under humidified conditions, with the highly oriented film exhibiting a larger memory capacity and lower NARMA2 error than the less-oriented film. These results indicate that structural orientation, together with humidity-dependent electrical dynamics, contributes to the nonlinear and temporal information-processing capabilities of SWCNT-PmPV nanofilms. These findings provide a practical strategy for modulating in-material reservoir computing performance through structural control of anisotropic material networks.
Authors
- Yuya Ishizaki (ORCID: https://orcid.org/0000-0002-1243-0441)
- Jun Matsui (ORCID: https://orcid.org/0000-0003-4767-4507)
- Hinako Ebe (ORCID: https://orcid.org/0000-0001-6815-9833)
- Shusaku NAGANO
- Shunei Kimura
- Jumpei Igarashi
Institutions
- Yamagata University (JP)
- Rikkyo University (JP)
- Shinshu University (JP)
Publication Details
- Journal
- ACS Applied Materials & Interfaces
- Published
- 2026-11-14
- DOI
- https://doi.org/10.1021/acsami.6c17055
- Primary Topic
- Advanced Memory and Neural Computing
- Type
- article
- Field-Weighted Citation Impact
- 0.00