A wearable neuromorphic electronic system, that can learn and interpret the non-structural biometric information at extremely low-power, has been brought great attention because of its applicability as the intelligent device that can easily attach onto the human body or any rough surface [1-3]. With this reason, organic-based artificial synaptic devices have been proposed as a potential candidate for wearable neuromorphic applications due to its inherent mechanical flexibility and the material (or device form) variability for the desired functionalities [2,3]. In this study, we designed 1D fiber-shaped multi-synapses comprising ferroelectric organic transistors fabricated on a 100 mm Ag wire and utilized them as multi-synaptic channels in an e-textile neural network for wearable neuromorphic applications [4]. The device mimics diverse synaptic functions, including short- and long-term plasticity with 80 states and spike rate- and timing-dependent plasticity. It exhibited excellent reliability even under 6,000 repeated input stimuli and mechanical bending stress. Various NOR-type textile arrays are formed simply by cross-pointing 1D synapses with Ag wires, where each output from individual synapse can be integrated and propagated without undesired leakage. Notably, the 1D multi-synapses achieved up to ~90% and ~70% recognition accuracy for MNIST and electrocardiogram patterns, respectively, even in a single-layer neural network, and almost maintained regardless the bending conditions.
References
[1] Park, Y.; Park, M.-J.; Lee, J.-S. Adv. Funct. Mater. 2018, 1804123
[2] van de Burgt, Y.; Lubberman, E.; J. Fuller, E.; T. Keene, S.; C. Faria, G.; Agarwal, S.; J. Marinella, M.; Talin, A. A.; Salleo A. Nat. Mater. 2017, 16, 414-418
[3] Jang, S.; Jang, S.; Lee, E.-H.; Kang, M.; Wang, G.; Kim, T.-W. ACS Appl. Mater. Interfaces 2019, 11, 1071-1080
[4] *S. Ham et al. Sci. Adv. 2020, 6, eaba1178 [Published]