Chemically Reconfigurable Neuromorphic Electronics for Neurochemical Homeostasis
Abstract Neuromorphic devices excel at emulating synaptic and neuronal computation, yet their states are predominantly programmed by electrical signals, limiting their ability to directly perceive and adapt to the molecular signals that govern biological neural function. Here we introduce chemical reconfiguration into neuromorphic electronics and develop a chemically reconfigurable neuromorphic healer (CR-NeuroHealer), in which molecular inputs directly reprogram synaptic plasticity to couple neurochemical perception with adaptive regulation. The CR-NeuroHealer responds distinctly to dopamine (DA) and Fe 3+ , enabling the chemical environment to reconfigure the device between healthy-like and pathological synaptic states. When interfaced with neuron-like PC-12 cells, the device establishes a biohybrid synapse through molecular communication, sensing presynaptic DA while providing adaptive chemical feedback. Pathological iron accumulation weakens synaptic transmission and triggers deferiprone (DFP) release, which chelates excess iron and restores dopamine-responsive plasticity, completing a chemically reconfigurable regulatory cycle. We further demonstrate this chemical perception in vivo using a miniaturized neuromorphic OECT implanted in a 6-hydroxydopamine-induced Parkinsonian mouse model. The implanted devices respond to endogenous DA modulation and pathological iron accumulation with distinct synaptic-like outputs that discriminate neurochemical states and guide DFP-mediated intervention, accompanied by recovery of motor performance and device responses. These results establish chemical reconfiguration as a neuromorphic design principle for therapeutic brain–computer interfaces that operate through the molecular language of the nervous system.
Authors
- Lianhui Wang (ORCID: https://orcid.org/0000-0001-9030-9172)
- Li Hua Gao (ORCID: https://orcid.org/0000-0003-2465-5784)
- Ting Wang (ORCID: https://orcid.org/0000-0002-9143-7229)
- Lixing Weng (ORCID: https://orcid.org/0000-0002-7719-892X)
- Shijie Peng (ORCID: https://orcid.org/0009-0001-4418-0045)
- Ming Wang (ORCID: https://orcid.org/0000-0003-0976-9871)
- Jianlong Ji (ORCID: https://orcid.org/0000-0002-8705-3046)
- Gang Song (ORCID: https://orcid.org/0000-0003-2311-4097)
- Xiao Hong Zhao (ORCID: https://orcid.org/0009-0002-0539-8639)
- Xu Qian
- Haochen Zou
- Ting Li
- Zhuoye Tang
- Lv Xia
- fan zhang
- Weimin Hao
- Rongyan Zhou
- Zhiwei Chen
Institutions
- Fudan University (CN)
- Nanjing University of Posts and Telecommunications (CN)
- The Synergetic Innovation Center for Advanced Materials (CN)
- Taiyuan University of Technology (CN)
Publication Details
- Journal
- Research Square
- Published
- 2026-10-29
- DOI
- https://doi.org/10.21203/rs.3.rs-11182543/v1
- Primary Topic
- Advanced Memory and Neural Computing
- Type
- preprint
Funders
- National Natural Science Foundation of China