Congratulations to Postdoctoral Researcher Li Wang on the publication of his article in ACS Nano!

Recently, Pro. Kai Xiao's research group proposed a new strategy for simulating synaptic functions using the ionic potential relaxation effect in hydrogels. Their work, titled “Ionic Potential Relaxation Effect in a Hydrogel Enabling Synapse-Like Information Processing”, was published in the academic journal ACS Nano.

The human brain contains approximately 10¹⁵ synapses, which are responsible for both information storage and computation, making them the core components of brain information processing. Biological synapses regulate membrane potential (synaptic weight) through ion-selective transmembrane transport mediated by the postsynaptic membrane, thereby achieving storage and computation functions. This biological principle has inspired the development of artificial synaptic devices that use ions as information carriers, aiming to build low-energy, high-efficiency, and biocompatible brain-like intelligent systems and seamless human-machine interfaces. Soft and biocompatible ion-conductive hydrogels are ideal materials for constructing such devices, but controlling ion transport behavior in hydrogels to simulate postsynaptic electrical signals remains a major challenge.

In this study, we proposed the existence of an ionic potential relaxation effect in hydrogel devices, which can be used to simulate characteristic ion electrical signals related to synaptic functions, such as paired-pulse facilitation/inhibition (PPF/PPD), long-term potentiation/depression (LTP/LTD), and pulse-timing dependent plasticity. We also demonstrated the implementation of synapse-like information processing functions, including information storage, experiential learning, perception, and computation. The device structure, relaxation mechanism, and specific experimental results are as follows:

Figure 1. Ionic potential relaxation effect and ion transport control mechanism in sandwich-structured hydrogel (P-AXMY): (A) Bionic design and device structure schematic. (B) Device photo. (C) Chemical structure of the hydrogel. (D and E) Pressure (D) and current (E) responses to ion signals exhibiting relaxation behavior. (F) Potential distribution under pressure stimulus. (G) Ion hysteresis transport induced by pressure stimulus. (H) Simulated response signal under pressure stimulus. (I) Potential distribution under current stimulus. (J) Ion hysteresis transport induced by current stimulus. (K) Simulated response signal under current stimulus.

Figure 2. Pressure-stimulus-induced synaptic plasticity for simulating tactile perception: (A) Biological principle of skin tactile perception. (B) Simulation of the PPD effect. (C) Relationship between PPD index and pulse interval. (D) Simulation of the LTD effect.

Figure 3. Current-stimulus-induced synaptic plasticity for simulating learning and memory: (A) Biological basis of brain memory. (B) Simulation of the PPF effect. (C) Relationship between PPF index and pulse interval. (D) Simulation of the LTP effect. (E) Simulation of experiential learning. (F) Image storage and encryption functionality.

Figure 4. P-A₈₀M₂₀ for neuromorphic computing — handwritten digit classification: (A) Computational workflow. (B) Encoding sequence and response signal for digit “0.” (C) Classification results, showing an accuracy of 95%. (D) Relationship between accuracy and training epochs.

Figure 5. Flexible design of the device: (A) Flexible design and physical image. (B and C) Device photos under 180° bending (B) and 100% stretching (C). (D) Relationship between PPF index and pulse interval under four states. (E) LTP electrical signals under four states.

Summary: We have developed a cost-effective, flexible, and universal new artificial synaptic device that uses the ionic potential relaxation effect in sandwich-structured hydrogels to use ions as information carriers. The special ion transport behavior induced by the middle anion-selective layer, including anion-selective permeation driven by external stimulation and cation diffusion hysteresis after stimulation, contributes to the ionic potential relaxation effect. This relaxation behavior closely resembles the decay dynamics of postsynaptic membrane potential. Based on this, the device can simulate multiple electrical signal patterns in synapses and further achieve synapse-like information processing functions, including tactile perception, learning, memory, and computation. This study helps overcome the construction bottlenecks of ionic neuromorphic devices in terms of flexibility, stability, and large-scale integration, providing new methods and ideas for building advanced brain-like intelligent systems.

This paper was authored by Dr. Wang Li (Postdoctoral Researcher) as the first author, with Associate Professor Xiao Kai as the corresponding author, and it was supported by the National Key R&D Program, the National Natural Science Foundation, Guangdong Provincial Key Laboratory Projects, the Shenzhen Science and Technology Innovation Committee, and the China Postdoctoral Science Foundation.

Link: https://doi.org/10.1021/acsnano.4c09154