Hydrophobically gated memristive nanopores for neuromorphic applications

Gonçalo Paulo,Ke Sun,Giovanni di Muccio,Alberto Gubbiotti,Blasco Morozzo della Rocca,Jia Geng,Giovanni Maglia,Mauro Chinappi,Alberto Giacomello

DOI: https://doi.org/10.1038/s41467-023-44019-y

2023-11-28

Abstract:Brain-inspired computing has the potential to revolutionise the current von Neumann architecture, advancing machine learning applications. Signal transmission in the brain relies on voltage-gated ion channels, which exhibit the electrical behaviour of memristors, resistors with memory. State-of-the-art technologies currently employ semiconductor-based neuromorphic approaches, which have already demonstrated their efficacy in machine learning systems. However, these approaches still cannot match performance achieved by biological neurons in terms of energy efficiency and size. In this study, we utilise molecular dynamics simulations, continuum models, and electrophysiological experiments to propose and realise a bioinspired hydrophobically gated memristive nanopore. Our findings indicate that hydrophobic gating enables memory through an electrowetting mechanism, and we establish simple design rules accordingly. Through the engineering of a biological nanopore, we successfully replicate the characteristic hysteresis cycles of a memristor \tr{and construct a synaptic device capable of learning and forgetting}. This advancement offers a promising pathway for the realization of nanoscale, cost- and energy-effective, and adaptable bioinspired memristors.

Mesoscale and Nanoscale Physics,Soft Condensed Matter

What problem does this paper attempt to address?

The problem that this paper attempts to solve is to achieve more efficient and miniaturized storage and processing units in neuromorphic computing. Specifically, using molecular dynamics simulations, continuous models, and electrophysiological experiments, the researchers proposed and implemented a bio - inspired hydrophobically gated memristive nanopore (HyMNs). This nanopore realizes the memory function through the electrowetting mechanism and follows simple rules in design. By modifying a biological nanopore (FraC), the researchers successfully replicated the characteristic hysteresis loop of the memristor and constructed a synaptic device capable of learning and forgetting. This progress provides a promising path for realizing bionic memristors with nano - scale, high cost - and energy - efficiency, and strong adaptability.

### Core problems of the paper:
1. **Improving the energy efficiency and size of neuromorphic computing**: Although the existing semiconductor - based neuromorphic technologies have demonstrated their effectiveness in machine - learning systems, they still cannot be compared with biological neurons in terms of energy efficiency and size. This research aims to solve this problem by developing new memristive nanopores.
2. **Realizing nano - scale bionic memristors**: Through molecular dynamics simulations and experimental verification, the researchers proposed a hydrophobically gated memristive nanopore (HyMNs) based on the electrowetting mechanism and successfully realized the function of this nanopore in practice.

### Key technologies and methods:
- **Molecular dynamics simulations**: Used to reveal the microscopic mechanisms inside the nanopore, especially how electrowetting realizes the memory function.
- **Electrophysiological experiments**: Verify theoretical predictions through experiments and show the conductance behavior of hydrophobically gated nanopores.
- **Biological nanopore modification**: Modify biological nanopores (such as FraC) through point - mutation techniques to make them have hydrophobically gated characteristics.

### Main achievements:
- **Successful implementation of hydrophobically gated memristive nanopores**: Through molecular dynamics simulations and experimental verification, it is proved that hydrophobically gated memristive nanopores can realize the memory function through the electrowetting mechanism.
- **Construction of synaptic devices**: Based on the modified biological nanopore, a synaptic device capable of learning and forgetting is constructed, showing its potential in neuromorphic computing.
- **Proposal of design criteria**: Four design criteria are proposed to guide how to design efficient hydrophobically gated memristive nanopores, including the optimization of parameters such as pore size, length, and hydrophobicity of materials.

### Potential impacts:
- **Promoting the development of neuromorphic computing**: This new type of memristive nanopore is expected to play an important role in future neuromorphic computing, especially in surpassing the existing semiconductor - based technologies in terms of energy efficiency and size.
- **Promoting interdisciplinary research**: Combining the knowledge of biology, physics, and engineering, it provides new ideas and methods for developing new nano - fluidic devices.

In conclusion, by proposing and implementing hydrophobically gated memristive nanopores, this paper provides a new and efficient solution in the field of neuromorphic computing, which has important scientific and application values.