# Polyacid Solutions as an Analogue of a Neural Network

## TL;DR

This paper explores how single-component polyacid solutions can mimic neural networks through charge fluctuations, offering a new approach to neuromorphic materials.

## Contribution

The novelty is proposing a single-component polyacid solution as a spontaneous neural network analogue without requiring multiple molecule types or direct cluster interactions.

## Key findings

- Fluctuations in charge distribution in polyacid solutions can lead to collective responses similar to neural networks.
- The system's behavior is modeled using the Poisson–Boltzmann equation to describe its neuromorphic properties.
- Such solutions could serve as prototypical materials for organic electronics and prebiological evolution studies.

## Abstract

Despite the increased interest in neuromorphic materials—a physical implementation of neural networks that could overcome the so-called von Neumann architecture’s limitations—most studies have been performed on the basis of systems specially constructed for this purpose. It has previously been shown that analogues of neural networks can spontaneously arise in solutions of hydrophilic polymers, but these systems involved molecules of different natures or required direct interaction between macromolecular clusters. The present paper proposes a theory that indicates the possibility of an analogue of neural network formation even in a single-component solution of a relatively weak polyacid. A model is suggested based on the account of heterogeneous distribution of polymer ionogenic groups within the volume leading to the fluctuations of electric fields and, as a result, to the local changes in the degree of ionisation of functional groups. Theoretical description of the system shows how it was reduced to a solution of the analogue based on the Poisson–Boltzmann equation. The results obtained showed that it is just fluctuations in the distribution of charges that provide the collective response of the system to external influences and serve as an argument in favour of analogy of such a solution within a neural network. The results are discussed in the context of a potential simple hydrophilic polymer system as a prototypical neuromorphic and evolving material that is relevant for organic electronics, metamaterials, and studies on prebiological evolution.