# Identification of Sarin Simulant DMMP Based on a Laminated MOS Sensor Using Article Swarm Optimization-Backpropagation Neural Network

## TL;DR

Researchers developed a sensor that can detect a Sarin simulant using a combination of materials and a neural network, improving detection accuracy.

## Contribution

A novel laminated MOS sensor with optimized neural network for identifying dimethyl methylphosphonate is introduced.

## Key findings

- The laminated MOS sensor showed good sensing performance for dimethyl methylphosphonate under temperature modulation.
- The neural network improved the identification accuracy of the sensor with increasing gas concentration.
- The sensor can effectively determine if a test gas contains dimethyl methylphosphonate.

## Abstract

A Pt@CeLaCoNiOx/Co@SnO2 laminated MOS sensor was prepared using Co@SnO2 as the gas-sensitive film material and Pt@CeLaCoNiOx as the catalytic film material. The sensor was verified to exhibit good sensing performances for dimethyl methylphosphonate, a simulant of Sarin, under a temperature modulation, and characteristic peaks appeared in the resistance response curves only for dimethyl methylphosphonate. The Article Swarm Optimization-Backpropagation Neural Network had a good ability to identify the resistance response data of dimethyl methylphosphonate. The identification accuracy increased as the concentration of dimethyl methylphosphonate increased. This scheme can effectively identify whether the test gas contained dimethyl methylphosphonate or not, which provided a reference for achieving the high selectivity of the MOS sensor for Sarin.