Microplastics, which are smaller than 5 mm, are widespread and pose growing environmental and health concerns. Raman spectroscopy combined with neural networks offers a promising solution for their identification, but the accuracy is often limited by overlapping signals in complex samples and low-quality data collected in real-world environments.

In a study published in Analytical Chemistry, a research team led by Prof. GAO Xiaoming and LIU Kun from the Hefei Institutes of Physical Science of the Chinese Academy of Sciences developed a cascaded neural network for analyzing mixed microplastic Raman spectra.

The cascaded neural network consists of three key components: a channel and spatial attention module for enhanced feature extraction, a reconstruction and classification module for robust spectral analysis, and a hybrid physical loss function to guide training and improve model convergence. It can efficiently reconstruct spectra, classify different microplastic types, and separate overlapping signals.

This network was tested on 20-25 mixed microplastic samples collected under varying conditions. The results showed that under low laser power (~50 mW) and short integration time (~3 seconds), classification accuracy increased from 52% using conventional methods to 91% with the new network.

This work provides a promising solution for accurate microplastic identification in complex and low-quality measurement environments, supporting more reliable environmental monitoring and research.