Identifying the charge density and dielectric environment of graphene using Raman spectroscopy and deep learning

Abstract

The impact of the environment on graphene’s properties such as strain, charge density, and dielectric environment can be evaluated by Raman spectroscopy. These environmental interactions are not trivial to determine...

Discriminative feature analysis of dairy products based on machine learning algorithms and Raman spectroscopy

“…The ELM algorithm has obvious advantages in computational speed and high recognition accuracy. The CNN algorithm has its own feature extraction ability, the highest recognition accuracy and moderate computation time ( Chen et al, 2022 ).…”

Recent Progresses in Machine Learning Assisted Raman Spectroscopy

“…In addition, parameters such as peak position and peak width from Raman spectra of graphene are frequently used to evaluate the strain and charge doping levels. [103,104] Chen et al [105] used CNN to classify graphene samples with slightly different charge densities or dielectric environments and enhanced the spectra data by adding noise and peak shifting. Figure 1a,b,c illustrates the prediction of graphene doping levels using a CNN model, and the flow chart of this experimental design is shown in Fig- ure 1d.…”

“…Experiments showed that the CNN model can classify the Raman spectra of graphene with different charge doping levels with 99% accuracy and even detect subtle differences in the spectra of graphene on SiO 2 and graphene on silanized SiO 2 . [105] Machado et al [106] proposed two approaches, one is a deep neural network with an autoencoder architecture, and another consists of a fully convolved autoencoder. They were used to remove noise from Raman spectra and improve graphene spectral data quality.…”

“…Adapted with permission. [105] Copyright 2022, Royal Society of Chemistry. [108] Copyright 2022, ACS Publications.…”

“…After preprocessing, five different (conventional) ML models and four different deep learning models were implemented. Adapted with permission [105]. Copyright 2022, Royal Society of Chemistry.…”