Dual-Principal Component Analysis of the Raman Spectrum Matrix to Automatically Identify and Visualize Microplastics and Nanoplastics

Abstract

As emerging contaminants, microplastics are challenging to characterize, particularly when their size is at the nanoscale. While imaging technology has received increasing attention recently, such as Raman imaging, decoding the scanning spectrum matrix can be difficult to achieve result digitally and automatically via software and usually requires the involvement of personal experience and expertise. Herewith, we show a dual-principal component analysis (PCA) approach, where (i) the first round of PCA analysis focuses on the raw spectrum data from the Raman scanning matrix and generates two new matrices, with one containing the spectrum profile to yield the PCA spectrum and the other containing the PCA intensity to be mapped as an image; (ii) the second round of PCA analysis merges the spectrum from the first round of PCA with the standard spectra of eight common plastics, to generate a correlation matrix. From the correlation value, we can digitally assign the principal components from the first round of PCA analysis to the plastics toward imaging, akin to dataset indexing. We also demonstrate the effect of the data pretreatment and the wavenumber variations. Overall, this dual-PCA approach paves the way for machine learning to analyze microplastics and particularly nanoplastics.

Raman imaging combined with an improved PCA/algebra-based algorithm to capture microplastics and nanoplastics

“…Another algorithm we developed recently incorporates two rounds of PCA, a dual-PCA algorithm. 24 In the 1st round PCA, as shown above, the PCA spectrum and the PCA intensity image are extracted. Afterwards, in order to assign the suspected items (PC1, PC2 and PC3) to plastics, we need to compare the PCA spectrum with the Raman spectrum to justify the similarity or difference.…”

“…For an accurate comparison, the PCA spectrum needs pre-treatments, including smoothening, baseline correction, interpolation, normalisation etc. 24 The pre-treatment for PC1's PCA spectrum is shown in Fig. 7(a).…”

“…Most of these pre-treatments can be conducted using either the operational software (WITec Project, such as cosmic ray removal) or Origin. 24 In the latter case, for example, we employ adjacent-averaging (points of window: 5) to smoothen the curves. To correct the baseline, we can use asymmetric least squares, with parameters of asymmetric factor, 0.001; threshold, 0.05; smoothing factor, 6; number of iterations, 10; etc.…”

“…In this case, we need a further step to follow up the PCA analysis. 22 Recently, we suggested that a follow-up PCA, or dual-PCA, can digitally justify and assign the PCA spectrum from the first round PCA, 24 and feedback the PCA intensity image's assignment. However, when the first round of the PCA spectrum is far away from the virgin Raman spectrum, such as with the upside-down peaks, it is not clear how to further extract the meaningful information for assignment.…”

“…Accordingly, when we further extract the PCA spectrum, we also need to further treat the PCA intensity images towards merging for cross-examination, such as using an algebra-based algorithm. 24 In addition to PCA, other techniques have also been employed to identify microplastics and nanoplastics from the complex spectral data, such as k-nearest neighbour classification, 25 Otsu's algorithm (automatic thresholding method), 26 and pattern recognition algorithm. 27 However, all the unsupervised analysis meet the similar decoding challenge.…”