Abstract:Raman imaging has advanced recently to be able to directly visualise microplastics and even nanoplastics. However, the generated scanning spectrum matrix, akin to a hyperspectral matrix, is challenging to decode....
“…2 a. To accurately identify it, we pre-teat the spectrum, including curve smoothening to remove the random noise, baseline correction to off-set the possible fluorescence background, wavenumber interpolation to standardise the x -axis, intensity normalisation (to 0–1) to standardise y -axis [ 23 ]. Fig.…”
Section: Resultsmentioning
confidence: 99%
“…3 , although the algebra-based algorithm is employed to catch multi peaks in the spectrum and merge them as one image, the signal beyond the selected peaks is still missed, meaning a low signal–noise ratio. A chemometrics of PCA has been recently employed to directly decode the scanning Raman spectrum matrix, for the whole set of spectrum rather than a solo peak, to participate the imaging process [ 21 , 23 ]. That is, ideally, PCA can orthogonally and directly decompose the data matrix to two new matrices, one containing the spectrum profile information while another containing the intensity information.…”
Section: Resultsmentioning
confidence: 99%
“…Two or more images, no matter mapping the Raman intensity or the PC loading coefficients (to be discussed below), can be merged using algebra functions with an increased flexibility, using Origin software [ 23 ]. Before merge, we normalise the colour value to 0–255 or the loading coefficient to 0–1, to avoid bias.…”
Section: Methodsmentioning
confidence: 99%
“…PCA can also identify the target via correlation by combining its spectrum with the standard spectra, akin to indexing [ 23 , 25 ]. To avoid the comparison bias and to increase the accuracy, we pre-treat all the Raman spectra, using techniques including cosmic ray removal, smoothing, baseline correction, interpolating to adjust the wavenumber ( x -axis) and normalising the intensity ( y -axis).…”
Background
COVID-19 pandemic is not yet over, and it has been generating lots of plastic wastes that become a big concern. To catch the virus, for example, no matter via antigen or PCR test, swab is generally used for sampling. Unfortunately, the swab tip is commonly made of plastics, and thus it can be a potential source of microplastics. This study aims to propose and optimise several Raman imaging to identify the microplastic fibres released from different COVID-19 test swabs.
Results
The results show that Raman imaging can effectively identify and visualise the microplastic fibres released from the swabs. In the meantime, on the surface of the fibres, additives such as titanium oxide particles are also captured for some brands of swabs. To increase the result certainty, scanning electron microscope (SEM) is first employed to get the morphology of the released microplastic fibres, along with Energy-dispersive X-ray spectroscopy (EDS) to confirm the presence of titanium element. Then, Raman imaging is advanced to identify and visualise the microplastics and titanium oxide particles, from different characteristic peaks in the scanning spectrum matrix. To further increase the imaging certainty, these images can be merged and cross-checked using algorithms, or the raw data from the scanning spectrum matrix can be analysed and decoded via chemometrics, such as principal component analysis (PCA). Beyond the advantages, the disadvantages of the confocal Raman imaging (affected by focal height) and algorithms (non-supervised calculation) are also discussed and intentionally corrected. In brief, the imaging analysis (particularly the combined SEM with Raman) is recommended to avoid the possible result bias that might be generated from the single spectrum analysis at a selective but random position.
Conclusions
Overall, the results indicate that Raman imaging can be a useful tool to detect microplastics. The results also send us a strong warning that, if we worry about the potential microplastics contamination, we should be cautious to select the suitable COVID-19 testing kits.
“…2 a. To accurately identify it, we pre-teat the spectrum, including curve smoothening to remove the random noise, baseline correction to off-set the possible fluorescence background, wavenumber interpolation to standardise the x -axis, intensity normalisation (to 0–1) to standardise y -axis [ 23 ]. Fig.…”
Section: Resultsmentioning
confidence: 99%
“…3 , although the algebra-based algorithm is employed to catch multi peaks in the spectrum and merge them as one image, the signal beyond the selected peaks is still missed, meaning a low signal–noise ratio. A chemometrics of PCA has been recently employed to directly decode the scanning Raman spectrum matrix, for the whole set of spectrum rather than a solo peak, to participate the imaging process [ 21 , 23 ]. That is, ideally, PCA can orthogonally and directly decompose the data matrix to two new matrices, one containing the spectrum profile information while another containing the intensity information.…”
Section: Resultsmentioning
confidence: 99%
“…Two or more images, no matter mapping the Raman intensity or the PC loading coefficients (to be discussed below), can be merged using algebra functions with an increased flexibility, using Origin software [ 23 ]. Before merge, we normalise the colour value to 0–255 or the loading coefficient to 0–1, to avoid bias.…”
Section: Methodsmentioning
confidence: 99%
“…PCA can also identify the target via correlation by combining its spectrum with the standard spectra, akin to indexing [ 23 , 25 ]. To avoid the comparison bias and to increase the accuracy, we pre-treat all the Raman spectra, using techniques including cosmic ray removal, smoothing, baseline correction, interpolating to adjust the wavenumber ( x -axis) and normalising the intensity ( y -axis).…”
Background
COVID-19 pandemic is not yet over, and it has been generating lots of plastic wastes that become a big concern. To catch the virus, for example, no matter via antigen or PCR test, swab is generally used for sampling. Unfortunately, the swab tip is commonly made of plastics, and thus it can be a potential source of microplastics. This study aims to propose and optimise several Raman imaging to identify the microplastic fibres released from different COVID-19 test swabs.
Results
The results show that Raman imaging can effectively identify and visualise the microplastic fibres released from the swabs. In the meantime, on the surface of the fibres, additives such as titanium oxide particles are also captured for some brands of swabs. To increase the result certainty, scanning electron microscope (SEM) is first employed to get the morphology of the released microplastic fibres, along with Energy-dispersive X-ray spectroscopy (EDS) to confirm the presence of titanium element. Then, Raman imaging is advanced to identify and visualise the microplastics and titanium oxide particles, from different characteristic peaks in the scanning spectrum matrix. To further increase the imaging certainty, these images can be merged and cross-checked using algorithms, or the raw data from the scanning spectrum matrix can be analysed and decoded via chemometrics, such as principal component analysis (PCA). Beyond the advantages, the disadvantages of the confocal Raman imaging (affected by focal height) and algorithms (non-supervised calculation) are also discussed and intentionally corrected. In brief, the imaging analysis (particularly the combined SEM with Raman) is recommended to avoid the possible result bias that might be generated from the single spectrum analysis at a selective but random position.
Conclusions
Overall, the results indicate that Raman imaging can be a useful tool to detect microplastics. The results also send us a strong warning that, if we worry about the potential microplastics contamination, we should be cautious to select the suitable COVID-19 testing kits.
“…In Figure c, the solo characteristic peak's image can be improved, using chemometrics, such as PCA. As said, PCA can orthogonally compress the scanning Raman matrix to two ones: one contains the spectrum profile toward identification, and another contains the intensity information for mapping. , In this case, the whole set of spectra will take part in the imaging process, not just the solo peak demonstrated above. The PCA results of the raw data from the Raman scanning spectrum matrix are shown in the bottom row in Figure .…”
Most teenagers experience orthodontic treatment, but we do not know the possible adverse effect of the released microplastics and nanoplastics that are recently categorized as emerging contaminants. Herein, we test the rubber band that has been employed to improve the biting of teeth during the orthodontic process to confirm the release of microplastics and nanoplastics. We improve the characterization of microplastics and nanoplastics by (i) Raman imaging, to extract and map the signal from the scanning spectrum matrix or the hyperspectral matrix and to enhance the signal-to-noise ratio statistically. To effectively extract the signal, (ii) chemometrics such as principal component analysis (PCA) are explored to convert the hyperspectral matrix to an image with an increased certainty. The nonsupervised PCA is intentionally corrected, via (iii) the algebra-based algorithm, to further increase the certainty to image the microplastics and nanoplastics. Once the signal is weak, (iv) an additional algorithm of image reconstruction via deconvolution is developed to average the background noise and smooth the image. By doing so, we estimate that millions of microplastics and nanoplastics are released daily in potential from a rubber band applied in a teenager's mouth, which might be a big concern. Overall, our approach provides a suitable option to characterize the microplastics and nanoplastics from a complex background.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
