Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2022 Oct 21;22(20):8032.
doi: 10.3390/s22208032.

A Deep Learning Approach to Organic Pollutants Classification Using Voltammetry

Mario Molinara  1 , Rocco Cancelliere  2 , Alessio Di Tinno  2 , Luigi Ferrigno  1 , Mikhail Shuba  3 , Polina Kuzhir  4 , Antonio Maffucci  1   5 , Laura Micheli  2
Affiliations

A Deep Learning Approach to Organic Pollutants Classification Using Voltammetry

Mario Molinara et al. Sensors (Basel). .

Abstract

This paper proposes a deep leaning technique for accurate detection and reliable classification of organic pollutants in water. The pollutants are detected by means of cyclic voltammetry characterizations made by using low-cost disposable screen-printed electrodes. The paper demonstrates the possibility of strongly improving the detection of such platforms by modifying them with nanomaterials. The classification is addressed by using a deep learning approach with convolutional neural networks. To this end, the results of the voltammetry analysis are transformed into equivalent RGB images by means of Gramian angular field transformations. The proposed technique is applied to the detection and classification of hydroquinone and benzoquinone, which are particularly challenging since these two pollutants have a similar electroactivity and thus the voltammetry curves exhibit overlapping peaks. The modification of electrodes by carbon nanotubes improves the sensitivity of a factor of about ×25, whereas the convolution neural network after Gramian transformation correctly classifies 100% of the experiments.

Keywords: carbon nanotubes; convolutional neural networks; cyclic voltammetry; pollutant detection; screen-printed electrodes.

PubMed Disclaimer

Conflict of interest statement

The authors declare no conflict of interest.

Figures

Figure 1
Figure 1
(a) A conceptual scheme of a screen-printed electrode; (b) example of current-potential curves as the output of a cyclic voltammetry for different values of the analyte concentration (here hydroquinone); (c) extrapolated calibration curve (blue dots: experimental current peaks).
Figure 2
Figure 2
Peak potential and currents measured on several cyclic voltammograms executed on the hydroquinone pollutant (5 mM), with different platforms: Bare SPEs (green), MWCNT SPEs (blue), and SWCNT SPEs (red).
Figure 3
Figure 3
Scanning electron microscope (SEM) images of the produced (a) SWCNT and (b) MWCNT.
Figure 4
Figure 4
Raman spectra of (a) single-walled and (b) multiwalled CNTs.
Figure 5
Figure 5
Comparison of the cyclic voltammograms; (a) simultaneous determination of the 3 analytes; (b–d) PF, HQ, and BQ were analyzed separately.
Figure 6
Figure 6
CV voltammograms obtained with different platforms (SWCNTs, MWCNTs, and bare SPEs), referred to: (a) the same concentration of HQ (5 mM), and (b,c) several HQ concentrations (from 0.1 µM to 1 mM).
Figure 7
Figure 7
Gramian angular field (GAF) transformation of an I–V cycle into an RGB image.
Figure 8
Figure 8
RGB images generated by GAF transformation of CV cycles, for the detection of: (a) Potassium ferricyanide (PF) in water, 0.25 mM, bare SPE; (b) potassium ferricyanide (PF) in water, 100 mM, bare SPE; (c) benzoquinone (BQ) in water, 2.5 mM, SWCNT SPE; (d) benzoquinone (BQ) in water, 80 mM, SWCNT SPE.
Figure 8
Figure 8
RGB images generated by GAF transformation of CV cycles, for the detection of: (a) Potassium ferricyanide (PF) in water, 0.25 mM, bare SPE; (b) potassium ferricyanide (PF) in water, 100 mM, bare SPE; (c) benzoquinone (BQ) in water, 2.5 mM, SWCNT SPE; (d) benzoquinone (BQ) in water, 80 mM, SWCNT SPE.
Figure 9
Figure 9
Evolution with the epochs of the accuracy during the training phase on the training set (red dotted curve) and the validation set (blue solid line) for the first fold.
Figure 10
Figure 10
Evolution with the epochs of the loss during the training phase on the training set (red dotted curve) and validation set (blue solid curve) for the first fold.
Figure 11
Figure 11
Confusion matrix evaluated on the test set.
Figure 12
Figure 12
A global overview of the entire workflow of the proposed system.
See this image and copyright information in PMC

References

    1. Avino P., Russo M.V. A Comprehensive Review of Analytical Methods for Determining Persistent Organic Pollutants in Air, Soil, Water and Waste. Curr. Org. Chem. 2018;22:939–953. doi: 10.2174/1385272822666180404144834. - DOI
    1. Moldovan Z., Popa D.E., David I.G., Buleandra M., Badea I.A. A Derivative Spectrometric Method for Hydroquinone Determination in the Presence of Kojic Acid, Glycolic Acid, and Ascorbic Acid. J. Spectrosc. 2017;2017:6929520. doi: 10.1155/2017/6929520. - DOI
    1. Dursunoğlu B., Yuca H., Güvenalp Z., Gözcü S., Yılmaz B., Yilmaz B. Simultaneous determination of arbutin and hydroquinone in different herbal slimming products by Gas Chromatography-Mass Spectrometry. Turk. J. Pharm. Sci. 2018;15:298. doi: 10.4274/tjps.38258. - DOI - PMC - PubMed
    1. Zhang M., Ge C., Jin Y., Hu L., Mo H., Li X., Zhang H. Sensitive and Simultaneous Determination of Hydroquinone and Catechol in Water Using an Anodized Glassy Carbon Electrode with Polymerized 2-(Phenylazo) Chromotropic Acid. J. Chem. 2019;2019:2327064. doi: 10.1155/2019/2327064. - DOI
    1. Barton J., García M.B.G., Santos D.H., Fanjul-Bolado P., Ribotti A., McCaul M., Diamond D., Magni P. Screen-printed electrodes for environmental monitoring of heavy metal ions: A review. Microchim. Acta. 2016;183:503–517. doi: 10.1007/s00604-015-1651-0. - DOI

LinkOut - more resources