A Deep Learning Approach to Organic Pollutants Classification Using Voltammetry
Abstract
:1. Introduction
2. Experimental Characterization
2.1. Detection via Cyclic Voltammetry
2.2. Carbon Nanotube Modified Platform
2.3. Characterization of Hydroquinone and Benzoquinone and Dataset Generation
3. Classification via Machine Learning
3.1. Gramian Angular Fields Transformations
3.2. Deep Learning Model
3.3. Dataset
4. Results and Discussion
- Epochs: 400
- Patience: 100
- Optimizer: Stochastic Gradient Descent
- Learning rate: 0.0001
- Momentum: 0.9
- Loss: categorical cross-entropy
- Metrics: accuracy
- Batch size: 16
5. Conclusions
Author Contributions
Funding
Conflicts of Interest
References
- 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. [Google Scholar] [CrossRef]
- 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. [Google Scholar] [CrossRef] [Green Version]
- 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. [Google Scholar] [CrossRef] [PubMed]
- 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. [Google Scholar] [CrossRef]
- 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. [Google Scholar] [CrossRef]
- Martín-Yerga, D.; Costa Rama, E.; Costa Garcia, A. Electrochemical Study and Determination of Electroactive Species with Screen-Printed Electrodes. J. Chem. Educ. 2016, 93, 1270–1276. [Google Scholar] [CrossRef]
- Komoda, M.; Shitanda, I.; Hoshi, Y.; Itagaki, M. Instantaneously usable screen-printed silver/silver sulfate reference electrode with long-term stability. Electrochem. Commun. 2019, 103, 133–137. [Google Scholar] [CrossRef]
- Meyer, A.M.; Klein, C.; Fünfrocken, E.; Kautenburger, R.; Beck, H.P. Real-time monitoring of water quality to identify pollution pathways in small and middle scale rivers. Sci. Total Environ. 2019, 651, 2323–2333. [Google Scholar] [CrossRef]
- Anichini, C.; Czepa, W.; Pakulski, D.; Aliprandi, A.; Ciesielski, A.; Samorì, P. Chemical sensing with 2D materials. Chem. Soc. Rev. 2018, 47, 4860–4908. [Google Scholar] [CrossRef] [Green Version]
- Beitollahi, H.; Mohammadi, S.Z.; Safaei, M.; Tajik, S. Applications of electrochemical sensors and biosensors based on modified screen-printed electrodes: A review. Anal. Methods 2020, 12, 1547–1560. [Google Scholar] [CrossRef]
- Di Tinno, A.; Cancelliere, R.; Mantegazza, P.; Cataldo, A.; Paddubskaya, A.; Ferrigno, L.; Kuzhir, P.; Maksimenko, S.; Shuba, M.; Maffucci, A.; et al. Sensitive Detection of Industrial Pollutants Using Modified Electrochemical Platforms. Nanomaterials 2022, 12, 1779. [Google Scholar] [CrossRef] [PubMed]
- Cancelliere, R.; Di Tinno, A.; Di Lellis, A.M.; Tedeschi, Y.; Bellucci, S.; Carbone, K.; Signori, E.; Contini, G.; Micheli, L. An inverse-designed electrochemical platform for analytical applications. Electrochem. Commun. 2020, 121, 106862. [Google Scholar] [CrossRef]
- Cancelliere, R.; Tinno, A.D.; Cataldo, A.; Bellucci, S.; Micheli, L. Powerful Electron-Transfer Screen-Printed Platforms as Biosensing Tools: The Case of Uric Acid Biosensor. Biosensors 2022, 12, 2. [Google Scholar] [CrossRef] [PubMed]
- Cancelliere, R.; Albano, D.; Brugnoli, B.; Buonasera, K.; Leo, G.; Margonelli, A.; Rea, G. Electrochemical and morphological layer-by-layer characterization of electrode interfaces during a label-free impedimetric immunosensor build-up: The case of ochratoxin A. Appl. Surf. Sci. 2021, 567, 150791. [Google Scholar] [CrossRef]
- Wang, J. Analytical Electrochemistry; John Wiley & Sons, Inc.: Hoboken, NJ, USA, 2006. [Google Scholar]
- Kuselman, I.; Pennecchi, F.; Fajgelj, A.; Karpov, Y. Human errors and reliability of test results in analytical chemistry. Accredit. Qual. Assur. 2012, 18, 3–9. [Google Scholar] [CrossRef]
- Kuselman, I.; Pennecchi, F. Human Errors in a Routine Analytical Laboratory—Classification, Modeling and Quantification: Overview of the IUPAC/CITAC Guide. Chem. Int. 2016, 38, 27–30. [Google Scholar] [CrossRef]
- Alévêque, O.; Levillain, E. A generalized lateral interactions function to fit voltammetric peaks of self-assembled monolayers. Electrochem. Commun. 2016, 67, 73–79. [Google Scholar] [CrossRef]
- Grdeń, M. Semi-differential analysis of irreversible voltammetric peaks. J. Solid State Electrochem. 2016, 21, 1045–1058. [Google Scholar] [CrossRef] [Green Version]
- Dean, S.N.; Shriver-Lake, L.C.; Stenger, D.A.; Erickson, J.S.; Golden, J.P.; Trammell, S.A. Machine Learning Techniques for Chemical Identification Using Cyclic Square Wave Voltammetry. Sensors 2019, 19, 2392. [Google Scholar] [CrossRef] [Green Version]
- Li, Z.; Liu, F.; Yang, W.; Peng, S.; Zhou, J. A Survey of Convolutional Neural Networks: Analysis, Applications, and Prospects. In IEEE Transactions on Neural Networks and Learning Systems; IEEE: Piscataway, NJ, USA, 2021; pp. 1–21. [Google Scholar]
- Hoar, B.B.; Zhang, W.; Xu, S.; Deeba, R.; Costentin, C.; Gu, Q.; Liu, C. Electrochemical mechanistic analysis from cyclic voltammograms based on deep learning. ACS Meas. Sci. 2022. [Google Scholar] [CrossRef]
- Cui, F.; Yue, Y.; Zhang, Y.; Zhang, Z.; Zhou, H.S. Advancing Biosensors with Machine Learning. ACS Sens. 2020, 5, 3346–3364. [Google Scholar] [CrossRef] [PubMed]
- Martynko, E.; Kirsanov, D. Application of Chemometrics in Biosensing: A Brief Review. Biosensors 2020, 10, 100. [Google Scholar] [CrossRef] [PubMed]
- Fónagy, O.; Szabó-Bárdos, E.; Horváth, O. 1,4-Benzoquinone and 1,4-hydroquinone based determination of electron and superoxide radical formed in heterogeneous photocatalytic systems. J. Photochem. Photobiol. A Chem. 2020, 407, 113057. [Google Scholar] [CrossRef]
- Giner, R.M.; Ríos, J.L.; Máñez, S. Antioxidant Activity of Natural Hydroquinones. Antioxidants 2022, 11, 343. [Google Scholar] [CrossRef] [PubMed]
- Zuo, Y.-T.; Hu, Y.; Lu, W.-W.; Cao, J.-J.; Wang, F.; Han, X.; Lu, W.-Q.; Liu, A.-L. Toxicity of 2,6-dichloro-1,4-benzoquinone and five regulated drinking water disinfection by-products for the Caenorhabditis elegans nematode. J. Hazard. Mater. 2017, 321, 456–463. [Google Scholar] [CrossRef]
- Fu, K.Z.; Li, J.; Vemula, S.; Moe, B.; Li, X.F. Effects of halobenzoquinone and haloacetic acid water disinfection byproducts on human neural stem cells. J. Environ. Sci. 2017, 58, 239–249. [Google Scholar] [CrossRef]
- Baah, M.; Rahman, A.; Sibilia, S.; Trezza, G.; Ferrigno, L.; Micheli, L.; Maffucci, A.; Soboleva, E.; Lähderanta, E.; Svirko, Y.; et al. Electrical Impedance sensing of benzoquinone with ultrathin graphitic membranes. Nanotechnology 2022, 33, 075207. [Google Scholar] [CrossRef]
- Miele, G.; Bellucci, S.; Cataldo, A.; Di Tinno, A.; Ferrigno, L.; Maffucci, A.; Micheli, L.; Sibilia, S. Electrical Impedance Spectroscopy for Real-Time Monitoring of the Life Cycle of Graphene Nanoplatelets Filters for Some Organic Industrial Pollutants. IEEE Trans. Instrum. Meas. 2021, 70, 1503912. [Google Scholar] [CrossRef]
- Bria, A.; Cerro, G.; Ferdinandi, M.; Marrocco, C.; Molinara, M. An IoT-ready solution for automated recognition of water contaminants. Pattern Recognit. Lett. 2020, 135, 188–195. [Google Scholar] [CrossRef]
- Ferdinandi, M.; Molinara, M.; Cerro, G.; Ferrigno, L.; Marroco, C.; Bria, A.; Di Meo, P.; Bourelly, C.; Simmarano, R. A Novel Smart System for Contaminants Detection and Recognition in Water. In Proceedings of the 2019 IEEE International Conference on Smart Computing (SMARTCOMP), Washington, DC, USA, 12–15 June 2019; pp. 186–191. [Google Scholar] [CrossRef]
- Molinara, M.; Ferdinandi, M.; Cerro, G.; Ferrigno, L.; Massera, E. An End to End Indoor Air Monitoring System Based on Machine Learning and SENSIPLUS Platform. IEEE Access 2020, 8, 72204–72215. [Google Scholar] [CrossRef]
- Bellucci, S.; Maffucci, A.; Maksimenko, S.; Micciulla, F.; Migliore, M.D.; Paddubskaya, A.; Pinchera, D.; Schettino, F. Electrical Permittivity and Conductivity of a Graphene Nanoplatelet Contact in the Microwave Range. Materials 2018, 11, 2519. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Wang, Z.; Oates, T. Imaging time-series to improve classification and imputation. In Proceedings of the 24th International Conference on Artificial Intelligence (IJCAI’15), Buenos Aires, Argentina, 25–31 July 2015; pp. 3939–3945. [Google Scholar]
- LeCun, Y.; Bengio, Y.; Hinton, G. Deep learning. Nature 2015, 521, 436–444. [Google Scholar] [CrossRef] [PubMed]
- LeCun, Y.; Boser, B.; Denker, J.S.; Henderson, D.; Howard, R.E.; Hubbard, W.; Jackel, L.D. Handwritten digit recognition with a back-propagation network. In Advances in Neural Information Processing Systems 2; Touretzky, D.S., Ed.; Morgan Kaufmann: Burlington, MA, USA, 1989; pp. 396–404. [Google Scholar]
- Lecun, Y.; Bottou, L.; Bengio, Y.; Haffner, P. Gradient-based learning applied to document recognition. Proc. IEEE 1998, 86, 2278–2324. [Google Scholar] [CrossRef] [Green Version]
- Fukushima, K. Artificial vision by multi-layered neural networks: Neocognitron and its advances. Neural Netw. 2013, 37, 103–119. [Google Scholar] [CrossRef]
- Keras. Chollet, Francois, 2015. GitHub. Software, Keras 2.4. Available online: https://github.com/fchollet/keras (accessed on 1 October 2022).
- Abadi, M.; Agarwal, A.; Barham, P.; Brevdo, E.; Chen, Z.; Citro, C.; Corrado, G.S.; Davis, A.; Dean, J.; Devin, M.; et al. TensorFlow: Large-Scale Machine Learning on Heterogeneous Systems. 2015. Software, Tensorflow 2.4. Available online: www.tensorflow.org (accessed on 1 October 2022).
Parameter/Electrode | Bare | SWCNT | MWCNT |
---|---|---|---|
LOD (μM) | 334.5 | 80.3 | 13.7 |
Reproducibility (RSD%) | 17 | 8 | 9 |
Layer (Type) | Output Shape | Size | Param # |
---|---|---|---|
input | (None, 224, 224, 3) | ||
conv2d (Conv2D) | (None, 222, 222, 48) | 3 | 1.344 |
max_pooling2d (MaxPooling2D) | (None, 111, 111, 48) | 2 | 0 |
conv2d (Conv2D) | (None, 109, 109, 48) | 3 | 20.784 |
max_pooling2d (MaxPooling2D) | (None, 54, 54, 48) | 2 | 0 |
conv2d (Conv2D) | (None, 52, 52, 32) | 3 | 13.856 |
max_pooling2d (MaxPooling2D) | (None, 26, 26, 32) | 2 | 0 |
conv2d (Conv2D) | (None, 24, 24, 32) | 3 | 9.248 |
max_pooling2d (MaxPooling2D) | (None, 12, 12, 32) | 2 | 0 |
conv2d (Conv2D) | (None, 10, 10, 16) | 3 | 4.624 |
max_pooling2d (MaxPooling2D) | (None, 5, 5, 16) | 2 | 0 |
conv2d (Conv2D) | (None, 3, 3, 8) | 3 | 1.160 |
max_pooling2d (MaxPooling2D) | (None, 1, 1, 8) | 2 | 0 |
Flatten | (None, 8) | 0 | |
Dense | (None, 64) | 576 | |
Dropout (0.5) | (None, 64) | 0 | |
Dense | (None, 3) | 195 | |
Batch Normalization | (None, 3) | 12 | |
Activation Softmax | (None, 3) | 0 | |
Total param # | 51,799 |
Benzoquinone | Sensors | ||||
Bare | MWCNT | SWCNT | |||
Number of voltammetry cycles | 2 | 3 | 3 | ||
Concentrations (mM) | 80 | 80 | |||
50 | 50 | 50 | |||
25 | 25 | 25 | |||
12.5 | 12.5 | 12.5 | |||
5 | 5 | 5.1 | |||
2.5 | 2.5 | 5 | |||
1 | |||||
Total images: | |||||
Number of images | 12 | 18 | 18 | 48 | |
Hydroquinone | Sensors | ||||
Bare | MWCNT | SWCNT | |||
Number of voltammetry cycles | 3 | 3 | 3 | ||
Concentrations (mM) | 100 | 100 | 100 | ||
50 | 50 | 50 | |||
25 | 25 | 25 | |||
12.5 | 12.5 | 12.5 | |||
5 | 5 | 5 | |||
2.5 | 2.5 | 2.5 | |||
1 | 1 | 1 | |||
0.5 | 0.5 | 0.5 | |||
0.25 | 0.25 | 0.25 | |||
Total images: | |||||
Number of images | 27 | 27 | 27 | 81 | |
Potassium ferricyanide | Sensors | ||||
Bare | MWCNT | SWCNT | |||
Number of voltammetry cycles | 6 | 6 | 6 | ||
Concentrations (mM) | 100 | 100 | 100 | ||
50 | 50 | 50 | |||
25 | 25 | 25 | |||
12.5 | 12.5 | 12.5 | |||
5 | 5 | 5 | |||
2.5 | 2.5 | 2.5 | |||
1 | 1 | 1 | |||
0.5 | 0.5 | 0.5 | |||
0.25 | 0.25 | 0.25 | |||
Total images: | |||||
Number of images | 54 | 54 | 54 | 162 |
R | G | B | |
---|---|---|---|
Average | 2.4596 | 2.7999 | 2.4832 |
STD | 0.1590 | 0.2940 | 0.0057 |
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. |
© 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
Molinara, M.; Cancelliere, R.; Di Tinno, A.; Ferrigno, L.; Shuba, M.; Kuzhir, P.; Maffucci, A.; Micheli, L. A Deep Learning Approach to Organic Pollutants Classification Using Voltammetry. Sensors 2022, 22, 8032. https://doi.org/10.3390/s22208032
Molinara M, Cancelliere R, Di Tinno A, Ferrigno L, Shuba M, Kuzhir P, Maffucci A, Micheli L. A Deep Learning Approach to Organic Pollutants Classification Using Voltammetry. Sensors. 2022; 22(20):8032. https://doi.org/10.3390/s22208032
Chicago/Turabian StyleMolinara, Mario, Rocco Cancelliere, Alessio Di Tinno, Luigi Ferrigno, Mikhail Shuba, Polina Kuzhir, Antonio Maffucci, and Laura Micheli. 2022. "A Deep Learning Approach to Organic Pollutants Classification Using Voltammetry" Sensors 22, no. 20: 8032. https://doi.org/10.3390/s22208032
APA StyleMolinara, M., Cancelliere, R., Di Tinno, A., Ferrigno, L., Shuba, M., Kuzhir, P., Maffucci, A., & Micheli, L. (2022). A Deep Learning Approach to Organic Pollutants Classification Using Voltammetry. Sensors, 22(20), 8032. https://doi.org/10.3390/s22208032