chemosensors Article Quantitative Estimation of COD Values from an Array of Metal Nanoparticle Modified Electrodes and Artificial Neural Networks † Qing Wang, Xavier Cetó and Manel del Valle * Sensors and Biosensors Group, Department of Chemistry, Universitat Autònoma de Barcelona, Bellaterra, 08193 Barcelona, Spain * Correspondence: manel.delvalle@uab.cat † This paper is an extended version of the conference paper: Wang, Q.; del Valle, M. Determination of Chemical Oxygen Demand (COD) Using Nanoparticle-Modified Voltammetric Sensors and Electronic Tongue Principles. In Proceedings of the 1st International Electronic Conference on Chemical Sensors and Analytical Chemistry (CSAC2021), 1–15 July 2021. Abstract: Water quality monitoring has become critical in modern societies in multiple areas and at different stages. In this regard, chemical oxygen demand (COD) has become a key index in water testing, as it readily allows the determination of its overall quality and the presence of organic contaminants. However, conventional COD determination presents several drawbacks in view of the use of toxic reagents and possible interferences. The electrochemical determination of COD can be an alternative with many advantages, especially if using an array of sensors. Herein, the use of an electronic tongue (ET) for the estimation of COD was explored. The proposed ET was formed by an array of five voltammetric electrodes modified with different metal nanoparticles. An artificial neural network (ANN) model was built based on the responses of the array towards glucose and glycine as standards. This model was then used with real and spiked water samples, and the results compared Citation: Wang, Q.; Cetó, X.; Valle, to the electrochemical calibration and the commercial COD colorimetric methods. While the COD M.d. Quantitative Estimation of COD values of the real samples were low and outside the range of the ANN model, a satisfactory prediction Values from an Array of Metal for the spiked samples was achieved, showing a good agreement with the reference colorimetric Nanoparticle Modified Electrodes method, that was better than the performance of the conventional electrochemical calibration method. and Artificial Neural Networks. Chemosensors 2022, 10, 504. https:// Keywords: chemical oxygen demand; wastewater; metal nanoparticles; voltammetric sensors; artificial doi.org/10.3390/chemosensors10120504 neural networks; electronic tongue Academic Editor: Nicole Jaffrezic-Renault Received: 28 October 2022 Accepted: 23 November 2022 1. Introduction Published: 28 November 2022 Chemical oxygen demand (COD) is a widely used parameter in the evaluation of the Publisher’s Note: MDPI stays neutral total organic compounds present in water, especially in wastewater. The value of COD is with regard to jurisdictional claims in expressed as the amount of molecular oxygen (in milligrams of O2 per liter) required to published maps and institutional affil- oxidize all organic compounds to carbon dioxide and water according to Equation (1). The iations. amino nitrogen will be converted to nitrate in the presence of an oxidizing agent [1–6]. Cn Ha Ob Nc + (n + a/4 + 3/2c − b/2) O2 →nCO2 + (a/2 − 3/2c) H2 O + cNO3 − (1) Copyright: © 2022 by the authors. To ensure that all organic matter is fully oxidized according to the above reaction, Licensee MDPI, Basel, Switzerland. the use of a strong oxidizer under acidic conditions and reflux temperature is normally This article is an open access article used; being potassium dichromate the one used in the traditional standard procedure distributed under the terms and of COD determination. However, it involves hexavalent chromium as reagent, which conditions of the Creative Commons in addition to being poisonous, is also not environment-friendly [7]. As an alternative, Attribution (CC BY) license (https:// potassium permanganate can also be employed, as it is less toxic than potassium dichromate. creativecommons.org/licenses/by/ Nevertheless, since it is a weaker oxidant, it is not able to oxidize all the organic compounds 4.0/). in water [8]. Moreover, the traditional COD test consists of several steps and complicated Chemosensors 2022, 10, 504. https://doi.org/10.3390/chemosensors10120504 https://www.mdpi.com/journal/chemosensors Chemosensors 2022, 10, 504 2 of 16 operations, which are very time-consuming and troublesome [9]. To avoid the difficulties derived from the traditional measurement, substantial efforts have been directed towards the development of more friendly methods. Because of their good sensitivity, low cost and effortless operation, electrochemical methods have attracted much attention as an alternative for the determination of COD [10–14]. Electrochemical methods offer a rapid and simple estimation of COD, from the po- tential equivalence of biochemical and electrochemical oxidation. Furthermore, they do not require the use of those toxic and expensive reagents used in the traditional method. Instead, calibration curves are obtained based on the reaction of organic compounds ox- idized by radicals generated electrochemically on the electrode [15]. Because of the ease of operation, they offer the possibility of implementing online COD determination sys- tems. For example, Duan and coworkers developed a compact analytical platform that integrates an electrochemical deposited copper nanoparticle-based (Cu NPs) carbon/silicon dioxide(C/SiO2 ) thin-film electrochemical sensor, which achieved on-line monitoring of COD in urban wastewater [16]. Over the last few years, many materials have been reported as electrocatalysts in the amperometric determination of COD, essentially with use of metal nanoparticles (NPs) [8,9,12–15], or oxide films [16–20]. For the calibration of the sensors, most often a single standard organic compound is used to simulate the wastewater components (e.g., glucose or potassium hydrogen phthalate), and the calibration curves of current intensity vs. the COD values are built, based on the concentration of the standard compound taking the Equation (1) mentioned above. Such a method has shown a satisfactory performance in the COD determination towards wastewater samples of simple composition [9,14,15,21–23]. However, the accuracy of this method is quite limited in the analysis of wastewater that consists of mixtures of organic components, each with different reactivity. In light of this consideration, a more powerful and flexible technique for nonlinear regression attracted us, namely, artificial neural networks (ANNs) [24,25]. ANNs are bio-inspired mathematical models and have been employed in the field of chemometrics for qualitative and quan- titative multivariate analysis [26–29]. What makes them different from the classic linear regression methods for data treatment is their capacity to find non-linear relationships be- tween different variables in the data, and this is also their most outstanding advantage [30]. As a case of principle, and to hypothesize the resolution of complex mixtures composed of several different organic compounds with different oxidation kinetics, we have designed an experiment that employs glucose and glycine as the organic standard substances, several voltammetric sensors as the mean to generate sufficient information [31], and an ANN response model to estimate the COD values. A total of five sensors were used, four of which were made of carbon modified with different metal NPs, plus an additional CuO sensor made of NPs electrogenerated over a Cu block. The system was trained to quantitatively predict the organic load (COD value) of mixtures of two compounds—in this case, glucose and glycine. The generated model has been validated with a separate set of test samples, consisting of other mixtures of the same two compounds. In summary, this paper presents a new method for predicting organic load properties in wastewaters, based on cyclic voltammograms and ANNs. 2. Experimental 2.1. Chemicals and Reagents The reagents used in this work were of analytical reagent grade and were used as received without any further purification. All solutions were prepared using Milli-Q water purified by a Milli-Q system (Millipore, Billerica, MA, USA). The following metallic NPs were used as modifiers during the fabrication of the electrodes: copper (Cu NPs, 40~60 nm) and copper (II) oxide (CuO NPs, <50 nm) which were purchased from Sigma-Aldrich (St. Louis, MO, USA), and nickel copper alloy (Ni Cu NPs, <80 nm) purchased from Nanografi Nano technology (Ankara, Turkey). Alumina powder (γ type), which was used for polishing the surface of the copper block, was purchased from CH Instruments (Austin, Chemosensors 2022, 10, 504 3 of 16 TX, USA). Glycine, sodium hydroxide and sodium chloride were purchased from Sigma- Aldrich (St. Louis, MO, USA). Glucose was obtained from Panreac (Barcelona, Spain). Graphite powder (<50 µm) and the epoxy resin kit, both used for the construction of the electrodes, were purchased from BDH (BDH Laboratory supplies, Poole, UK) and Resineco Green Composites (Barcelona, Spain), respectively. 2.2. Apparatus Voltammetric measurements were conducted in a PGSTAT 30 Autolab potentiostat (EcoChemie, Utrecht, The Netherlands) with GPES 4.7 version software (EcoChemie). Electrochemical measurements were carried out in a conventional three-electrode cell configuration composed of the 5 fabricated electrodes as the working electrodes; and a metallic platinum electrode (Crison 5261, Barcelona, Spain) and a Ag/AgCl electrode, as the auxiliary and reference electrodes, respectively. In the manner described below, a cyclic voltammogram was recorded for each of the electrodes by sweeping the potential between −0.1 V and +0.8 V, with a step potential of 0.01 V and a scan rate of 100 mV/s. For the morphological characterization of the electrodes, images were collected using scanning electron microscope (SEM) EVO® MA10 from Zeiss (Oberkochen, Germany). The commercial cuvettes for the sealed tube colorimetric COD determination were bought from HACH (HACH LANGE GMBH, Dusseldorf, Germany). Measurements were completed with the DR3900 laboratory spectrophotometer (HACH). 2.3. Electrode Fabrication Two distinct kinds of electrodes were used in this work: a surfaced-modified copper block electrode and 4 bulk-modified graphite-epoxy composite (GEC) electrodes. The electrodes were fabricated according to the previously reported procedures [14,21]. Briefly, in the case of the copper electrode, the electrode surface was first polished using alumina powder, and cleaned with Milli-Q water; copper (II) oxide (CuO) nanoparticles were then deposited on its surface by cycling the potential between −0.1 to +0.8 V in a 0.05 M NaOH solution for 50 cycles; this electrode will be referred as CuO/Cu in this report. As for the GEC electrodes, these were prepared by mixing 54% of graphite, 41% of the epoxy resin and 5% of the modifier. Next, this mixture was inserted into a PVC tube, where a connector had previously been introduced, and allowed to harden for 2 days at 40 ◦ C. In this manner, 4 different electrodes were prepared employing Cu NPs, CuO NPs, Ni Cu NPs and Ni NPs as the modifiers, which will be used to refer to the respective electrodes from now. 2.4. Data Analysis Collected voltammograms were first analyzed employing the GPES Multichannel 4.7 software package (EcoChemie, Utrecht, The Netherlands) to accomplish the analytical characterization of the different sensors’ response. In this manner, calibration plots were built from which the usual parameters such as the sensitivity, linearity, limit of detection (LOD), repeatability, etc., were assessed. For the estimation of the COD index, a chemometric model was built using ANNs as the modeling tool. Briefly, recorded voltammograms were unfolded, and causal index (CI) analysis was used to reduce the dimensionality of the recorded voltammograms [32]. Next, the ANN topology was optimized to find the best configuration that optimizes its performance. This data processing was done in Matlab 7.1 (MathWorks, Natick, MA, USA) by specific routines developed by the authors using the Neural Network toolbox. 2.5. Samples under Study For the characterization of the electrodes’ response, glucose and glycine stock solutions were prepared in 0.05 M NaOH, from which solutions of increasing concentration of each compound were prepared and measured with the different sensors. For the estimation of COD values, a set of samples consisting of mixtures of glucose and glycine were also prepared, with concentrations ranging from 0.25 mM to 4.00 mM Chemosensors 2022, 10, 504 4 of 16 of glucose and 0.40 mM to 6.25 mM of glycine, to simulate a real case in which different compounds with different reactivity are present. To obtain the ANN response model, a training set was used (16 samples); a further testing set (10 samples) was used to assess the model’s performance. The concentration of the former was based on a tilted factorial design with 2 factors and 4 levels (24 ), while the latter had the concentrations randomly distributed along the experimental domain defined by the former. Lastly, 5 real water samples were collected from different rivers and streams through- out Catalonia (Spain). The first sample (R1) was from a stream alongside a farm in Ripollet. Two more samples (R2 and R3) were from the two tributary streams of the river Llobregat in Gavà, flowing alongside the farms. R4 was from the river across Sant Boi de Llobre- gat. Finally, R5 was from a river near to a farm in Lleida. Additionally, these 5 samples were spiked with different concentrations of glucose and/or glycine (S1–S5) to assess the recovery values from the values predicted by the ANN. These real samples were analyzed both electrochemically with the proposed sensor array as well as with the colorimetric COD reference method. For the electrochemical measurements, samples were mixed with 0.1 M NaOH (1:1, v/v) to have the same supporting electrolyte. The voltammetric responses of the electrode array towards these samples were recorded in a cycled potential range from 0 to 1.0 V with a scan rate of 50 mV/s. The reference COD determinations of these samples were also conducted using dichro- mate reagent with the sealed tube colorimetric method. After placing the sample in the tube with pre-dosed reagents, these were heated for 2 h at 148 ◦ C. Cuvettes of two different measuring ranges were used, for 5~60 mg/L O2 and 100~2000 mg/L O2 , respectively. COD values were obtained after colorimetric reading, with factory-defined calibration curves. 3. Results and Discussion 3.1. Preparation of the CuO/Cu Electrode Copper has been reported as a powerful electrocatalyst for the oxidation of carbohy- drates and amino acids, which added to its good electrical conductivity [4], makes it an interesting material for the electrochemical estimation of COD [21–23]. Moreover, alkaline media have also been reported to improve the performance of copper-based electrodes for the oxidation of organic compounds [23]. Their electrocatalytic behavior lies in the action of radicals generated on the sensor surface [4]. As reported in previous studies [21–23], the performance of this kind of electrode can be affected by the hydroxide concentration and the pre-formation of a specific layer of CuO. Furthermore, Cu(III) species also take part in the reaction as an important electron transfer mediator [33]. The electrocatalytic behavior of CuO involved in the oxidation of organic compounds was proposed as follows by Silva et al. [21]: CuII O + OH− →CuIII OOH· + e− (electronic step) (2) CuIII OOH· + organic(reduced)→CuII O + organic(oxidized) + H2 O (chemical step) (3) For this reason, in this work, the use of a copper disk electrode modified with CuO NPs (CuO/Cu), which was also incorporated in the sensor array, was the first choice. CuO NPs were formed and deposited electrochemically as described in Section 2.3. As illustrated in Figure 1, the deposition showed three differentiated stages: The initial step was the formation of a first layer of copper (I) oxide (Cu2 O), corresponding to the anodic peak at around −0.30 V (vs. Ag/AgCl) observed in Figure 1a. The decrease of the current intensity at this potential occurred as the copper (I) transformed to Cu (II) after losing electrons, as indicated with down arrows in Figure 1a,b. The formation of a second mixed layer that evolves during preparation of the electrode was indicated by a broader anodic peak appearing at −0.10 V, as the right arrow indicates in Figure 1a,b. Finally, once stabilized, the broad anodic peak at around −0.05 V (vs. Ag/AgCl) increased as indicated in Figure 1c, Chemosensors 2022, 10, 504 5 of 16 suggesting the formation of a durable layer of a mixture of copper (II) oxide and copper (II) hydroxide. In all cases, the two cathodic peaks observed at −0.60 V and −0.90 V (vs. Ag/AgCl) correspond to the reduction of Cu(II)/Cu(I) and Cu(I)/Cu(0), respectively. Figure 1. Cyclic Figure voltammograms 1. Cyclic corresponding voltammograms corresponding to thetoelectrochemical the electrochemical formation offormation CuO on the of CuO surface of the copper, that alters the electrode surface in three distinct stages surface of the copper, that alters the electrode surface in three distinct stages (a–c), (a–c), observed during observed its preparation. Electrochemical measurements were performed at 100 mV/s in 0.05 M NaOH for its preparation. Electrochemical measurements were performed at 100 mV/s in 0.05 M NaO 50 cycles. cycles. During the electrochemical deposition, the color of the electrode surface changed from the initial glossy yellow of copper to dark brown, characteristic of copper oxide. To further During the electrochemical deposition, the color of the electrode surface ch confirm the surface modification, SEM images before and after the modification of the from the initialwere electrode glossy taken,yellow ofincopper as shown Figure 2.toAsdark brown, the CuO characteristic NPs formed, of copper it spread over the ox furtherelectrode, confirmwhich the surface modification, made the surface SEM look rougher. Thisimages beforeisand rougher surface after helpful the modifica to increase the electro-catalytic oxidation activity. The SEM images of the other the electrode were taken, as shown in Figure 2. As the CuO NPs formed, it sprea four NPs-modified electrodes are also shown in Figure 2c–f. the electrode, which made the surface look rougher. This rougher surface is hel 3.2.the increase Voltammetric Response of the electro-catalytic Electrodes activity. The SEM images of the other fou oxidation After the fabrication modified electrodes are alsoof the electrodes, shown the next 2c~f. in Figure step was the evaluation of their voltam- metric responses towards glucose and glycine. These were selected as model compounds prior to proceeding with the determination of the total COD. For this, 0.05 M NaOH was chosen as the supporting electrolyte, which provided the required mild alkaline condition [22]. The corresponding measurements were conducted with standard solutions of glucose and glycine (with COD value 400 mg/L O2 ) employing the prepared sensor array. As shown in Figure 3, glucose and glycine can be oxidized at the surface of all the considered electrodes, but with the oxidation occurring at different potentials for each of them. For example, the oxidation of glucose and glycine on the CuO/Cu electrode occurs at 0.5~0.6 V and 0.7~0.8 V, respectively, while in the case of the Ni NPs modified GEC electrode, oxidation of both occurred at around 0.6~0.7 V. Furthermore, it should be noted that the associated current intensities for the CuO/Cu electrode are much higher than those of the other electrodes, possibly attributable to the generated NPs on the electrode surface. The fact that the different electrodes each show different behavior towards the different analytes is highly desirable when developing ET applications. To complete the characterization, voltammetric responses of the different electrodes in the array were recorded against increasing theoretical COD values, for glucose and for glycine, respectively. The responses of the electrodes towards glucose are shown in Figure 4, with a COD range from 0 to 1000 mg/L O2 , while those towards glycine are shown in Figure 5, with a COD range from 0 to 400 mg/L O2 . Table 1 summarizes the fitted calibration lines obtained for the CuO/Cu electrode. mosensors 2022, 10, 504 2 of Chemosensors 2022, 10, 504 6 of 16 Figure 2. Scanning electron microscopy (SEM) characterization of (a) bare polished Cu block; (b) elec- Figure 2. Scanning electron microscopy (SEM) characterization of (a) bare polished Cu block; trodeposited electrode CuO/Cu and the four modified GECs with: (c) Cu NPs, (d) CuO NPs, (e) Ni electrodeposited electrode CuO/Cu and the four modified GECs with: (c) Cu NPs, (d) CuO NPs, Cu NPs, (f) Ni NPs. Ni Cu NPs, (f) Ni NPs. 3.2. Voltammetric Response of the Electrodes After the fabrication of the electrodes, the next step was the evaluation of their vo ammetric responses towards glucose and glycine. These were selected as model com tion, voltammetric responses of the different electrodes in the array were recorded ag increasing theoretical COD values, for glucose and for glycine, respectively. The respo of the electrodes towards glucose are shown in Figure 4, with a COD range from 0 to mg/L O2, while those towards glycine are shown in Figure 5, with a COD range from Chemosensors 2022, 10, 504 400 mg/L O2. Table 1 summarizes the fitted calibration lines obtained for the 7 ofCuO/Cu 16 trode. Chemosensors 2022, 10, 504 4 of 13 Figure 3. Voltammetric responses of the different electrodes obtained for (a) glucose and (b) glycine Figure solutions, 3. Voltammetric responses of the different electrodes obtained for (a) glucose and (b) gl with a theoretical COD value of 400 mg/L O2 in 0.05 M NaOH electrolyte with a scan rate solutions, of 50with mV/s.a theoretical COD value of 400 mg/L O2 in 0.05 M NaOH electrolyte with a scan of 50 mV/s. Figure 4. Voltammetric responses of the different electrodes—(a) CuO/Cu, (b) Cu NPs, (c) CuO NPs, Figure 4. Voltammetric responses of the different electrodes—(a) CuO/Cu, (b) Cu NPs, (c) CuO NPs (d) Ni Cu NPs and (e) Ni NPs—to glucose solutions of increasing theoretical COD values from 0 to (d) Ni Cu NPs and (e) Ni NPs—to glucose solutions of increasing theoretical COD values from 0 to 1015 mg/L O in 0.05 M NaOH with a scan rate of 50 mV/s. 1015 mg/L O2 in20.05 M NaOH with a scan rate of 50 mV/s. Table 1. Calibration lines of the electrode CuO/Cu towards the two organic compounds, glucose and glycine, at the fixed potentials of +0.6 and +0.7 V, respectively. Compound Equation Linear Range Limit of Detection Glucose y = 1.0281x + 85.90, R2 = 0.9996 30~800 mg/L O2 15.7 mg/L O2 Glycine y = 3.3315x + 167.19, R2 = 0.9991 10~180 mg/L O2 4.6 mg/L O2 Chemosensors 2022, 10, 504 5 of 13 Chemosensors 2022, 10, 504 8 of 16 Figure 5. Voltammetric responses of the different electrodes—(a) CuO/Cu, (b) Cu NPs, (c) CuO NPs, Figure 5. Voltammetric responses of the different electrodes—(a) CuO/Cu, (b) Cu NPs, (c) CuO NPs (d) Ni Cu NPs and (e) Ni NPs—to glycine solutions of increasing theoretical COD values from 0 to (d) Ni Cu NPs and (e) Ni NPs—to glycine solutions of increasing theoretical COD values from 0 to 400 mg/L O in 0.05 M NaOH with a scan rate of 50 mV/s. 400 mg/L O2 in20.05 M NaOH with a scan rate of 50 mV/s. 3.3. Repeatability Table 1. Calibration lines of the electrode CuO/Cu towards the two organic compounds, glucose To complete and glycine, thepotentials at the fixed analyticalofcharacterization +0.6 and +0.7 V,ofrespectively. the sensors’ response, the next step was studying its repeatability. Ensuring that repetitive measurements can be obtained when a new sensing device Compound is developed is always important, Equation but this is especially Linear Range Limit ofcritical Detection when developing ET applications, as the sensors need to withstand the larger number of Glucose y = 1.0281x + 85.90, R² = 0.9996 30~800 mg/L O2 15.7 mg/L O2 measurements, as required to build the chemometric model in comparison to a univariate Glycine y = 3.3315x + 167.19, R² = 0.9991 10~180 mg/L O2 calibration model. 4.6 mg/L O2 To this aim, the voltammetric responses of the five electrodes towards a mixed solution 3.3. Repeatability of glucose and glycine (equivalent to a COD value of 300 mg/L O2 ) were evaluated for 10 consecutive measurements, during which a blank (NaOH 0.05M) was also measured To complete the analytical characterization of the sensors’ response, the next step was in between. The corresponding measurements are shown in Figure 6, where the over- studying lappingitsbetween repeatability. Ensuring the different curvesthat repetitive clearly evidences measurements the good response canofbethe obtained sensors. when a newFurthermore, sensing device is developed is always important, but this is especially the relative standard deviation (RSD) was calculated over 10 measurements critical when developing based on ETthe applications, current intensity asatthe 0.6 sensors V and 0.7need V. As to canwithstand the larger be seen in Table number of meas- 2, the percentage urements, as required to build the chemometric model in comparison to athe RSD change was found to be smaller than 5% in all the cases, thus indicating good univariate cal- repeatability ibration model. of the electrodes’ response. To this aim, the voltammetric responses of the five electrodes towards a mixed solu- Table 2. Repeatability as the percentage of the RSD (RSD%) over 10 consecutive measurement cycles tionofofblank/standard glucose andforglycine (equivalent each of the electrodes. to a COD value of 300 mg/L O2) were evaluated for 10 consecutive measurements, during which a blank (NaOH 0.05M) was also meas- Sensor RSD% 1 RSD% 2 ured in between. The corresponding measurements are shown in Figure 6, where the over- CuO/Cu 2.91% lapping between the different curves clearly evidences the good response of the sensors 2.64% Cu NPs 3.60% 2.21% Furthermore,CuO the NPs relative standard deviation 3.95% (RSD) was calculated2.69% over 10 measurements based on theNi current Cu NPsintensity at 0.6 V and 0.7 V. As can be seen in 2.88% 0.69% Table 2, the percentage RSD change was Ni NPs found to be smaller than 3.19% 5% in all the cases, thus 2.80% indicating the good 1, 2 Calculated based on the current intensity at E = 0.6 V and E = 0.7 V, respectively. repeatability of the electrodes’ response. Chemosensors 2022, 10,2022, Chemosensors 504 10, 504 9 of 16 6 of 13 Figure 6. Ten repeated scans of cyclic voltammetric responses of the different electrodes—(a) CuO/Cu, Figure 6. Ten repeated scans of cyclic voltammetric responses of the different electrodes—(a) (b) Cu NPs, (c) CuO NPs, (d) Ni Cu NPs and (e) Ni NPs—to the mixture of glucose and glycine in CuO/Cu, (b) Cu NPs, (c) CuO NPs, (d) Ni Cu NPs and (e) Ni NPs—to the mixture of glucose and 0.05 M NaOH electrolyte with a scan rate of 50 mV/s. glycine in 0.05 M NaOH electrolyte with a scan rate of 50 mV/s. 3.4. Interference of Chloride Ions Table 2. Repeatability as the percentage of the RSD (RSD%) over 10 consecutive measurement cycles Unlike in for of blank/standard the each conventional COD test, the presence of halides does present a problem of the electrodes. when carrying out the estimation of COD electrochemically [3,6,34]. This problem features in the conventional Sensor test in that, on the one hand,1 halides can react with dichromate RSD% RSD%2 to produce the elemental form of the halogen and chromium(III), as shown in Equation (4), CuO/Cu 2.91% 2.64% leading to an overestimation of the COD value. On the other hand, to ensure the efficient oxidationCu of NPs 3.60% straight-chain aliphatic compounds, 2.21% silver sulfate is added as a catalyst which, CuO NPs in the presence of chloride or bromide, will 3.95% 2.69% lead to the precipitation of the corresponding silver salt, Ni Cu thus deactivating the catalyst and NPs leading to an underestimation of 0.69% the actual 2.88% COD value. With the intention of overcoming such interference, mercuric sulfate is used to Ni NPs 3.19% 2.80% complex chloride before the reflux. 1, 2 Calculated based on the current intensity at E = 0.6 V and E = 0.7 V, respectively. 6 Cl− + Cr2 O7 2− + 14 H+ → 3 Cl2 + 2 Cr3+ + 7 H2 O (4) 3.4. Interference of Chloride Ions Therefore, because of its common existence, chloride ion is one of the major factors Unlike in affecting CODthedetection. conventional In thisCOD test,the research, theinterference presence of of halides chloride does ion topresent a problem the voltam- when carrying metric outwas response the investigated estimation towards of CODincreasing electrochemically [3,6,34]. concentrations Thisadded of NaCl problem into afeatures mixture of glucose and glycine (equivalent to a COD value of in the conventional test in that, on the one hand, halides can react 300 mg/L O ). From Figure 2 with dichromate 7, to the voltammetric signals did not show any influence in the presence of Cl − , as could be produce the elemental form of the halogen and chromium(III), as shown in Equation (4), expected. Furthermore, the RSD% values were also calculated as carried out in Section 3.4, leading to an overestimation of the COD value. On the other hand, to ensure the efficient where the low values again indicate the good behavior of the system (Table 3). oxidation of straight-chain aliphatic compounds, silver sulfate is added as a catalyst which, in the presence of chloride or bromide, will lead to the precipitation of the corre- sponding silver salt, thus deactivating the catalyst and leading to an underestimation of the actual COD value. With the intention of overcoming such interference, mercuric sul- fate is used to complex chloride before the reflux. 6 Cl− + Cr2O72− + 14 H+→ 3 Cl2 + 2 Cr3+ + 7 H2O (4) Therefore, because of its common existence, chloride ion is one of the major factors affecting COD detection. In this research, the interference of chloride ion to the voltam- metric response was investigated towards increasing concentrations of NaCl added into Chemosensors 2022, 10, 504 7 of 1 Chemosensors 2022, 10, 504 10 of 16 Figure 7. Cyclic Figure 7. Cyclicvoltammetric responses voltammetric responses ofdifferent of the the different electrodes—(a) electrodes—(a) CuO/Cu, (b) CuO/Cu, (b)CuO Cu NPs, (c) Cu NPs, (c CuONPs, NPs, (d)(d) NiNPs Ni Cu Cu and NPs(e)and (e) Ni NPs—to Ni NPs—to the mixturethe mixture of glucose andofglycine glucose and in 0.05 Mglycine in 0.05 M NaOH NaOH electrolyte electrolyte with an with an increasing increasing of Cl− . Scan rate: concentration concentration of Cl50−. mV/s. Scan rate: 50 mV/s. 3. Change in the variation − - as the Table Table 3. Change in the variation of of thethe voltammetric signal voltammetric signal due due to to the the presence presenceofofClCl as the percen percentage of the RSD (RSD%) for each of the age of the RSD (RSD%) for each of the electrodes. electrodes. Sensor RSD% 1 RSD% 2RSD%2 Sensor RSD%1 CuO/Cu 1.82% 1.43% CuO/Cu 1.82% 1.43% Cu NPs 1.63% 2.22% CuCuONPsNPs 1.63% 2.84% 1.97% 2.22% CuO Ni Cu NPsNPs 2.94% 2.84% 1.00% 1.97% Ni NPs 4.58% 1.55% Ni Cu NPs 2.94% 1, 2 Calculated based on the current intensities at E = 0.6 V and E = 0.7 V, respectively. 1.00% Ni NPs 4.58% 1.55% Calculated 1, 2 based Analysis 3.5. Quantitative on the current intensities at E=0.6 V and E=0.7 V, respectively. of Real Samples 3.5.1. Calibration Curve 3.5. Quantitative VoltammetricAnalysis of Real responses Samples of the CuO/Cu electrode towards increasing concentrations of glucose and glycine have been shown in Figure 4. As can be seen, the oxidation peaks 3.5.1. Calibration Curve corresponding to glucose and glycine increase with their increasing concentrations, and Voltammetric also shift to higher responses potentials.ofNevertheless, the CuO/Cuthe electrode towards calibration increasing curves relating concentration the current of glucose intensityand at theglycine have been peak maximum withshown in Figure the theoretical COD4. As can(calculated values be seen, the oxidation according to peak I (µA) = m ∗ c (mg/L O ) + b) were built, showing a linear behavior as corresponding to glucose and glycine increase with their increasing concentrations, 2 shown in Figure 8. an The analytical parameters derived from these are summarized in Table 4, where the close- also shift to higher potentials. Nevertheless, the2calibration curves relating the current in ness to 1.0 of the correlation of determination (R ) confirms the good linearity within tensity at the peak the specified maximum range with theThe of concentrations. theoretical COD values limit of detection (calculated was calculated according using the to (μA)3σ=IUPAC m∗c (mg/L O2) + b) were built, showing a linear behavior as shown in Figure 8. Th criterion. analytical parameters derived from these are summarized in Table 4, where the closenes to 1.0 of the correlation of determination (R2) confirms the good linearity within the spec ified range of concentrations. The limit of detection was calculated using the 3σ IUPAC criterion. Chemosensors 2022, 10, 504 8 of 13 Chemosensors 2022, 10, 504 11 of 16 Figure 8. Calibration Figure 8. Calibrationcurves obtained curves (black from dotted the obtained curve) peak heights of peak from the the voltammetric responses of the heights of the voltammetric CuO/Cu electrodes responses (shown in of the CuO/Cu Figures (shown electrodes 4a and 5a) to increasing in Figures 4a and (a) 5a)glucose and (b) to increasing (a) glycine. glucose and (b) glycine. Red lines display the fitted least squares fittings to the linear zone. Table 4. Calibration lines of the CuO/Cu electrode based on measurement to glucose or glycine, from currents Table at oxidation 4. Calibration peak lines of maxima.electrode based on measurement to glucose or glycine, from the CuO/Cu currents at oxidation peak maxima. Sensitivity Intercept Linear Range Limit of Detection Compound R2 Sensitivity Intercept Linear Range Limit of Detection Compound (μA/mg/L O2) (μA) R2 (mg/L O2) (mg/L O2) (µA/mg/L O2 ) (µA) (mg/L O2 ) (mg/L O2 ) Glucose 0.9732 1.48 0.9997 20–800 13.6 Glucose 0.9732 1.48 0.9997 20–800 13.6 Glycine 2.662 5.94 0.9998 20–200 4.1 Glycine 2.662 5.94 0.9998 20–200 4.1 Despite the best behavior in terms of sensitivity was obtained for the CuO/Cu elec- Despite the best behavior in terms of sensitivity was obtained for the CuO/Cu elec- trode, a good behavior was also obtained for the other ones (calibrations not shown). An- trode, a good behavior was also obtained for the other ones (calibrations not shown). other remarkable Another detail, remarkable observable detail, from observable fromthethegraphs graphsininFigures Figures 44 and and 5,5,isisthe thecross- cross-re- sponse between the different electrodes, not only from the different potential response between the different electrodes, not only from the different potential associ- associated to the atedvarious oxidation to the various peaks,peaks, oxidation but also fromfrom but also the the different sensitivities different sensitivitiesobtained obtainedfor for each of the eachelectrodes and compounds. of the electrodes Overall, and compounds. this isthis Overall, a situation highly is a situation desirable highly desirable when when devel- developing oping ETs. ETs. 3.5.2. Artificial Neural Network Method 3.5.2. Artificial Neural Network Method Upon completing the characterization of the sensors, the next step was building the Upon actual completing model that willthe characterization allow the estimation of of the COD sensors, values the next in real step was samples. Thebuilding aim is the actual model that will allow the estimation of COD values in real samples. that the use of an artificial intelligence response model will allow a better estimation The aimasis that theituse of an artificial combines intelligence the responses response from different model sensors andwill allow a better compensates estimationinasthe the differences it com- bines the responses voltammetric from towards responses differentdifferent sensorscompounds and compensates (e.g., peakthe differences position, in theetc.). sensitivity, voltam- In the interest metric responses of proving towards our hypothesis, different compounds glucose (e.g.,and peakglycine were used position, as standards. sensitivity, etc.). Next, with the conditions described in Section 2.5, the samples consisting In the interest of proving our hypothesis, glucose and glycine were used as of binary mixtures standards. of glucose and glycine were prepared and analyzed employing the sensor array. As Next, with the conditions described in Section 2.5, the samples consisting of binary mix- evidenced from Figure 6, upon mixing the two organic compounds, there are no longer tures of glucose and glycine were prepared and analyzed employing the sensor array. As two separate peaks, but a broader peak is obtained instead. However, different sensitivity evidenced from is obtained for Figure 6, upon mixing each compound, thus not the two being organic compounds, as straightforward there are into the interpolation no longer a twotheseparate previouspeaks, simplebut a broader univariate peak iscurve. calibration obtained instead. For this reason,However, ANNs were different chosen sensitivity as the is obtained for the tool to build each compound, multivariate thus notmodel. calibration being as straightforward the interpolation into a the previous simple univariate calibration curve. After a trial-and-error process in which For this the topology reason, of the ANN ANNs were chosen was optimized, the as thefinal tool model to buildhadthe91 multivariate neurons in thecalibration input layer model. (corresponding to the selected points of the voltammograms from the five sensors), After a trial-and-error process in which the seven neurons topologyand logsig of thetransfer ANN was function in the the optimized, hidden layer, and one neuron and purelin transfer function in the output layer. During the final model had 91 neurons in the input layer (corresponding to the selected points of the ANN optimization, the inputs were pruned by means of CI analysis, while the number of voltammograms from the five sensors), seven neurons and logsig transfer function in the hidden layer, and one neuron and purelin transfer function in the output layer. During the ANN optimization, the inputs were pruned by means of CI analysis, while the number of neurons and transfer functions were systematically varied to identify the configuration that led to the best performance employing only the data of the training subset. Next, the Chemosensors 2022, 10, 504 12 of 16 Chemosensors 2022, 10, 504 9 of 13 neurons and transfer functions were systematically varied to identify the configuration that led to the best performance employing only the data of the training subset. Next, the predicted vs. expected COD were built (Figure 9). As can be seen, a satisfactory perfor- samples of the test subset were interpolated into the model and the comparison graphs mance was obtained for both subsets, with regression lines close to the ideal identity ones. of predicted vs. expected COD were built (Figure 9). As can be seen, a satisfactory With the aim was performance of numerically obtained for quantifying the performance, both subsets, with regression linesthe regression close to the idealparameters identity (as well as With ones. the confidence intervals for the aim of numerically them) and quantifying the the root mean performance, thesquare error regression (RMSE) of the parameters fitting were calculated (Table 5). From the observed values, a good trend (as well as the confidence intervals for them) and the root mean square error (RMSE) is confirmed, of as thetheideal values fitting of slope and were calculated intercept (Table 5). Fromwere within values, the observed the confidence a good trendintervals 1 and 0, re- is confirmed, as the ideal spectively; values of slope furthermore, theand intercept of coefficient were within the confidence determination was close intervals to 1 and1 and the0,RMSE respectively; furthermore, value was relatively low. the coefficient of determination was close to 1 and the RMSE value was relatively low. Figure Figure9.9.Modeling Modeling ability of the ability of theoptimized optimized CI-ANN CI-ANN for for the the estimation estimation of COD of COD value.value. Comparison Comparison graphs graphsofofpredicted predictedvs. vs.expected expectedforforthe the(a) (a)training trainingand and (b) (b) testing testing subsets. subsets. The The dashed line corre- dashed line sponds to the to corresponds ideal comparison the ideal comparisonlineline (y=x). (y = x). Table Table5.5.Fitted regressionlines Fitted regression lines forfor the the comparison comparison between between obtained obtained vs. expected vs. expected values values for the trainfor the train and test subsets. and test subsets. DataData SetSet Slope Slope Intercept Intercept (mg/L (mg/L O2 ) O 2) R2 R2 RMSE RMSE (mg/L(mg/L O2 ) O2) Training Training 0.997 ± 0.027 0.997 ± 0.027 6 ±618 ± 18 0.9980.998 11.3 11.3 Testing Testing 0.996 0.996 ± 0.077 ± 0.077 4 ±445 ± 45 0.9910.991 16.6 16.6 Finally, upon Finally, upon construction construction and validation and validationof theofmodel with the the model standard with samples, the the standard samples, last the step last waswas step its application to the analysis its application of real samples. to the analysis of realThe analysisThe samples. of the real samples analysis of the real didn’t require samples didn’tany pre-treatment require apart from the any pre-treatment dilution apart from 1:1the in the measuring dilution media 1:1 in the as we measuring need to preserve the same experimental conditions as in the calibration. In this manner, media as we need to preserve the same experimental conditions as in the calibration. In a total of 10 samples were analyzed, five (samples R1~R5) of them corresponding to the this manner, a total of 10 samples were analyzed, five (samples R1~R5) of them corre- above described in Section 2.5, plus those with spiked concentrations of glucose and glycine sponding (samples to the above described in Section 2.5, plus those with spiked concentrations of S1~S5). glucose andinterpolating After glycine (samples S1~S5). responses into the model, the predicted COD values were com- After pared withinterpolating those obtained responses employingintothe thereference model, method, the predicted and withCOD thevalues were com- direct inter- pared withinthose polation, obtained this case usingemploying the CuO/Cu theelectrode reference method, alone (Tableand with the 6). From thosedirect interpo- values, two remarks lation, have using in this case to be made. On the one the CuO/Cu side, it alone electrode can be(Table seen how 6). the From COD values those for two values, the realhave remarks samples aremade. to be close to Onthethe LODs oneof the individual side, it can be seensensors, howand theconsequently COD valuesout forofthe real the range of the model. On the other side, it should be taken into account that the various samples are close to the LODs of the individual sensors, and consequently out of the range of the model. On the other side, it should be taken into account that the various spiked samples correspond to the different real samples considered, thus meaning that the good behavior is obtained even when different samples are being considered. Chemosensors 2022, 10, 504 13 of 16 spiked samples correspond to the different real samples considered, thus meaning that the good behavior is obtained even when different samples are being considered. Table 6. Results obtained for the real and spiked samples using different methods for estimating the COD values. Spiked COD y = 0.9732x + 1.4830 a y = 2.662x − 5.937 b ANN Colorimetric Sample (mg/L O2 ) COD (mg/L O2 ) COD (mg/L O2 ) COD (mg/L O2 ) COD (mg/L O2 ) R1 – 99.32 41.88 – 12.10 R2 – 102.7 43.12 – 15.10 R3 – 98.06 41.43 – 9.21 R4 – 100.8 42.43 – 5.24 R5 – 105.6 44.18 – 11.70 S1 187.3 (glycine) 481.8 178.9 141.9 216.0 S2 376.5 (glucose) 402.1 149.8 284.7 416.0 S3 320.0 (glucose + glycine) 361.1 134.8 286.7 342.0 S4 234.4 (glucose + glycine) 466.7 173.4 211.0 260.0 S5 686.5 (glucose + glycine) 868.3 320.2 659.7 719.0 aequation corresponding to the calibration of the Cu/CuO disk electrode at the first oxidation peak; b equation corresponding to the calibration of the Cu/CuO disk electrode at the second oxidation peak. With the aim of evaluating which method displayed better agreement with what, an exploratory analysis of the different methodologies was done, in which a comparison of the results from paired samples was performed between the two methodologies, and the regression line calculated. In order to avoid bias in the lowest COD values, results below the LOD were discarded. These results are summarized in the Table 7. In light of these, there are a few observations to be made. First, it is noteworthy how the standard interpolation with the voltammetric data is not capable of obtaining correct results, with limited linearities, and errors in defect. Second, the spiked concentrations could be reasonably predicted, both with the colorimetric reference method (as was expected), but also with the ET method. Third, the best linearity found throughout the comparison, also remarkably close to the identity line, was the one of the colorimetric results vs. those of the ET methodology. This best comparison line is shown in Figure 10, and as can be seen, a good agreement was again obtained between both values, confirming the suitability of the approach. Table 7. Exploratory analysis of the COD results (mg/L O2 ) obtained for the spiked samples with the different methodologies being compared. Results Compared Comparison Line R2 (n = 5) ET vs. Spiked Y = 1.0132 × X − 48.9 0.9383 Colorimetric vs. Spiked Y = 1.0134 × X + 24.8 0.9553 Volt 1 a vs. Spiked Y = 0.8532 × X + 208 0.5384 Volt 2 b vs. Spiked Y = 0.8701 × X + 240 0.7413 ET vs. Colorimetric Y = 0.9949 × X − 71.9 0.9723 c Volt 1 a vs. Colorimetric Y = 0.3119 × X + 78.8 0.5385 Volt 2 b vs. Colorimetric Y = 0.3181 × X + 90.7 0.7414 aCurrent intensity corresponding to the first oxidation peak. b Current intensities corresponding to the second oxidation peak. c Best determination coefficient. Lastly, although it is true that other ET systems have already been reported for the es- timation of COD values in water [35], one of the key advantages of the approach proposed herein is that the calibration is carried out using standard solutions of model compounds instead of requiring the analysis of some samples previously analyzed with a reference method. Moreover, despite the comparison not being straightforward given disparities in the nature of the considered samples and chemometric tools, a better correlation be- tween the estimated and actual values seems to be obtained in our case than in previous Chemosensors 2022, 10, 504 14 of 16 systems [36–40]. This can be attributed to the nature of the sensors used here. Previous approaches relied on chemo-sensors that were based on bare metallic electrodes, whereas sensors modified with different NPs have been considered herein, with the well-known advantages of NPs in sensors. In addition, some previous approaches relied on the use of enzymatic biosensors, and while it is true that these might typically show a better perfor- mance, the presence of contaminants in wastewater can also have a harmful effect on them. Chemosensors 2022, 10, 504 This was already stated by Czolkos et al., who reported that when only COD needs11 of to 13 be estimated, the use of a single carbon electrode can yield similar results than an array of enzyme-coated electrodes, obviously a much simpler and cheaper option [39]. Figure 10. 10. Figure Comparison Comparisonof results obtained of results for for obtained the the spiked realreal spiked samples, between samples, the the between proposed ET ET proposed approach and the reference colorimetric method. approach and the reference colorimetric method. 4. Lastly, Conclusions although it is true that other ET systems have already been reported for the estimation of COD In this work,values in water one surface [35], one modified of thedisk Cu/CuO keyelectrode advantages andof themetal four approach pro- nanoparticle- modified posed hereingraphite electrodes is that the wereisfabricated calibration carried out to using be employed standard in solutions the research for the com- of model analysis of quantitative pounds instead of organic requiringload. the Glucose analysis and glycine of some werepreviously samples chosen to be standard analyzed substances with a ref- representative erence of organicdespite method. Moreover, contaminants. Based on the comparison notthe electrochemical being straightforward voltammetric given dis- re- sponses parities of these in the natureelectrodes, a one-sensor of the considered samples calibration curve method and chemometric tools,was built, a better and two correla- tioncalibration between the curves were obtained estimated and actualfromvalues peak heights seems to of be voltammetric obtained inresponses to glucose our case than in and glycine, previous systems respectively, [36–40]. Thiswhich can beshowed goodtocorrelations attributed the nature in of their linear range the sensors of each used here. one-analyte Previous system. approaches However, relied the application on chemo-sensors that of both were calibration based on bare curves metallicinelectrodes, the practical detection was limited because of the multi-peak responses in multi-analyte whereas sensors modified with different NPs have been considered herein, with the well- samples. Nev- ertheless, an artificial intelligence response model allows a better known advantages of NPs in sensors. In addition, some previous approaches relied on the estimation by combining usethe of responses enzymaticfrom differentand biosensors, sensors whileandit iscompensating true that these the differences might typically in the showvoltammetric a better responses to performance, thedifferent presencecompounds. of contaminantsTherefore, a five-sensor in wastewater ET array can also have aemployed with an harmful effect ANN model was built. Finally, both methods were applied on them. This was already stated by Czolkos et al., who reported that when only COD to real freshwater samples for analysis needs and prediction. to be estimated, the use of Bya comparing single carbon results detected electrode can from yield those similarwith a reference results than an array of enzyme-coated electrodes, obviously a much simpler and cheaper optionthan colorimetric method, the proposed ET model showed a much better performance [39].the single-sensor calibration curve method. With this research, we have demonstrated a high value application with neural net- 4. Conclusions works as useful tools to solve a difficult problem in the water and wastewater field [41], In this work, one surface modified Cu/CuO disk electrode and four metal nanoparti- here in relation to the prediction of pollution from sensor data. cle-modified graphite electrodes were fabricated to be employed in the research for the analysis Authorof quantitative Conceptualization, Contributions: organic load. GlucoseM.d.V.; and glycine M.d.V.; methodology, were chosen software,toX.C.; be standard investigation, substances representativedraft Q.W.; writing—original of organic contaminants. preparation, Q.W. andBased on the electrochemical X.C.; writing—review voltam- and editing, M.d.V.; metric responses funding of these acquisition, M.d.V.electrodes, All authors ahave one-sensor calibration read and agreed to the curve method published versionwas built, of the and manuscript. two calibration curves were obtained from peak heights of voltammetric responses to glu- Funding: This research was funded by the Spanish Ministry of Science and Innovation, MCINN cose and glycine, respectively, which showed good correlations in their linear range of (Madrid), grant number PID2019-107102RB-C21/AEI/10.13039/5001100011033. Q.W. acknowledges eachtheone-analyte concession ofsystem. However, a PhD grant the application of the Chinese Scholarshipof both Council calibration (China). curves in the prac- tical detection was limited because of the multi-peak responses in multi-analyte samples. Nevertheless, an artificial intelligence response model allows a better estimation by com- bining the responses from different sensors and compensating the differences in the volt- ammetric responses to different compounds. Therefore, a five-sensor ET array employed with an ANN model was built. Finally, both methods were applied to real freshwater Chemosensors 2022, 10, 504 15 of 16 Institutional Review Board Statement: Not applicable. Informed Consent Statement: Not applicable. Data Availability Statement: Not applicable. Acknowledgments: We acknowledge the help provided by Diana Losantos Velasco, from the Depart- ment of Chemical, Biological and Environmental Engineering, Universitat Autònoma de Barcelona, in the colorimetric determination of COD of the real samples. Conflicts of Interest: The authors declare no conflict of interest. References 1. Alves, N.A.; Olean-Oliveira, A.; Cardoso, C.X.; Teixeira, M.F.S. Photochemiresistor sensor development based on a bismuth vanadate type semiconductor for determination of chemical oxygen demand. ACS Appl. Mater. Interfaces 2020, 12, 18723–18729. [CrossRef] [PubMed] 2. Hao, N.; Dai, Z.; Xiong, M.; Han, X.; Zuo, Y.; Qian, J.; Wang, K. Rapid potentiometric detection of chemical oxygen demand using a portable self-powered sensor chip. Anal. Chem. 2021, 93, 8393–8398. [CrossRef] [PubMed] 3. Ramon, R.; Valero, F.; del Valle, M. Rapid determination of chemical oxygen demand using a focused microwave heating system featuring temperature control. Anal. Chim. Acta 2003, 491, 99–109. [CrossRef] 4. Huang, X.; Zhu, Y.; Yang, W.; Jiang, A.; Jin, X.; Zhang, Y.; Yan, L.; Zhang, G.; Liu, Z. A self-supported CuO/Cu nanowire electrode as highly efficient sensor for COD measurement. Molecules 2019, 24, 3132. [CrossRef] 5. Gutiérrez-Capitán, M.; Baldi, A.; Gómez, R.; García, V.; Jiménez-Jorquera, C.; Fernández-Sánchez, C. Electrochemical nanocomposite-derived sensor for the analysis of chemical oxygen demand in urban wastewaters. Anal. Chem. 2015, 87, 2152–2160. [CrossRef] 6. Sawyer, C.N.; McCarty, P.L.; Parkin, G.F. Chemistry for Environmental Engineering and Science, 5th ed.; McGraw-Hill Education: New York, NY, USA, 2002; pp. 625–626. 7. Yu, H.; Ma, C.; Quan, X.; Chen, S.; Zhao, H. Flow injection analysis of chemical oxygen demand (COD) by using a boron-doped diamond (BDD) electrode. Environ. Sci. Technol. 2009, 43, 1935–1939. [CrossRef] 8. Diksy, Y.; Rahmawati, I.; Jiwanti, P.K.; Ivandini, T.A. Nano-Cu modified Cu and nano-Cu modified graphite electrodes for chemical oxygen demand sensors. Anal. Sci. 2020, 36, 1323–1330. [CrossRef] 9. Zhou, Y.; Jing, T.; Hao, Q.; Zhou, Y.; Mei, S. A sensitive and environmentally friendly method for determination of chemical oxygen demand using NiCu alloy electrode. Electrochim. Acta. 2012, 74, 165–170. [CrossRef] 10. Zhang, Z.; Chang, X.; Chen, A. Determination of chemical oxygen demand based on photoelectrocatalysis of nanoporous TiO2 electrodes. Sens. Actuators B Chem. 2016, 223, 664–670. [CrossRef] 11. Jin, W.; Maduraiveeran, G. Electrochemical detection of chemical pollutants based on gold nanomaterials. Trends Environ. Anal. Chem. 2017, 14, 28–36. [CrossRef] 12. Orozco, J.; Fernández-Sánchez, C.; Mendoza, E.; Baeza, M.; Céspedes, F.; Jiménez-Jorquera, C. Composite planar electrode for sensing electrochemical oxygen demand. Anal. Chim. Acta 2008, 607, 176–182. [CrossRef] [PubMed] 13. Cheng, Q.; Wu, C.; Chen, J.; Zhou, Y.; Wu, K. Electrochemical tuning the activity of nickel nanoparticle and application in sensitive detection of chemical oxygen demand. J. Phys. Chem. C 2011, 115, 22845–22850. [CrossRef] 14. Wang, Q.; del Valle, M. Determination of chemical oxygen semand (COD) using nanoparticle-modified voltammetric sensors and electronic tongue principles. Chemosensors 2021, 9, 46. [CrossRef] 15. Duan, W.; Gunes, M.; Baldi, A.; Gich, M.; Fernández-Sánchez, C. Compact fluidic electrochemical sensor platform for on-line monitoring of chemical oxygen demand in urban wastewater. Chem. Eng. J. 2022, 449, 137837. [CrossRef] 16. Hou, Y.; Hu, J.; Liu, L.; Zhang, J.; Cao, C. Effect of calcination temperature on electrocatalytic activities of Ti/IrO2 electrodes in methanol aqueous solutions. Electrochim. Acta 2006, 51, 6258–6267. [CrossRef] 17. Wang, J.Q.; Wu, C.; Wu, K.B.; Cheng, Q.; Zhou, Y.K. Electrochemical sensing chemical oxygen demand based on the catalytic activity of cobalt oxide film. Anal. Chim. Acta 2012, 736, 55–61. [CrossRef] 18. Ai, S.; Gao, M.; Yang, Y.; Li, J.; Jin, L. Electrocatalytic sensor for the determination of chemical oxygen demand using a lead dioxide modified electrode. Electroanalysis 2004, 16, 404–409. [CrossRef] 19. Li, J.; Li, L.; Zheng, L.; Xian, Y.; Jin, L. Rh2 O3 /Ti electrode preparation using laser anneal and its application to the determination of chemical oxygen demand. Meas. Sci. Technol. 2006, 17, 1995–2000. [CrossRef] 20. Ma, C.; Tan, F.; Zhao, H.; Chen, S.; Quan, X. Sensitive amperometric determination of chemical oxygen demand using Ti/Sb– SnO2 /PbO2 composite electrode. Sens. Actuators B Chem. 2011, 155, 114–119. [CrossRef] 21. Silva, C.R.; Conceição, C.D.C.; Bonifácio, V.G.; Fatibelho Filho, O.; Teixeira, M.F.S. Determination of the chemical oxygen demand (COD) using a copper electrode: A clean alternative method. J. Solid State Electrochem. 2009, 13, 665–669. [CrossRef] 22. Carchi, T.; Lapo, B.; Alvarado, J.; Montero, P.J.E.; Llorca, J.; Fernández, L. A nafion film cover to enhance the analytical performance of the CuO/Cu electrochemical sensor for determination of chemical oxygen demand. Sensors 2019, 19, 669. [CrossRef] [PubMed] 23. Hassan, H.H.; Badr, I.H.A.; Abdel-Fatah, H.T.M.; Elfeky, E.M.S.; Abdel-Aziz, A.M. Low cost chemical oxygen demand sensor based on electrodeposited nano-copper film. Arab. J. Chem. 2018, 11, 171–180. [CrossRef] Chemosensors 2022, 10, 504 16 of 16 24. Hanrahan, G. Computational neural networks driving complex analytical problem solving. Anal. Chem. 2010, 82, 4307–4313. [CrossRef] 25. Liu, J.; Xu, Y.; Liu, S.; Yu, S.; Yu, Z.; Low, S.S. Application and progress of chemometrics in voltammetric biosensing. Biosensors 2022, 12, 494. [CrossRef] [PubMed] 26. Aroca-Santos, R.; Cancilla, J.C.; Matute, G.; Torrecilla, J.S. Identifying and quantifying adulterants in extra virgin olive oil of the picual varietal by absorption spectroscopy and nonlinear modeling. J. Agric. Food Chem. 2015, 63, 5646–5652. [CrossRef] 27. Gutés, A.; Calvo, D.; Céspedes, F.; del Valle, M. Automatic sequential injection analysis electronic tongue with integrated reference electrode for the determination of ascorbic acid, uric acid and paracetamol. Microchim. Acta 2007, 157, 1–6. [CrossRef] 28. Cetó, X.; Gonzalez-Calabuig, A.; Capdevila, J.; Puig-Pujol, A.; del Valle, M. Instrumental measurement of wine sensory descriptors using a voltammetric electronic tongue. Sens. Actuators B Chem. 2015, 207, 1053–1059. [CrossRef] 29. González-Calabuig, A.; del Valle, M. Voltammetric electronic tongue to identify Brett character in wines. On-site quantification of its ethylphenol metabolites. Talanta 2018, 179, 70–74. [CrossRef] 30. Serrano, N.; González-Calabuig, A.; del Valle, M. Crown ether-modified electrodes for the simultaneous stripping voltammetric determination of Cd(II), Pb(II) and Cu(II). Talanta 2015, 138, 130–137. [CrossRef] 31. del Valle, M. Sensor arrays and electronic tongue systems. Int. J. Electrochem. 2012, 2012, 986025. [CrossRef] 32. Cetó, X.; Céspedes, F.; del Valle, M. Comparison of methods for the processing of voltammetric electronic tongues data. Microchim. Acta 2013, 180, 319–330. [CrossRef] 33. Dhara, K.; Mahapatra, D.R. Electrochemical nonenzymatic sensing of glucose using advanced nanomaterials. Microchim. Acta 2018, 185, 49. 34. American Public Health Association; American Water Works Association; Water Environment Federation. Standard Methods for the Examination of Water and Wastewater, 23rd ed.; American Public Health Association: Washington, DC, USA; American Water Works Association: Washington, DC, USA; Water Environment Federation: Washington, DC, USA, 2017. 35. Cetó, X.; del Valle, M. Electronic tongue applications for wastewater and soil analysis. iScience 2022, 25, 104304. [CrossRef] [PubMed] 36. Tønning, E.; Sapelnikova, S.; Christensen, J.; Carlsson, C.; Winther-Nielsen, M.; Dock, E.; Solna, R.; Skladal, P.; Nørgaard, L.; Ruzgas, T.; et al. Chemometric exploration of an amperometric biosensor array for fast determination of wastewater quality. Biosens. Bioelectron. 2005, 21, 608–617. [CrossRef] 37. Gutés, A.; Cespedes, F.; del Valle, M.; Louthander, D.; Krantz-Rülcker, C.; Winquist, F. A flow injection voltammetric electronic tongue applied to paper mill industrial waters. Sens. Actuators B Chem. 2006, 115, 390–395. [CrossRef] 38. Campos, I.; Alcañiz, M.; Aguado, D.; Barat, R.; Ferrer, J.; Gil, L.; Marrakchi, M.; Martínez-Mañez, R.; Soto, J.; Vivancos, J.-L. A voltammetric electronic tongue as tool for water quality monitoring in wastewater treatment plants. Water Res. 2012, 46, 2605–2614. [CrossRef] 39. Czolkos, I.; Dock, E.; Tønning, E.; Christensen, J.; Winther-Nielsen, M.; Carlsson, C.; Mojzíková, R.; Skládal, P.; Wollenberger, U.; Nørgaard, L.; et al. Prediction of wastewater quality using amperometric bioelectronic tongues. Biosens. Bioelectron. 2016, 75, 375–382. [CrossRef] 40. Legin, E.; Zadorozhnaya, O.; Khaydukova, M.; Kirsanov, D.; Rybakin, V.; Zagrebin, A.; Ignatyeva, N.; Ashina, J.; Sarkar, S.; Mukherjee, S.; et al. Rapid evaluation of integral quality and safety of surface and waste waters by a multisensor system (electronic tongue). Sensors 2019, 19, 2019. [CrossRef] 41. Bui, H.M.; Bui, H.N.; Le, T.M.; Karri, R.R. Application of artificial neural networks on water and wastewater prediction: A review. In Soft Computing Techniques in Solid Waste and Wastewater Management; Karri, R.R., Ravindran, G., Dehghani, M.H., Eds.; Elsevier: Amsterdam, The Netherlands, 2021; pp. 95–109.