**Simultaneous Quantification of Four Principal NSAIDs through Voltammetry and Artificial Neural Networks Using a Modified Carbon Paste Electrode in Pharmaceutical Samples †**

**Guadalupe Yoselin Aguilar-Lira 1, Prisciliano Hernandez <sup>2</sup> , Giaan Arturo Álvarez-Romero <sup>1</sup> and Juan Manuel Gutiérrez 3,\***


**Abstract:** This work describes the development of a novel and low-cost methodology for the simultaneous quantification of four main nonsteroidal anti-inflammatory drugs (NSAIDs) in pharmaceutical samples using differential pulse voltammetry coupled with an artificial neural network model (ANN). The working electrode used as a detector was a carbon paste electrode (CPE) modified with multiwall carbon nanotubes (MWCNT-CPE). The specific voltammetric determination of the drugs was performed by cyclic voltammetry (CV). Some characteristic anodic peaks were found at potentials of 0.446, 0.629, 0.883 V related to paracetamol, diclofenac, and aspirin. For naproxen, two anodic peaks were found at 0.888 and 1.14 V and for ibuprofen, an anodic peak was not observed at an optimum pH of 10 in 0.1 mol L−<sup>1</sup> Britton–Robinson buffer. Since these drug's oxidation process turned out to be irreversible and diffusion-controlled, drug quantification was carried out by differential pulse voltammetry (DPV). The Box Behnken design technique's optimal parameters were: step potential of 5.85 mV, the amplitude of 50 mV, period of 750 ms, and a pulse width of 50 ms. A data pretreatment was carried out using the Discrete Wavelet Transform using the db4 wavelet at the fourth decomposition level applied to the voltammetric records obtained. An ANN was built to interpret the obtained approximation coefficients of voltammograms generated at different drug concentrations to calibrate the system. The ANN model's architecture is based on a Multilayer Perceptron Network (MLP) that employed a Bayesian regularization training algorithm. The trained MLP achieves significant R values for the test data to simultaneous quantification of the four drugs in the presence of aspirin.

**Keywords:** carbon paste electrode; voltammetry; artificial neural network; quantification; nonsteroidal anti-inflammatory

#### **1. Introduction**

Nonsteroidal anti-inflammatory analgesics NSAIDs are important drugs worldwide due to their low cost and easy accessibility. Most of these drugs can be purchased without a prescription; they are widely used to relieve pain, reduce inflammation and reduce high temperatures. Standing out from this large group of NSAIDs are paracetamol, diclofenac, naproxen, aspirin, and ibuprofen, which are the most frequently used. The pharmacological action of these drugs is that they block the enzyme cyclooxygenase and break down the prostaglandins produced by the cells of the body that increase inflammation, pain, and

**Citation:** Aguilar-Lira, G.Y.; Hernandez, P.; Álvarez-Romero, G.A.; Gutiérrez, J.M. Simultaneous Quantification of Four Principal NSAIDs through Voltammetry and Artificial Neural Networks Using a Modified Carbon Paste Electrode in Pharmaceutical Samples. *Chem. Proc.* **2021**, *5*, 3. https://doi.org/ 10.3390/CSAC2021-10450

Academic Editor: Frederic Melin

Published: 30 June 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 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/).

fever. Although NSAIDs are commonly used, they are not suitable for everyone and can sometimes cause adverse side effects if their use is constant: peptic ulceration, digestive disorders, temporary deafness. Recent studies mention that they may be related to heart attacks [1–3]. Due to the high demand for NSAIDs in pharmaceutical samples, many analytical methods have been proposed for their quantification, the most common being liquid chromatography (HPLC) [4]. This method has some disadvantages, such as the need for sample preparation by chemical reaction or extraction. Some cases include previous derivatization, long analysis times, and a high cost associated with the use and maintenance of the equipment.

An alternative to traditional analysis methods in areas such as the food industry, pharmaceuticals, and environmental monitoring is known as Electronic Tongues (ETs) [5,6]. These systems combine electrochemical techniques (e.g., potentiometry, voltammetry, and impedance spectroscopy) with sophisticated multivariate analysis tools to classify or quantify samples [7]. Their main advantage compared to traditional methods is that they allow quick and low-cost measurements, avoiding the pretreatment of samples in most cases. Although ETs using potentiometric and voltammetric techniques have been reported in the literature [8,9], voltammetric methods are usually the most widespread due to their advantages such as short analysis time and high sensitivity [10,11]. In addition to this, data processing techniques based on artificial neural networks (ANNs), principal component analysis (PCA), and partial least squares (PLS) are popular for decoding the acquired voltammograms of aqueous solutions containing mixtures of different chemical species giving results favorable in the quantification of these species [12,13]. In this work, a methodology based on voltammetric methods is proposed together with ANNs as a modeling and calibration tool to quantify NSAIDs simultaneously.

#### **2. Methods and Materials**

#### *2.1. Instrumentation and Reagents*

The chemical reagents used are of an analytical grade. All solutions used were prepared using high purity deionized water (18.2 MΩ cm). The experiments were carried out using a potentiostat Autolab PGSTAT302N (Metrohm, Utrecht, The Netherlands) connected to a computer. A three-electrode system was used consisting of a reference electrode with saturated Ag/AgCl (BASi), a graphite bar as the auxiliary electrode, and a carbon paste electrode (CPE) modified with multi-wall carbon nanotubes (MWCNT-CPE) OD × L 6–9 nm × 5 μm> 95% (Sigma-Aldrich, CAS 308068-56-6, Mexico) as working electrode. The pH measurements were performed on a Corning pH/Ion meter 450 digital pH meter. A Sartorius CPA224S brand analytical balance was used. All the potentials referred to in this work are referred to the Ag/AgCl electrode. Standard solutions of diclofenac sodium, naproxen sodium, paracetamol, aspirin, and ibuprofen (Sigma-Aldrich, <sup>M</sup>éxico) were flushed with high purity nitrogen. A Britton–Robinson (BR) 0.1 mol·L−<sup>1</sup> buffer solution was used in a range of pH 7–11. Buffer BR was prepared by mixing appropriate volumes of acids (phosphoric acid, boric acid, and acetic acid) and adjusted with concentrated NaOH to the desired pH.

#### *2.2. Electrochemical Characterization*

#### 2.2.1. Electrode Preparation

The paste mixture for the proposed working electrode consisted of a 3:2 ratio for mineral oil and multi-walled carbon nanotube graphite powder (MWCNT). The graphite mix is made up of 30% graphite powder and 10% MWCNT. The paste obtained is placed with a spatula in a 1 mL syringe tube (30 mm long by 6 mm wide) and compacted with the syringe's plunger, placing it on a flat surface until excess air is eliminated. At one end of the syringe with the paste, the electrical contact (copper wire) is placed. Finally, the working electrode is built.

#### 2.2.2. Electrochemical Analysis of the NSAIDs in the Proposed Working Electrode

Cyclic voltammetry (CV) is performed for the supporting electrolyte (0.1 molL−<sup>1</sup> BR buffer at pH 7) and the NSAIDs-BR system (5 × <sup>10</sup>−<sup>4</sup> mol·L−<sup>1</sup> for each drug), starting at the zero-current potential, in anodic direction and cycling in a potential window from 0 to 1.3 V considering a scan rate of 0.1 Vs−1. At different scan rates, anodic and cathodic CV peaks were analyzed to determine the mechanism that controls the oxidative processes. Moreover, a pH study for the NSAIDs-BR system was performed to choose the maximum anodic current intensity for analytical quantification.

Quantification of the drug is carried out by DPV. The optimization of the parameters of the technique is carried out, with the Box Behnken (BBD) four-factor design, step potential "*X*1" (V), interval time "*X*2" (s), modulation time "*X*3" (s) y modulation amplitude "*X*4" (V) and three levels for each factor so that the highest intensity of the anodic current is obtained. The design matrix considers 27 experimental units at random that include the three replicas of the central point. Using a polynomial regression, the response variable was predicted as a function of the independent variables and their interactions. The prediction of the model is described in Equation (1).

$$Y = \beta\_0 + \sum\_{i=1}^{K} \beta\_i X\_i + \sum\_{i=1}^{k} \beta\_{ii} X\_i^2 + \sum\_{i=1}^{k} \sum\_{j=1}^{k} \beta\_{ij} X\_i X\_j + \varepsilon,\tag{1}$$

where *Y* is the response variable (maximum anode current intensity), *Xi* and *Xj* are coded independent variables, and *β*0, *βi*, *βii*, and *βij* are coefficients of intercept, linear, quadratic, and interaction terms, respectively. k is the number of independent variables (*k* = 4 in this study), ε is an experimental error [14]. Determination and regression coefficients were estimated using Minitab® Statistical software version 18.

#### *2.3. Quantification of NSAIDs by ANN*

#### 2.3.1. Data Processing

Having the optimal parameters of the DPV, the DPV's are carried out at different concentrations of the NSAIDs (ranging from 5 × <sup>10</sup>−<sup>7</sup> to 7 × <sup>10</sup>−<sup>5</sup> mol·L−1). A matrix of peak intensities of dimensions [189 × 27] (intensities x number of samples) was built with the recorded voltammogram records. The pretreatment of the data was carried out using the Discrete Wavelet Transform using the 4th level wavelet decomposition of a Daubechies function (db4). In this pretreatment, only approximation coefficients were chosen considering the degree of similarity between the original voltammogram and the one recovered with these coefficients [13]. Thus, the final input matrix to feed the ANN model has a dimension of [18 × 27] (approximation coefficients x number of samples). The ANN calibration model is based on a Multilayer Perceptron Network (MLP). The MLP input layer was established considering the number of approximation wavelet coefficients. In contrast, the output layer was defined using one neuron for each of the analytes to be quantified since it is associated with the matrix of concentrations of dimension [4 × 27] (i.e., paracetamol, diclofenac, naproxen, and ibuprofen), aspirin was not quantified in the model as it was considered only as an interferer. The hidden layers were established through a trial-and-error process, modifying the number of neurons in the layers until an appropriate number of neurons were found that favored obtaining a satisfactory linear regression coefficient. In this way, the final MLP model was 18 × 10 × 8 × 4 (18 input neurons, 10 neurons in the first hidden layer, 8 neurons in the second hidden layer and 4 output neurons).

#### 2.3.2. ANN Modeling

The described data set of 27 samples, were selected for the training set. The testing set was conformed using an external set of 10 additional samples randomly generated within the concentration range described above. All data sets were normalized in the interval of [−1, 1] to favor the training process. The activation functions established were: purelin for the input layer, tansing for the two hidden layers, and purelin for the output layer. In the same way, the chosen training algorithm was Bayesian regularization, with a training error set at a value of 0.001, together with a learning rate of 0.01. The MLP models were programmed on the MATLAB® R2021a (MathWorks, Natick, MA, USA) platform using the Deep Learning and Wavelet Toolboxes.

#### **3. Results and Discussion**

#### *3.1. Electrochemical Characterization*

The electrochemical study of the NSAIDs was carried out by CV in a solution of <sup>4</sup> × <sup>10</sup>−<sup>4</sup> mol·L−<sup>1</sup> of the drug standards, in BR buffer at pH 7 (Figure 1a). In the window from 0 to 1.3 V in the first sweep in the anodic direction, the oxidation peak corresponding to paracetamol, diclofenac, at a potential of 0.446 and 0.629, respectively, was observed. Naproxen presented two oxidation peaks at a potential of 0.888 and 1.14 V. In the case of ibuprofen, no anodic signal was observed, but the baseline was modified compared to the blank, so we believe that if it is carried out by the oxidation of the drug. Reversing the sweep in the cathodic direction, only one reduction peak was observed with a maximum peak potential at 0.263 V, which corresponds to the reduction of paracetamol. Moreover, a CV was performed in a mixture of the five NSAIDS and the BR Buffer (blank) using the proposed working electrode (Figure 1b), under the same conditions of the systems shown in Figure 1a. In Figure 1b, only three oxidation peaks can be observed in the anodic sense, since some signals overlap in the drugs, and it should also be noted that in the cathodic sweep the reduction signal of paracetamol is not observed, and this could be because the product that was reduced reacted chemically with some oxidation products of the other drugs and the product is no longer electroactive in reduction.

**Figure 1.** CVs obtained for the systems containing the NSAIDs and the supporting electrolyte, using the proposed working electrode and in presence of aspirin. Potential window from 0 to 1.3 V and at a scan rate of 100 mVs<sup>−</sup>1, (**a**) individual for each NSAIDs, (**b**) mixture of the four NSAIDs.

Considering the anodic wave corresponding to the oxidation of the NSAIDs for the mixture of the drugs. A pH study was carried out by CV, in a 0.1 mol L−<sup>1</sup> BR buffer with a concentration of 5 × <sup>10</sup>−<sup>4</sup> mol L−<sup>1</sup> for each drug, in Figure <sup>2</sup> the anodic sweep is shown. It can be observed that the highest current intensity is obtained at pH 10 for the mixture of drugs; it can also be observed that as the pH increases the anodic peak potentials of the drugs shift to lower values. Different CV scan rates (10 to 300 mV·s−1) were studied, and the maximum anodic peak current was plotted vs. the square root of the scan rate for each NSAID. A correlation coefficient greater than 0.99 was obtained after the proper statistical analysis, which suggests that the diffusion of the electroactive species to the surface of the electrode governs the oxidation processes.

**Figure 2.** CVs obtained for the NSAIDs mixture at different pHs (range of 7–10) in a 0.1 mol·L−<sup>1</sup> BR buffer, using a potential window of 0 to 1.3 V and a scan rate of 100 mV·s<sup>−</sup>1, and a concentration of <sup>4</sup> <sup>×</sup> <sup>10</sup>−<sup>4</sup> mol·L−<sup>1</sup> for the NSAIDs.

A BBD with three levels was used for the optimization of the four variables related to the DPV technique to maximize the anodic current peak of paracetamol (this NSAID has the highest peak current intensity). Twenty-seven experiments were carried out, generating the corresponding voltammograms of NSAIDs using the MWCNT-CPE at pH 10. The parameters of the DPV technique were chosen considering the capabilities of the potentiostat used. The proposed second-order model regression that correlates the current response and the DPV factors is shown in Equation (2).

$$\begin{array}{c} Y = 1.747 + 0.22X\_1 + 0.352X\_2 - 0.935X\_3 + 1.082X\_4 - 0.476X\_2^2 - 0.431X\_2^2 - 0.308X\_3^2 - 0.012X\_4^2 \\ - 0.309X\_1X\_2 + 0.211X\_1X\_3 + 0.209X\_1X\_4 + 0.297X\_2X\_3 + 0.66X\_2X\_4 + 1.507X\_3X\_4. \end{array} \tag{2}$$

Table 1 shows the theoretical response (Y) after the optimization of the DPV parameters, obtained by using the Response Optimizer function in the Minitab® V.18 software. The maximum anodic current for paracetamol was determined as 5.24 μA under optimal conditions.

**Table 1.** Optimal DPV parameters found with the Box Behnken design.


#### *3.2. Quantification of NSAIDs Using ANN*

Using the optimal parameters of the DPV to analyze the 27 samples considering different concentrations of the NSAIDs, and they were performed using a 35−<sup>2</sup> fractional facto-rial design. The trained MLP was used to determine the performance in the quantification task of drugs and the relationship between the concentrations obtained and those expected was evaluated, both for the training and the test phases. In this sense, the linear regression obtained from the comparison was a measure of the model's goodness. Given ideal conditions, the line must have a slope equal to 1 and its intersection equal to 0. The comparative graphs between the real concentrations of paracetamol, diclofenac, naproxen, and ibuprofen and those predicted with the MLP model for the training and test data set (Figures 3 and 4, respectively). The high level of linearity allows a linear regression coefficient of the data obtained very close to one (R = 0.98) for paracetamol and diclofenac, while for the naproxen and ibuprofen, the correlation value was 0.87 and 0.78 respectively; aspirin was present as an interfering agent in the mixture.

**Figure 3.** Comparison between the expected NSAIDs concentration and those obtained after MLP training phase.

**Figure 4.** Comparison between the expected NSAIDs concentration and those obtained during MLP test phase.

#### **4. Conclusions**

In this work, a potential tool for voltammetric determination is presented. A combination of DPV and DWT-ANN allowed us to obtain satisfactory results for quantifying paracetamol, diclofenac, naproxen, and ibuprofen in the presence of aspirin. The use of DWT is helpful to compact voltammograms, preserving the analytical information of the original records. Multivariable models created with ANN correctly describe the complexity in voltammograms caused by overlapping peaks without the need for a pretreatment step on the samples. Finally, carbon paste electrodes with nanotubes are low-cost and easy-to-make devices that allow us to determine the drugs in the order of microgram per liter.

**Supplementary Materials:** The following are available online at https://www.mdpi.com/article/10 .3390/CSAC2021-10450/s1.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

**Acknowledgments:** The authors thank the Mexican National Council of Science and Technology for financial support through the National System of Researchers (SNI) and the distinction of their memberships.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


### *Proceeding Paper* **An Inkjet-Printed Amperometric H2S Sensor for Environmental Applications †**

**Franc Paré 1,2, Rebeca Castro 3, Xavier Guimera 3, Gemma Gabriel 4,5 and Mireia Baeza 1,2,\***


**Abstract:** Hydrogen sulfide (H2S) is a highly toxic chemical capable of causing severe health issues. Due to its environmental impact, it is critical to create effective methods for its monitoring. Inkjet printing technology has become an alternative for sensor fabrication because it is an economic, fast, and reproducible method for mass producing micro-electrodes. Herein, a miniaturized 25 mm2 inkjet-printed amperometric sensor is presented. A gold electrode coupled with a silver track was modified with two inks: single-walled carbon nanotubes (SWCNTs) and a mixture of SCWCNTs and poly(vinyl alcohol) (PVA). Morphological and electrochemical properties were studied, as well as H2S sensor performance. This approach is a suitable option for environmental H2S tracking.

**Keywords:** electrochemical sensor; amperometric sensor; H2S sensor; single-walled carbon nanotubes

#### **1. Introduction**

Environmental equilibrium is a hard-to-preserve resource, dangerously impacted by human heavy industrial activities [1,2]. It is naturally regulated through biogeochemical cycles. Among these, the sulfur cycle is of crucial importance, since it is vital for maintaining the composition of both the atmosphere and soils, as well as most living beings. Nonetheless, even more dangerous is an excess of highly toxic compounds such as water-soluble hydrogen sulfide (H2S) gas. This is a poisonous, inflammable, and corrosive chemical, hazardous to human health at concentrations as low as 20 ppm (1.1 μM) for prolonged exposure [3]. Even though it generally appears as a gas, it has labile hydrogens, meaning it coexists as different species in aqueous media. Hydrogen sulfide can appear as different species depending on the pH of its medium, being capable of losing both its protons and transitioning from a gas to ions. It has a pka1 of approximately 7 and a pka2 of about 13.5, meaning that HS- predominates between pH 7.5 and 13. Due to the dangerous nature of H2S and its frequent appearance in gas streams, a great need has recently arisen for many biotechnological processes to remove it [4], which require adequate systems for quick and easy tracking.

Nonetheless, many of these removal processes occur in aqueous media, requiring consideration of the environmental pH for adequate quantification. Thus, it is appropriate to incorporate a simultaneous pH measurement with the H2S tracking.

**Citation:** Paré, F.; Castro, R.; Guimera, X.; Gabriel, G.; Baeza, M. An Inkjet-Printed Amperometric H2S Sensor for Environmental Applications. *Chem. Proc.* **2021**, *5*, 4. https://doi.org/10.3390/ CSAC2021-10462

Academic Editor: Núria Serrano

Published: 30 June 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 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/).

In recent years, printed electronics have steadily replaced more traditional electrode fabrication methods. Among these, inkjet printing has the main advantage of not requiring any mask preparation, greatly reducing the time and cost of device fabrication.

Moreover, its low drop volume and precision, added to its capacity for printing metal-based inks, allow for the fabrication of highly reproducible micro-electrodes.

For biotechnological applications, microsensors can be printed on different substrates and their designs can be adapted to the shapes of bioreactors with little cost impact [5]. Therefore, inkjet printing technology is an interesting alternative for the development of a microsensors platform for H2S and pH measurements.

Moreover, electrochemical sensors have the advantages of high sensitivity, in situ application, and a broad range of applicable materials. However, H2S determination using electrochemical sensors has many design and implementation challenges to solve, such as the pH influence on measurements [6] and electrode passivation by accumulation of S0 produced from H2S oxidation [7].

Among the authors who have developed H2S sensors, Yang et al. (2018) [8] fabricated a sensor using Nafion for H2S measurement in gaseous samples. This membrane was added to carbon fibers modified with platinum and rhodium nanoparticles. The H2S was adsorbed on the platinum, releasing protons that crossed the Nafion membrane while electrons moved through an external circuit. The H2S concentration was proportional to the circulating charge. This sensor had a linear range from 2.9 μM to 5.9 mM, and the minimum detectable signal was 2.9 μM. Brown et al. (2019) [9] deposited a thin layer of S0 on a glassy carbon electrode (GCE) and covered it with electro-polymers to avoid passivation of the electrode via accumulation of S0. The H2S was measured by constant potential amperometry (CPA), using a potential of 0.3 V vs. Ag/AgCl. This sensor exhibited a high degree of selectivity and a linear range between 0 μM and 15 μM, with lowest and highest detection limits of (9 ± 6) nM and (79 ± 51) nM, respectively.

Furthermore, new carbon materials with excellent electronic properties have become suitable options for electrode development. As an example, Lawrence et al. (2004) [10] modified a GCE with carbon nanotubes (CNTs) deposited on the surface by drop casting a solution of CNTs in dimethylformamide (DMF). The main advantage found was the catalytic capacity that decreased the oxidation potential from 0.4 V to −0.3 V (vs. Ag/AgCl). This allowed amperometric measurements at 0.1 V (vs. Ag/AgCl) in a range between 1.25 μM and 112.5 μM, with a detection limit of 0.3 μM. Li et al. (2017) [6] used cobalt to magnetically attach MoS2 monolayer sheets to CNTs, which were deposited on glassy carbon electrodes using a Nafion membrane as an adhesive. Analysis was performed by amperometry obtaining a linear range of application from 0.05 μM to 0.6 μM, with a detection limit of 7.6 nM.

Despite authors applying carbon materials for H2S sensor development, there are no reports of the use of SWCNTs ink for an inkjet-printed H2S sensor. Moreover, several studies have reported poly(vinyl alcohol) (PVA) addition to conductive materials to improve its mechanical and adhesive properties [11–13].

In this study, the fabrication of a H2S amperometric microsensor has been studied using both a SWCNTs ink and a SWCNTs–PVA ink for Au electrode modification. Morphological and electrochemical characterizations were carried out for electrode performance analysis.

#### **2. Materials and Methods**

#### *2.1. Inkjet-Printed H2S Sensor Fabrication*

The printing process for the electrode was performed using Ag NPs ink (SI-J20x Nanosilver Inkjet Printing Ink from Agfa, Mortsel, Belgium), Au NPs ink (Drycure Au-JB 1010B from C-ink, Japan), and SWCNTs ink (Carbon nanotube, single-walled, conductive aqueous ink, SWCNT 1.00 mg/mL from Sigma-Aldrich, Spain). To passivate the electrode, SU8 ink (XP PriElex SU-8 1.0 Inkjettable Dielectric from Kayaku, Westborough, MA, USA) was used. All inks were printed over polyethylene teraphtalate (PET) sheets (Q65HA,

Du Pont Teijin Films, Dumfries, UK), using a Dimatix printer (DMP-2831 from FUJIFILM Dimatix, Santa Clara, CA, USA). The SWCNTs–PVA composite ink was prepared by mixing a commercial SWCNTs ink, and a 5 wt.% PVA solution, both acquired from Sigma-Aldrich. For SWCNTs ink and SWCNTs–PVA ink deposition, the drop casting technique was performed on a thermal plate. The applied temperature was a studied parameter. The final electrode dimensions were a 1 mm diameter gold disk with a total length of 26 mm and surface of 25 mm2, and an approximately 2 mm diameter SWCNTs disk.

#### *2.2. Sensor Characterization*

Morphological characterization of the electrodes was carried out via optical microscopy. Images were obtained with a digital microscope (USB microscope AM4815ZTL from DinoLite, Alemere, The Netherlands).

Electrochemical characterization was performed with a PalmSens potentiostat– galvanostat (PalmSens4 from PalmSens, The Netherlands). A three-electrode configuration was used for the electrochemical cell. An Ag/AgCl (1 M KCl) reference electrode (reference electrode with Ag/AgCl in aqueous KCl from ItalSens, PalmSens, Houten, The Netherlands) and a platinum wire counter electrode (counter electrode made of platinum wire from ItalSens, PalmSens, Houten, The Netherlands) were used. The fabricated electrode functioned as the working electrode. Cyclic voltammetry (CV) measurements were carried out using the redox pair K3[Fe(CN)6]/K4[Fe(CN)6] (0.01 M) at 0.01 V/s from −0.1 V to 0.5 V. Intensity peaks (Ip) resulting from redox reactions allowed for electrochemical characterization.

To study the effect of deposition temperature in the SWCNTs–PVA ink, ten layers of PVA were drop casted over a gold inkjet-printed electrode at 90 ◦C and at room temperature. Intensity-peak values were obtained.

For H2S calibration, a stock H2S solution (0.1 M) was prepared by dissolving NaS·9H2O and NaOH (both from Sigma-Aldrich, Spain) in deionized water (Milli-Q from Millipore Corporation, Burlington, MA, USA). Standardization was carried out according to standard methods [14]. The stock solution was diluted in phosphate-buffered saline solution (PBS) (from Sigma-Aldrich, Madrid, Spain) to obtain a 0.02 M H2S solution.

#### **3. Results and Discussion**

#### *3.1. Inkjet-Printed H2S Miniaturized Sensor Fabrication*

Inkjet-printed micro-electrodes were fabricated by first printing the Ag tracks and pads, drying at 100 ◦C in an oven, then printing the Au surfaces and connecting tracks, which were later dried at 120 ◦C. Subsequently, inks were sintered at 150 ◦C for 60 min to improve their conductivity. Finally, micro-electrodes were passivated by printing an SU8 layer over the tracks, preventing the short circuiting of the electrodes.

Both SWCNTs and SWCNTs–PVA inks were drop casted on a gold electrode, covering it completely (Figure 1). Both inks showed good adhesion to gold, allowing morphological and electrochemical characterization of the microsensors.

**Figure 1.** Photographs of (**A**) inkjet-printed Au electrode, (**B**) SWCNTs drop casted over Au electrode, and (**C**) drop-casted SWCNTs–PVA over Au electrode.

#### *3.2. Microsensor Electrochemical Characterization*

Cyclic voltammetry measurements of Au, SWCNTs, and SWCNTs–PVA electrodes were carried out. These allowed verification of whether the applied modifications yielded an electrode with properties of interest that could be used for an H2S sensor. The main interest was in fabricating a sensor capable of oxidizing H2S at low voltages and with a high tolerance for sulfur poisoning since it is well known that S0 is highly insoluble in water [7] and has a high affinity for Au atoms. Thus, the electrode would gradually deteriorate by accumulating non-conductive layers of sulfur atoms.

Modified electrodes presented a similar current density and a smaller potential gap than bare Au electrodes (Figure 2). The smaller peak separation, specifically due to reducing the potential necessary for ferrocyanide oxidation, meant that SWCNTs and SWCNTs–PVA are both favorable for use as an H2S sensor, with the added benefit of lowered rates of sulfur deposition on their surfaces due to the less-favorable S–C interaction compared with S–Au.

**Figure 2.** Cyclic voltammetry of Au, SWCNTs, and SWCNTs-PVA electrodes in hexacyanoferrate/hexacyanoferrate (0.01 M).

#### *3.3. PVA Deposition Temperature Study*

The effect of PVA deposition temperature was studied through the drop casting of 10 layers of PVA on an Au inkjet-printed electrode at 90 ◦C and at room temperature. Cyclic voltammetry was performed, and current-peak (Ip) values were obtained. The results showed that at room temperature, PVA deposition had a smaller passivation effect than at 90 ◦C, in which case the electrode's current values dropped significantly (Table 1).

**Table 1.** Effect of polymer deposition temperature on Au inkjet-printed electrodes.


#### *3.4. Microsensor Calibration and Analytical Response*

Calibration of SWCNTs and SWCNTs–PVA microsensors was performed by adding different volumes of H2S standard to a PBS solution, measuring a concentration range between 0 and 600 μM. The chronoamperometry method was used for H2S oxidation (Equation (1)) at a polarization voltage of 50 mV, and the resulting current was measured.

$$\text{NH}\_2\text{S} \rightarrow \text{S}^0 + 2\text{e}^- + 2\text{H}^+ \tag{1}$$

Calibrations were performed for both SWCNTs and SWCNTs–PVA under the same conditions (Figure 3). The results showed good correlation between the measured current and the analyte concentration, with a similar slope for both sensors but different linear ranges.

**Figure 3.** Calibrations of two different sensors made of Au electrodes modified with SWCNTs and SWCNTs-PVA by drop casting. Inset shows a reduced range of H2S concentration for the SWCNTs sensor calibration.

While the SWCNTs sensor showed a higher sensitivity (19.3 ± 0.4 mA/M) compared to SWCNTs–PVA (9.4 ± 0.2 mA/M), it also had a reduced working range. SWCNTs can measure H2S concentrations from 8 μM to 60 μM, with a limit of detection (LD) of 4.3 μM. On the other hand, SWCNTs–PVA is capable of measuring from 52 μM to 512 μM, with a LD of 34 μM.

#### **4. Conclusions**

We demonstrated a novel miniaturized inkjet-printed amperometric H2S sensor fabricated by modification of a gold electrode with SWCNTs ink. The results showed that the addition of a stabilizing polymer allowed for an increased range of H2S concentration measurements. Due to the low versatility of the SWCNTs sensor, the addition of PVA was considered, to improve the H2S sensor performance. However, it was necessary to adapt the fabrication temperature conditions to avoid electrode passivation, limiting the SWCNTs–PVA ink-deposition temperature to 25 ◦C. This sensor offers an approach for H2S tracking with environmental and biotechnological applications.

**Supplementary Materials:** The following are available online at https://www.mdpi.com/article/10 .3390/CSAC2021-10462/s1.

**Author Contributions:** Conceptualization, M.B. and G.G.; methodology, F.P.; validation, M.B. and X.G.; formal analysis, F.P. and R.C.; investigation, F.P. and R.C.; resources, G.G. and M.B.; data curation, F.P. and R.C.; writing—original draft preparation, F.P. and R.C.; writing—review and editing, M.B., G.G., and X.G.; visualization, G.G. and M.B.; supervision, M.B., G.G., and X.G.; project administration, M.B.; funding acquisition, M.B. and G.G. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the Spanish Government, Ministerio de Economía y Competitividad, through projects RTI2018-099362-B-C21 and RTI2018-099362-B-C22 MINECO/FEDER, EU.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** The data presented in this study are available on request from the corresponding author. The data are not publicly available due to the repository that is used to keep the data is a private one provided by the University.

**Acknowledgments:** The authors wish to acknowledge David Gabriel and Xavier Gamisans for their coordinated direction of the ENSURE project.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


### *Proceeding Paper* **Molecular Emitters as a Tunable Light Source for Optical Multisensor Systems †**

**Anastasiia Surkova 1,2,\* , Aleksandra Paderina <sup>1</sup> , Andrey Legin <sup>1</sup> , Elena Grachova <sup>1</sup> and Dmitry Kirsanov <sup>1</sup>**


**Abstract:** In this study, optical multisensor systems based on molecular emitters as a light source are introduced. To obtain such light sources, cyclometalated Ir(III) complexes and Cu(I)-based complexes were synthetized and investigated. Since each complex has its own emission spectrum in the visible range, it is possible to choose an appropriate set of emitters for specific analytical tacks. The developed analytical device was successfully applied for fluoride and phosphate quantification in surface water.

**Keywords:** optical multisensor systems; molecular emitters; iridium complexes; copper complexes; photoluminescence; chemometrics; water analysis; fluoride; phosphate

**Citation:** Surkova, A.; Paderina, A.; Legin, A.; Grachova, E.; Kirsanov, D. Molecular Emitters as a Tunable Light Source for Optical Multisensor Systems. *Chem. Proc.* **2021**, *5*, 5. https://doi.org/10.3390/CSAC2021- 10611

Academic Editor: Elena Benito-Peña

Published: 6 July 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 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/).

#### **1. Introduction**

The development of portable and inexpensive analyzers allowing fast determination of the integral sample characteristics is a current trend in analytical chemistry. Optical spectroscopy in the visible and near infrared (NIR) range has a great potential due to the advances in modern optical engineering. Optical multisensor systems (OMS) are devices working on the principle of optical spectroscopy but optimized for a specific analytical task and composed of cheaper elements: light-emitting diodes (LEDs), optical fibers, 3D-printed parts, stamped optics, etc. Such specialization enables essential reduction of analyzers' price, size, and weight, thus making the analysis widely available for both real-time application and in-field measurements. There are many examples of OMS applications for various analytical problems in the recent literature [1–5].

In the present work, a novel platform for construction of OMS was suggested. The idea is to use a combination of molecular emitters as a multichannel light source with tunable intensity and wavelength range. Cyclometalated Ir(III) complexes [6] and Cu(I) based complexes [7] were synthetized and tested in order to obtain such a light source. Each individual complex has its own emission spectrum in the visible range. This enables the selection and optimization of the light source for a specific analytical application. Several optical setup designs of OMS were developed. The proposed prototype was tested to analyze the metal ions in aqueous mixtures. The practical application of the OMS was demonstrated for the quantification of fluoride and phosphate in real surface and tap waters.

#### **2. Methods and Materials**

To obtain light sources for OMS, two sets of molecular emitters were synthesized. One of them consisted of eight cyclometalated Ir(III) compounds: firstly synthesized

[Ir(dfppy)2(bpbpy)]PF6 (**1**), [Ir(ppy)2(bpbpy)]PF6 (**2**), [Ir(pybt)2(bpbpy)]PF6 (**3**), and synthesized according to the standard procedure: (**4**) [8], (**5**) [9], (**6**) [10,11], (**7**) [12], and (**8**) [13]. The second set consists of six Cu(I)-based complexes: [Cu(MePPy3)I]2[Cu2I4] (**9**) [14], [Cu4I4(py)4] (**10**) [15], [Cu(Tpdp)I] (**11**) [16], [CuCl(PPh3)2(py)] (**12**) [17], [Cu(PPh3)3(4- Mepy)]Br (**13**), and [CuI(PPh3)2(4-Mepy)] (**14**) [18]. The mixture of each series gives multiband light source in the region from about 400 to 800 nm.

A different optical setup was constructed for each set of emitters. A homogeneous mixture of molecular emitter powders (mixture of iridium(III) or copper(I) complexes) was placed on a glass substrate under the sample. The solution under study was placed either in a glass cup (1 cm in diameter) for the setup with Ir(III)-complexes or in a polystyrene Petri dish (3.5 cm in diameter) for the setup with Cu(I)-complexes. Initially, photoluminescence of the molecular emitters was initiated using the laser diode (λexct = 365 nm and λexct = 385 nm for the first and the second experimental setup, respectively). Further, the laser diode was replaced by a UV flashlight with λexct = 365 nm. The light that passed through the sample was recorded by a fiber-optic UV−vis spectrometer AvaSpec-ULS2048CL-EVO (Avantes, Apeldoorn, the Netherlands). The optimal geometry of the device and the required amount and placement of sample solution were chosen experimentally to obtain a stable and reproducible analytical signal.

To compare Ir(III)-based and Cu(I)-based OMS, the optical setups were tested on two separate calibration series prepared from aqueous solutions of Co(II) and Cu(II) nitrates. Each calibration set consisted of seven samples with different concentrations ranging from 0.01 to 0.1 M with a 0.015 M step for Ir(III)-based setup and ten samples with different concentrations ranging from 0.01 to 0.1 M with a 0.01 M step for Cu(I)-based setup.

The practical application of the Cu(I)-based OMS was demonstrated for quantitative determination of fluoride and phosphate in surface and tap waters. A total of 5 samples for analysis were collected from the tap, rivers, and a lake. The standard photometric methods were used as a reference for the quantification of phosphate [19] and fluoride [20] in water samples. Sample preparation for analysis was carried out in accordance with the procedures described in [19,20]. The mass concentrations of phosphate and fluoride in the calibration solutions were 0–0.96 mg/L with a 0.12 mg/L step and 0–0.4 mg/L with a 0.04 mg/L step, respectively.

To relate the analytical signal response of OMS to the analyte content, the calibration models were built using partial least-squares (PLS) regression [21]. The model performance was estimated by full cross-validation (CV) and validation with a test set. Root mean-square error (RMSE) of calibration (RMSEC), prediction (RMSEP, for the test set) or cross-validation (RMSECV) and the respective coefficients of determination (R2) were used to compare the models.

#### **3. Results and Discussion**

In order to choose a set of appropriate emitters for multiband light sources, several requirements were established: the absorption spectrum must not overlap with the emission spectrum; excitation radiation at 365/385 nm must fit into the absorption maximum; the emission should be intense enough to be employed as the light source in the solid phase. Ir(III) luminescent complexes that have bright controlled emission fit these requirements. Another set of molecular emitters based on Cu(I) complexes is a cost-effective alternative to Ir(III) complexes and their application as a light source in OMS was also tested. The resulting mixture of each set of complexes upon excitation with the laser diode has shown emission spectra in the range of 450–800 nm (Figure 1). Emission of the mixture of eight Ir(III) compounds is higher at the right end of the spectrum from 600 to 800 nm (black line in Figure 1), while emission of the mixture of six Cu(I) complexes is higher at the left end of the spectrum (red line in Figure 1). Despite this difference, both sets are suitable as a multiband light source in the visible region.

**Figure 1.** Normalized emission spectra of: mixture of eight Ir(III) complexes is marked in black and mixture of six Cu(I) complexes is marked in red.

The block diagram of the experimental setup is shown in Figure 2. In accordance with the proposed optical setup, the emission of molecular emitters is excited by the laser diode. The light passing through a sample is registered with a fiber-optic cable connected with a spectrometric detector. The laser diode can be replaced by inexpensive UV flashlight. Another advantage of UV flashlight is that its power is higher than that of the laser diode. Therefore, two optical setups were constructed and compared: with laser diode and with UV flashlight as excitation light.

**Figure 2.** Block diagram of the experimental setup, where 1—exciting light; 2—emitted light; 3 transmitted light.

In order to compare Ir(III) and Cu(I) compounds as a light source in OMS, the series of colored aqueous solutions of copper and cobalt nitrates were chosen because these solutions absorb the light in the visible range (Co(NO3)2 in the range of 400–600 nm, Cu(NO3)2— 700–900 nm). Individual PLS models were built to relate the registered optical signals with copper and cobalt content, and the respective modeling statistics are presented in Table 1. Despite the simplicity of the developed analytical platform, PLS models for copper and cobalt exhibit a good performance for both OMSs (with Ir(III) and Cu(I) complexes). The parameters of the model for copper quantification are somewhat worse in Cu(I)-based OMS than in Ir(III)-based OMS, while for the cobalt model they are better. A set of copper(I) based molecular emitters was chosen for further research because it is more efficient both economically and environmentally.

At the next stage, the laser diode was replaced by a UV flashlight. The experimental results showed that the recorded optical density of the initial emission of Cu(I) complexes upon excitation with the UV flashlight is much higher than that upon the laser diode. The PLS results for OMS with UV flashlight are better for both Cu and Co models (Table 1).


**Table 1.** PLS modeling and validation statistics.

<sup>a</sup> 7 samples, <sup>b</sup> 10 samples, <sup>с</sup> mixture of 8 Ir(III) complexes, <sup>d</sup> mixture of 6 Cu(I) complexes. The interval range for the modeling: 450–800 nm for Co, Cu, and F−, 600–750 nm for PO4 <sup>3</sup>−.

The practical application of the developed Cu(I)-based OMS was demonstrated for the determination of fluoride and phosphate content in real surface and tap waters. Fluoride and phosphate are essential components for living cells; however, excess concentrations in surface water can lead to various human diseases and a general reduction in water quality. The evaluation of fluoride and phosphate by simple and inexpensive analytical methods is an important task for timely environmental monitoring.

The colored complexes of fluoride and phosphate absorb light in the regions of 450–800 (absorption maximum at around 590 nm) and 550–800 nm (absorption maximum at around 700 nm), respectively. Since these analytes are determined using an individual calibration sample set, a single molecular emitter was chosen for each of the anions based on its the emission properties: complex (**9**) with emission maximum of 659 nm for fluoride determination and complex (**10**) with emission maximum of 584 nm for phosphate determination.

The RMSE and R<sup>2</sup> for the full cross-validation are similar for both phosphate and fluoride PLS models (Table 1 and Figure S1 in Supporting Information). Further, the calibration models were used to predict the content of fluoride and phosphate in five water samples. The content of anions evaluated by a standard spectrophotometric technique was employed for model precision assessment. The prediction performance of the phosphate model is pretty good: RMSEP = 0.073 mg/L with R<sup>2</sup> = 0.97. The PLS model for fluoride has also relatively low RMSEP (0.074 mg/L), but R2 is 0.7. This can be caused by the fact that water samples were taken in different regions and from different sources (several from rivers, one from a lake, and one from a tap). Each sample may contribute strongly to the model, and more samples are needed to make more accurate predictions. Mean squared errors (MSE) that show the average squared difference between the estimated by reference method values and predicted by OMS values were 0.34% and 0.54% for phosphate and fluoride, respectively.

#### **4. Conclusions**

The proposed approach to OMS development allows reducing analysis time and does not require additional sample preparation. Moreover, OMS based on molecular emitters can be adopted for the particular analytical task by selecting the appropriate wavelength region. Despite the relative technical simplicity of OMS, its application in combination with modern chemometric methods provides high accuracy of analysis, comparable with that of full-featured spectrometers. Both synthetized sets of Ir(III) and Cu(I)-based complexes are suitable as a light source in OMS. However, Cu(I) complexes are easier to produce, cheaper, and environmentally friendly. The demonstrated application of OMS based on molecular emitters for the determination of fluorides and phosphates in surface water proves their high practical significance.

**Supplementary Materials:** The following are available online at https://www.mdpi.com/article/ 10.3390/CSAC2021-10611/s1, Figure S1: Predicted versus measured values of cross-validation for quantification of (A) PO4 <sup>3</sup><sup>−</sup> and (B) F−.

**Author Contributions:** Conceptualization, D.K.; methodology, A.S.; formal analysis, A.S.; investigation, A.S. and A.P.; writing—original draft preparation, A.S.; writing—review and editing, E.G.; supervision, E.G. and D.K.; project administration, A.L. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the RSF, grant number 19-79-00076.

**Conflicts of Interest:** The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

#### **References**


*Proceeding Paper*
