**Implementation of Metallic Film Electrodes for Catalytic Adsorptive Stripping Voltammetric Determination of Germanium(IV) †**

**Agnieszka Królicka \*, Jerzy Zar ˛ebski and Andrzej Bobrowski**

Department of Building Materials Technology, Faculty of Materials Science and Ceramics, AGH University of Science and Technology, Mickiewicza 30, 30-059 Krakow, Poland; zarebski.jerzy@gmail.com (J.Z.); abobrow@agh.edu.pl (A.B.)

**\*** Correspondence: krolicka@agh.edu.pl

† Presented at the 1st International Electronic Conference on Chemical Sensors and Analytical Chemistry, 1–15 July 2021; Available online: https://csac2021.sciforum.net/.

**Abstract:** In the present work, it has been shown that bismuth film electrodes deposited on screenprinted carbon supports could be successfully used to provide well-shaped, sensitive and reproducible catalytic adsorptive stripping signals of Ge(IV) in the presence of catechol and V(IV)-HEDTA (HEDTA-N-hydroxyethyl-ethylene diamine-triacetic acid) complex.

**Keywords:** film electrodes; germanium; stripping voltammetry

**Citation:** Królicka, A.; Zar˛ebski, J.; Bobrowski, A. Implementation of Metallic Film Electrodes for Catalytic Adsorptive Stripping Voltammetric Determination of Germanium(IV). *Chem. Proc.* **2021**, *5*, 7. https:// doi.org/10.3390/CSAC2021-10484

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/).

**1. Introduction**

Germanium represents a unique group of elements known as Critical Raw Materials (EU). Although such elements typically constitute only a small percentage of a material by weight, they are essential to its performance [1]. Among the numerous analytical techniques available, voltammetry seems to have much to offer in this regard, as voltammetric techniques are insensitive to the presence of inorganic salts and, at the same time, they offer low detection limits. Among voltammetric methods, catalytic adsorptive stripping voltammetry (CAdSV) plays an essential role in trace analysis due to its remarkable sensitivity. To induce a catalytic effect which gives the method its outstanding sensitivity, ions with oxidizing properties must be added to the examined solution, e.g., nitrate, nitrite, bromate or chlorate. Unfortunately, many electrode materials, both metallic sensing layers as well as auxiliary polymers, are damaged under the influence of these oxidants.

As was reported earlier, using HMDE or silver-amalgam working electrodes and the supporting electrolyte containing V(IV)-HEDTA (HEDTA-N-hydroxyethyl-ethylene diamine-triacetic acid) complex (Figure 1), catechol or its derivatives, the well-developed germanium signals could be recorded at nM level [2,3]. To develop a workable analytical procedure for the detection of trace amounts of germanium, and to meet the current guidelines that impose limitations on the application of mercury in chemical and analytical laboratories, an environmentally friendly alternative should be employed instead. The metallic film electrodes, such as bismuth (BiFE) and lead (PbFE) electrodes are among the most widely used sensors in the field of stripping voltammetry [4].

This work aims to assess the applicability of lead and bismuth film electrodes deposited electrochemically or by physical deposition from gaseous phase on different supports to provide catalytic adsorptive stripping signals of Ge(IV) in the presence of catechol and V(IV)-HEDTA complex.

**Figure 1.** The structure of V(IV)-HEDTA complex.

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

Electrochemical study was performed on a Autolab 204 analyzer (Metrohm Autolab B.V., Utrecht, The Netherlands). Disposable screen-printed electrodes (4 mm diameter) (DropSens, Oviedo, Spain) or disc electrodes (3 mm diameter) made of glassy carbon or gold (Mineral, Łomianki-Sadowa, Poland) were used as supports for bismuth films. Platinum wire and Ag/AgCl (3 M KCl) were applied as the anode and reference electrodes. All applied reagents were analytical grade.

Lead and bismuth films were plated just prior to use by means of potentiostatic deposition. Before plating, the disc substrates were polished using an Al2O3 suspension (0.3 and 0.05 μm) applied onto a polishing cloth. Screen-printed electrodes did not require any preparation or processing other than 2 min of soaking in the plating solution immediately prior to electrolysis. The plating process performed in quiescent 0.34 M HClO4 containing 0.043 M of Bi(III) or stirred 0.2 M acetate buffer containing 0.003 M Pb(II) was monitored by recording chronoamperometric curves and stopped when the charge reached the defined threshold (Eplat <sup>=</sup> −0.9 V, Qplat = 0.8 mC per mm2). Pre-plated electrodes were rinsed with 0.1 M acetate buffer (PbFE) or 0.34 M HClO4 (BiFE) and water. The PVD deposition followed the protocol described in the previous work [5].

The supporting electrolyte contained 0.05 M acetate buffer (pH of 4.4), 1 mM of catechol, 1 mM of V(IV) and 1.5 mM of HEDTA. CAdSV voltammograms were recorded after 30 s of accumulation performed at the potential of −0.6 V (PbFE) or −0.4 V (BiFE) by differential pulse mode.

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

#### *3.1. Support Selection*

To deposit the metallic sensing layers, the following supports were considered: (1) carbon-based electrodes: glassy carbon, carbon paste, impregnated carbon, screen printed carbon and (2) gold-based electrodes: bulk disc, PDV deposited gold, gold screen printed. The sensing layers were deposited electrochemically by ex situ plating or by PVD deposition. The preliminary tests of freshly prepared films rinsed with water were performed by microscopic inspection. In the case of carbon-based supports and bismuth films, the visual changes were not pronounced, as the plating only gave the surface a black, velvety appearance. In the case of lead deposits, the electrode surface was turned gray. In the case of gold supports, the results were more complex. While the expected finding was that the type of the support (gold monolithic disc, gold screen printed layer, or PVD deposited

gold) does not play the key role, this study showed that the method of support preparation plays an important role. The films deposited on the gold disc electrode were very stable and adhesive. The lead layers deposited on the SPE or PVD gold were oxidized within minutes once removed from the plating solution (Figure 2). In the case of bismuth, the oxidation at AuSPE was slower but unrelenting. Information on the mechanism of oxidation of the metallic layer was provided by analysis of the bismuth films deposited on the microscratched Au PVD layer (Figure 3a–c). The microscopic images revealed the bismuth crystallites formed near microscratches are susceptible to oxidation. Gradually the gold support was exposed, but after 10 min, further changes were not observed as the bismuth crystallites adjacent to the scratches were depleted. The color histograms of microscopic images of Bi deposited on AuPVD shown in Figure 3d have not demonstrated any further changes for pictures taken 10 and 12.5 min after removal from the plating solution. The microanalysis performed by SEM XRF, and XRD studies of AuSPE as well as AuPVD electrodes did not reveal other elements than gold. Some insight was provided by contact angle measurements, displaying a significant difference in wettability of PVD gold (68.0 ± 0.2◦) and SPE (92.0 ± 0.3◦) electrodes. Since the surface microstructure of materials correlates closely with the apparent contact angle at the boundary between the liquid and the surface, the different surface properties of AuSPE and AuPVD were confirmed. The AuPVD contact angle value corresponds well to earlier reported values [6], while those obtained for AuSPE are substantially higher, placed on the threshold of hydrophobicity. The high contact angle can be explained by the rough three-dimensional structure of the gold layer, preventing the access of water molecules to the electrode surface by trapped air or by organic compounds of the ink used for screen printing. Regardless of cause, the gold screen printed electrodes cannot be used for bismuth or lead ex situ plating.

**Figure 2.** Microscopic images of Pb plated gold SPE recorded after 0 (**a**), 2.5 (**b**), 5 (**c**) and 7.5 (**d**) min after plating. Gray regions represent the lead layer while yellow the exposed gold support.

**Figure 3.** Microscopic images of Bi plated gold sputtered electrode recorded 0 (**a**), 2.5 (**b**) and 12 (**c**) min after plating (150x; images brightened to make the details more visible). Color histogram analysis of the bismuth layer plated on the sputtered gold electrode (**d**) (https://www.dcode.fr/ image-histogram, accessed on 8 June 2021).

#### *3.2. Stability Studies of Metallic Films in Contact with V(IV)-HEDTA Solution*

The externally plated bismuth and lead films as well as PVD deposited bismuth and lead layers were exposed to the supporting electrolyte containing the V(IV)-HEDTA complex, catechol and acetate buffer. Employing bismuth film electrodes of any type, it was possible to record well-shaped germanium signals in solutions containing from a few to several hundred nM of germanium. In the case of lead film electrodes, only PbFE deposited on glassy carbon, carbon paste and impregnated carbon provided measurable germanium signals. Lead layers deposited on SPE supports of any type did not deliver any germanium signals. Although the signals recorded by the most promising lead electrode, namely, PbFE/GC, in the solution containing 30 nM of Ge(IV), were initially quite pronounced, there was no Ge(IV) signal after recording ten or so voltammograms. In this way, the unsuitability of PbFEs for their intended use in stripping voltammetry became apparent.

#### *3.3. Analytical Performance of Bismuth Plated Screen Printed Electrodes*

It was shown [2] that when the HMDE electrode was used as the working electrode, the catalytic activity of HEDTA vanadium complexes towards the germanium complex was correlated with the redox behavior of V(IV)-HEDTA. In the case of pyrogallol germanium complexes, the intensity of the germanium signal correlated with the characteristics of the voltammetric signal representing the reduction of the V(IV) complex [3]. The cyclic voltammograms shown in Figure 4a were recorded by means of bismuth plated glassy carbon and carbon SPE electrodes and compared with the curves obtained when CGMDE and GC were used. Both CGMDE and GC delivered well-shaped reversible voltammograms of V(IV)-HEDTA complex, while at BiFE electrodes only reduction signals were observed. Such behavior indicates that the catalytic process at BiFEs proceeds following a different mechanism than that observed at CGMDE.

The catalytic signals of Ge(IV) recorded at BiFE/SPE in the presence of V(IV)-HEDTA were pronounced and highly reproducible, as it is shown in Figure 4b. The calibration curve is given by the equation y = (0.083 ± 0.001)x + (0.06 ± 0.01) (r<sup>2</sup> = 0.9989), where y and x denote the peak current (μA) and Ge(IV) concentration (nM). The Ip = f(cGe(IV)) is linear within the range from 2 to 30 nM (LOD = 1.5 nM). Finally, BiFE/SPEs were applied for Ge(IV) determination in Ge(IV) spiked snow water (10 nM) via the standard addition method and the concentration of 10.05 ± 0.11 nM was determined.

**Figure 4.** CV voltammograms recorded in an 0.05 M acetate buffer containing 1 mM of V(IV) and 2 mM of HEDTA using glassy carbon (GC), bismuth plated glassy carbon (BiFE/GC), bismuth plated screen printed carbon electrode (BiFE/SPE) and a controlled-growth mercury drop electrode (CGMDE). Scan rate = 50 mVs−<sup>1</sup> (**a**). Ten consecutive differential pulse voltammograms recorded by Bi/SPE electrode in the solution containing 3 or 30 nM of Ge(IV) and 0.05 M acetate buffer (pH of 4.4), 1 mM of catechol, 1 mM of V(IV) and 1.5 mM of HEDTA (**b**).

#### **4. Conclusions**

The results reported here demonstrate that the properties of lead and bismuth film electrodes differ considerably and only bismuth plated electrodes enable germanium analytical signals to be obtained when Ge(IV)-catechol-V(IV)-HEDTA system is employed.

**Author Contributions:** Conceptualization, A.K.; methodology, A.K.; software, A.K.; writing original draft preparation, A.K.; writing—review, A.B. and J.Z.; visualization, A.K. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was supported by the subsidy of the Ministry of Education and Science for the AGH University of Science and Technology in Kraków (Project No 16.16.160.557).

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

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** The data are available on request from the corresponding author, (A.K.).

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

#### **References**


### *Proceeding Paper* **Tunable Electrochemical Sensors Based on Carbon Nanocomposite Materials towards Enhanced Determination of Cadmium, Lead and Copper in Water †**

**Laia L. Fernández 1,2, Julio Bastos-Arrieta <sup>3</sup> , Cristina Palet <sup>1</sup> and Mireia Baeza 2,\***


**Abstract:** Many carbon materials are well-known conductive materials, widely used in the fabrication of composite electrodes. In this work, diverse allotropic forms of carbon such as graphite, MWCNTs and rGO were tested. Furthermore, these materials allow the construction of cheaper, smaller, portable, reliable and easy-to-use devices, which can be easily modified. The above-mentioned composite electrodes were developed for metal analysis in water such as Cu, Cd and Pb that, at a high concentration, can have consequences on human health. SWASV is the selected technique. It would be ideal to exploit the potential properties of mercury for metal detection by tuning the electrode's surface. Due to mercury's hazardous properties and to reduce the amount of this substance used in polarography, the use of nanoparticles is a good option due to their properties. Mercury nanoparticles were used to modify the surface of the composite electrodes to improve electroanalytical sensor response. For this reason, using these modified composite electrodes can lower detection limits and widen the linear range that can be achieved for Cd (0.05–1 mg·L−1) and Pb (0.045–1 mg·L−1). However, for Cu (0.114–1.14 mg·L−1), meaningful variations were not observed compared to the bare electrode.

**Keywords:** electrochemistry; Hg nanoparticles; graphite; composite electrodes; metal analysis; SWASV

### **1. Introduction**

Water is fundamental for all Earth's living forms, and a key issue for social and economic development. Currently, water analysis is a vital topic, for because monitoring some parameters is important to prevent some health problems. One of the parameters that has become important involves determining the concentration of heavy metals in water. To do this, several analysis techniques are used, such as atomic absorption spectroscopy (AAS) [1], inductively coupled plasma (ICP) [2], high-performance liquid chromatography (HPLC) [3], etc. Some of the metals that can be found in water are Cu, Cd and Pb and, at high concentrations, can have consequences on human health [4–6].

In this work, a voltametric technique has been chosen, known as square-wave anodic stripping voltammetry (SWASV) [7,8]. SWASV consists of two steps: first, applying a potential to preconcentrate the analyte on the surface of the electrode; second, taking a measurement by applying staircase potential to record the current generated.

**Citation:** Fernández, L.L.; Bastos-Arrieta, J.; Palet, C.; Baeza, M. Tunable Electrochemical Sensors Based on Carbon Nanocomposite Materials towards Enhanced Determination of Cadmium, Lead and Copper in Water. *Chem. Proc.* **2021**, *5*, 8. https://doi.org/ 10.3390/CSAC2021-10456

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/).

To use this technique, composite electrodes were construct using different carbon materials and a non-conductor epoxy. The behavior of graphite, reduced graphene oxide (rGO) and carbon nanotubes (CNTs) were tested in the detection of Cd, Pb and Cu. However, we work with the bare electrode; the modification of their surface with mercury nanoparticles (Hg-NPs) was also tested [9]. Mercury was used, a long time ago, in polarography, and it is well-known for its ability to form amalgams with some metals, reducing the potential where they appear [10,11]. Hence, taking advantage of these properties, the aim of this work is to reduce the amount of mercury used in polarography for the determination of Cd, Pb, and Cu.

#### **2. Composite Electrodes Construction, Characterization, and Modification**

#### *2.1. Composite Electrode Construction*

Composites were constructed using three different carbon materials: graphite, CNTs and rGO. The first step is to weld a copper sheet to a commercial connector; after that, it is placed in a PVC tube. A mixture of one of the carbon materials and Epotek H77 is prepared, and the PVC tube (2.1 cm, ∅6 mm) is filled with this mixture. Then, it is cured for 2 days at 80 ◦C. Then, the surface must be polished.

The percentages tested of carbon materials are shown in Table 1. These percentages were optimized previously, and they are related to their respective improvement in the electroanalytical properties of developed sensors, in terms of detection limit and sensitivity [12].


**Table 1.** Percentages used in the construction of each electrode.

#### *2.2. Composite Electrode Chacaterization*

Electrodes were characterized using Cyclic Voltammetry (CV) and Electrochemical Impedance Spectroscopy (EIS) using a computer-controlled Multi-AUTOLAB M101 (Eco Chemie, Utrecht, The Netherlands) with a three-electrode cell: a platinum-based electrode 53–671 (Crison Instruments, Alella, Barcelona, Spain) as a counter electrode, an Ag/AgCl handmade electrode as a reference electrode and the constructed composite electrodes as working electrodes. The characterization was performed in solution composed of 0.01 M K4Fe(CN6), 0.01 M K3Fe(CN6) and 0.1 M KCl. For CV, the scan rate was 10 mV·s−<sup>1</sup> and the rate of frequencies used in EIS was 0.01 to 104 Hz.

The behavior of the 15% rGO electrode was unusual, possibly related to the orientation of the layers in the Epotek H77 matrix, and its characterization using CV and EIS was not successful. In Figure 1, the characterization of the rest of the carbon electrodes, with graphite or CNTs, can be observed. The most notable difference is showed in EIS, where the 20% graphite presents the lower charge transference resistance. Thus, a highly conductive surface is then available for the preconcentration of cationic metals.

**Figure 1.** CV (**a**) and EIS (**b**) characterization of the different electrodes.

#### *2.3. Composite Electrode Modification with Hg-Nps*

After electrode characterization, the surface of the electrode is modified with mercury nanoparticles (Hg-NPs) following the synthesis from [9]. In the synthesis, 78 mg Hg2(NO3)2·2H2O is used, 1 mL 1 M HNO3 is added and then 0.5 mL of a solution of 3.5 g of PVA (Polyvinyl Alcohol) added to 16 mL of Milli-Q water. All the steps of the synthesis were performed at 25 ◦C and under stirring conditions.

A total of 20 μL of the nanoparticle solution is drop casted on the electrode surface and dried in the oven at 80 ◦C for 2 h. The modified electrodes were characterized using scanning electron microscopy (SEM) (MerlinFe-SEM, Carl Zeiss, Germany) and the Hg-NPs were characterized using transmission electron microscopy (TEM) (JEM-2011 200 kV, Jeol, Peabody, MA, USA) (see Figure 2).

**Figure 2.** (**a**) Retrodispersive (**left**) and secondary electron (**right**) SEM images; (**b**) 20% graphite electrode drop casted with Hg-NPs image; (**c**) TEM image of the Hg-NPs.

#### *2.4. Metal Solution Preparation and Determination*

The metal solutions were prepared using certified stock standards of 37 mg·L−<sup>1</sup> Pb(NO3)2 (≥99%, supplied from Sigma-Aldrich), 11,438 mg·L−<sup>1</sup> Cu(NO3)2 (99.5 %, purchased from Merck) and 1000 mg·L−<sup>1</sup> Cd(NO3)2 (99 %, obtained from Panreac). They were added to a 0.1 M acetic acid (CH3COOH, 99.9% acquired from J.T.Baker, HPLC

reagent)/0.1 M Ammonium acetate (NH4CH3COO, 97 % purchased from Panreac) buffer with Milli-Q water at pH 4.6 [13].

#### *2.5. Bare Composite Electrodes*

For metal determination, the technique chosen was SWASV. This consists of applying a potential (−1.4 V) for 7 min that reduces the metal ions on the electrode surface; then, staircase potential is applied and the current generated is recorded. This process is performed under N2 bubbling. Moreover, a modification in the electrochemical cell is used. Instead of using a handmade reference electrode, the one used for the measurements is Orion 900 electrode (Thermo Scientific, Beverly, MA, USA).

Firstly, the bare electrodes were used for the electrochemical detection of Cd, Pb and Cu. The results for all electrodes studied are shown in Figure 3.

**Figure 3.** Calibration curves for Cd (**a**), Pb (**b**) and Cu (**c**) for each raw material.

As can be seen, 20% graphite electrodes showed the best response, as it has a better sensitivity compared with 15% graphite and 10% CNTs composite electrodes for three metal cations analyzed.

#### *2.6. Hg-NPs Drop Casted Electrodes*

The next step is to modify the surface of the 20% graphite electrode with Hg-NPs, as mentioned above. Once the surface is modified, the electrode is tested for Cd, Pb and Cu determination using SWASV. The corresponding results are shown in Figure 4.

**Figure 4.** Calibration curves for Cd (**a**), Pb (**b**) and Cu (**c**) for 20% graphite (black) and 20% graphite plus Hg-NPs (blue).

With this modified 20% graphite electrode, lower quantification limits can be achieved. In Table 2, all the parameters of the calibration curves are summarized.



#### **3. Conclusions**

Carbon composite electrodes are very versatile, robust, and reliable electrodes to work with for Cd, Pb and Cu detection. The well-known properties of mercury to form an

amalgam with other metals can be taken advantage of to modify the surface of the carbon composite electrode in order to decrease the limit detection of the bare electrode. To emulate the polarography, the use of Hg-NPs reduces the amount of mercury used without losing its properties. In this case, Cd and Pb form an amalgam with Hg, reducing the detection limit (Cd = 0.05 mg·L−1; Pb = 0.045 mg·L−1) in comparison with the bare electrode. The Cu metallic cation does not exhibit this behavior. Although the bare electrode has higher sensitivity because its electroactive area is not modified, when the electrode was modified with Hg-NPs, its electroactive area decreases. We added a polymer (from the synthesis of the NPs) over the electrode's surface that is not as good as a conductor as graphite. On the other hand, we improved the detection limit due to the specific interaction of mercury with metals cations.

**Supplementary Materials:** The following are available online at https://www.mdpi.com/article/10 .3390/CSAC2021-10456/s1.

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

**Funding:** Authors are thankful for the financial support from the RTI2018-099362-B-C21 research project from the Spanish Ministerio de Economía y Competitividad y Fondo Europeo de Desarrollo Regional (MINECO/FEDER, UE).

**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:** Laia L. Fernández acknowledges Universitat Autònoma de Barcelona (UAB) for the PIF grant. She thanks Servei de Microscopia from UAB for the assistance in electron microscopy characterization.

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

#### **References**


### *Proceeding Paper* **Colorimetric Determination of Nitrate after Reduction to Nitrite in a Paper-Based Dip Strip †**

**Amer Charbaji \* , Hojat Heidari-Bafroui , Nasim Rahmani, Constantine Anagnostopoulos and Mohammad Faghri \***

> Microfluidics Laboratory, Department of Mechanical, Industrial and Systems Engineering, University of Rhode Island, 2 East Alumni Avenue, Kingston, RI 02881, USA; h\_heidari@uri.edu (H.H.-B.); nara7@uri.edu (N.R.); anagnostopoulos@uri.edu (C.A.)

**\*** Correspondence: charbaji@uri.edu (A.C.); faghrim@uri.edu (M.F.)

† Presented at the 1st International Electronic Conference on Chemical Sensors and Analytical Chemistry, 1–15 July 2021; Available online: https://csac2021.sciforum.net/.

**Abstract:** Paper-based microfluidic technology is a relatively new field of research that provides low-cost platforms and sensors for point-of-care diagnostics. While the majority of research in this field has been for biomedical applications, more and more paper-based devices and platforms are being designed and developed for environmental applications, such as water quality monitoring and assessment. One such application is the detection of nitrate in water samples. Colorimetric detection of nitrate by paper-based devices using the Griess assay requires the reduction of nitrate to nitrite before undergoing the reaction. In this paper, we measured the performance of a paper-based dip strip for detecting nitrate and nitrite by calculating its limit of detection and limit of quantification. We also calculated the reduction efficiency of vanadium (III) chloride in the dip strip for detecting nitrate. Our results show that the reduction time of nitrate via vanadium (III) chloride is much longer than that when using zinc microparticles. Our results also show that the performance of the dip strip using vanadium (III) chloride for nitrate detection is not as good as more intricate paper-based devices that have a separate reaction zone with zinc microparticles. The limits of detection and quantification calculated were 3.352 and 7.437 ppm, and the nitrate reduction efficiency varied over the range of nitrate concentrations tested.

**Keywords:** nitrate reduction; zinc microparticles; vanadium (III) chloride; materials for chemical sensing; nitrate detection; Griess reaction; colorimetric assay; paper-based devices; paper microfluidics; point-of-care diagnostics

#### **1. Introduction**

Paper-based microfluidic technology has been gaining a lot of attention over the past several years for the many advantages it provides. Most importantly, paper-based microfluidic technology allows the development of low-cost, portable and easy-to-use devices and sensors that can be easily disposed of. These devices can also provide qualitative or quantitative results and data at the point of care without the need for specialized equipment or power sources. Several paper-based devices have been developed for various applications, such as for water analysis [1–4], biomedical applications [5,6], food analysis [7–10], soil analysis [11] and many other miscellaneous applications [12–15]. The field of paper-based microfluidics is expected to continue garnering greater attention as more applications are sought after or the performance improved for the ones already developed [16].

Paper-based devices are generally made up of several different sections that serve different purposes. While more complex devices may include valves and actuators to manipulate fluids and perform multistep reactions [17,18], simpler devices generally include a sample port, transport channels, reactions zones and a detection zone [19]. The majority of paper-based devices use colorimetric detection since it is the simplest technique

**Citation:** Charbaji, A.; Heidari-Bafroui, H.; Rahmani, N.; Anagnostopoulos, C.; Faghri, M. Colorimetric Determination of Nitrate after Reduction to Nitrite in a Paper-Based Dip Strip. *Chem. Proc.* **2021**, *5*, 9. https://doi.org/10.3390/ CSAC2021-10459

Academic Editor: Manel del Valle

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/).

to produce a quantifiable signal [20,21]. Properties of the material used in paper-based devices influence assay performance and have a substantial impact on the development of paper-based sensors [22]. Therefore, proper material selection and optimization is critical to enhancing the performance of assays in paper-based devices [23]. This is usually an iterative and an ongoing process to learn and adapt different advancements in the field of paper-based technology to check for the possibility of improving the output and performance of paper-based sensors. An example is the selection of a suitable reducing agent to be used in a paper-based device meant for detecting nitrate in water.

Nitrate is part of the nitrogen cycle [24] and is an essential nutrient needed for plant growth; however, it plays a significant role in water nutrient pollution when present in excessive amounts [25,26]. Nitrate is also the most stable form of nitrogen in oxygenated systems, and all other forms of nitrogen-containing compounds can become a source for it [27,28]. Ingesting nitrate has been linked to colorectal cancer, thyroid disease and central nervous system birth defects [29]. Therefore, it is important to measure nitrate levels in water for environmental monitoring purposes and to ensure its safety for consumption. Different techniques are readily available to measure nitrate concentrations in water but are either costly, time-consuming or may require trained personnel [30,31]. Several paper-based sensors have been developed for the rapid and inexpensive detection of nitrate in water, food and human saliva, and their limits of detection (LOD) and limits of quantification (LOQ) are given in Table 1.


[35] Food Sample 0.4 1.4 [36] Human Saliva 4.96 16.74

**Table 1.** Performance of paper-based sensors developed for detecting nitrate in different media.

<sup>1</sup> NA, not available.

All of the paper-based devices developed thus far for measuring nitrate levels have used the Griess assay for detection since it is the most commonly used spectrophotometric method for quantifying concentrations of nitrate and nitrite [37,38]. However, this assay is specific to nitrite molecules and, therefore, nitrate molecules have to be reduced to nitrite first before detection. There are several different reducing agents that can reduce nitrate to nitrite, such as cadmium, copperized cadmium, zinc, nitrate reductase, irradiation by ultraviolet light, hydrazine sulfate, titanium (III) chloride, vanadium (III), hydroxylamine, tin chloride or ascorbic acid [36,39,40]. Some of these reducing agents are not suitable for use in paper-based devices, while others have been tested and used in this type of sensors.

Nitrate reductase, irradiation by ultraviolet and hydrazine require lengthy reduction times [41], which may not be suitable for paper-based sensors due to concerns of sample evaporation. Titanium (III) chloride is violet in color and absorbs light in the same range as the azo dye product of the Griess assay [41]. Ferreira et al. [36] tested tin chloride, hydroxylamine, ascorbic acid and zinc microparticles. They used zinc microparticles in their paper-based nitrate sensor since the other agents tested did not extensively reduce nitrate to nitrite. Experimental results by Jayawardane et al. [33] showed that cadmium and zinc microparticles produced similar results for nitrate reduction in their paper-based device. They opted for zinc microparticles due to the higher toxicity of cadmium. Thongkam et al. [35] developed a very simple paper-based device for measuring nitrate and nitrite concentrations in food samples, and they used vanadium (III) chloride to reduce nitrate before detection.

We had previously developed a sensitive paper-based nitrate sensor by testing different device architectures and optimizing the different components of the device [32]. The final device adopted a folding architecture with part of the detection chemistry immobilized

at the detection zone. This improved the quality and uniformity of the signal developed. The device also incorporated a new composite material made-up of zinc microparticles and cellulose fibers to enhance nitrate reduction. A nitrate conversion efficiency of 27% was achieved using this new composite material called Zinculose [42]. However, the results obtained by Thongkam et al. [35] for nitrate detection in food samples by using vanadium (III) chloride as a reducing agent are very promising. In this paper, we measure the performance of a dip strip using vanadium (III) chloride for reducing nitrate by calculating its limits of detection and quantifications. We also calculate the nitrate reduction efficiency of vanadium (III) chloride and compare the results to those obtained when using Zinculose.

#### **2. Methods**

Thongkam et al. [35] studied the effect of the different parameters on nitrate detection. They tested different concentrations of sulfanilic acid and N-(1-Naphthyl)ethylenediamine dihydrochloride used in the Griess assay for detection. They also examined the effect of different concentrations of vanadium (III) chloride and reaction times on the intensity of the color produced in the detection zone. In this paper, we use the optimum concentrations they have found when preparing the reagents to be used in our experiments.

#### *2.1. Materials*

The items below were used in preparing and running the experiments presented in this paper. Whatman grade 1 filter paper (GE Healthcare Whatman 1-1001824), backing cards (DCN Dx MIBA-050), sulfanilamide (98%, Alfa Aesar-A1300136), N-(1-Naphthyl)ethylenediamine dihydrochloride (Alfa Aesar-J6321414), hydrochloric acid (Fisher Chemical-A142-212), sodium nitrate (≥99.5%, Honeywell Fluka-31440), sodium nitrite (≥99%, Honeywell Fluka-31443) and ASTM Type 1 deionized water (resistivity > 18 MΩ/cm, LabChem-LC267405).

#### *2.2. Methods*

Strips 1 × 8 cm were cut out from a 30 × 8 cm backing card using a guillotine paper cutter. Three circles, 6 mm in diameter each, were punched out using a tissue biopsy from the Whatman filter paper and stuck onto the backing card, Figures 1 and S1.

**Figure 1.** (**a**) Schematic showing the components and dimensions of the dip strip used. (**b**) The yellow circle shows the color analysis zone used in ImageJ to quantify the color intensity of one of the detection zones; the diameter of the circle is about 125 pixels, which is approximately 5.3 mm.

Nitrate and nitrite solutions at concentrations of 1000 ppm were freshly prepared on the day of testing by dissolving the required amount of nitrate or nitrite salt in deionized water. These solutions were then diluted using deionized water into the following concentrations 0.5, 1, 2.5, 5, 10, 20 and 40 ppm. We followed the procedure outlined by Thongkam et al. [35] in preparing the detection chemistry for nitrate and nitrite. For nitrite detection,

the solution was called reagent "A" and consisted of equal parts (1:1 ratio) volume of sulfanilic acid and NED solution. For nitrate detection, the solution used was called reagent "B" and consisted of equal parts (1:1:1 ratio) volume of the above sulfanilic acid, NED solution and the reducing reagent solution. The sulfanilic acid used in reagents "A" and "B" was prepared by dissolving 0.1 g of sulfanilamide in 100 mL of 2 mol L−<sup>1</sup> hydrochloric acid. The NED solution used in reagents "A" and "B" was prepared by dissolving 0.1 g of N-(1-Naphthyl)ethylenediamine dihydrochloride in 100 mL of deionized water. The reducing reagent solution used in reagent "B" was prepared by dissolving 3 g of vanadium (III) chloride in 100 mL of 6 mol L−<sup>1</sup> hydrochloric acid. 2 μL of reagent A or B was pipetted onto each circle and allowed to air-dry for at least 30 min, Figure S2. Each dip strip was then submerged into the appropriate nitrate or nitrite solution for 1 s, shaken to remove excess fluid and then scanned using a desktop scanner (Canon TS6020) at a resolution of 600 DPI. The nitrate dip strips were scanned after 10 min, and the nitrite dip strips were scanned after 5 min following the optimized scan times previously found by Thongkam et al. [35]. The detection zones were analyzed using ImageJ in RGB mode, similar to how they analyzed their results. We have previously shown that the green component of the measured color intensity shows the largest difference in value over the concentration of nitrate or nitrite for paper-based devices using the Griess assay [43]. Therefore, the data for the different color intensities were provided in the supplementary file, Tables S1–S4. A MATLAB code was used to fit the data to an exponential decay function of the form y=a × exp (−x/b) + c, and the symbolic toolbox was used to calculate the limits of detection and quantification. The limits of detection and quantification were obtained by finding the analyte concentrations corresponding to *yLOD* or *yLOQ* on the calibration curves developed. *yLOD* or *yLOQ* were calculated using the following equations [44]:

$$y\_{LOD} = \overline{y}\_B - 3\,\sigma\_B$$

$$y\_{LOQ} = \overline{y}\_B - 10\,\sigma\_B$$

where *yB* corresponds to the mean color intensity of the blank solution (0 ppm) and *σ<sup>B</sup>* is its respective standard deviation.

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

The detection zones of the nitrate dip strips showed little to no color change after 10 min, Figure S3, but color started to form after a much longer wait time, so the dip strips were scanned after 1 h as well, Figure S4. The following section shows the results obtained for the nitrate and nitrite dip strips.

#### *3.1. Nitrate and Nitrite Analysis*

Figure 2 shows the calibration curves developed for the detection of nitrate in deionized water after a reaction time of 10 min and 1 h. The limits of detection and quantification for nitrate after 10 min are 37.03 and 121 ppm, respectively. The limits of detection and quantification for nitrate after 1 h are 3.352 and 7.437 ppm, respectively.

Figure 3 shows the calibration curves developed for the detection of nitrite in deionized water after a reaction time of 5 min and 1 h. The limits of detection and quantification for nitrite after 5 min are 0.522 and 0.854 ppm, respectively. The limits of detection and quantification for nitrite after 1 h are 0.889 and 1.823 ppm, respectively.

**Figure 2.** (**a**) An exponential decay calibration curve in the formy=a × exp (−x/b) + c, where a = 2741, b = 41,430 and c = −2492 was established for nitrate after a reaction time of 10 min. (**b**) An exponential decay calibration curve in the form y=a × exp (−x/b) + c, where a = 80,230, b = 147,700 and c = −79,980 was established for nitrate after a reaction time of 1 h. The error bars represent the standard deviation.

**Figure 3.** (**a**) An exponential decay calibration curve in the formy=a × exp (−x/b) + c, where a = 54.94, b = 14.57 and c = 197.4 was established for nitrite after a reaction time of 5 min. (**b**) An exponential decay calibration curve in the form y=a × exp (−x/b) + c, where a = 45.56, b = 15.65 and c = 205.2 was established for nitrite after a reaction time of 1 h. The error bars represent the standard deviation.

#### *3.2. Reduction Efficiency*

The reduction efficiency of vanadium (III) chloride was calculated using the data obtained in the above experiments used to calculate the LOD and LOQ for nitrate and nitrite. First, the results obtained from the nitrite experiment after 1 h were used to establish the calibration curve using the method outlined in Section 2.2. Then the results obtained from the nitrate experiment after 1 h were used to calculate the intersection of the measured result with the calibration established for nitrite using the symbolic toolbox. Table 2 gives the nitrate conversion efficiency calculated. As can be seen from the table, the conversion efficiency varies between almost 0% and 27%.


**Table 2.** Calculated nitrate conversion efficiency.

<sup>1</sup> This concentration is normalized by subtracting the intensity calculated for 0 ppm from all other concentrations.

#### *3.3. Discussion*

The limits of detection and quantification obtained for nitrate and nitrite in our analysis were much higher than those obtained by commercial dip strips using the Griess assay. This can be attributed to one or more of the following reasons: using the RGB mode in data analysis, not depositing enough reagent volume for reaction or using hydrochloric acid since it evaporates completely without producing acidic conditions when rewet. The reaction with the Griess assay should take place under acidic conditions [45].

A maximum reduction efficiency of 27% was obtained by vanadium (III) chloride. This is similar to the reduction efficiency obtained by Zinculose (27%). However, this reduction efficiency was only obtained for a high nitrate concentration of 40 ppm, while lower concentrations resulted in a much lower reduction efficiency. This raises the question of repeatability and uniformity of vanadium (III) chloride nitrate reduction when used in paper-based devices.

Each of the two reducing agents, zinc microparticles and vanadium (III) chloride, has its own set of advantages and should be used in specific applications with an appropriate device design. Zinculose is a composite material that can be incorporated into any paperbased device. The zinc microparticles in Zinculose are held in place by the matrix, which allows the passage of more sample volume through the material and the reduction of more molecules as they pass through it. This allows for signal amplification as more molecules become available to be captured and detected. However, vanadium (III) chloride is not immobilized and would wash away in any lateral flow paper-based device design. Nitrate reduction using vanadium (III) chloride takes much longer than that by zinc microparticles. That is why commercial dip strips generally use zinc microparticles in the detection zone to reduce nitrate to nitrite before detection, Figure S5. Vanadium (III) chloride allows for the development of simple dip strips since the reducing reagent can be mixed with the detection chemistry and easily deposited in the detection zone. However, the limits of detection and quantification achieved by dip strips utilizing vanadium (III) chloride are not as good as those obtained in more intricate designs using zinc microparticles.

#### **4. Conclusions**

Paper-based microfluidic technology is a relatively new field of research that is gaining a lot of attention and is producing a lot of innovation. In this paper, we measured the performance of a dip strip utilizing vanadium (III) chloride to reduce nitrate before detection. We observed that vanadium (III) chloride has some drawbacks that make it impractical for use in paper-based devices meant for detecting nitrate. These include long reduction times required and low limits of detection and quantification obtained. Therefore, we recommend using zinc microparticles as the reducing agent for nitrate detection in paper-based devices. Future work will include developing a suitable lightbox, similar to [46], that emits green light for measuring nitrate and nitrite concentrations using paper-based devices utilizing the Griess assay in the field.

**Supplementary Materials:** The following are available online at https://www.mdpi.com/article/ 10.3390/CSAC2021-10459/s1, Figure S1. Dip strip used in experiments, Figure S2. Dip strip used in experiments after the solutions are dried on the detection zones, Figure S3. Color formed in the detection zone vs. nitrate or nitrite concentrations after several minutes, Figure S4. Color formed in the detection zone vs. nitrate or nitrite concentrations after 1 h, Table S1. ImageJ analysis of nitrate detection zones after 10 min. Test order was randomized, Table S2. ImageJ analysis of nitrate detection zones after 1 h. Test order was randomized, Table S3. ImageJ analysis of nitrite detection zones after 5 min. Test order was randomized, Table S4. ImageJ analysis of nitrite detection zones after 1 h. Test order was randomized, Figure S5. Zinc microparticles observed using an electron scanning microscope with EDS analysis in the nitrate test fields of commercial dip strips (a) Quantofix 91313 (b) Quantofix 91351.

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

**Funding:** This research was funded by the National Science Foundation under EPSCoR Cooperative Agreement #OIA-1655221.

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

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Data is contained within the article or supplementary material. Additional data not presented in this article is available on request from the corresponding author.

**Acknowledgments:** The authors would like to acknowledge the support from the Rhode Island EPSCoR, which is funded by the National Science Foundation under Award #OIA-1655221. The authors would also like to acknowledge the students, research scientists and visiting scholars at the Microfluidics Laboratory at the University of Rhode Island for their help and support. SEM and EDS data were acquired at the RI Consortium for Nanoscience and Nanotechnology, a URI College of Engineering core facility partially funded by the National Science Foundation EPSCoR, Cooperative Agreement #OIA-1655221.

**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results.

#### **References**

