Voltammetric Sensor Based on Titania Nanoparticles Synthesized with Aloe vera Extract for the Quantification of Dithiophosphates in Industrial and Environmental Samples

Abstract

In this work, TiO₂ spherical nanoparticles with a mean diameter of 10.08 nm (SD = 4.54 nm) were synthesized using Aloe vera extract. Rutile, brookite, and anatase crystalline phases were identified. The surface morphology of a carbon paste electrode does not change in the presence of nanoparticles; however, the surface chemical composition does. The voltammetric response to dicresyl dithiophosphate was higher when the electrode was modified with TiO₂ nanoparticles. After an electrochemical response study from pH 1.0 to 12.0, pH 7.0 was selected for the electroanalysis. The electroactive area of the modified sensor was 0.036 cm², while it was 0.026 cm² for the bare electrode. The oxidation process showed mixed adsorption-diffusion control. The charge transfer resistance of the modified sensor (530.1 Ω, SD = 4.08 Ω) was much lower than that of the bare electrode (4298 Ω, SD = 8.53 Ω). The linear quantitative range by square wave voltammetry was from 5 to 150 μmol/L, with a limit of detection of 1.89 μmol/L and a limit of quantification of 6.26 μmol/L under optimal pulse parameters of 50 Hz frequency, 1 mV step potential, and 25 mV pulse amplitude. The sensor response was repeatable and reproducible over 30 days. The results on real flotation and synthetically contaminated soil samples were statistically equivalent to those obtained by UV-vis spectrophotometry. A dithiocarbamate showed an interfering effect on the sensor response to dithiophosphate.

1. Introduction

Organic dithiophosphates are a group of substances containing as a functional group a dithiophosphate substituted by alkyl and/or aryl chains. They are generally presented as alkali and transition metal salts where the anion has the general formula (RO)₂PS₂⁻. Organic dithiophosphates have several industrial applications, one of the best known being their use as reagents in the flotation process of sulfide and platinum group minerals [1,2,3]. In addition, these substances have other important applications, mainly due to their ability as chelating ligands. This allows them to be used as highly selective extractive reagents of some metals. Dialkyl dithiophosphates do not form stable complexes with alkali and alkaline earth metals nor with Ge(IV), Mn(II), Pd(II and IV), Ru(III), Th(IV), or Fe(II), among others. In this way, it is possible to separate with high selectivity ions such as Ag⁺, Cu⁺, Cd²⁺, or Tl⁺ from matrices in which they are present. They are also used as additives in catalysts, pesticides, organic light-emitting diodes, pharmaceuticals, as reagents for synthesis, among others [4].

There is growing concern about the effects of organophosphates on human health as well as on flora and fauna [4]. The toxicity of dithiophosphates to microorganisms appears to be greater than that of other sulfhydryl-type flotation reagents. In a study focused on mineral bioprocessing, dithiophosphates were found to be more toxic to acidophilic bacteria and archaea than xanthates, dithiocarbamates, and thionocarbamates [5]. This demonstrates the potential toxic effect on fauna in mining ecosystems. In humans, dialkyl dithiophosphates reduce the proliferation of peripheral blood mononuclear cells due to their cytotoxic effect. There is also some evidence that these substances are immunosuppressive and may cause autoimmune diseases [4].

Different methods have been used to quantify organic dithiophosphates, mainly focusing on their toxicity due to their classification as organophosphates. One of the most widely used methods is chromatography, which has been performed by the ion exchange technique, gas chromatography, and hybrid techniques such as gas chromatography or liquid chromatography–tandem mass spectrometry [4]. This method is very sensitive and selective, but an important disadvantage is its high cost due to the instrument and ultrapure reagents. This is more relevant for low-performance applications such as the control of industrial reagents in technological processes such as mineral flotation. There are other less expensive alternatives, such as UV-vis spectrophotometry [6]. However, optical methods have an important drawback for samples with high solids content, such as those from mineral processing, because dispersive and absorptive effects on the incident radiation cause serious analytical deviations [7]. Electroanalytical methods are the best alternative to these problems. Dithiophosphates have been quantified by polarography [8,9], but this technique is no longer recommended due to the toxicity of mercury. The best current option is voltammetry with solid electrodes, especially those using materials with electrocatalytic effect and high surface area, such as metallic nanoparticles of Pt and Au [10] or oxides such as TiO₂, CuO, Fe₂O₃, and NiO [11,12].

Nanometric metal oxides are of particular interest because there are synthesis methods that allow the construction of economically viable electrochemical sensors. Among the various types, carbon paste electrodes (CPEs) are one of the better options as they allow for easy and inexpensive modification of the electrochemical interface to improve the electrochemical response and, thus, analytical performance. Metal oxide nanomaterial-modified CPEs have been used for the determination of both inorganic and organic analytes in samples of environmental, pharmaceutical, and clinical interest, among others [13,14,15,16,17,18]. However, there has been no progress in the development of electrochemical sensors, whether CPEs or not, for the determination of dithiophosphates in industrial and environmental samples using nanometric metal oxides as modifiers.

In this context, TiO₂ nanoparticles may be a suitable modifier material for a CPE for dithiophosphates due to their easy synthesis, high electrocatalytic performance, low toxicity, and biocompatibility [19,20]. This nanomaterial has mostly been synthesized by the sol-gel method, which consists of hydrolyzing a titanium precursor, such as titanium alkoxides, to form hydroxide particles with colloidal dimensions (sol) that eventually grow in the form of a polymeric network to form a gel that, when dried and often calcined, allows obtaining the TiO₂ nanomaterial. However, there is a wide variety of methods, such as solvothermal synthesis, microwave-assisted synthesis, anodization, and laser-assisted synthesis, among others [20,21,22]. Nevertheless, among their main drawbacks, traditional methods use toxic reagents and several stages involving high temperatures, and although they are effective in obtaining titania of nanometric dimensions (<100 nm), there is currently a great interest in the so-called green methods, which are more environmentally friendly. Among them, biological approaches stand out, in which organisms or bioproducts are used to obtain a substance or material. In the case of TiO₂, plant extracts such as Jatropha curcas, Azadirachta indica, and Psidium guajava have been used [20]. The use of Aloe vera extract for the synthesis of titania is relatively new; this extract has been used to synthesize other nanomaterials, such as silver nanoparticles, but it is only recently that Olcay et al. [23] have applied it to TiO₂. In the process, the aloin extracted from Aloe vera acts as a Ti⁴⁺ reducing agent during the synthesis and as a stabilizer of the nanoparticles [23]. According to Alarif et al. [24], the synthesis using hydroxylated extracts consists of a first reductive stage to form metallic titanium, from which TiO₂ is formed through successive stages of growth, stabilization, drying, and calcination. An important advantage is that this green synthesis minimizes the application of high temperatures compared to the traditional sol-gel method [25]. The TiO₂ nanoparticles obtained using the Aloe vera extract have not been investigated for their photocatalytic and electrocatalytic applications, including the construction of sensors.

This paper presents the development of a voltammetric sensor for dialkyl/aryl dithiophosphate salts. For the construction of the sensor, TiO₂ nanoparticles obtained by a green method with Aloe vera were used. The sensor was validated on a mineral flotation sample and soil contaminated with dithiophosphate. The novelty of the work lies in the fact that a new voltammetric sensor with a low detection limit is reported for the quantification of dithiophosphates in industrial and environmental samples, making the quantification of these analytes cost-effective compared to other more expensive methods. In addition, titania nanoparticles obtained using Aloe vera extract are applied to electrochemical sensing for the first time.

2. Materials and Methods

2.1. Reagents and Instruments

All reagents used in this work were of high purity (ACS grade) and purchased from Sigma-Aldrich (St. Louis, MA, USA). Deionized water (0.055 µS/cm) from the Hydronix purification system (Ontario, CA, USA) was used. All instruments and materials were certified for use.

A VersaSTAT 3F potentiostat/galvanostat (Ametek, Berwyn, PA, USA) was used for the electrochemical experiments. An Autolab PGSTAT30 potentiostat/galvanostat (Metrohm AG, Herisau, Switzerland) was used only in the case of the square wave voltammetry technique. The reference electrode was Ag/AgCl/KCl(sat), while the auxiliary electrode was Pt. Either an unmodified CPE or one modified with TiO₂ nanoparticles (TiO₂/CPE) was always used as the working electrode, depending on the experiment. Before each electrochemical experiment, nitrogen gas was used for 3 min to purge the dissolved gases in the solution. The supporting electrolyte was a 0.1 mol/L phosphate buffer containing 0.1 mol/L KNO₃. To maintain a constant temperature throughout each experiment, a jacket cell with a temperature-controlled water circulation system was employed. For open potential circuit (OCP) stability, 5 min were waited before recording all voltammograms.

For the X-ray diffraction (XRD) method, a Bruker D8-Advance instrument (Bruker, Billerica, MA, USA) operated at 30 kV with CuKα 1.5406 Å radiation was used. For transmission electron microscopy (TEM), a HR-TEM FEI Tecnai F30 instrument (Thermo Fisher Scientific Inc., Waltham, MA, USA) operated at 300 kV was employed, whereas a JEOL JSM-6610LV instrument (JEOL, Tokyo, Japan) operated at 20 kV with secondary electron and backscattered electron detectors was used for the scanning electron microscopy (SEM).

2.2. Methodology

2.2.1. Synthesis and Characterization of Nanoparticles and Electrode Surfaces

The synthesis of TiO₂ nanoparticles was carried out using Aloe vera extract and 2 mol/L TiCl₄ as precursors. The synthesis was conducted in 1.0 L of deionized water where Aloe vera extract was added dropwise under constant stirring at 400 rpm and a constant pH of 7.0, which was maintained with drops of HCl and NaOH, both 2 mol/L. The synthesis product was recovered by centrifugation and washed with plenty of ethanol and water. The nanomaterial was then dried at 100 °C for 7 h, which yielded a white-colored material, a typical appearance of TiO₂ [23].

The crystallinity of the synthesized nanomaterial and the composition of the crystalline phases were investigated by XRD. The morphology and size of the nanomaterial were characterized by TEM. The histogram of the nanomaterial size was generated using SPSS Statistics 22.0 software, with measurements taken from micrographs using Image J 1.54 software. The qualitative elemental composition of the nanomaterial was analyzed in the TEM instrument by energy dispersive X-ray spectroscopy (EDS).

The morphology and surface composition of the CPE containing the nanomaterial in comparison with the unmodified paste electrode were studied by SEM. The EDS spectra were recorded for both electrodes, and elemental mapping was performed to study the distribution of titanium on the surface after the modification.

2.2.2. Preparation of Carbon Paste Electrodes

The electrochemical sensor consisted of a CPE modified with TiO₂ nanoparticles. The modified electrode was prepared by first mixing 0.5 g of graphite (<20 μm) with 0.1 g of nanometric TiO₂ in a porcelain capsule. This mixture was homogenized, and 0.85 mL of mineral oil was added to act as a binder. A homogeneous paste was mixed and introduced into an insulin-type syringe, previously modified to include a copper wire as a conductive contact between the paste and the connector of the potentiostat. The electrode was allowed to rest for 24 h before being used for electrochemical measurements. A CPE without nanomaterial was fabricated following the same methodology but using 0.6 g of graphite to prepare the paste [26].

2.2.3. Study of the Electrochemical Response to Dithiophosphate

The electrochemical response to 5.4 mmol/L sodium dicresyl dithiophosphate (DCDTP) was investigated using the carbon paste electrodes with and without modification with the nanomaterial. Cyclic voltammograms were recorded from 0 to 1000 mV vs. OCP at a scan rate of 50 mV/s. The sensor response to DCDTP was studied for different pH values from 1.0 to 12.0 to select the highest peak current. This was performed to maximize the electroanalytical response to develop the quantitative voltammetric method using the electrochemical sensor.

The electroactive surface area of the CPE, both with and without nanoparticles, was evaluated by cyclic voltammetry. Potassium ferrocyanide at 10 mmol/L in 1 mol/L KNO₃ supporting electrolyte was used as the electroactive system. Cyclic voltammograms were obtained by varying the scan rate in the range of 10 to 80 mV/s. The current peak was plotted versus the square root of the scan rate according to the Randles-Ševčík equation [27]. The area was calculated using Equation (1):

$$ECSA={\displaystyle \frac{S}{\left(2.69·{10}^5\right)n^{3/2}{D}^{1/2}c}}$$

(1)

where ECSA is the electrochemical active surface area (cm²), S is the slope of the plot, n is the number of electrons transferred, D is the diffusion coefficient (cm²/s), and c is the concentration of the electroactive species (mol/cm³).

2.2.4. Electrochemical Study of the Electrodes

The controlling process when the peak current is reached in the voltammograms was determined by recording cyclic voltammograms for a 5.4 mmol/L DCDTP solution in a buffer medium at different scan rates ranging from 20 to 120 mV/s. A plot of the logarithm of the peak current versus the logarithm of the scan rate was constructed from these voltammograms. A slope close to 0.5 indicates diffusion control, a slope close to 0.75 indicates mixed control, and a slope close to 1.0 indicates adsorption control [28,29].

Electrochemical impedance spectroscopy (EIS) was used to determine the equivalent circuit of the electrochemical interface of both CPE and TiO₂/CPE. The method was performed in 10 mmol/L ferrocyanide prepared in a 1 mol/L KNO₃ solution as supporting electrolyte. To conduct the study, the electrochemical cell was placed in a Faraday cage to minimize interference during low frequency measurements. In addition, proper grounding was established. The DC potential was determined from the linear voltammograms. A potential amplitude of 10 mV was applied, and a frequency scan from 10⁻² to 10⁴ Hz was performed.

2.2.5. Linear Range, LOD, and LOQ

The quantitative linear range of the voltammetric method for dithiophosphate was evaluated by square wave voltammetry after optimizing the pulse parameters (frequency, potential step, and amplitude). For this purpose, a standard solution of 5.4 mmol/L was prepared, and dilutions were made to reach each of the concentrations evaluated. A linear regression of the peak current intensity as a function of concentration was performed to obtain the equation of the line and the statistical parameters of the regression.

The limit of detection (LOD) and the limit of quantification (LOQ) were calculated from the calibration curve. The LOD was calculated using Equation (2), while the LOQ was calculated using Equation (3):

$$LOD= 3{\displaystyle \frac{{SD}_i}{S}}$$

(2)

$$LOQ=3.3LOD$$

(3)

where SD_i is the standard deviation of the intercept of the calibration curve, and S is the slope of the calibration curve.

2.2.6. Repeatability and Reproducibility

The repeatability of the electroanalytical method with the TiO₂/CPE was evaluated by quantifying 10 times a 0.1 mmol/L DCDTP solution prepared by dilution of a 5.4 mmol/L standard solution and calculating the relative standard deviation (RSD) of the results. The reproducibility of the sensor was evaluated over 30 days by measuring the peak current of a 5.4 mmol/L DCDTP solution. An analysis of variance (ANOVA) for multiple sample comparison was used to determine whether there were statistically significant differences between the results for each of the days evaluated.

2.2.7. Trueness

The trueness of the voltammetric method for dithiophosphate using the TiO₂-based sensor was evaluated on a mineral flotation sample and on a soil sample synthetically contaminated with DCDTP. Following the approach of comparing two different analytical methods to assess trueness [30], the quantification by voltammetry and UV-vis spectrophotometry was performed, and a statistical comparison was conducted. The spectrophotometric determination was carried out at 267 nm by constructing a calibration curve from 0.05 to 1 mmol/L.

The flotation sample was obtained in the laboratory from a microflotation experiment of an ore containing pyrite. For this, 1 g of ore was conditioned with 30 mL of DCDTP solution in a beaker under stirring at 400 rpm for 6 min. Within this conditioning time, the pulp was transferred to the microflotation cell of a Hallimond tube, and the tube was placed to perform microflotation for 5 min, also at 400 rpm stirring and a nitrogen flow of 159 cm³/min. The flotation tail, containing suspended solids, was analyzed to quantify the DCDTP content. Dilutions were prepared for both voltammetric and spectrophotometric quantification. For the voltammetric method, no separation of the solids was performed, but for the spectrophotometric quantification, a scheme of filtration, centrifugation, and decantation was followed.

The synthetically contaminated soil sample was prepared by mixing a few drops of DCDTP with sufficient water and a quantity of natural soil. The mixture was homogenized and allowed to stand for 72 h. This soil was then continuously leached with 50 mL of deionized water to extract the DCDTP. This soil extract was used to prepare dilutions for voltammetric and spectrophotometric quantification. As in the case of the flotation sample, the soil extract sample was subjected to solids separation by filtration, centrifugation, and decantation for the spectrophotometric analysis but not for the voltammetric determination.

The voltammetric and spectrophotometric results for both samples were compared using a paired sample t-test. Analytical results with no significant differences were indicative of trueness according to international analytical regulations [30].

2.2.8. Analytical Interferences

The possible interfering effect of other flotation reagents on the voltammetric method for dithiophosphate was investigated. The cyclic voltammogram was recorded for a 5.4 mmol/L DCDTP solution and two solutions of the same concentration but containing 1 mmol/L potassium isopropyl xanthate (NaIPX) and 0.05 mL sodium dialkyl dithiocarbamate (DADTC), respectively. The second reagent was handled by volume because it is a concentrated solution sold for mineral flotation. Each experiment was performed in quintuplicate. A multiple-sample comparison was used to detect interference from any of the substances.

3. Results and Discussion

3.1. Characterization of Nanoparticles and Electrode Surfaces

The X-ray diffractogram of the TiO₂ obtained in this work is shown in Figure 1. The three most common crystalline phases of titania are identified, namely anatase, rutile, and brookite. It has been found that rutile can provide a higher surface area, while anatase has a better bulk transport of excitons to the surface, leading to a higher catalytic efficiency [31,32]. Although studies explaining the difference in catalytic efficiency have focused on the photocatalytic effect, all evidence suggests that the polymorphic composition may be favorable [33], which can be confirmed in Figure S1 for the case of the electrochemical oxidation of DCDTP, where the peak current intensity was higher for the nanoparticles obtained in this work compared to anatase phase TiO₂ of similar size obtained according to a sol-gel method with heat treatment reported by Bagheri et al. [25]. The higher electrochemical response with the polymorphic material avoids the prolonged heat treatment of at least 5 h for crystalline phase transformation, which contributes to the green nature of the method by saving energy consumption. The broad peaks in the diffractogram indicate that while the nanomaterial is predominantly crystalline, there is some amorphous content.

A TEM image of the TiO₂ obtained is shown in Figure 2a. The presence of abundant nanoparticles with spherical morphology can be observed, indicating that the synthesis leads to a high yield of nanomaterial. Figure 2b shows a higher magnification micrograph, in which the spherical morphology can be better defined. The frequency histogram with the size distribution curve is shown in Figure 2c, where it can be observed that the nanoparticles have a low size dispersion. The diameter range is comparable to that previously obtained by Olcay et al. [23], which extended from 5 to 35 nm. The mean diameter of the nanoparticles obtained in our work is 10.08 nm (SD = 4.54 nm). Figure 2d shows the EDS spectrum of the prepared nanomaterial, where peaks associated with titanium and oxygen are observed, corresponding to TiO₂. No peaks associated with other major elements are observed, indicating that a high-purity material was synthesized.

Figure 3a presents the SEM secondary electron micrograph of the CPE surface, showing a typical graphite flake morphology associated with the electrode composition. Figure 3b shows the SEM secondary electron micrograph of the TiO₂/CPE surface, where no differences are observed with respect to the flake morphology in Figure 3a. However, when backscattered electrons are used, no differences in brightness are observed in Figure 3c, which is due to the homogeneous carbon composition of the CPE as can be corroborated in the EDS spectrum, but in Figure 3d, distinct brightnesses can be seen related to the composition of graphite (dark) and TiO₂ (light), which is confirmed by the EDS spectrum. Figure 3e shows an EDS mapping of the TiO₂/CPE surface, where the distribution of the nanomaterial can be observed all over the surface, although at some points, a greater amount of titanium signals is concentrated due to the presence of some larger particles together with the nanometric ones, which could not be visualized in Figure 3d, but give a signal in the mapping. The nanometer sites are the ones that could cause an increase in the electrochemical response due to a combination of the electrocatalytic effect and the increase of the electroactive area.

3.2. Study of the Electrochemical Response to Dithiophosphate

Figure 4 shows the cyclic voltammograms of a 5.4 mmol/L DCDTP solution at pH 12.0 using the CPE and the TiO₂/CPE in the 0 to 1000 mV vs. OCP potential window. An anodic peak is observed at 437.98 mV for the CPE and 572.93 mV for the TiO₂/CPE, associated with the oxidation of the dithiophosphate. No cathodic peak is observed for either electrode, indicating that the oxidation is an irreversible process. Note that for the modified electrode, the OCP decreased (−110.41 mV) with respect to the CPE (−68.40 mV) due to the presence of the modifying nanomaterial in the carbon paste.

The electrochemical oxidation reaction of alkyl (or aryl) dithiophosphates occurs according to Equation (4), which is similar to that of xanthates, where the mechanism involves monoelectronic adsorptive steps leading to the alkyl dixanthogen [34]. In the case of a dithiophosphate, the product is a dimer with chemical formula ((RO)₂PS₂)₂ [35]. However, in the particular case of TiO₂/CPE, more research is required to understand the mechanism:

$$2{\left(\mathrm{R}\mathrm{O}\right)}_2{\mathrm{P}\mathrm{S}}_2^{-}\to {\left({\left(\mathrm{R}\mathrm{O}\right)}_2{\mathrm{P}\mathrm{S}}_2\right)}_2+2{\mathrm{e}}^{-}$$

(4)

Figure 5 shows the peak current intensity (cyclic voltammetry) versus pH in the range of 1.0 to 12.0 for a 5.4 mmol/L DCDTP solution. The peak current is virtually constant between pH 1.0 and 3.0, then increases at pH 4.0 and remains virtually constant until pH 7.0. At pH 8.0, the peak current decreases until it becomes constant at pH 11.0 and 12.0. Based on these results, pH 7.0 was chosen to develop the electroanalytical method.

Figure 6a shows the cyclic voltammograms for a 5.4 mmol/L DCDTP solution at different scan rates. The peak current for the oxidation of dithiophosphate increases with increasing scan rate from 20 to 120 mV/s. Figure 6b shows the plot of the logarithm of the peak current versus the logarithm of the scan rate. Such plots provide insight into the controlling process as the peak current is reached in the voltammograms. In this case, since the slope of the regression line is 0.73, a value very close to 0.75, it can be concluded that this is a process with mixed adsorption-diffusion control [28]. This means that the rate of oxidation of DCDTP at the electrochemical interface depends on the rate of diffusion of the analyte and the rate of adsorption on the electrode surface.

3.3. Electrochemical Study of the Electrodes

Figure 7a shows the cyclic voltammograms for a 10 mmol/L potassium ferrocyanide solution at different scan rates. An increase in the anodic and cathodic peak currents is observed with an increasing scan rate. Figure 7b shows the linear relationship between the peak current and the square root of the scan rate according to the Randles-Ševčík equation. This plot allowed the ECSA to be determined from the slope, which was found to be 0.036 cm². This area is larger than that obtained by the same procedure for the unmodified CPE, which was 0.026 cm². This indicates that the TiO₂ nanoparticles cause an increase in the electroactive area of the electrode. Nanoparticles, in general, are known to have a high surface area, including those synthesized by green methods [36].

The geometric area of both electrodes, i.e., without and with nanoparticles, was 0.033 cm², so the roughness factor (RF) can be calculated as the division of the ECSA by the geometric area. The RF for the CPE is 0.79, while for the TiO₂/CPE it is 1.09. The RF for the electrode without nanomaterial is less than 1, something commonly observed for carbon electrodes using non-conductive materials as a binder, in this case, mineral oil, which creates inactive sites at the electrochemical interface. This ultimately results in an electroactive area that is smaller than the geometric area [37]. However, when TiO₂ nanoparticles are present, the RF increases despite the effect that the binder may have.

Figure 8 shows Nyquist plots resulting from the EIS studies for both the CPE and TiO₂/CPE. Within the graph, the equivalent circuit has been drawn, the modeling of which is shown as a continuous line. The circuit consisted of the electrolyte resistance (R_s) in series with a parallel combination of a constant phase element (Q_dl), generally associated with the capacitance of the electrical double layer, and the charge transfer resistance (R_ct). In addition, a Warburg element (Z_W) was included in the faradic branch of the circuit, which represented the resistance to mass transfer. The use of a constant phase element instead of a pure capacitor is since, in the former case, non-ideal effects such as electrode surface roughness are taken into account [38,39]. Figure 8 shows a high degree of agreement between the experimental and simulated data for both electrodes.

Based on the results of the application of the EIS, the electrochemical parameters in

Table 1. Electrochemical parameters and statistics of the simulation of the equivalent circuit for CPE and TiO₂/CPE.

Electrode	CPE	TiO2/CPE
Rs (Ω)	0.009873 (SD = 1.25·10−6)	0.1563 (SD = 0.049)
Qdl (S·secn)	3.49·10−6 (SD = 6.12·10−8)	2.18·10−8 (SD = 0.73)
n	0.70	0.73
Rct (Ω)	4298 (SD = 8.53)	530.1 (SD = 4.08)
ZW S·sec0.5	6.53·10−5 (SD = 1.42)	3.62·10−4 (SD = 0.54)
C (F)	5.77·10−7	1.79·10−7
λ2	0.003	0.00031
$\mathrm{SD}: \mathrm{standard} \mathrm{deviation}; \mathrm{capacitance} \mathrm{calculated} \mathrm{as} \mathrm{C}=(\mathrm{Q} \times {\mathrm{R}}_{\mathrm{ct}}{)}^{1/\mathrm{n}}/{\mathrm{R}}_{\mathrm{ct}}$

were calculated. It can be verified that the value of R_ct is drastically reduced with the modification of the electrode with TiO₂ nanoparticles. This indicates that the nanoparticles enhance electron transfer at the electrochemical interface due to an electrocatalytic effect. Thus, the faradaic process involved in the determination of dithiophosphate is favored from the kinetic point of view, considering that the current intensity expresses the amount of charge per unit time [40,41]. This behavior can also be observed in the Nyquist plot generated in the EIS (Figure 8) since the diameter of the semicircle is equivalent to R_ct. It can be noted that the semicircle recorded for the TiO₂/CPE presents a smaller diameter than that recorded using the unmodified electrode.

From Table 1, it can also be noted that the value of n associated with the constant phase element in both electrodes is close to 1, indicating that it is a capacitive element [42]. Therefore, Q can be effectively associated with the capacitance of the electrical double layer. Table 1 shows other electrochemical parameters, such as the magnitude of the Warburg element and the capacitance calculated from Q_dl. In addition, the chi-square value validates the calculated parameters for both electrodes due to its low value, indicating low error in the fit.

3.4. Analytical Performance of the Sensor

3.4.1. Linear Range, LOD and LOQ

The quantitative response range is one of the most important analytical performance parameters for a chemical method. Pulse voltammetry techniques are known for their ability to discriminate the capacitive current, which significantly increases the electrochemical response at low concentrations. For this reason, calibration for DCDTP quantification was performed using the square wave voltammetry technique. First, a response optimization was performed by applying a one-factor-at-a-time methodology considering the pulse parameters. Figure 9a shows the peak current intensities by square wave voltammetry for 10, 25, and 50 Hz pulse frequency, keeping the potential step (1 mV) and amplitude (5 mV) constant. In this case, the highest peak current intensity was obtained for the 50 Hz frequency, which was used for the following tests. Figure 9b shows the peak current intensities for the 1, 2, and 3 mV potential steps, keeping the above frequency and the pulse amplitude at 5 mV. The highest current was obtained when 1 mV was used, so this value was selected. Figure 9c shows the test for 5, 10, and 25 mV pulse amplitude, where it can be observed that the best response was obtained for 25 mV. From these results, a frequency of 50 Hz, a potential step of 1 mV, and a pulse amplitude of 25 mV were selected as the pulse parameters for the calibration and electroanalytical determination of dithiophosphate by square wave voltammetry.

Figure 10 shows the square wave voltammograms in the range of 5 to 150 μmol/L, where the peak current increases with increasing concentration of DCDTP. The inset shows the calibration curve (I_p = 2.80 × [DCDTP] − 12.76), with an R² of 0.9939, which is acceptable for an electroanalytical method. The p-value of the ANOVA for this curve is 1.41 × 10⁻⁵, well below the 0.05 significance level, indicating that the calibration is statistically significant at 95% confidence.

The linear range of the TiO₂/CPE is wider than that achieved by other electrochemical methods for dithiophosphates using classical techniques such as polarography. For example, using differential pulse polarography, the quantification of diethyl dithiophosphate was achieved in a limited linear range from about 30 to 80 µmol/L at −400 mV, although the response range was extended to slightly more than 120 µmol/L in a nonlinear relationship; in any case, the response range is more limited than that obtained in our work. For the quantification of diphenyl dithiophosphate, the best linear relationship existed from 10 to 40 μmol/L at −850 mV, although the range may extend non-linearly to 100 μmol/L [8]. Even in the case of the most extensive quantitative range, it is shorter than that obtained by square wave voltammetry in this work.

The traditional method for the rapid and inexpensive determination of organic dithiophosphates is the UV-vis molecular absorption spectrophotometry. In this work, the quantification of DCDTP by direct UV-vis spectrophotometry could be achieved in the linear range of 50 μmol/L to 1 mmol/L, with the proposed sensor 10 times more sensitive for the detection and quantification of dithiophosphate.

The LOD and the LOQ are two of the most important performance parameters for a chemical analytical method. The former expresses the minimum amount of analyte that can be detected with 95% confidence, while the latter expresses the minimum amount that can be quantified. The LOD for the square wave voltammetry technique using the proposed sensor was 1.89 μmol/L, whereas the LOQ was 6.26 μmol/L. No results have been reported on the development of solid-state voltammetric sensors for dithiophosphates used in mineral processing and the environmental contamination resulting from this activity. Therefore, comparisons can only be made with polarographic quantifications, which are currently practically abandoned due to the toxicity of the mercury used in the drop electrodes. However, for the polarographic determination, only the LOD for the cathodic stripping technique has been reported, being 0.1 μmol/L [9]. This LOD is an order of magnitude lower than that obtained with the sensor by square wave voltammetry, and this is because stripping involves a preconcentration step by forming an Hg-dithiophosphate complex at the drop electrode at an imposed potential, which intentionally causes a significant decrease in the LOD. This type of preconcentration is not possible in the TiO₂/CPE; however, note that even without preconcentration, the LOD obtained by square wave voltammetry indicates the sensor’s high sensitivity to the detection of dithiophosphates.

Defining the common concentration for sulfhydryl collectors used in flotation processes is very difficult since in the mineral separation and concentration steps, reagents are dosed according to the amount of mineral in the pulp (g/ton). However, it is important to note that collectors are typically quantified in laboratory tests at concentrations in the order ranging from 10⁻² mol/L to 10⁻⁵ mol/L, with 10⁻⁴ mol/L being the most common [43,44,45]. The 10⁻² mol/L concentration is usually chosen to prepare stock solutions from which lower concentration solutions are obtained to dose the collector to the flotation pulp [44]. In the case of soil and wastewater samples, concentrations can vary widely.

3.4.2. Repeatability and Reproducibility

Table 2. Repeatability test for the quantification of DCDTP using the TiO₂/CPE.

Experiment	Ip (nA)	DCDTP (mmol/L)
1	276	0.095
2	310	0.107
3	281	0.097
4	304	0.105
5	278	0.096
6	279	0.096
7	306	0.105
8	279	0.096
9	331	0.114
10	296	0.102
Mean		0.101
SD		0.006
RSD (%)		6.08

shows the repeatability test of the DCDTP quantification using the sensor modified with TiO₂ nanoparticles. It can be observed that for 10 determinations under the same conditions, the concentration was always close to 0.10 or 0.11 mmol/L, which shows that the quantification using the sensor is repeatable. However, the statistical parameter that describes the repeatability is the RSD, and in this case, it can be observed that its value is 6.08%. According to the Association of Official Analytical Collaboration (AOAC) International, for a concentration range between 10 and 100 ppm, the RSD associated with the repeatability of the method should be between 7.3% and 5.3%, respectively [46]. The average concentration in the repeatability test is 0.101 mmol/L, which corresponds to 33.57 ppm (M = 332.4 g/mol), and it can be noted in Table 2 that the RSD is within the established acceptance range. This demonstrates that the voltammetric method using the TiO₂/CPE is repeatable in accordance with international analytical standards.

Intra-laboratory reproducibility is one of the most important analytical performance parameters since it describes long-term precision. Therefore, its evaluation allows to determine to what extent the results obtained are reproducible in days, months, and years. In this work, a reproducibility test was performed for a 5.4 mmol/L DCDTP solution using the same TiO₂/CPE for 30 days. Measurements at 1, 7, 14, 21, and 30 days resulted in mean peak currents (N = 5) of 27.54 μA (SD = 4.17 μA), 30.48 μA (SD = 2.25 μA), 28.76 μA (SD = 2.14 μA), 29.94 μA (SD = 2.29 μA), and 28.42 μA (SD = 2.88 μA), respectively. The ANOVA for multiple samples (groups) yielded an F-ratio of 0.86 and a p-value of 0.5027 > 0.05, indicating no statistically significant differences between the 5 days with 95% confidence, confirming the reproducibility of the sensor response.

3.4.3. Trueness

The trueness of the electroanalytical method for dithiophosphate using the sensor modified with TiO₂ nanoparticles was evaluated by comparison with another analytical method, in this case, UV-vis spectrophotometry with prior separation of suspended solids from the sample. Two samples were analyzed, one from a flotation experiment and the other from a liquid extract of soil synthetically contaminated with DCDTP.

Table 3. Determination of DCDTP in a flotation sample by UV spectrophotometry and voltammetry using the TiO₂/CPE.

	DCDTP (mmol/L)
Experiment	UV	Voltammetry
1	4.48	5.12
2	4.48	4.35
3	4.66	4.11
4	4.93	4.91
5	5.15	4.67
Mean	4.74	4.63
SD	0.29	0.41
t-statistic = 0.50     p-value = 0.6440

shows the results of five determinations for the flotation sample using both methods, which yielded very similar mean concentrations. A paired t-test for comparison of means yielded a p-value greater than the 0.05 significance level, indicating no statistically significant difference between the two results at 95% confidence.

On the other hand,

Table 4. Determination of DCDTP in a liquid extract of a synthetically contaminated soil sample by UV spectrophotometry and voltammetry using the TiO₂/CPE.

	DCDTP (mmol/L)
Experiment	UV	Voltammetry
1	2.99	2.51
2	3.15	1.61
3	2.82	1.46
4	2.79	3.19
5	2.80	2.47
Mean	2.91	2.25
SD	0.16	0.71
t-statistic =   1.86   p-value = 0.1362

shows the analysis of the soil extract contaminated with DCDTP. In this case, very similar mean concentration values were obtained. The paired t-test yielded a p-value greater than 0.05, indicating no statistically significant differences between the analytical results obtained by the two methods at 95% confidence. The results of the DCDTP analysis for both the flotation sample and the aqueous soil extract demonstrate that the modified sensor is reliable for determining dithiophosphate in both flotation process control and water and soil contamination assessment.

An important aspect to highlight is that the samples analyzed in both cases (Table 3 and Table 4) presented a high amount of solids in suspension, as can be appreciated in Figure 11. For the UV-vis molecular absorption spectrophotometric analysis, it was necessary to perform solid separation steps until the solution was completely transparent to the light beam. However, the voltammetric method using the TiO₂/CPE did not require these steps; that is, the analysis was performed directly on both samples with prior dilution. This means that the electrochemical method eliminates sample preparation steps, which reduces the cost of analysis by eliminating the need for pre-separation infrastructure. In addition, the analysis is much faster, allowing more analytical determinations to be performed in a given unit of time.

3.4.4. Analytical Interferences

Interferences are one of the main problems in analytical determinations since they force the establishment of procedures for the correction of the effect on the analytical signal of the target analyte(s). In this work, the interfering effect of two substances used as mineral flotation reagents, sometimes together with dithiophosphates, has been evaluated. Figure 12a shows the cyclic voltammograms for a solution containing 5.4 mmol/L DCDTP and two solutions containing, in addition to DCDTP, 1 mmol/L NaIPX, and 0.05 mL DADTC. Dithiocarbamate is shown as a volume because it is a liquid reagent. Xanthates are the most common collectors for sulfide group minerals, so their evaluation is particularly important because they are usually dosed with dithiophosphates. In this case, a peak appears close to 400 mV, which is related to the oxidation of the IPX⁻ species. The effect of NaIPX on the peak current of DCDTP oxidation can be seen more clearly in Figure 12b, where both peak currents are practically the same (without NaIPX: 40.78 μA, with NaIPX: 41.49 μA). In the case of the dithiocarbamate, it is observed in Figure 12a that a peak appears near 250 mV. In addition, the peak near 650 mV, which is that of DCDTP, increases significantly. Figure 12b shows that the peak current increases from 40.78 μA to 65.99 μA. Therefore, dithiocarbamate has an interfering effect on the peak current of dithiophosphate, resulting in an analytical result of DCDTP concentration far from the exact value. However, it is important to mention that it is most common to find the combination of xanthate + dithiophosphate in a mineral flotation scheme. In addition, if an interfering effect of dithiocarbamate is detected in a sample, calibration by the addition of a standard could be performed to correct the matrix effect.

4. Conclusions

The synthesis of TiO₂ nanoparticles by a green method using Aloe vera extract allows for obtaining a high-purity nanomaterial, mainly crystalline, with spherical morphology and an average particle diameter of 10.08 nm. The nanomaterial does not alter the surface morphology of a carbon paste electrode; however, it does change the surface chemical composition and generates nanometer active sites. Modification of this type of sensor with the nanoparticles enhances the voltammetric peak current for the oxidation of dithiophosphates. Above pH 8.0, the peak current for dithiophosphate oxidation on the modified electrode begins to decrease. The presence of TiO₂ nanoparticles obtained by the green method increases the electroactive area of the sensor with respect to the unmodified electrode. There is a mixed adsorption-diffusion control in the oxidation of dithiophosphates using the sensor. Charge transfer at the electrochemical interface is facilitated when nanoparticles are present on the carbon paste electrode. The sensor allows quantification in the linear range from 5 to 150 μmol/L by square wave voltammetry, with a detection limit of 1.89 μmol/L. The determination with the nanoparticle-modified sensor is repeatable and reproducible for at least 30 days. The sensor is reliable for use in mineral flotation samples and in contaminated soil extracts, where there is no need to separate suspended solids. Dithiocarbamates can interfere with the quantification of dithiophosphates using the modified sensor.