Electrochemical Oxidation of Glyphosate Using Graphite Rod Electrodes: Impact of Acetic Acid Pretreatment on Degradation Efficiency

Abstract

The uncontrolled use of herbicides such as glyphosate (GLY) (N-phosphonomethylglycine) in agricultural production has resulted in its presence in water bodies and in negative impacts on the environment and public health. On the frame of understanding the interaction between GLY and graphite rod surfaces, this contribution relies on the study of electrochemical responses of different GLY concentrations by cyclic voltammetry under both open and closed-circuit conditions. Furthermore, the effect of the electrodes’ electrochemical pretreatment with acetic acid on the double-layer capacitance and the subsequent surface functionalization of the graphite rod materials were evaluated. The increment in GLY concentration showed a decrease in the electrochemical oxidation response associated with the adsorption of the contaminant on the surface of the graphite rod electrode and the concomitant blockage of the active sites. Electrochemical pretreatment of the electrodes with acetic acid and GLY concentration play crucial roles in electric double-layer formation due to their ability to interact with both positive and negative electrical charges. By means of optical microscope observations and Fourier Transform Infrared Spectroscopy analysis, it was possible to detect the formation of oxygenated functional groups on the electrode surfaces after the electrochemical pretreatment. Through a 2³ factorial design analysis in repetition, the factors significant in the degradation of GLY were identified. The high degradation of GLY with the pretreated electrodes can be attributed to the preferential adsorption of the zwitterionic molecule at the interface, which allowed great direct oxidation of the contaminant on the anode’s surface.

1. Introduction

The use of herbicides has been of paramount importance in the efficient growth of plants cultivated for food production. Among these compounds, glyphosate (N-phosphonomethylglycine, GLY) is a broad-spectrum and non-selective herbicide synthetically produced to eliminate weeds, particularly perennials [1,2]. Its popularity is reflected by an estimated 126 × 10⁶ kg of GLY that was employed in 2014 in agricultural and non-agricultural applications [3]. In 2015, the International Agency for Research on Cancer classified GLY as probably carcinogenic to humans. Although scientific reports have associated this herbicide with effects such as alterations in the endocrine system or deoxyribonucleic acid damage [4,5], there is still controversy on the degree of its suggested toxicity [6]. Typically, the herbicide is dispersed by spraying it over the plants, and hence, while some GLY is metabolized, a relatively small fraction is eventually transported to the soil by the roots [7], where it is finally transferred to groundwater or surface water bodies. In this way, concentrations up to 430 µg L⁻¹ were determined in the USA [8], 36.7 µg L⁻¹ in Mexico [9], 105 µg L⁻¹ in Argentina [10], 2.77 mg L⁻¹ in Colombia [11], and 2.46 mg L⁻¹ in Portugal [12]. The main physicochemical properties of GLY are presented in Table 1.

Due to its controversial toxicity and its relatively long half-life in aquatic environments (2–91 d) [2], there are several reports that suggest a variety of physicochemical and biological techniques to remove GLY from aquatic matrices [2]. Among these, electrochemical advanced oxidation processes (EAOPs) stand out as a promising alternative since they are characterized by the electrochemical production of the highly reactive hydroxyl radical species (^•OH, E° = 2.8 V vs. SHE (standard hydrogen electrode)), a molecule that promotes the fast and efficient oxidation of a wide variety of contaminants in aqueous media [13,14]. Table 2 summarizes some research works that have reported the removal of GLY by EAOPs, and inspection of the relevant information reveals not only that GLY removal yields of higher than 70% can be obtained, but also that the anode material plays a pivotal role in the performance of the process. In this context, anode substrates are commonly classified on the basis of their overpotential of oxygen evolution and, therefore, are divided into two main groups: (i) active and (ii) non-active electrodes [15]. Active anodes are characterized by low oxygen evolution overpotentials (1.4–1.9 V vs. SHE), and typical examples in this category are electrodes made of mixed metal oxides (also known as dimensionally stable anodes (DSA^®)), RuO₂-TiO₂, IrO₂-Ta₂O₅, graphite, and Pt. On the other hand, non-active anodes show high oxygen evolution overpotential values, usually between 1.9 and 2.6 V vs. SHE. Electrodes made of PbO₂, SnO₂, and BDD are typical examples of non-active anodes that allow high-yield pollutant degradation efficiencies that result from the effective electrochemical production of ^•OH molecules that takes place when the electrode is polarized at large positive potential values. In these processes (electro-oxidation, EO), pollutant degradation takes place via (i) direct or (ii) indirect oxidation mechanisms [15,16,17]. The former occurs when the contaminant directly reacts on the anode’s surface or when it gets oxidized by physisorbed ^•OH molecules [13,18,19]. The indirect oxidation mechanism, on the other hand, takes place through the electrochemical generation of a mediator (such as active chlorine, persulfate, ozone, H₂O₂, among others), which reacts with the pollutant species in the bulk of the solution [20,21].

To promote the oxidation of the pollutants by electrogenerated ^•OH molecules, non-active anodes are highly effective but expensive, and therefore, their potential for large-scale applications is seriously limited [22]. The direct electron transfer upon pollutant adsorption on the anode’s surface, on the other hand, is another mechanism that also takes place in active anode materials (such as those made of carbon) when polarized below the oxygen evolution overpotential [17,23]. In this case, adsorption (Equation (1)) is usually the limiting slow step of the mechanism, followed by the charge transfer reaction (Equation (2)) and the desorption of the oxidized species (Equation (3), (in Equations (1)–(3), R is the pollutant and M is referred to as the anode surface) [24].

$$\mathrm{R}+\mathrm{M}\to \mathrm{M}-{\mathrm{R}}_{\mathrm{a}\mathrm{d}\mathrm{s}}$$

(1)

$${\mathrm{M}-\mathrm{R}}_{\mathrm{a}\mathrm{d}\mathrm{s}}+{z\mathrm{e}}^{-}\to {\mathrm{M}-\mathrm{R}}_{\mathrm{a}\mathrm{d}\mathrm{s},\mathrm{o}\mathrm{x}}$$

(2)

$${\mathrm{M}-\mathrm{R}}_{\mathrm{a}\mathrm{d}\mathrm{s},\mathrm{o}\mathrm{x}}\to {\mathrm{R}}_{\mathrm{o}\mathrm{x}}$$

(3)

Therefore, a strategy to improve the electrochemical properties of carbonaceous materials consists of chemical and electrochemical treatment with different compounds (such as HNO₃, H₂SO₄, citric acid, and oxalic acid, among others) so that adsorption properties are improved [25,26,27]. This treatment step, before electrochemical degradation, is sought to increase the active surface area of the electrode, thus improving the charge and discharge capacity of the substrate. Furthermore, it can also create new active sites on the electrode surface, which may promote a convenient interaction with electrolytes [28,29]. Due to its availability and affordability, electrodes made of carbon have been extensively used for EAOPs, mostly for the electrosynthesis of hydrogen peroxide (H₂O₂) via the 2 e⁻ oxygen reduction reaction (ORR). In addition, carbon is generally biocompatible, and its incorporation into the environment is not dangerous. Despite extensive research on the anodic oxidation of pollutants of emerging concern, a critical gap remains unaddressed on the surface-level interactions between carbon electrodes and GLY. This study aims to bridge this gap by systematically correlating faradaic and non-faradaic processes with GLY removal using graphite rod electrodes. A novel aspect of the present work is the comprehensive investigation of the impact of electrochemical surface pretreatment on both pollutant degradation and H₂O₂ production. To the best of our knowledge, no other work has explored the efficacy of acetic acid treatment on carbonaceous electrodes for GLY degradation, offering new insights into the performance of electrocatalytic processes.

Table 1. Physicochemical properties of GLY.

Chemical Structure
Molecular formula	C3H8NO5P
Molecular mass (g mol−1)	169.07
CAS number	1071-83-6
Solubility in water at 20 °C (g L−1)	1050
LogKow	−3.4
pKa1	2
pKa2	2.6
pKa3	5.6
pKa4	10.6
Henry’s Law Constant (atm m3 mol−1)	4.08 × 10−19

Table 1. Physicochemical properties of GLY.

Chemical Structure
Molecular formula	C3H8NO5P
Molecular mass (g mol−1)	169.07
CAS number	1071-83-6
Solubility in water at 20 °C (g L−1)	1050
LogKow	−3.4
pKa1	2
pKa2	2.6
pKa3	5.6
pKa4	10.6
Henry’s Law Constant (atm m3 mol−1)	4.08 × 10−19

Table 2. Literature reports about the removal of glyphosate by EAOPs.

Process	Experimental Conditions	Removal Efficiency (%)	Reference
Electrocoagulation	Al as anode and cathode, 7.5 mA cm−2, 2 L of 270 mg L−1 of commercial GLY, 116 mg L−1 NaCl, 1130 mg L−1 of Na2SO4, pH 5.5, flow 2 L min−1	>40% after 40 min of treatment	[30]
EO	BDD as anode and cathode, 20 mL of 100 mg L−1 GLY, 10 mA cm−2, pH 2	79% Total organic carbon (TOC) removal at 6.0 Ah L−1	[31]
EO	Ti/PbO2 anode, Ti cathode, 4.77 A, 1 L of 16.7 mg L−1 GLY, 0.01 M of Na2SO4	95% and 90% of GLY and TOC, respectively, after 173 min	[32]
EO	DSA® or Ti/Ru0.36Ti0.64O2 anodes, stainless steel cathode, 0.15 M of NaCl, 40 mA cm−2, pH 3	>90% mineralization of GLY after 3 h	[33]
EO	Pt-doped SnO2–Sb anode, stainless steel plate cathode, 200 mL of 200 mg L−1 GLY, 0.5 M H2SO4, flow 40.6 mL s−1, 1 A	62% of GLY degradation after 48 h of treatment	[34]
Electro-Fenton	Activated carbon fiber cathode, RuO2/Ti mesh anode, 0.1 M Na2SO4 electrolyte, pH 3, 1 mM Fe2+, 0.36 A	85% after 120 min	[35]
EO assisted by UV-C	DSA or BDD anode, stainless steel (AISI 304) cathode, 10 mA cm−2, NaCl of 2 g L−1, 1.5 L of 50 mg L−1 GLY, pH 4–5, 9 W UV-C lamp	Complete mineralization after 1 h of treatment	[36]
Photoelectrocatalytic process	TiO2 on BDD anode, Pt mesh cathode, 400 mL of 50 mg L−1 GLY, 3 mA cm−2, UV light, 0.05 M Na2SO4, pH 4	99.5% removal	[37]
Electro-Fenton, photoelectrooxidation, and photoelectro-Fenton processes	Ti/Ru0.3Ti0.7O2 anode, 0.05 M Na2SO4, pH 3, 100 mL of 100 mg L−1 GLY from Roundup®	70% GLY removal and >80% chemical oxygen demand and TOC removal using the electro-Fenton and photoelectro-Fenton processes	[38]

Table 2. Literature reports about the removal of glyphosate by EAOPs.

Process	Experimental Conditions	Removal Efficiency (%)	Reference
Electrocoagulation	Al as anode and cathode, 7.5 mA cm−2, 2 L of 270 mg L−1 of commercial GLY, 116 mg L−1 NaCl, 1130 mg L−1 of Na2SO4, pH 5.5, flow 2 L min−1	>40% after 40 min of treatment	[30]
EO	BDD as anode and cathode, 20 mL of 100 mg L−1 GLY, 10 mA cm−2, pH 2	79% Total organic carbon (TOC) removal at 6.0 Ah L−1	[31]
EO	Ti/PbO2 anode, Ti cathode, 4.77 A, 1 L of 16.7 mg L−1 GLY, 0.01 M of Na2SO4	95% and 90% of GLY and TOC, respectively, after 173 min	[32]
EO	DSA® or Ti/Ru0.36Ti0.64O2 anodes, stainless steel cathode, 0.15 M of NaCl, 40 mA cm−2, pH 3	>90% mineralization of GLY after 3 h	[33]
EO	Pt-doped SnO2–Sb anode, stainless steel plate cathode, 200 mL of 200 mg L−1 GLY, 0.5 M H2SO4, flow 40.6 mL s−1, 1 A	62% of GLY degradation after 48 h of treatment	[34]
Electro-Fenton	Activated carbon fiber cathode, RuO2/Ti mesh anode, 0.1 M Na2SO4 electrolyte, pH 3, 1 mM Fe2+, 0.36 A	85% after 120 min	[35]
EO assisted by UV-C	DSA or BDD anode, stainless steel (AISI 304) cathode, 10 mA cm−2, NaCl of 2 g L−1, 1.5 L of 50 mg L−1 GLY, pH 4–5, 9 W UV-C lamp	Complete mineralization after 1 h of treatment	[36]
Photoelectrocatalytic process	TiO2 on BDD anode, Pt mesh cathode, 400 mL of 50 mg L−1 GLY, 3 mA cm−2, UV light, 0.05 M Na2SO4, pH 4	99.5% removal	[37]
Electro-Fenton, photoelectrooxidation, and photoelectro-Fenton processes	Ti/Ru0.3Ti0.7O2 anode, 0.05 M Na2SO4, pH 3, 100 mL of 100 mg L−1 GLY from Roundup®	70% GLY removal and >80% chemical oxygen demand and TOC removal using the electro-Fenton and photoelectro-Fenton processes	[38]

2. Materials and Methods

2.1. Chemicals and Materials

Glacial acetic acid (CH₃COOH), sodium sulfate (Na₂SO₄), potassium phosphate monobasic (KH₂PO₄), and potassium phosphate dibasic (K₂HPO₄) reactive grade reagents were obtained from J.T. Baker (State of Mexico, Mexico) and used without further purification. Copper sulfate (CuSO₄, analytical grade) was purchased from FAGAlab (Sinaloa, Mexico) and TiOSO₄ was supplied by Sigma-Aldrich (Steinheim, Germany). Commercial formulation of GLY was obtained from Grupo Rivas, Nivela brand (74%, Guanajuato, Mexico). Graphite rods were supplied by Grupo Rooe (State of Mexico, México). Deionized water was obtained from the Millipore Direc-Q water purifier (resistivity ≥ 18.2 MΩ cm) purchased from Merck Millipore (Darmstadt, Germany) and used to prepare all solutions.

2.2. Experimental Procedures

2.2.1. Determination of the Electrochemically Active Surface Area of the Graphite Rod Electrode

The electrochemically active surface area (ECSA) was determined by imposing an electrode potential in the region where no redox processes occur (only the double-layer charging process takes place) [39]. To compute the capacitance, cyclic voltammograms were recorded at sweep rates ranging from 10 to 110 mV s⁻¹, in a potential range of ±10 mV relative to the open-circuit potential (OCP). By calculating the capacitive current (I_cap) using Equation (4) (where I_an and I_cat correspond to the anodic and cathodic currents of each voltammogram at the OCP) and by plotting I_cap against the scan rate (v_s), a straight line is obtained. The corresponding slope is equal to the double layer capacitance (C_DL) as expressed by Equation (5):

$$I_{\mathrm{c}\mathrm{a}\mathrm{p}}={\displaystyle \frac{I_{an}+I_{cat}}{2}}$$

(4)

$$I_{\mathrm{c}\mathrm{a}\mathrm{p}}=\left(C_{DL}\right)\left(v_{\mathrm{s}}\right)$$

(5)

To estimate the ECSA, the C_DL of the graphite rod electrode was divided by the specific capacitance (C_s) of graphite (3 µF cm⁻²) [40] (see Equation (6)).

$$ECSA={\displaystyle \frac{C_{DL}}{C_s}}$$

(6)

In these experiments, a three-electrode electrochemical cell was utilized along with a Pt wire and a Ag|AgCl electrode that were used as counter electrode and reference electrode, respectively. A graphite rod with a geometric area of 2.54 cm² served as a working electrode. The three electrodes were connected to an Epsilon potentiostat (BASi, model SP-300, West Lafayette, IN, USA). A 0.5 M Na₂SO₄ solution was employed as a supporting electrolyte, and to prevent any interference with the 2 e⁻ ORR reaction, the electrolyte was purged with N₂ gas for 10 min prior to the tests. Additionally, before the measurements, the working electrode was allowed to remain in contact with the electrolyte for 30 min to stabilize the OCP.

2.2.2. Effect of GLY Concentration: Open-Circuit Adsorption Experiments

To evaluate the effect of GLY concentration on the electrochemical response of the graphite rod electrode, open-circuit adsorption experiments were performed using the cyclic voltammetry technique. To that end, the graphite rod electrode was subjected to an open-circuit adsorption process for 30 min in a 0.5 M Na₂SO₄ solution containing different concentrations of GLY (from 5 to 100 mg L⁻¹). Subsequently, the electrode was removed from the solution and one voltammetric cycle was performed at 100 mV s⁻¹ in a solution containing only the supporting electrolyte. Cyclic voltammograms were then recorded over a potential range between −1.7 and 1.7 V vs. Ag|AgCl. In order to understand the influence of the acetic acid electrochemical pretreatment of the electrodes on the capacitance of the working electrodes, measurements of the C_DL of the substrates were performed in the absence and in the presence of different GLY concentrations. As it was previously noted for capacitance measurement experiments, cyclic voltammetry responses in the presence of 5, 50, and 100 mg L⁻¹ of GLY were recorded over a potential range of ±10 mV relative to the OCP in the 2 to 110 mV s⁻¹ scan rate range.

2.3. Experimental Setup for GLY Degradation

An undivided electrochemical cell (100 mL of effective volume) containing two graphite rod electrodes was used for the electrolysis assays. The geometric area for both electrodes was 6.29 cm², with an inter-electrode spacing of 10 mm (Figure 1). The electrodes were polarized using a Wanptek DPS 3010U power supply (30 V, 10 A, Shenzhen, China) working in a constant current mode. A magnetic stirrer was used to agitate the solutions. The anodic (E_an) and cathodic (E_cat) potentials were determined using a reference electrode (Ag|AgCl (KCl 3 M)). After each experiment, the electrodes were sanded and washed with deionized water to remove any adsorbed material.

2.4. Effect of Operational Parameters and Electrode Pretreatment on Glyphosate Degradation

In the traditional one-factor-at-a-time approach, the effect of the variables is investigated independently, varying in a controlled way one parameter while keeping the others at constant values. However, factorial design (FD) methodologies may be used to evaluate variable response surfaces and the interactions between variables that define the optimal operational conditions [41]. From the FD, it is therefore possible to determine the contribution of each factor and their interactions. Since only two levels are investigated in the 2^k FD system, the experimental response is linear over the studied interval [42]. To determine the effect of the operational parameters—current density (factor A), electrolysis time (factor B), and electrochemical pretreatment of the electrodes (factor C)—on GLY degradation by the EO process, a full 2³ FD system (in duplicate (16 randomized experiments)) was performed. The FD was used to create an empirical model for the response, as shown by Equation (7), in which Y_j is the response variable, X_i is the independent variable, and β and β_i stand for the linear model coefficients and the constant, respectively [43,44].

$${\mathrm{Y}}_{\mathrm{j}}={\mathsf{\beta }}_0+\sum {\mathsf{\beta }}_{\mathrm{i}}{\mathrm{X}}_{\mathrm{i}}$$

(7)

Galvanostatic EO assays were conducted at current density values (A) of 5 and 15 mA cm⁻², electrolysis time (B) of 20 and 60 min, and electrochemical pretreatment of the electrodes (C), at constant stirring (coded factors are shown in Table S1 in Supplementary Materials). The electrochemical pretreatment of the electrodes consisted of imposing a current density of 5 mA cm⁻² for 10 min in a 0.05 M acetic acid solution. After that, the electrodes were washed with deionized water and sonicated for 10 min to remove the excess acetic acid on the surface. Subsequently, the electrodes were placed in the electrochemical cell. Analysis of variance (ANOVA) of the experimental data was carried out using the Design-Expert^® v11 software (Stat-Ease Inc., Minneapolis, MN, USA). Before each galvanostatic experiment, the electrodes were sanded, sonicated for 10 min, and rinsed with deionized water to remove deposits from the surface. The specific energy consumption in kWh g⁻¹ GLY (SEC) and the operating cost (OC, expressed as USD m⁻³) were calculated using Equations (8) and (9) [45].

$$\mathrm{S}\mathrm{E}\mathrm{C}={\displaystyle \frac{EI{t}_e}{Vol{R}_{em}{C}_0}}$$

(8)

$$\mathrm{O}\mathrm{C}=\mathrm{a}{\displaystyle \frac{EI{t}_e}{Vol}}\times 1000$$

(9)

where t_e is the treatment time (h), Vol is the volume of the electrolytic solution (L), R_em is the removal percentage of GLY (decimal), C₀ is the initial concentration of GLY (mg L⁻¹), and a is the cost of electric energy, considering a value of 0.091 USD kWh⁻¹ according to the medium voltage high-demand tariff in México.

2.5. Analytical Methods

The electrochemically produced H₂O₂ concentration was determined by assessing the absorption of the titanium (IV) oxysulfate (TiOSO₄) complex at 406 nm [46]. A Hach UV-Vis spectrophotometer (DR-6000 model) was used for the analysis. The morphology of the graphite rod surfaces, on the other hand, was observed using an optical digital microscope (Keyence VHX-5000 digital microscope). In addition, Fourier transform infrared (FTIR) spectra of the surfaces were recorded using a Perkin Elmer spectroscopy apparatus (model Spectrum 100) with an attenuated total reflection module. GLY concentration was indirectly measured using the anodic stripping voltammetry method reported by Pintado et al., 2012. Due to the interaction of GLY with Cu ions, an oxidation peak current increase (reflecting the Cu(I) to Cu(II) transformation) is proportional to the GLY concentration that forms the GLY-Cu complex [47,48]. A three-electrode electrochemical cell was employed for this analysis. The cell was equipped with a glassy carbon, a Ag|AgCl, and a Pt wire as working, reference, and counter electrodes, respectively. The electrochemical cell was filled with 50 mL of a 0.1 M phosphate buffer, also containing 0.4 mM of a Cu²⁺ solution at pH 7.2. The Cu(II) pre-concentration step was performed at a constant potential of −0.8 V vs. Ag|AgCl during 180 s under constant stirring. After 30 s of resting, linear voltammetry scanning from −0.3 to 0.8 V vs. Ag|AgCl was carried out. While Figure S1 shows the copper electrochemical response with and without glyphosate, Figure S2 depicts linear voltammograms at different glyphosate concentrations and the resulting linear calibration curve, respectively.

3. Results and Discussion

3.1. Electrochemically Active Surface Area of the Graphite Rod Electrode

Figure 2 depicts the cyclic voltammograms and linear regression used to calculate the C_DL of the graphite rod electrode. Due to the double layer of the electrode–solution interface becoming charged, a higher I_cap was obtained as the scan rate was increased (see Figure 2a). In this experiment, the OCP corresponded to 0.105 V vs. Ag|AgCl.

Following Equation (5), a C_DL with a value of 14.7 µF was obtained from the straight line in the I_cap vs. scan rate plot displayed in Figure 2b. The ECSA that was determined using Equation (6) was 4.89 ± 0.1 cm², which is characterized by a roughness factor of 1.89 (which is the ratio of electrochemically active surface area to geometric surface area (2.54 cm²)).

3.2. Effect of Glyphosate Concentration: Open-Circuit Adsorption and Double-Layer Capacitance Measurements

The cyclic voltammograms recorded in the open-circuit adsorption experiments are shown in Figure 3a. As can be observed from inspection of the corresponding data, a GLY concentration of 5 mg L⁻¹ is reflected by an irreversible oxidation peak at a potential near 1 V vs. Ag|AgCl, which is attributed to the GLY oxidation process. Oliveira et al., 2018 [49] evaluated the electrochemical response of GLY in a carbon paste electrode at a concentration of 1000 µM. These authors reported an anodic peak current at 0.95 V vs. Ag|AgCl, which may be related to the EO of the isopropylamine group of the GLY molecule [50,51]. On the other hand, at concentrations higher than 5 mg L⁻¹ of GLY, no oxidation peak was identified. This behavior can be attributed to the adsorption of GLY to the graphite rod surface during the oxidation process, which results in the effective blocking of the active sites of the electrode and the consequent inhibition of the charge transfer reaction. The performance of porous carbon electrodes is greatly dependent on several factors, including the material properties (porosity, surface chemistry, electrical conductivity, and preparation method), as well as the electrolyte characteristics (ion dimensions, dielectric constants, etc.) [52]. The electrical double layer region is that portion of the electrochemical interface where the charge distribution is different from most of the ionic conductor and consists of an electrolyte layer of a specific thickness that is in contact with the electrode surface. Both the structure of the electrochemical interface and, to some extent, the kinetics of faradaic events, are determined in this region by long- and short-range electrostatic forces [53]. In this context, the effect of GLY adsorption on its oxidation at various concentrations at the electrode–electrolyte interface, as well as the role of the acetic acid electrochemical pretreatment on the functionalization of the graphite surface and its interaction with GLY could both be elucidated by C_DL measurements.

Figure 3b shows the C_DL of the graphite surface in the presence of 0, 5, 50, and 100 mg L⁻¹ of GLY. As can be seen from the corresponding captions, these measurements were conducted with electrodes previously treated electrochemically with acetic acid (red line) and raw electrode substrates without treatment (black line). As previously pointed out, the C_DL of the electrode without treatment was characterized by a value of about 14.5 µF (OCP = 0.105 V vs. Ag|AgCl). In the presence of 5 mg L⁻¹ of GLY, the C_DL increased to 89 µF (OCP = 0.120 V vs. Ag|AgCl). GLY contains three functional groups in its structure (amino, carboxylate, and phosphonate), which makes it a zwitterionic molecule [54,55]. Zwitterions are molecules with both cationic and anionic groups that result in a zero net charge [56,57]. This particular feature plays a crucial role in electrical double-layer formation and stability due to their ability to interact with both positive and negative electrical charges [58]. According to Ridwan et al., 2022 [58], adsorbed zwitterions at electrified interfaces in dilute solutions do not electrostatically screen interfaces. In this regard, the presence of GLY at low concentration values increases the number of dipoles at the interface, which increases C_DL.

It is interesting to note that the C_DL of the graphite electrode without electrochemical treatment decreases linearly with the concentration of GLY between 50 and 100 mg L⁻¹. While at the first concentration value, C_DL decreases to 61.1 µF (OCP = 0.132 V vs. Ag|AgCl), at the highest concentration of GLY, the capacitance is reduced to 39.8 µF (OCP = 0.142 V vs. Ag|AgCl). It can, therefore, be inferred that at high concentrations, the GLY dipoles cancel each other, decreasing C_DL. A similar trend was reported by Zouaoui et al., 2022 [59], for the electroanalysis of GLY in water using a chitosan-imprinted graft on 4-aminophenylacetic acid microsensor by means of electrochemical impedance spectroscopy responses. From this study, it was pointed out that the capacitance of the electrode decreases as the concentration of GLY increases; an observation that is attributed to the discharging of charged imprinted sites when GLY molecule interacts through electrostatic forces. In another study, Tolman et al., 2019 [60] evaluated the effect of physical adsorption of organic compounds on the capacitance of activated carbon electrodes. More than 70% of the capacitance of the activated carbon was lost due to the adsorption of toluene or chloroform vapor, and there was a 20–30% loss of capacitance due to the adsorption of organics that are miscible with the aqueous electrolyte. The blockage of the carbonaceous material’s surface sites was the reason given for this behavior. These results highlight the complex interplay that exists between C_DL, the electrode surface chemical composition, and the organic nature and concentrations of chemical species in solution for charge transfer process applications.

On the other hand, the graphite rod electrode electrochemically pretreated with acetic acid showed much higher capacitance values (in the range of 6000 to 9200 µF) when compared with the value of the carbonaceous surface (red line in Figure 3b). As can be seen in the figure, the electrochemically pretreated graphite electrode in the absence of GLY showed a C_DL of 9189 µF (OCP = 0.528 V vs. Ag|AgCl). Following the previously observed trend, C_DL decreased when GLY concentration increased, reaching values on average of 8215 µF (OCP = 0.547 V vs. Ag|AgCl), 6690 µF (OCP = 0.550 V vs. Ag|AgCl), and 6394 µF (OCP = 0.480 V vs. Ag|AgCl) for 5, 50, and 100 mg L⁻¹ of GLY, respectively. These results may be related to the increase in the porosity of the treated electrode, as well as to the formation of functional groups on the surface that probably have a positive contribution to the electrode capacitance [52]. The specific capacitance enhancement of carbon materials can be achieved by electrochemical polarization, e.g., using a solution of HNO₃ or H₂SO₄ significantly enriches the surface functionality and, in some cases, enhances the surface area [61]. Surface functional groups could, therefore, modify the potential distribution across the carbon–electrolyte interface. Accordingly, the potential at the electrical double layer’s outer Helmholtz plane and the potential of zero charge may be affected by the oxygen-containing groups [62]. Moreover, the surface carbon can be modified from hydrophobic to hydrophilic as a consequence of oxidation processes. As a result, depending on their oxygen content, these surfaces may exhibit selective adsorption properties [62]. Bleda-Martínez et al., 2005, explored the role of surface chemistry on the C_DL of various carbon materials under different physical (gasification with CO₂ and steam) and chemical (KOH and NaOH) activation methods [52]. These authors demonstrated that the capacitance of these materials is closely related to the porosity of the surface, as well as to the content of oxygenated groups of the CO-type. In addition, the authors suggest that the conductivity of the material positively contributes to the capacitance.

3.3. Superficial Analysis of the Electrodes

At this point, to further understand the role of electrochemical pretreatment of the electrodes with acetic acid, a superficial analysis of the materials under study was carried out. Figure 4 shows digital optical microscopy images of 1000× magnification of graphite rod electrode surfaces before and after acetic acid electrochemical pretreatment.

Figure 4a shows a graphite rod electrode surface with slight porosities and deformations. Nevertheless, the electrode that performed as an anode during the electrochemical pretreatment with acetic acid, shown in Figure 4b, presented more porosities compared with the previous one (observed in the dark regions). This difference may be related to corrosion, which should also be related to an increment in the surface area. This, in turn, can explain the higher capacitance of the electrode after its treatment. In the same line of thought, the cathode substrate after the electrochemical pretreatment (Figure 4c) shows low pore formation. On the other hand, the chemical composition of the surface could be another important factor in the performance of the electrochemical process. In this sense, functional groups containing opposed electrical charges on the carbon surface may interact with zwitterionic molecules in a particularly strong fashion, leading to selective adsorption of these molecules [63]. In order to investigate this possibility, the chemical structures on the surface of the graphite rod electrodes that were produced during the electrochemical pretreatment were studied by FTIR analysis (Figure 5).

Figure 5, for instance, does not show peaks related to functional groups located on the surface of a raw graphite rod electrode on which there was no previous acetic acid treatment. In contrast, some absorption bands corresponding to oxygenated functional groups could be observed on both the anode and cathode after the electrochemical pretreatment was carried out. In this regard, the first band at 1075 cm⁻¹ (marked in the gray column) can be attributed to the stretching of the C-O bond (between 1050 and 1150 cm⁻¹) that belongs to the ether functional group. The second band (marked in blue) can be observed around 1540 cm⁻¹ and is attributed to the stretching vibration of the C=C group, which is derived from quinones and ketones [64]. Overlapping bands can be observed in the 1750–1550 cm⁻¹ region (marked in green), which are associated with carbonyl (C=O, carboxylic acid, etc.) and aromatic carbon stretching [65]. A broadband is also observed in the 2100 and 2200 cm⁻¹ region (marked in violet), which is associated with alkyne functional groups [66]. FTIR spectra clearly confirmed the formation of oxygenated functional groups on the graphite electrodes by the action of electrochemical pretreatment with acetic acid. Therefore, it can be inferred that the presence of these groups on the surface contributes to the increase in the capacitance of the graphite rod electrode.

3.4. Effect of Electrochemical Pretreatment with Acetic Acid on GLY Oxidation and H₂O₂ Electrogeneration

The electrochemical response of graphite rod electrodes with and without acetic acid pretreatment in the presence of 5 mg L⁻¹ of GLY was evaluated to determine its effect on the oxidation of the pollutant. The responses of the graphite rod electrodes obtained from cyclic voltammetry at a scan rate of 10 mV s⁻¹ are shown in Figure 6.

As can be noted from the voltammetric responses in Figure 6, there is a clear anodic peak associated with the direct oxidation of GLY on the electrode surfaces of both rod electrodes. Notably, a higher current density is recorded for the signal corresponding to the electrode treated with acetic acid, which could be an indication that the oxidation of the pollutant is carried out more efficiently due to the increased charge transfer that is taking place, compared with the rod electrode without the pretreatment. This behavior can be related to the presence of oxygenated functional groups on the surface of the graphite rod (as shown in the preceding section), which increases the double layer capacitance by enhancing ion adsorption, thus increasing the charge storage in the electrode. In turn, higher capacitance allows for higher current density, as the electrode can handle a greater amount of charge per unit area, thus improving its electrochemical performance. Inspection of the corresponding data for reduction processes shows that a cathodic signal on the previously pretreated rod electrode appears at around −0.8 V vs. Ag|AgCl, which can be associated with the 2 e⁻ ORR signal for the generation of H₂O₂ (see Equation (10), E° = 0.695 V vs. SHE) [67]:

$${\mathrm{O}}_2+{2\mathrm{H}}^{+}+2{\mathrm{e}}^{-}\to {\mathrm{H}}_2{\mathrm{O}}_2$$

(10)

Carbonaceous materials effectively reduce oxygen to H₂O₂ [68], and among these, carbon nanostructures [69,70], carbon felts [71,72], and carbon cloths have been successfully applied and studied. In addition, they can also be embedded with some metals for the electroanalytical detection of H₂O₂ through its reduction in the presence of Fe²⁺ (Fenton reaction) [73,74].

In order to confirm the 2 e⁻ ORR on the graphite rod electrodes, the concentration of the electrogenerated H₂O₂ was quantified. Figure 7 shows the cumulative H₂O₂ concentration using the raw graphite rod electrode and the graphite rod electrode after acetic acid electrochemical pretreatment in the undivided cell at different cell potential (E_c) values. As shown in Figure 7a, by increasing E_c from 1.5 V to 2.5 V with the raw graphite rod electrode, the production rate rose up to a linear accumulation of H₂O₂, which reached concentrations of up to 11.2 mg L⁻¹ over the 60 min of electrolysis. Nevertheless, the production rate declined when a 3 V potential was applied, with an average concentration of 0.94 mg L⁻¹ after the 60 min test. This may be associated with the 4 e⁻ oxygen reduction reaction to water at largely negative cathodic potentials (E° = 1.229 V vs. SHE) (see Equation (11)) [68]:

$${\mathrm{O}}_2+{4\mathrm{H}}^{+}+4{\mathrm{e}}^{-}\to {\mathrm{H}}_2\mathrm{O}$$

(11)

Inspection of Figure 7b shows that the graphite electrode that was previously treated with acetic acid showed an effective H₂O₂ production rate, with E_c values as high as 3 V at an average H₂O₂ concentration of 10.8 mg L⁻¹. These results indicate that the graphite cathode electrochemically pretreated with acetic acid is characterized by stability since H₂O₂ production did not decrease when the potential was increased, as in the case of the raw graphite rod. The higher stability of this electrode can be correlated with its surface chemical and physical properties. In this regard, the C=O and C-O functional groups, as observed in the FTIR spectrum in Section 3.3, have a great influence on the 2 e⁻ ORR for H₂O₂ electrosynthesis, as reported by Chu et al., 2023 [75]. Based on electrochemical tests and density functional theory calculations, it was revealed that the C=O group is the most active component in H₂O₂ electrosynthesis, followed by the -COOH group.

3.5. Factorial Design: GLY Electrochemical Oxidation Assays

The operational parameters of the 2³ FD, the calculated SEC, and the experimental and predicted responses are shown in

Table 3. Summary of the experimental and predicted degradation of GLY and SEC calculations.

Exp.	Operational Parameters
	A	B	C	GLY Degradation (%)			SEC	OC
	Current Density (mA cm−2)	Electrolysis Time (min)	Electrochemical Pretreatment	Actual Response 1	Actual Response 2	Predicted Response	(kWh g−1 GLY)	(USD m−3)
1	5	20	No	17	13	16	1.25 × 10−2	0.047
2	15	20	No	16	17	15	5.43 × 10−2	0.181
3	5	60	No	37	28	31	1.83 × 10−2	0.116
4	15	60	No	18	13	17	1.10 × 10−1	0.549
5	5	20	Yes	14	19	15	1.03 × 10−3	0.035
6	15	20	Yes	14	19	18	5.31 × 10−2	0.141
7	5	60	Yes	50	39	46	1.20 × 10−2	0.104
8	15	60	Yes	34	37	34	6.48 × 10−2	0.426

. The ANOVA results are presented in

Table 4. Analysis of variance (ANOVA) for the removal of GLY.

Source	Sum of Squares	Degree of Freedom	Mean Square	F-Value	p-Value
Model	1870.88	6.00	311.81	16.12	0.0002
A	150.06	1.00	150.06	7.76	0.0212
B	1008.06	1.00	1008.06	52.12	<0.0001
C	280.56	1.00	280.56	14.51	0.0042
AB	189.06	1.00	189.06	9.78	0.0122
AC	10.56	1.00	10.56	0.55	0.4787
BC	232.56	1.00	232.56	12.02	0.0071
Residual	174.06	9.00	19.34
Cor Total	2044.94	15	R2 =	0.9149
			R2 adjusted =	0.8581

, and the Pareto plot for GLY degradation is presented in Figure S3.

According to the results in Table 3, the pretreated electrodes enabled greater degradation of GLY compared with the electrodes without the pretreatment, showing a reasonable improvement when lower current density (5 mA cm⁻²) was applied, reaching efficiencies of 39% and 50% for the best conditions. The obtained degradation yields are slightly lower than those already reported in the literature using EAOPs (Table 2), which may be due to the short treatment time evaluated in the present work and the use of carbonaceous material as anodes. In addition, SEC and OC were also lower with the pretreated electrodes; for instance, values of 1.10 × 10⁻¹ and 6.48 × 10⁻² kWh g⁻¹ GLY were obtained for the raw graphite and pretreated graphite electrodes, respectively, at 60 min and 15 mA cm⁻².

ANOVA provides a way to justify the significance and adequacy of the statistical model [41]. Thus, the results of the ANOVA (Table 4) and the Pareto plot (Figure S3) indicate that the model’s F-value of 16.12 implies that it is significant. In this regard, there is only a 0.02% chance that an F-value this large could occur due to noise. Also, p-values less than 0.05 indicate the significant model terms; in this way, the main factors, A, B, and C, and the AB and BC interactions had significant effects. The most statistically significant factors influencing GLY degradation (above the Bonferroni limit in the Pareto plot) were the main effects of B and C, followed by BC, AB, and A (between the Bonferroni and t-value limits of the Pareto plot). As expected, the results showed that electrolysis time is an important factor in the degradation process as it defines the treatment period [14]. Moreover, surface modification by the electrochemical treatment has a great influence on the degradation of GLY, suggesting that oxygen-containing functional groups during electrolysis in the acetic acid medium play an important role in the degradation mechanisms of the contaminant.

An indicator of the concordance of the data experimentally obtained with the ones predicted by the model is the determination coefficient (R²). Values for this coefficient higher than 0.80 are typically regarded as appropriate for a good fit [76]. Hence, the statistical model offers a reliable prediction for GLY degradation with graphite rod electrodes within the suggested experimental range, as evidenced by the R² and the adjusted determination coefficient (R² adj), with values of 0.9149 and 0.8581, respectively.

On the other hand, the empirical correlation between the GLY degradation percentage and the process factors is stated by the following polynomial expression (Equation (12)):

$${\mathrm{Y}}_{\mathrm{G}\mathrm{L}\mathrm{Y}}=24.06-3.06 \mathrm{A}+7.94 \mathrm{B}+4.19 \mathrm{C}-3.44 \mathrm{A}\mathrm{B}+0.8125 \mathrm{A}\mathrm{C}+3.81 \mathrm{C}\mathrm{D}$$

(12)

Here, the independent coefficient indicates the average degradation of GLY after the 16 assays (24.06%). From this equation, the predicted degradation of GLY was obtained, as shown in Table 3. As can be seen, the values predicted by the statistical model fit the experimental data adequately, properly explaining the effect of current density, electrolysis time, and electrochemical pretreatment on GLY degradation. The 3D surface response plots for the degradation of GLY graphite rod electrodes are shown in Figure 8.

According to Figure 8a, the highest GLY degradation with raw graphite rod electrodes was obtained at 60 min of electrolysis and when applying the lowest current density (5 mA cm⁻²), reaching an average degradation of 33%. In other words, GLY decreased from an average concentration of 216 mg L⁻¹ to 146.5 mg L⁻¹. With similar behavior in Figure 8b, the degradation of GLY with the pretreated graphite electrodes showed the highest degradation under these same conditions; however, an efficiency of 45% was obtained. Therefore, the concentration of GLY was reduced to 117.5 mg L⁻¹.

In view of the results obtained, it can be considered that there is an important effect of pretreatment of graphite rod electrodes with acetic acid that can be attributed to the surface changes that occur on the electrodes during the procedure, which influence the physical and chemical properties of the electrodes, including porosity, surface area, and double-layer capacitance. This is closely related to the formation of oxygenated functional groups and, consequently, has a positive effect on the adsorption and direct oxidation of the pollutant under study.

4. Conclusions

This investigation explored the application of graphite rod electrodes in the EO of GLY, aiming to elucidate the effect of electrochemical pretreatment of the electrodes with acetic acid on the degradation of the pollutant. The results demonstrate the significant improvement in graphite electrodes for GLY degradation, which is attributed to its electrochemical pretreatment. When dissolved zwitterionic molecules, such as GLY, are placed in the electrode–solution interface, functional groups with opposite charges can interact with the polarized surface, resulting in preferential adsorption of the zwitterionic molecule at the interface. Moreover, this selective adsorption can considerably modify the properties of the electrical double layer, such as its thickness, structure, and stability, which explains the variation in C_DL as a function of GLY concentration. Additionally, the graphite rod cathode also presented a surface functionalization that impacted the stability of the electrochemical production of H₂O₂, promoting the indirect oxidation pathway of the pollutant in the electrochemical assays.

The EO of GLY using carbonaceous rod electrodes represents a promising approach for the efficient degradation of this widely used herbicide. Understanding the interaction of the pollutant with the electrodes, optimizing process parameters, and evaluating environmental implications are crucial for the development and implementation of sustainable remediation strategies.