Efficient Removal of Tartrazine Yellow Azo Dye by Electrocoagulation Using Aluminium Electrodes: An Optimization Study by Response Surface Methodology

Abstract

This study investigates the efficiency of electrocoagulation (EC) in removing Tartrazine Yellow (TY) azo dye from synthetic wastewater using aluminium electrodes. The effects of current density, i (0.008–0.024 A cm⁻²), initial solution pH (3.0–7.0), and treatment time, t (10–50 min) on key process parameters, including pH, temperature (T), TY dye concentration (c) and removal efficiency (R), anode consumption, and sludge characterisation were studied. The experiments were conducted in a batch reactor according to the experimental plan developed in Design-Expert software, which was also used for the evaluation of the obtained results. As the EC process progresses, the removal efficiency of the TY dye increases, while the removal dynamics and the final value of R (ranging from about 28% to 99%) depend on the experimental conditions (i, initial pH, and t). A high R-value is reached faster with the application of higher current densities and lower initial pH. This is associated with a higher proportion of carbon and sulphur in the sludge (from the TY dye) after the EC process. Additionally, a mathematical model was developed to predict the experimental data. A numerical optimisation method using response surface methodology (RSM) was applied to determine the optimal operating conditions for TY dye removal. This resulted in the following conditions: pH = 3.37, t = 18.74 min, and i = 0.016 A cm⁻², achieving a removal efficiency of ≈70%.

1. Introduction

Azo dyes represent the largest and most versatile class of dyes, accounting for over 50% of global dye production [1,2,3,4,5,6,7,8,9]. About 70% of dyes used in industrial applications belong to this category. With a staggering variety of over 2000 distinct types, azo dyes are a staple in colour manufacturing, boasting an impressive global production volume that surpasses 700,000 tons annually [1]. Azo dyes are widely used in various industries, including textiles, leather, chemicals, food, pharmaceuticals, cosmetics, etc., due to their strong colour intensity, chemical stability, thermal stability and cost-effectiveness [3,4,5,6,7,8,9,10,11,12,13,14]; one of these is Tartrazine Yellow (TY) dye [4,5]. TY, along with erythrosine (E127) is one of the most commonly used synthetic dyes in the food industry, used in the production of soft drinks, cakes, sauces, yoghurts, chips, candies, etc. It is also used in the production of non-food consumables, like soaps, cosmetics, and shampoos, in the pharmaceutical industry, for colouring medicinal capsules and prescription drugs, and in the textile industry [15,16,17].

However, the presence of azo dyes in industrial effluents poses significant environmental concerns due to their resistance to biodegradation, leading to their accumulation in natural water bodies. Their release into the environment causes water coloration, reduced light penetration, and the disruption of aquatic ecosystems [18]. Additionally, under certain conditions, azo dyes can degrade into aromatic amines, which are associated with mutagenic and carcinogenic effects [19,20,21,22,23]. Aquatic organisms are particularly sensitive to dye effluents. The presence of these chemicals affects their development and behaviour and induces oxidative stress. The negative influence of TY on aquatic organisms was shown in the study by Yu and associates [24], who found that TY exposure significantly reduces embryo fertilisation and hatching rates, inflicts histopathological damage to tissue, triggers oxidative stress, disrupts innate immune function, and causes dysbiosis of gut microbiota in juvenile crucian carp. Investigations of the ecotoxicity of TY have shown high embryo malformation and larval mortality rates in aquatic biota, polluted in concentrations considered to be safe for humans [17]. The acceptable daily intake of TY for humans is 7.5 mg/kg body weight per day, but investigations have shown adverse effects, like nasal congestion, rhinorrhea, wheezing, itchy skin, and urticaria, for people with higher sensitivity to this dye and those with aspirin intolerance, estimated at 3% of population [24,25,26,27]. Health hazards, including allergic reactions, negative behavioural effects in children, and genotoxic potential, make its removal a priority objective in wastewater treatment [28,29,30,31].

Given their persistence and negative effects, developing effective remediation techniques is essential. Various physical, chemical, biological, and electrochemical methods have been explored for azo dye removal from wastewater. Advanced oxidation processes (AOPs), including Fenton oxidation, ozonation, and photocatalysis, provide efficient dye degradation but often require high-energy inputs and specialised reagents [32]. Ziembowicz and Kida [33] investigated the suitability of advanced oxidation processes for the degradation of Indigo Carmine, Tartrazine, and Allura Red AC dyes through single processes (oxidation using H₂O₂, the Fenton process, and O₃) and hybrid processes (O₃ + H₂O₂ and O₃ + Fenton). They found that the use of hydrogen peroxide (H₂O₂) alone was not effective for the degradation of the analysed dyes (max. efficiency was 9.38%), the ozonation process was effective in removing only Indigo carmine (97%), while for Tartrazine and Allura Red AC, only 8.46% and 4.32% were removed, and the Fenton process led to the complete decolorization of all dyes. The total cost of these processes was from 92.28 PLN m⁻³ at minimum to 1632 PLN m⁻³ at maximum, which is equivalent to 20 euros to 375 euros per m³. Pieczynska et al. [34] studied the electrochemical oxidation of five azo dyestuffs using a highly boron-doped diamond (BDD) electrode in a flow reactor. They discovered that a higher degradation rate was achieved at increased current densities, with a maximum removal efficiency of 98.9%. The composition of the wastewater had a significant impact on the rate of dyestuff removal. Notably, the presence of Cl^- ions greatly accelerated the removal process, while the addition of NaOH inhibited it. However, the high cost associated with BDD electrode material limits their practical application on an industrial scale [35].

Microbial degradation and enzymatic treatment offer environmentally friendly alternatives, but they are often hindered by slow degradation rates and the specificity of microbial consortia [36,37]. Thus, Peres-Diaz and McFeethers [38] found that Lactobacillus casei LA1133, Lactobacillus Paracasei LA0471, and other lactic acid bacteria have the ability to modify TY under anaerobic conditions to a significant extent. The study by Wan and associates confirmed that Pseudomonas aeruginosa has the ability to remove several dyes, including TY, from various industrial wastewaters. The maximum dye degradation of 72.65% was observed under aerobic conditions within 5 h at a pH of 9.12 and a temperature of 28.5 °C [39]. Adsorption onto activated carbon and membrane filtration effectively remove azo dyes, but these approaches can be costly and generate secondary waste [40]. The costs associated with membrane filtration treatment of wastewater from the dye industry can largely be attributed to membrane replacement expenses, which are directly related to the required membrane area and, consequently, the size of the treatment plant being utilised. utilising membrane technology presents a challenge in maintaining low costs, primarily due to the frequent need for membrane replacements associated with these applications. In industrial settings, membrane costs typically represent 10 to 20% of the total equipment expenses. For the membrane nanofiltration method, the replacement cost is between USD 30 and USD 60 per m² [41].

Electrocoagulation has gained recognition as a cost-effective and efficient alternative for azo dye removal. The process involves the in situ generation of coagulant species from sacrificial electrodes, facilitating the destabilisation and aggregation of dye molecules. The materials used as sacrificial anodes are usually different aluminium alloys, iron, steel, ductile iron, and zinc, which dissolve anodically and produce cations like Al³⁺, Fe²⁺, and Zn²⁺ [42,43,44,45,46]. In aqueous solutions, these ions form the corresponding hydroxides Al(OH)₃, Fe(OH)₂, Fe(OH)₃, and Zn(OH)₂, as well as various monomeric, polymeric, and hydroxo-complexes. The specific products that form depend on the pH of the solution and the electrode potential. These hydrolysis products help destabilise contaminants in the solution, facilitating their agglomeration and enabling their separation from the solution through settling or flotation.

Electrocoagulation offers several advantages over conventional treatment methods, making it an efficient and scalable solution for wastewater remediation. It effectively removes a wide range of contaminants, including dyes, heavy metals, and organic pollutants, while minimising the need for external coagulant dosing. Additionally, EC is adaptable for industrial-scale applications with minimal infrastructure modifications and generates compact sludge, which is easier to handle compared to traditional chemical coagulation [47,48]. Despite its advantages, EC presents certain challenges, including electrode passivation and energy consumption, which require further research and optimisation. In EC, optimising parameters such as pH, current density, and electrolysis time for enhanced dye removal are emphasised [47,49,50]. Recent studies have demonstrated the effectiveness of aluminium electrodes in EC, further underscoring the importance of parameter optimisation for improved performance [51,52,53].

Several studies have previously explored the removal of TY using EC under various operating conditions and electrode configurations. Modirshahla et al. [54] examined TY removal using iron and aluminium electrodes, achieving nearly 100% colour and 90% COD removal at pH 5.78, within just 6 min of electrolysis, at a current density of 120 A m⁻² and an interelectrode distance of 1.5 cm. They also reported that the Fe/Al electrode pair (anode/cathode) was more effective than the Fe/Fe configuration. Tahir and Rauf [55] investigated TY removal using iron and steel electrodes, obtaining 50% removal efficiency after 70 min of treatment at 5 V, for initial dye concentrations ranging from 30 to 68 mg/L. More recently, Zhian and Noroozi Pesyan [56] found that EC cells equipped with iron and aluminium anodes and an aluminium cathode in a series connection achieved higher removal efficiency compared to simple EC cells using only an iron anode and aluminium cathode. Husein et al. [57] focused on the influence of sodium chloride (NaCl) concentration on the EC process for simultaneous TY removal and hydrogen generation.

Their study recorded the highest removal efficiency of 93% at NaCl concentrations of 0.8 and 1 g/L after 240 min, underscoring the method’s promise for future environmental applications.

Although several studies have explored the EC process for TY dye removal using various electrode combinations and operating conditions, the present work provides a more comprehensive and integrative investigation focused specifically on the performance of aluminium electrodes. This study stands out by simultaneously monitoring key solution parameters (pH, temperature, and removal efficiency), while also conducting a microscopic assessment of anode consumption, detailed sludge characterisation using SEM/EDS and FTIR analysis, and evaluation of the settling behaviour of post-EC suspensions. Furthermore, the economic aspect of the EC process is addressed through operational cost analysis. The optimisation of process variables is conducted using response surface methodology (RSM), offering a statistically robust and efficient experimental design. This integrated approach delivers novel insights into the mechanistic, practical, and economic dimensions of TY removal with EC and advances the development of cost-effective and environmentally sound water treatment strategies.

2. Materials and Methods

Tartrazine Yellow (TY) (chemical formulae C₁₆H₁₀Na₂O₇S₂N₂), obtained from Glentham Life Sciences, Chorsham, UK, was used to prepare a test solution with a concentration of 0.05 mmol L⁻¹ (26.72 mg L⁻¹). Sodium chloride (Kemika, Zagreb, Croatia) was used as a supporting electrolyte in a concentration of 1 g L⁻¹. Hydrochloric acid and sodium hydroxide (Kemika, Zagreb, Croatia) were used to adjust the pH. All chemicals were analytical-reagent grade and were used without further purification. A 600 mL of solution was placed in a glass reactor for the EC process. The anode and cathode were made of aluminium alloy AA 2007 (surface area of 47 cm²) with the composition shown in

Table 1. Chemical composition of alloy AA2007.

Element	Al	Cu	Mg	Fe	Si	Mn	Pb	Bi	Zn
wt.%	93.50	3.84	0.88	0.67	0.21	0.67	0.11	0.07	0.04

.

Before each measurement, the rectangle-shaped electrodes were ground with sandpaper grades from P280 to P800, cleaned ultrasonically with ethanol and deionised water, dried, weighed on the analytical balance, and then immersed in the test solution at a mutual distance of 3 cm. Electrodes were connected to the Wanptek DPS605U digital rectifier (60 V, 5 A) (Wanptek Electronic Co., Shenzhen, China), used for current adjustment. The solution was mixed with the Dlab MS-H-Pro+ magnetic stirrer (Dlab Scientific Co., Ltd., Beijing, China) at a constant speed of 150 rpm.

The influence of current density (0.008, 0.016, and 0.024 A cm⁻²) on the efficiency of TY dye removal was measured. The duration of the EC process varied between 10 and 50 min, while the initial pH value of the solution ranged from 3.0 to 7.0. The experimental study was conducted through 20 separate EC processes, following a predetermined experimental plan presented in

Table 2. EC measurements for TY dye removal working plan.

Experiment	A pH	B t (min)	C i (A cm−2)
RUN 1	7.0	32	0.008
RUN 2	5.0	41	0.016
RUN 3	7.0	10	0.016
RUN 4	7.0	50	0.016
RUN 5	7.0	50	0.024
RUN 6	5.0	19	0.016
RUN 7	5.7	50	0.008
RUN 8	4.8	50	0.024
RUN 9	5.7	10	0.024
RUN 10	5.7	10	0.024
RUN 11	3.0	23	0.008
RUN 12	3.0	37	0.024
RUN 13	3.0	23	0.008
RUN 14	5.7	50	0.008
RUN 15	3.0	10	0.016
RUN 16	7.0	10	0.008
RUN 17	7.0	10	0.016
RUN 18	4.7	10	0.008
RUN 19	3.0	50	0.016
RUN 20	3.0	37	0.024

using the Design-Expert 13 software (State-Ease. Inc., Minneapolis, MN, USA). During the EC process, changes in pH value, temperature (T), and concentration (c) of the TY solution were monitored. The pH value of the TY solution was measured every 5 min using a Mettler Toledo Seven Multi pH meter (Mettler Toledo, Greifensee, Switzerland). The temperature was measured using a Testo 925 thermometer (Testo SE & Co. KGaA, Titisee-Neustadt, Germany) at 5 min intervals, with an accuracy of 0.1 °C. The dye concentration in the solution was determined spectrophotometrically using a Perkin Elmer Lambda 25 UV/VIS spectrophotometer (PerkinElmer U.S. LLC, Shelton, CT, USA) at 426.28 nm. Samples (10 mL) for concentration determination were taken at the beginning of the measurement (initial solution) and after 5, 10, 15, 30, 40, and 50 min, depending on the measurement duration.

The change in the mass of the anode was determined by weighing it before and after the EC process using a Kern ALJ 160-4NM analytical balance (Kern & Sohn GmbH, Balingen, Germany). The surface condition of the aluminium anode after the EC processes was examined using an MXFMS-BD optical microscope (Ningbo Sunny Instruments Co., (Yuyao City, Zhejiang Province, China) at a magnification of 200×.

Also, after the electrocoagulation, the settling ability of the suspensions was determined by settling the suspension in a 500 mL cylinder and measuring the supernatant height at set time intervals. The solution was then filtered, and the sludge was dried and weighed. Electrogenerated sludge was characterised using a scanning electron microscope, Thermo Scientific Quattro S (FEG SEM, Hillsboro, OR, USA), with an EDS SDD detector Ultim^®Max, Oxford Instruments (Abingdon, Oxfordshire, UK). Also, Fourier transform infrared spectroscopy (FT-IR) spectra of dye and sludge were recorded on a Perkin Elmer Spectrum Two FT-IR spectrometer ((PerkinElmer U.S. LLC, Shelton, CT, USA) by the Attenuated Total Reflectance (ATR) technique with a diamond reflection crystal. The spectra were collected in 10 scans at a resolution of 4 cm⁻¹ and in the range of 4000–450 cm⁻¹ at room temperature.

3. Results

3.1. Change in Process Parameters During the EC Process

During the execution of each EC process (under the conditions listed in Table 2), pH, temperature (T), concentration (c), and TY removal efficiency (R) change. For clarity, selected examples of the temporal change in process parameters are shown graphically (Figure 1, Figure 2, Figure 3 and Figure 4), with an attempt to include the experiments with the most extended duration generally.

3.1.1. Change in pH Value

An increase in the pH value is observed at all current densities, and the dynamics of this increase depend on the applied current intensity (Figure 1a).

At the lowest current density, the solution’s pH value rises gradually over the first 20 min and then stabilises at around 8.4. The higher the current density, the faster the final pH value is reached. In addition to the current density, the initial pH value also affects its change during the EC process (Figure 1b). The higher the initial pH value, the faster the final (stable) value is reached. Regardless of the initial conditions, the final pH of the TY solution is around 8.4 in all experiments, which is consistent with the literature data [58,59,60]. During the EC process, the pH of the solution shifts from acidic (or neutral) to alkaline, and a minimum pH of 8.0 is expected at the end [61]. The main cause of the pH increase under acidic conditions is the reduction of water at the cathode, which leads to the production of gaseous hydrogen and hydroxide ions (OH)⁻ [59,60,62].

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

(1)

The pH change is also influenced by the electrochemical dissolution of the anode, during which Al³⁺ ions are released:

$$\mathrm{A}\mathrm{l}\to {\mathrm{A}\mathrm{l}}^{3+}+3{\mathrm{e}}^{-}$$

(2)

If the anode potential is sufficiently high, secondary reactions can also occur, in particular oxygen evolution [47,62,63,64]:

$$2{\mathrm{H}}_2\mathrm{O}\to {\mathrm{O}}_2+4{\mathrm{H}}^{+}+4{\mathrm{e}}^{-}$$

(3)

The Al³⁺ ions produced at the anode undergo spontaneous hydrolysis reactions, which lead to acidification of the anode environment and the formation of various monomeric species, according to the following order:

$${\mathrm{A}\mathrm{l}}^{3+}+{\mathrm{H}}_2\mathrm{O}\to {\mathrm{A}\mathrm{l}\left(\mathrm{O}\mathrm{H}\right)}^{2+}+{\mathrm{H}}^{+}$$

(4)

$${\mathrm{A}\mathrm{l}\left(\mathrm{O}\mathrm{H}\right)}^{2+}+{\mathrm{H}}_2\mathrm{O}\to {\mathrm{A}\mathrm{l}\left(\mathrm{O}\mathrm{H}\right)}_2^{+}+{\mathrm{H}}^{+}$$

(5)

$${\mathrm{A}\mathrm{l}\left(\mathrm{O}\mathrm{H}\right)}_2^{+}+{\mathrm{H}}_2\mathrm{O}\to {\mathrm{A}\mathrm{l}\left(\mathrm{O}\mathrm{H}\right)}_3+{\mathrm{H}}^{+}$$

(6)

This is a simplified scheme, as dimeric, trimeric, and polynuclear hydrolysis products of aluminium can also be formed, e.g., Al₂(OH)₂⁴⁺, Al₃(OH)₄⁵⁺, Al₆(OH)₁₅³⁺, Al₇(OH)₁₇⁴⁺, Al₈(OH)₂₀⁴⁺, Al₁₃O₄(OH)₂₄⁷⁺, and Al₁₃(OH)₃₄⁵⁺ [62,65,66,67]. The resulting cationic aluminium hydrolysis products can further react with OH^- ions and finally transform into amorphous Al(OH)₃ in the bulk solution through complex precipitation kinetics [62,65,66,67]. The mode of action of the coagulant (Al³⁺) in the EC process is generally explained by two interrelated mechanisms: the charge neutralisation of negatively charged colloidal particles by cationic hydrolysis products and adsorption of impurities on the amorphous hydroxide precipitate (the so-called “sweep flocculation”). Based on theoretical predictions and data from the literature, the formation of the Al species mentioned strongly depends on the pH value [68]. At a pH below 3.5, Al³⁺ is the predominant species. However, at a pH of 4−9, the predominant chemical Al species is Al(OH)₃(s) (and its polymerised species), which has a large surface area useful for the rapid adsorption of soluble organic compounds and for the trapping of colloidal particles. However, at a higher pH (>9), the predominant species is Al(OH)_4⁻, which does not coagulate pollutants [68,69,70].

3.1.2. Temperature Change

It is known that temperature significantly affects the rate of electrode reactions, floc formation, hydroxide solubility, and the conductivity of the electrolyte solution during the EC process [48,49]. During the removal of TY dye, an increase in temperature is observed, with the current density having the greatest influence (Figure 2).

The higher the applied current, the faster the solution temperature rises, which is consistent with data from the literature [71]. Specifically, the temperature increases by about 4 °C at the lowest current density and even by about 25 °C at the highest current density. On the other hand, the initial pH of the solution has no significant effect on the dynamics of the temperature change during the removal of the TY dye, as can be seen from the partial overlap of the curves in Figure 2b. Regardless of the initial pH of the solution, the temperature consistently increases by about 25 °C at a current density of 0.024 A cm⁻².

3.1.3. TY Dye Removal Efficiency

During the EC process, 10 mL of the TY solution was sampled every 5–10 min. After centrifugation, the dye concentration was determined spectrophotometrically. The dye removal efficiency (R) was determined using the following expression:

$$R={\displaystyle \frac{c_0-c_{\mathrm{t}}}{c_0}}\times 100$$

(7)

where c₀ represents the initial concentration (0.05 mmol L⁻¹), and c_t is the concentration at sampling time. As EC progresses, the removal efficiency of TY dye increases from zero to very high values (Figure 3). The growth dynamics and the final value of R depend on the experimental conditions (i, initial pH, and t).

This process is accompanied by a decrease in the intensity of the yellow colour, which is particularly pronounced at higher current densities and lower pH values. At a current density of 0.008 A cm⁻², for example, the visible colour intensity began to change after 20 min (at R ≈ 50%, which at the end of the process reached ≈75 (i.e., after 32 min). At 0.024 A cm⁻², the colour change was already visible after 10 min, and after 40 min, the solution was completely colourless, with an R of about 99.40%.

A similar effect can be observed when the initial pH is lowered. The solution with a pH of 3.0 showed the best results and became colourless after only 15 min (removal efficiency of 99.40%). The solution with a pH of 4.8 became colourless after 30 min and showed similar removal efficiency, while the solution with a pH of 7.0 lost its colour after 40 min, with R ≈ 99.40%.

3.1.4. Three-Dimensional Representation of Process Parameter Changes

With constant input parameters (i and initial pH), the final values of the process responses (T, pH value, and c) can be represented as a function of time (i.e., the process duration) in a three-dimensional (3D) coordinate system. Such a representation results in 3D surfaces with different shapes, contours, and colours (Figure 4).

Any changes in the final values of the system responses at a constant current density (T, pH, and c = f (t, initial pH) can be more easily observed by following the colour change over a flat or curved surface in the 3D diagram. As the colour changes from orange to yellow and from green to blue, the final values of the output parameters (T, pH, and c) decrease. A reverse colour transition indicates an increase in the output parameter’s final value (the transition’s limits are outlined on the grey base of the diagram).

The surface describing the change in the final pH value as a function of its initial value and time is more complex but remains the same for all applied currents. The final solution pH increases with EC duration at lower initial pH values. As the initial pH of the solution increases, the observed surface curves bend, and its colour changes from green to yellow to orange (with a slight decrease in intensity towards the end of the process). In other words, as the initial pH of the solution increases, its final value increases to a maximum and then gradually decreases. The greater the current density, the less pronounced these changes are.

The most important parameter is the concentration of the solution, as its decrease over time indicates the rate of TY dye removal. Under all experimental conditions, the concentration decreases with time and, in some cases, reaches very low values (close to zero). Colour changes accompany these changes on a flat rectangular area in the 3D diagram. As the colour changes from yellow to green and then to blue, the concentration of the TY dye decreases, and its removal efficiency increases. Accordingly, the lowest concentration (represented by the most intense blue colour) is reached at higher current densities and lower initial pH values. Thus, at the highest current density, depending on the initial pH, the dye concentration drops to zero after 20–25 min (which means that R is ≈99%). At the medium current density, this occurs after 35–40 min, while at the lowest current density, the concentration does not reach zero and is 0.01 mmol L⁻¹ (and R ≈ 80%) after 50 min (at pH = 3).

3.2. Anode Consumption and Sludge Characterisation

To understand the impact of different parameters on the TY dye removal efficiency during the EC process, the consumption of the anode material (determined as the difference in electrode mass before and after the EC process) and collected sludge mass (after the EC process, the solution was filtered, and the sludge was dried and weighed) were monitored. The results are compared in

Table 3. Anode material consumption and sludge mass under different process conditions.

Experiment	pH	t (min)	i (A cm−2)	Anode Consumption (g)	Sludge Mass (g)
RUN 1	7.0	32	0.008	0.0860	0.1853
RUN 4	7.0	50	0.016	0.2574	0.6047
RUN 5	7.0	50	0.024	0.3733	1.1764
RUN 8	4.8	50	0.024	0.3973	1.5023
RUN 12	3.0	37	0.024	0.2895	1.1693

. It can be seen that under all conditions, there is consumption of anode material, which increases with increasing current density and does not depend significantly on the initial pH value. On the other hand, the sludge mass increases with increasing current density, but also with decreasing pH value. It should be noted that the durations of RUN1 and RUN12 (pH = 3) were about 20 min shorter.

Anode consumption with increasing current density is expected and occurs through electrode kinetics (i.e., the Butler–Volmer equation), according to which the rate of metal dissolution increases exponentially with the rise in anode polarisation [72]. The dissolution of Al anodes and the formation of Al³⁺ ions are necessary for the further conduction and progression of the EC process. However, when the anodes dissolve, the state of their surfaces changes. Microscopic examination of the electrodes after the EC process (Figure 5) reveals the nature of their dissolution and their corrosion patterns. Under all test conditions, a general pattern of corrosion attack can be observed. The current density and the solution’s initial pH value have no significant influence on the corrosion form.

It is known that TY belongs to the group of anionic azo dyes, which means that it is negatively charged in aqueous solutions [73,74] and easily combines with positively charged coagulants (Al³⁺ ions formed during the dissolution of the anode). Due to strong electrostatic attraction, their coagulation and flocculation are very effective, forming stable Al(OH)₃ flocs (of different surface areas) that settle quickly. A comparison of the results (Figure 1 and Figure 3) shows that the best efficiency is achieved under neutral to slightly alkaline conditions (pH 6–8), where flocs are easily formed. These conditions are most effectively produced during the EC process at higher current densities and a lower initial pH.

After the EC process, the settling ability of the resulting milky, gelatinous precipitate was determined, and the results are shown in Figure 6. During sedimentation, a sharp boundary forms between the clear solution and the sludge, which facilitates the determination of settling ability (Figure 6a). In all investigated cases, the “interface height between the solid and liquid phases” (h) decreases over time. A curve exhibiting a steeper slope, i.e., a more pronounced decline in h, indicates superior settling performance and more efficient solid–liquid phase separation.

Sedimentation ability is closely related to the mechanism of dye removal, which is described in the literature by five basic steps: (1) oxidation of the aluminium anode and formation of metal ions [75,76,77]; (2) formation of oxygen and hydrogen bubbles through water hydrolysis; (3) hydrolysis of metal cations and formation of various coagulating species that destabilise the electrostatic forces between the particles; (4) aggregation of coagulants with dye molecules; and (5) coupling of aggregates with gas bubbles and their ascent to the surface by flotation.

The kinetics of sludge formation during EC and its subsequent sedimentation strongly depend on the experimental conditions. An increase in the current density leads to a more intensive dissolution of the Al anode, which results in a larger mass of the generated sludge [47,48]. Since the amount of Al dissolved depends primarily on the level of current flow and not on the initial pH value, the mass of sludge formed is almost independent of the pH value [47,48]. On the other hand, sedimentation ability is not determined by the total mass of the sludge, but by the physicochemical properties of the flocs formed [47,78]. Experimental results show that at higher pH values, larger, more compact, and denser aluminium hydroxide flakes are likely to form, leading to faster separation and settling [16]. Conversely, at lower pH values, smaller and more dispersed aggregates are likely to form, resulting in slower settling, even if the sludge mass is larger [78].

It is important to note that pH also significantly affects removal efficiency. At a lower pH, there is an increased concentration of highly reactive Al³⁺ ions and simple hydroxyl species, which effectively destabilise dye molecules through charge neutralisation and adsorption mechanisms [47,78]. As a result, azo dyes are removed faster and more efficiently at lower pH values than at higher pH values, although the sludge settles more slowly [78]. In addition, lowering the pH enhances the cathodic reduction of water (Equation (3)), resulting in a greater amount of hydrogen microbubbles. These bubbles significantly improve flocculation as they aggregate more easily with coagulant particles and adsorbed dye molecules, allowing them to rise to the surface more easily during EC [47].

In summary, although sludge settles faster at higher pH values, probably due to the formation of larger and denser flocs, overall dye removal efficiency is higher at lower pH values due to the higher reactivity of the aluminium ions, higher hydrogen bubble production, and better flocculation [47,78]. Therefore, the choice of optimal electrocoagulation conditions requires a careful balance between the efficiency of pollutant removal and the dynamics of sludge deposition.

Table 4. SEM images and elemental composition (on marked spots) of sludge under different process conditions.

RUN 1 (pH = 7.0; i = 0.008 A cm−2)	Spectrum
	Element	1	2	3	4
	C	5.14	9.35	5.64	7.77
	O	57.31	55.53	56.81	56.00
	Na	0.13	0.19	0.10	0.15
	Mg			0.27	0.09
	Al	36.37	33.88	35.98	34.61
	Si	0.10	0.15	0.13	0.17
	S	0.10	0.12	0.09	0.11
	Cl	0.85	0.78	0.98	1.10
	Total	100.00	100.00	100.00	100.00
RUN 4 (pH = 7.0; i = 0.016 A cm−2)	Spectrum
	Element	1	2	3
	C	9.10	11.69	11.88
	O	56.79	58.20	57.31
	Na	0.26	0.34	0.22
	Mg	0.14	0.04	0.06
	Al	32.97	29.01	29.65
	Si	0.19	0.07	0.13
	S	0.14	0.15	0.16
	Cl	0.41	0.50	0.59
	Total	100.00	100.00	100.00
RUN 7 (pH = 7.0; i = 0.024 A cm−2)			Spectrum
	Element	1	2	3	4
	C	17.83	50.46	51.15	29.97
	O	48.65	38.00	35.61	36.66
	Na	0.76	0.26	0.11	0.05
	Al	30.78	10.78	12.65	32.61
	S	0.22	0.31	0.37	0.25
	Cl	1.76	0.19	0.11	0.46
	Total	100.00	100.00	100.00	100.00
RUN 12 (pH = 3.0; i = 0.024 A cm−2)	Spectrum
	Element	1	2	3
	C	60.84	56.14	54.86
	O	28.49	31.23	30.01
	Al	10.14	12.21	14.73
	S	0.53	0.42	0.40
	Cl			0.10
	Total	100.00	100.00	100.00

shows the characterisation of sludge using SEM/EDS analysis. The main elements in all cases are Al and O (from Al(OH)₃) and C and S (from the removed dye); their content varies depending on the EC process conditions. In some cases, the sludge also contains varying amounts of other elements (Na, Mg, Cl, etc.).

The experimental observations regarding removal efficiency suggest that the EC process can be optimised by adjusting parameters such as current density and pH. For example, when the current density increases from 0.008 to 0.024 A cm⁻² at a pH of 7.0, the C content increases from 5% to 50%. When the initial pH is lowered to 3.0, the C content increases to 55–60%. In contrast to C, the content of S in the sludge is significantly lower, but it also increases.

The FTIR spectrum of Tartrazine (E102), a synthetic azo dye, was recorded in the range of 4000–450 cm⁻¹ (Figure 7a). The spectrum reveals characteristic absorption bands corresponding to the principal functional groups of the molecule, including sulfonate groups, azo linkages, aromatic rings, and a heterocyclic pyrazolone moiety. The analysis of the major peaks is presented below. A broad absorption band centred at 3443.7 cm⁻¹ is attributed to the stretching vibration of O–H bonds, most likely due to adsorbed moisture or hydrogen bonding interactions involving sulfonate groups. A series of intense peaks at 1633.8–1557.0 cm⁻¹ corresponds to the C=C stretching vibrations of aromatic rings (1621.9 and 1599.1 cm⁻¹) and N=N azo stretching modes (1557.0 cm⁻¹), reflecting the extensive conjugation between azo linkage and aromatic systems. The peak at 1478.9 cm⁻¹ is also associated with C–N stretching or symmetric azo bond vibrations. Sulfonate functional groups (–SO_3⁻), which are important features of Tartrazine structure, are evidenced by the 1344.8 cm⁻¹ peak (S=O asymmetric stretching). An additional peak at 1123.1 cm⁻¹ further supports the presence of S–O stretching modes, commonly observed in sulfonated aromatic compounds. Bands in the region of 1174.2–1106.1 cm⁻¹ are primarily attributed to para-substituted aromatic compounds by sulphoxyl groups, while adsorption lines from 869.2 to 564.5 cm⁻¹ are related to di-substituted aromatic compounds [5].

FTIR spectra of sludge after TY removal through electrocoagulation (Figure 7b) show spectroscopic changes. Electrocoagulation causes destabilisation of tartrazine due to the formation of Al³⁺ ions, which hydrolyze to Al(OH)₃. This hydroxide adsorbs dye molecules, resulting in shifts and/or the disappearance of characteristic vibrations (particularly those of azo and sulfonate groups). FTIR spectra indicate that interactions occur between the functional groups of tartrazine and aluminium hydroxide, including the possibility of complexation, adsorption, and hydrophobic interactions. The disappearance of the azo bond (–N=N–) suggests that, in addition to adsorption, partial degradation of the dye molecule also takes place. Similar results for azo dye sludge were found in the literature [8,12].

Despite the organic content of the resulting sludge, its classification as hazardous waste cannot be definitively determined without further ecotoxicological assessment in accordance with EU Directive 2008/98/EC and Regulation No. 1357/2014 [79,80]. Given its composition—rich in aluminium hydroxide and organic residues—there is theoretical potential for the reuse of EC sludge as a secondary coagulant in other water or wastewater treatment applications, similar to studies investigating the recovery of flocculants from industrial sludge [81,82]. In a review paper, Rajaniemi and associates showed the possibilities of using EC sludge as a fertiliser (mainly as struvite), pigment, construction material (mainly as blocks), adsorbent, and catalyst [82]. Agraw [83] used raw and calcined EC sludge collected from a textile wastewater treatment plant as an adsorbent for the removal of direct red 28 dye. EC sludge adsorbent was prepared through wet treatment with deionized water and calcination at 500 °C. So, there is a possibility of using waste sludge generated through the electrocoagulation process, but after additional toxicological analysis.

3.3. Mathematical Model Development

The experimental design was created using Design-Expert software, which was used for mathematical modelling to determine the optimal conditions for carrying out the EC process. The pH, T, and c of TY were monitored during all EC processes. Their final values and the initial conditions of each experiment are presented in

Table 5. Initial and final values of the selected variables during the EC process to remove the TY dye (initial concentration 0.05 mmol L⁻¹).

Experiment	Input Parameters			Output Parameters			R (%)
	A pH	B t (min)	C i (A cm−2)	pH	T (°C)	c (mmol L−1)
RUN 1	7.0	32	0.008	8.48	27.5	0.0126	74.8
RUN 2	5.0	41	0.016	8.64	38.9	0.0030	94.0
RUN 3	7.0	10	0.016	8.58	27.1	0.0257	48.6
RUN 4	7.0	50	0.016	8.49	41.6	0.0014	97.2
RUN 5	7.0	50	0.024	8.17	50.0	0.0012	97.6
RUN 6	5.0	19	0.016	8.36	24.8	0.0097	80.6
RUN 7	5.7	50	0.008	8.56	28.5	0.0071	85.8
RUN 8	4.8	50	0.024	8.25	50.9	0.0009	98.2
RUN 9	5.7	10	0.024	8.51	32.1	0.0188	62.4
RUN 10	5.7	10	0.024	8.61	35.0	0.0192	61.6
RUN 11	3.0	23	0.008	5.805	27.3	0.0213	57.4
RUN 12	3.0	37	0.024	8.563	48.5	0.0004	99.2
RUN 13	3.0	23	0.008	6.14	26.8	0.0204	59.2
RUN 14	5.7	50	0.008	8.72	29.8	0.0114	77.2
RUN 15	3.0	10	0.016	5.41	29.3	0.0245	51.0
RUN 16	7.0	10	0.008	8.75	23.4	0.0367	26.6
RUN 17	7.0	10	0.016	8.40	27.6	0.0265	47.0
RUN 18	4.7	10	0.008	8.00	24.1	0.0230	54.0
RUN 19	3.0	50	0.016	8.00	39.3	0.0013	97.4
RUN 20	3.0	37	0.024	7.74	49.0	0.0003	99.4

.

Using the Design-Expert software [84,85,86,87,88], the relationship between the input parameters and the output values was represented by three different empirical models, enabling the evaluation of the correlation between responses and experimental parameters. The first response analysed was the change in electrolyte T, where the minimum value was 23.4 °C, the maximum value was 50.9 °C, the mean value was 34.1 °C, and the standard deviation was 9.41 °C. The second response was the change in electrolyte pH, where the minimum value was 5.415, the maximum was 8.746, the mean was 8.01, and the standard deviation was 0.9995. The third response analysed was the change in final TY c, with a maximum value of 0.0367 mmol L⁻¹, a minimum of 0.0003 mmol L⁻¹, a mean of 0.0133 mmol L⁻¹, and a standard deviation of 0.0111 mmol L⁻¹. Based on the results obtained and the recommendations of Design-Expert software, the final values of the model were presented as coded factors as follows:

T = + 34.58 − 0.9280 × A + 5.77 × B − 7.67 × C[1] − 0.9070 × C[2] + 0.9045 × AB + 0.5897 × AC[1] + 1.26 × AC[2] − 3.47 × BC[1] + 0.9552 × BC[2]

(8)

pH = +7.98 + 0.7864 × A + 0.3638 × B − 0.2084 × C[1] − 0.1012 × C[2] − 0.5337 × AB + 0.3629 × AC[1] + 0.0696 × AC[2] + 0.0740 × BC[1] + 0.2685 × BC[2]

(9)

ln(c) = − 5.10 + 0.4167 × A − 1.16 × B + 0.9810 × C[1] − 0.0698 × C[2] − 0.1206 × AB − 0.4700 × AC[1] − 0.3907 × AC[2] + 0.6025 × BC[1] − 0.2519 × BC[2]

(10)

A Pareto analysis of variance (ANOVA) was performed to analyse the experimental results. The higher F-values of the model (42.66, 5.88, and 48.88 for the measured electrolyte T, pH, and c) and the associated lower p-values show the validity of the developed mathematical models (

Table 6. Analysis of variance (ANOVA) for all measured responses.

	T (°C)		pH		c (mmol L−1)
	F Value	p-Value	F Value	p-Value	F Value	p-Value
MODEL	42.66	<0.0001	5.88	0.0053	48.88	<0.0001
A–pH	1.67	0.2247	29.16	0.0003	13.36	0.0044
B–t (min)	118.28	<0.0001	2.46	0.1481	195.89	<0.0001
C–i (A cm−2)	95.57	<0.0001	1.23	0.3342	51.02	<0.0001
AB	0.9867	0.3440	4.87	0.0518	0.7386	0.4103
AC	1.87	0.2042	1.48	0.2744	15.06	0.0010
BC	9.30	0.0052	1.42	0.2855	11.33	0.0027
R2	0.9746		0.8412		0.9778
Adj-R2	0.9518		0.6982		0.9578
Pre-R2	0.8639		−0.1692		0.8070
C.V. (%)	6.07		6.86		6.32
Std. Dev.	2.07		0.5491		0.3187
Mean	34.08		8.01		−5.04
Adeq. Pre.	19.3520		8.1390		18.9734

). Furthermore, all p-values below 0.0500 indicate that the model terms are significant, while p-values above 0.1000 indicate that the model terms are insignificant. For temperature, B, C, and BC are significant model terms (AB and AC are not significant); for electrolyte pH, A and AB are significant model terms (B, C, AC, and BC are not significant, which is to be expected since electrolyte pH depends on electrolyte temperature); and for the change in electrolyte c, A, B, C, AC, and BC are significant model terms (AB is not significant).

The goodness of fit statistics for T show that the predicted R² (Pre-R²) of 0.8639 is in reasonable agreement with the adjusted R² (Adj-R²) of 0.9518, i.e., the difference is less than 0.2. Adequate precision (Adeq. Pre.) measures the signal-to-noise ratio. A ratio of more than 6 is desirable. The achieved ratio of 19.352 indicates a reasonable signal, which means that the model can be used for navigation in the design space. Similarly, for the change in c, the Pre-R² of 0.8070 is in reasonable agreement with the Adj-R² of 0.9578 (the difference is less than 0.2). The Adeq. Pre. ratio of 18.9734 indicates a reasonable signal. However, the negative Pre-R² of −0.1692 for electrolyte pH suggests that the overall mean may predict the response better than the current model. The higher-order models may also provide better prediction in some cases. The Adeq. Pre. of 8.1390 is a reasonable signal, indicating that the model can be used to navigate the design space.

3.3.1. Diagnostic of Model Adequacy

In general, the model must be validated to ensure that it provides an accurate approximation of the actual (experimental) values. Investigating and optimising parameters without assessing the model’s fit could lead to misleading and unreliable results. Therefore, diagnostic plots such as the normalised plot and the parity plot between predicted and actual values are essential for verifying the model’s fit and analysing the relationship between actual and predicted values. The left column of Figure 8 shows a graph comparing the observed (actual) response values with the predicted response values.

This comparison helps identify observations that the model poorly predicts. Ideally, the data points should be evenly distributed around the 45-degree line. The right-hand column of Figure 8 displays the normal probability plot of the residuals, which indicates whether the residuals follow a normal distribution. If the residuals are normally distributed, the points should align along a straight line, although a moderate amount of scatter is expected even for normal data. No formal statistical tests are required for this diagnosis; visual inspection of the plot is sufficient. Externally standardised residuals provide the most reliable metric for this analysis.

3.3.2. Optimisation and Authentication of Process Parameters and Responses

The final step in the modelling process was to optimise the model to determine the optimal input factors based on the desired objectives for each response (

Table 7. Constraint table with goals and limits for all factors.

Name	Goal	Lower Limit	Upper Limit	Lower Weight	Upper Weight	Importance
A–pH	is in range	3	7	1	1	3
B–t (min)	is in range	10	50	1	1	3
C–i (A cm−2)	is in range	0.008	0.024	1	1	3
T (°C)	target (30)	22	35	1	1	3
pH	target (7)	6	8	1	1	4
c (mmol L−1)	minimise	0.0003	0.0367	1	0.1	5

). The available objectives include maximising, minimising, targeting a specific value, staying within a range, none (for responses only), and setting an exact value (for factors only). A minimum and maximum value must be specified for each parameter. In addition, a weight can be assigned to each goal to customise the shape of the respective desirability function. The “importance” of each goal can be adjusted in relation to the others. By default, all goals are considered equally important, with a setting of 3 plus points (+++). If a goal is more important, it can be set to 5 plus points (+++++).

The goals are combined into a general desirability function that the programme tries to maximise. The optimisation process begins with a random starting point and proceeds via the steepest slope to reach the maximum. The search can start from several points within the design space to increase the chances of finding the “best” local maximum. The default is 30 starting points, which can be adjusted as required.

Desirability is an objective function that ranges from 0, indicating values outside the acceptable limits, to 1, representing the target value. Numerical optimisation searches for a point that maximises this desirability function. The characteristics of each objective can be adjusted by changing its weighting or importance. When multiple responses and factors are involved, all objectives are combined into a single desirability function. It is important to note that achieving a high desirability score is not always necessary. The desirability value depends entirely on how close the lower and upper limits are to the optimum. The goal of optimisation is to determine a set of conditions that satisfy all objectives, rather than aiming for a desirability value of 1.0. Desirability is a mathematical tool used to find the optimal conditions.

Table 8. Obtained solutions depending on set criteria and constraints.

Number	pH	t (min)	i (A cm−2)	T (°C)	pH	c (mmol L−1)	Desirability
1	3.3696	18.741	0.016	30.000	7.000	0.016	0.976	Selected
2	3.000	46.592	0.008	28.404	7.000	0.011	0.932
3	3.048	45.027	0.008	38.303	7.000	0.012	0.927

shows the number of solutions obtained based on the specified criteria and constraints. In this case, only three solutions fulfil the desirability criteria, and the highest solution is selected.

Figure 9 shows a graph of desirability-generated selected solutions obtained through numerical optimisation.

3.4. Operating Costs of the EC Process

The operating costs of the EC process include various components, including material and energy consumption, chemical reagents, sludge disposal, labour, maintenance, and equipment costs [89,90]. The most important cost factors are the electrode material and electrical energy consumption. Therefore, special emphasis is placed on their calculation as described in Equation (11):

Operating cost = a × C_energy + b × C_electrode

(11)

where C_energy is the energy consumption per cubic metre of treated wastewater (kWh m⁻³), C_electrode is the mass of the Al anode consumed per cubic metre of wastewater (kg m⁻³), and a and b approximate the cost of electrical energy (0.1035 EUR/kWh [91]) and Al material (2.25 EUR/kg Al sheet [92,93]), respectively. The energy consumption is calculated using Equation (12):

$$C_{\mathrm{energy}}={\displaystyle \frac{U\times I\times t}{V}}$$

(12)

where U is the average voltage, I is the current, t is the reaction time, and V is the volume of treated wastewater. Electrode consumption is determined by measuring mass loss during the EC process. The results are shown in

Table 9. The comparison of a calculated electrode and energy cost.

Experiment	U (V)	Cenergy (kWh m−3)	Celectrode (kg m−3)	Celectrode,Faraday (kg m−3)	Operating Cost (EUR m−3)
RUN 1	11.78	4.00	0.14	0.11	0.736
RUN 2	28.37	24.64	0.36	0.29	3.360
RUN 3	26.61	5.65	0.09	0.07	0.787
RUN 4	26.61	28.19	0.43	0.36	3.890
RUN 5	38.87	61.79	0.62	0.53	7.880
RUN 6	26.61	10.73	0.17	0.14	1.493
RUN 7	12.07	6.40	0.23	0.18	1.180
RUN 8	37.02	58.85	0.66	0.53	7.486
RUN 9	36.29	11.57	0.15	0.11	1.535
RUN 10	36.96	11.79	0.14	0.11	1.535
RUN 11	13.92	3.39	0.08	0.07	0.531
RUN 12	33.02	38.88	0.48	0.40	5.104
RUN 13	14.54	3.55	0.10	0.08	0.592
RUN 14	16.80	8.91	0.21	0.18	1.395
RUN 15	28.63	6.08	0.09	0.07	0.832
RUN 16	12.06	1.28	0.05	0.04	0.245
RUN 17	29.25	6.21	0.09	0.07	0.845
RUN 18	14.00	1.49	0.04	0.03	0.244
RUN 19	20.19	21.39	0.42	0.36	3.159
RUN 20	35.59	41.90	0.50	0.40	5.462

, which also contains theoretical values for electrode consumption calculated according to Faraday’s law for comparison. The energy costs depend on the experimental conditions. They generally increase with increasing current (at a constant initial pH), while they decrease with an increasing initial pH of the solution (at a constant current). The highest energy consumption (61.79 kWh m⁻³) was determined for RUN 5, with the lowest for RUN 16 (1.28 kWh m⁻³).

The total consumption of Al material is higher than the theoretical value, which is a consequence of electrochemical and chemical reactions that contribute significantly to the dissolution of the aluminium electrodes and generate Al³⁺ coagulant during the EC process. Similar results have been observed in previous studies, referred to as super-faradaic efficiency [62,66,89,94].

Previous research has shown that electrocoagulation can achieve comparable or better contaminant removal than conventional methods like chemical coagulation, often at lower operating costs, particularly when applied to sludge dewatering or greywater treatment [94,95,96] A comparative study bt Kobya at al. on the treatment of textile wastewater using the EC process and CC showed that the EC process was faster and more economical; it consumed less material and produced less sludge, and the pH of the medium was more stabilised than CC for similar COD and turbidity removal levels [97]. In the present study, the calculated operating costs for TY removal through the EC process are in this range, namely between 0.245 EUR m⁻³ (RUN 16) and 7.486 EUR m⁻³ (RUN 5).

Reducing energy costs can be achieved by coupling the EC process with renewable energy sources such as solar panels and wind turbines. This integration reduces dependency on conventional energy grids and lowers operational costs and carbon emissions. Studies have demonstrated that the incorporation of renewable energy into EC systems can lead to improved efficiency and reduced emissions, aligning sustainable treatment goals. By integrating renewable energy sources and utilising waste metal as an anode material, the electrocoagulation process can become more economically viable and environmentally sustainable. These strategies not only reduce operational costs but also contribute to resource conservation and pollution reduction, making EC a more attractive option for wastewater treatment applications [98,99,100].

4. Conclusions

The study investigated the possibility of removing TY dye from an aqueous solution using the EC process with aluminium electrodes. Response surface methodology was applied to evaluate the effects of the three main independent parameters—current density (0.008–0.024 A cm⁻²), initial solution pH (3.0–7.0), and treatment time (10–50 min)—on the process parameters, including pH, temperature, TY dye concentration, and removal efficiency, and to optimise the operating conditions of the treatment process.

As the EC process progresses, the removal efficiency of the TY dye increases, while the growth dynamics and final value (between approx. 28% and 99%) depend on the experimental conditions (current density, initial pH, and treatment time). High removal efficiency is achieved more quickly when current density increases and initial pH decreases. Contrary, sludge settles more rapidly at higher pH values due to the formation of denser flocs. Despite containing organic residues, EC-generated sludge has potential for reuse in environmental applications, such as adsorbents or construction materials, provided it undergoes thorough ecotoxicological assessment in accordance with EU regulations.

During the EC process, aluminium anodes undergo general corrosion, with consumption increasing as current density rises and initial pH decreases. At the end of the EC process, the sludge was found to contain Al and O (from Al(OH)₃) as well as C and S (from the removed dye), their content depending on the EC process conditions. EC not only removes TY from the solution but also alters its chemical structure through adsorption (on Al(OH)₃) and possible degradation.

Furthermore, the experimental data were statistically analysed, and mathematical models were developed for each response (pH, temperature, and dye concentration). The optimal conditions to treat TY using EC were derived and validated. The optimum current density, EC duration, and initial pH of the dye solution were found to be 0.016 A cm⁻², 18.741 min, and 3.3696, respectively. Under these optimal values of process parameters, the removal efficiency of TY should be 70% (or 0.016 mmol L⁻¹).

The results confirm that EC is a cost-effective treatment method for TY, with operating costs ranging from 0.245 to 7.486 EUR/m³ depending on process conditions. Key contributors to cost are energy and electrode consumption, both of which can be optimised through operational adjustments. Furthermore, integrating renewable energy and alternative electrode materials can enhance the economic and environmental sustainability of the EC process, making it a promising solution for wastewater treatment.