Adsorption Combined with Electrocoagulation Process for Ketoprofen Removal from Aqueous Solution: Optimization Using Central Composite Design

Abstract

The combination of electrocoagulation (EC) with complementary treatment methods has garnered increasing attention for wastewater remediation. This study aims to design and optimize a hybrid electrocoagulation–adsorption (EC/Ads) process for the removal of ketoprofen (KTP) from aqueous solutions. The adsorption of KTP onto activated carbon (AC) alone exhibited a low removal efficiency of approximately 27% under the following conditions: initial KTP concentration ([KTP]₀) = 23 mg·L⁻¹, pH = 6, adsorbent dose (q_AC) = 0.5 g, and contact time = 30 min. In contrast, the EC process alone achieved a removal efficiency of 59.69% under similar conditions (current density (i) = 18.6 mA·cm⁻², NaCl = 3.5 g·L⁻¹). The combined EC/Ads process significantly enhanced KTP removal, reaching 87.11% under the same operational parameters. The synergistic effect of the combined treatment was quantified with a synergy index of 1.37. Characterization techniques included FTIR analysis of both AC and KTP, as well as adsorption–desorption isotherms and pH_PZC determination for AC. To further optimize the EC/Ads process, a response surface methodology based on central composite design (CCD) was applied to assess the influence of four independent variables: pH, [KTP]₀, current density, and q_AC. Optimal conditions were identified as follows: q_AC = 0.63–0.99 g, i = 12.32–14.68 mA·cm⁻², pH = 6.5, and [KTP]₀ = 22.5 mg·L⁻¹; these conditions resulted in 100% KTP removal after 30 min of treatment. These findings demonstrate the potential of the EC/Ads hybrid process to be an efficient and sustainable alternative for pharmaceutical contaminant removal.

1. Introduction

Pollution is a multifaceted global problem, and one of its most concerning aspects is the contamination caused by pharmaceutical substances. Pollution generally refers to the presence of harmful substances in the environment that negatively affect ecosystems and human health. It can originate from various sources including industrial emissions, chemical waste and agricultural activities. However, pharmaceutical pollution is a specific environmental issue resulting from the improper use and disposal of medicines and related products. Once medications are consumed, a portion of their active ingredients is excreted through urine and feces, eventually ending up in sewage systems and natural aquatic environments. These substances are typically detected in concentrations ranging from nanograms per liter (ng·L⁻¹) to milligrams per liter (mg·L⁻¹) [1,2,3].

Among pharmaceutical substances, ketoprofen (2-(3-benzoylphenyl) propionic acid) is a non-steroidal anti-inflammatory drug (NSAID) that is widely used worldwide mainly due to its over-the-counter availability and low cost. It is commonly used to relieve muscle and joint issues, as well as to treat disorders such as arthritis, gout, rheumatoid osteoarthritis, and inflammation in general. However, excessive levels of ketoprofen (KTP) can lead to various adverse health effects, including digestive and gastrointestinal disorders [4,5,6]. KTP is continuously released into the environment through urine excretion and the limited efficiency of wastewater treatment plants. It has been detected in wastewater treatment plants (WTPs), surface water and groundwater in several studies [6].

Different treatment methods can be employed to reduce the organic matter present in aqueous media including chemical and physical adsorption [7], membrane filtration [8], advanced oxidation processes [9,10] and electrocoagulation (EC) [11,12,13]. Electrocoagulation is characterized by its operational safety and adaptability, particularly with respect to the chemical compositions of various industrial effluents making it one of the most suitable treatment methods in many cases [14,15]. However, this technology also presents certain limitations. The electrodes, typically composed of aluminum or iron, undergo progressive wear and must be replaced regularly, resulting in non-negligible operational costs. Furthermore, the process can be energy-intensive, especially when treating highly contaminated wastewater, thereby increasing large-scale treatment expenses. Lastly, to meet increasingly stringent environmental regulations, electrocoagulation is often coupled with complementary treatment processes to ensure optimal effluent quality.

The selection of the coupling between electrocoagulation and a complementary treatment process must undergo a rigorous evaluation, with the aim of optimizing pollutant removal efficiency while minimizing the overall energy consumption of the system. Few studies have thoroughly examined recent advances regarding the combination of electrocoagulation (EC) with adsorption processes. Among these, the work of Wang et al. [16] stands out for its investigation of the coupling of electrocoagulation with peanut shell adsorption for the removal of malachite green from water. The authors observed that an adsorbent dose of 5 g·L⁻¹ significantly enhanced removal efficiency, while reducing both the required current density and treatment duration, compared to electrocoagulation alone. Under optimal conditions, a removal efficiency of 98% was achieved within only 5 min, representing a 23% improvement over the EC-only process conducted over 60 min.

Furthermore, the combined process enabled a 94% reduction in both unit energy consumption and electrode material demand compared to standalone electrocoagulation. Sorayyaei et al. [17] investigated the efficiency of an integrated electrocoagulation/adsorption system for the removal of methyl orange from aqueous solutions. The system consisted of two electrodes spaced 2 cm apart with an iron electrode serving as the cathode and a stainless steel sieve box functioning as the anode, the latter being filled with modified pumice stones acting as the adsorbent. The effective surface area of the electrodes was 3 × 10 cm. The system achieved a removal efficiency of 93.1% for the dye. Another study conducted by Yang et al. [18] investigated the removal of micropollutants from real municipal wastewater using graphene adsorption combined with simultaneous electrocoagulation/electrofiltration. They determined that the primary mechanisms involved in the combined process were carbon adsorption, size exclusion, electrostatic repulsion, electrocoagulation and electrofiltration. Notably, the combined approach demonstrated significant removal efficiencies for various pollutants: di-n-butyl phthalate (89%), di-(2-ethylhexyl) phthalate (85%), acetaminophen (99%), caffeine (94%), cefalexin (100%) and sulfamethoxazole (98%).

This study aims to develop and optimize a hybrid electrocoagulation–adsorption (EC/Ads) process for the removal of ketoprofen (KTP) from aqueous solutions. Coupling the electrocoagulation (EC) process with adsorption (Ads) enhances the overall treatment efficiency by significantly reducing the time required for KTP elimination. This combined approach also contributes to lowering operational costs, particularly those related to electrode degradation and energy consumption, while maintaining high removal performance. A central composite design (CCD) was used to optimize the operating conditions for KTP removal.

2. Materials and Methods

2.1. Chemicals

A commercial pallet coconut shell-based granular activated carbon was used as received. KTP (98% purity), acetonitrile (99.9% purity), Hydrochloric acid (36% purity) and sodium hydroxide and (99.5% purity) were supplied by Sigma-Aldrich, France. The characteristics and chemical structure of KTP are shown in

Table 1. Characteristics and chemical structure of KTP.

Commercial Name	Raw Formula	λmax (nm)	Molar Weight (g·mol−1)	Purity (%)	pKa	Solubility in Water (mg·L−1) at 25 °C
ketoprofen	C16H14O3	260	254.28	≥98%	4.4	51
	Chemical structure

.

2.2. Activated Carbon Characterization

2.2.1. Infrared Spectroscopy

A BRUKER Alpha FTIR spectrometer (S/N: 100855, Bremen-Germany) was used for the identification of functional groups in ketoprofen, fresh activated carbon, activated carbon used in the adsorption process and activated carbon used in the electrocoagulation–adsorption coupling process.

2.2.2. Activated Carbon Texture Characterization

The structural heterogeneity of the AC was assessed through nitrogen adsorption–desorption isotherms at −196.290 °C using a Brunauer–Emmett–Teller (BET) and Barrett–Joyner–Halenda (BJH) method (micromeritics ASAP 2020 Plus Version 2.00, Georgia-US). The specific surface area was determined using the BET method, while the total pore volumes were calculated based on the liquid volume of nitrogen adsorbed measured over a relative pressure (P/Po) range from approximately 0.0009 to 0.9997 at a relative pressure of 0.147. The Dubinin–Radushkevich (D–R) equation is represented below [19,20,21]:

$$\log V=\log V_{mic}-D{\left(\log \left({\displaystyle \frac{P}{P_0}}\right)\right)}^2$$

(1)

where P₀ is saturation pressure, D is constant, and V_mic is the total micropore volume.

The mesopore volume (Vmes) was obtained by subtracting the micropore volume from the total pore volume. The average pore diameters were estimated from the BET surface area and total pore volume (D_p = 4V_tot/S_BET), assuming an open-ended cylindrical pore model without pore networks.

2.3. pH_PZC Determination

The pH at the point of zero charge (pH_PZC) for the AC sample was determined following the method described by Wang et al. [16] to perform the measurement, and aliquots of 50 mL of a 0.01 mol·L⁻¹ NaCl solution were prepared in separate flasks. The pH of each solution was adjusted to a range between 2 and 12 using 0.01 mol·L⁻¹ HCl or NaOH solutions. Then, 0.15 g of the AC sample was added to each flask and the mixtures were shaken for 48 h. The final pH values were measured using a pH meter after stabilization. The pH_PZC was identified as the point where the plot of the final pH versus the initial pH intersected with the line where pH_initial equaled pH_final [16].

2.4. Experimental Procedure

2.4.1. Adsorption Experiments

The activated carbon used in this study was a commercial adsorbent presented in the form of black rods. Prior to use, it was thoroughly washed with distilled water and then dried to remove any impurities that could affect its performance.

For the adsorption experiments, synthetic KTP solutions were prepared in the presence of salt (NaCl) at a concentration of 3.5 g·L⁻¹. The pH was adjusted to 6 using 0.1 mol·L⁻¹ HCl or 0.1 mol·L⁻1 NaOH. The initial KTP concentration ([KTP]₀) was 23 mg·L⁻¹, with an adsorbent dose (q_AC) of 0.5 g/500 mL of solution and a reaction time of 30 min. The solution was placed in a reactor, stirred at 240 rpm and maintained at a constant temperature of 25 ± 2 °C. Samples were collected at specific time intervals and filtered using 0.45 μm syringe filters prior to analysis.

2.4.2. Electrocoagulation

Before initiating the electrocoagulation experiments, two types of cathodes were evaluated to determine the one exhibiting the best performance (in the absence of activated carbon). The materials tested included a 304L stainless steel cathode (12 cm²) and a grid-type iron cathode (95.135 cm²). The anode (95.135 cm²), also in a grid configuration, remained the same for all experiments regardless of the cathode type. The experiments were conducted under the following operating conditions: an interelectrode distance fixed at 1 cm, a current density of 18.60 mA·cm⁻², a NaCl concentration of 3.5 g·L⁻¹, an initial pH of 6, a KTP concentration of 23 mg·L⁻¹, a temperature of 25 ± 2 °C and constant stirring at 240 rpm.

2.4.3. Combined System—Electrocoagulation/Adsorption

In the context of the combined treatment system (EC/Ads), experiments were carried out under the same operational conditions as those used in the electrocoagulation tests, using both an anode and a cathode made of iron in the form of a grid. Two experimental configurations were investigated. The first involved the use of activated carbon suspended in the solution during electrocoagulation (EC/Ads-ACS), while the second incorporated activated carbon immobilized directly within the anode (EC/Ads-ACT). The adsorbent was integrated into the folded grid structure to ensure the homogeneous distribution of the activated carbon. The edges of the grid were carefully sealed to keep the material in place. This configuration optimized the adsorption process by maximizing the contact surface between the adsorbent and the target molecules (Figure 1). All experiments were carried out under identical operating conditions to ensure a valid comparison.

2.4.4. Analysis Technique

High-Performance Liquid Chromatography (HPLC ACC 3000 HPLC) determined the residual concentrations of KTP. Samples were filtered using a 0.45 μm membrane syringe filter. The HPLC system comprised a UV/Vis detector (Model VWD 3400 RS, REUZEit, Berlin, Germany), an autosampler (Model LPG 3400 SD, Dionex Softron GmbH, Germering, Germany) and a standard degasser (LPG 3400 SD, Dionex Softron GmbH, Germering Germany). The separation process utilized a C18 reversed-phase column (5 μm; 4.6 × 150 mm) manufactured by Thermo Fisher Scientific (Dionex Softron GmbH, Germering, Germany). The detection of KTP was performed at 260 nm. The mobile phase was a mixture of acetonitrile and ultrapure water (30/70 v/v) at a flow rate of 0.5 mL·min⁻¹. The removal efficiency of KTP (y) was determined using Equation (2):

$$y\left(\%\right)={\displaystyle \frac{{\left[KTP\right]}_0-{\left[KTP\right]}_t}{{\left[KTP\right]}_0}}\ast 100$$

(2)

where [KTP]₀ is the initial concentration of KTP and [KTP]_t is the final concentration of KTP.

2.4.5. Electrodes

Two types of electrodes were tested in this study: a 304L stainless steel electrode (dimensions: 40 mm × 30 mm × 2 mm) and an iron grid (Figure 1). The 304L stainless steel corresponded to the standard designation X5CrNi18-10 (1.4301). Its chemical composition was as follows: C ≤ 0.07%, Si ≤ 1%, Mn ≤ 2%, p < 0.045%, S ≤ 0.015%, N ≤ 0.1%, Cr = 17–19.5%, and Ni = 8–10.5%, with the remainder being Fe.

The iron grid used was an iron–zinc alloy. To ensure that only iron participated in the electrochemical reactions during electrocoagulation, the grid was subjected to a pre-treatment process to remove the zinc layer. This treatment consisted of immersing the grid in a 0.1 mol·L⁻¹ hydrochloric acid solution for 5 min, followed by thorough rinsing with distilled water. The crystalline structure of the iron grid, both before and after the acid treatment, was analyzed by X-ray diffraction (Germany) using a vertical θ/θ diffractometer (CuKα radiation, λ = 0.15406 nm, 40 kV, 30 mA) over a 2θ range of 20–120°, with a step size of 0.02°.

2.5. Experimental Design

Statistical Design of Experiments (DOE) is a systematized and structured method of experimentation that determines the relationship between factors that affect the process response [22,23,24,25,26,27,28,29]. This method optimizes the process by simultaneously modifying all contributing factors; it also detects non-linear effects and complex interactions between factors that may be difficult to detect with conventional methods [22,23,24,25,26,27,28,29].

In this study, the EC/Ads process was optimized using a central composite design. The statistical experimental design was developed considering four factors, i.e., the pH ($x_1$), initial concentration of KTP ($x_2$), current density ($x_3$) and activated carbon dose ($x_4$). The real values of each factor and their respective levels are presented in

Table 2. Experimental range and levels of independent variables.

Variables	Levels
	−2	−1	0	1	2
pH: $x_1$	3.00	4.75	6.50	8.25	10.00
[KTP]0 (mg·L−1):$x_2$	5.00	13.75	22.5	31.25	40.00
i (mA·cm−2): $x_3$	2.10	5.25	8.41	11.56	14.71
qAC (g): $x_4$	0.10	0.325	0.550	0.775	1.00

. Other variables, such as experiment duration and temperature, were held constant at 30 min and 25 ± 2 °C, respectively. The central composite design (CCD) consisted of 27 experiments: 16 experiments (Runs 1–16) of the factorial design, 3 experiments realized in the central working range (Runs 17–19) and 8 start points (Runs 20–27) (

Table 3. Matrix of central composite design tests.

Run	Real Variables				Coded Variables					Response
	pH	[KTP]0 (mg·L−1)	i (mA·cm−2)	qAC (g)	${\mathit{x}}_0$	${\mathit{x}}_1$	${\mathit{x}}_2$	${\mathit{x}}_3$	${\mathit{x}}_4$	y (%)
1	4.75	13.75	5.25	0.325	1	−1	−1	−1	−1	40.91
2	4.75	13.75	5.25	0.775	1	−1	−1	−1	1	42.62
3	4.75	13.75	11.56	0.325	1	−1	−1	1	−1	47.72
4	4.75	13.75	11.56	0.775	1	−1	−1	1	1	77.53
5	4.75	31.25	5.25	0.325	1	−1	1	−1	−1	40.91
6	4.75	31.25	5.25	0.775	1	−1	1	−1	1	42.62
7	4.75	31.25	11.56	0.325	1	−1	1	1	−1	47.72
8	4.75	31.25	11.56	0.775	1	−1	1	1	1	77.53
9	8.25	13.75	5.25	0.325	1	1	−1	−1	−1	40.91
10	8.25	13.75	5.25	0.775	1	1	−1	−1	1	42.62
11	8.25	13.75	11.56	0.325	1	1	−1	1	−1	47.72
12	8.25	13.75	11.56	0.775	1	1	−1	1	1	77.53
13	8.25	31.25	5.25	0.325	1	1	1	−1	−1	40.91
14	8.25	31.25	5.25	0.775	1	1	1	−1	1	42.62
15	8.25	31.25	11.56	0.325	1	1	1	1	−1	47.72
16	8.25	31.25	11.56	0.775	1	1	1	1	1	77.53
17	6.50	22.50	8.41	0.55	1	0	0	0	0	45.70
18	6.50	22.50	8.41	0.55	1	0	0	0	0	46.57
19	6.50	22.50	8.41	0.55	1	0	0	0	0	45.96
20	3.00	22.50	8.41	0.55	1	−2	0	0	0	45.57
21	10.00	22.50	8.41	0.55	1	2	0	0	0	45.57
22	6.50	5.00	8.41	0.55	1	0	−2	0	0	45.57
23	6.50	40.00	8.41	0.55	1	0	2	0	0	45.57
24	6.50	22.5	2.10	0.55	1	0	0	−2	0	51.21
25	6.50	22.5	14.71	0.55	1	0	0	2	0	92.94
26	6.50	22.5	8.41	0.10	1	0	0	0	−2	29.81
27	6.50	22.5	8.41	1.00	1	0	0	0	2	61.33

).

Based on the data collected during the CCD experiment, which are summarized in Table 3, central composite design was implemented to establish the following quadratic polynomial equation representing the second-order model:

$$\widehat{y}=b_0+{\displaystyle \sum _{i=1}^4{b}_i{x}_i+{\displaystyle \sum _{i=1}^4{b}_{ii}{x}_i^2+}}{\displaystyle \sum _{i=1}^4{\displaystyle \sum _{j=i+1}^4{b}_{ij}{x}_i{x}_j}}$$

(3)

where the symbols have the following definitions:

• Predicted yield (ŷ);
• The coefficients of the response model (${\mathit{b}}_{\mathbf{1}},{\mathit{b}}_{\mathbf{2}},\dots ,{\mathit{b}}_{\mathit{n}}$);
• Linear terms, corresponding to the variables (${\mathit{x}}_{\mathbf{1}},{\mathit{x}}_{\mathbf{2}},\dots ,{\mathit{x}}_{\mathit{n}}$);
• Squared terms, corresponding to the variables (${\mathit{x}}_{\mathbf{21}}$, ${\mathit{x}}_{\mathbf{22}}$, …, ${\mathit{x}}_{\mathbf{2}\mathit{n}}$);
• First-order interaction terms for each paired combination (${{\mathit{x}}_{\mathbf{1}}\mathit{x}}_{\mathbf{2}}$, ${\mathit{x}}_{\mathbf{1}}{\mathit{x}}_{\mathbf{3}}$, …, ${\mathit{x}}_{\mathit{n}-\mathbf{1}}{\mathit{x}}_{\mathit{n}}$).

3. Results

3.1. Characterization Study

3.1.1. FTIR Analysis

The FTIR spectra of fresh activated carbon (AC), ketoprofen (KTP), activated carbon after adsorption (Ads) and activated carbon used in the electrocoagulation–adsorption coupling process (EC/Ads) were analyzed to confirm the occurrence of the adsorption process (Figure 2). The spectrum of KTP exhibited characteristic bands (

Table 4. Comparative summary of characteristic FTIR bands.

Band	Assignment	KTP	AC	EC/Ads and Ads
~1700 cm−1	(C=O)	strong	absent	appeared after adsorption
~1600–1500 cm−1	(C=C aromatic)	present	present	present, intensified
~1250–1000 cm−1	(C–O)	strong	weak	appeared/intensified

) around 1700 cm⁻¹ corresponding to C=O stretching (carbonyl group), bands in the 1250–1000 cm⁻¹ range attributed to C–O stretching and peaks between 1600 and 1500 cm⁻¹ associated with aromatic C=C vibrations. In contrast, the spectrum of fresh activated carbon displayed a broad O–H stretching band around 3400 cm⁻¹, a peak near 1580 cm⁻¹ due to aromatic C=C bonds and weak bands around 1000 cm⁻¹ likely related to C–O vibrations.

The spectra recorded after KTP adsorption onto activated carbon alone (Ads) and in combination with electrocoagulation (EC/Ads) exhibited nearly identical profiles, indicating that the integration of the electrocoagulation process did not alter the surface characteristics of the activated carbon. Notable spectral changes included the appearance of a C=O peak near 1700 cm⁻¹ and intensified C–O bands between 1250 and 1000 cm⁻¹, confirming the presence of KTP on the material surface. Additionally, changes observed in the aromatic region (1600–1500 cm⁻¹) suggested interactions between KTP and activated carbon.

3.1.2. BET Analysis

• Nitrogen isotherm and surface area: Figure 3 illustrates the nitrogen adsorption–desorption isotherms for the AC. The amount of N₂ adsorbed is plotted as a function of the relative pressure (P/P₀), where P is the vapor pressure and P₀ is the saturation vapor pressure of nitrogen. The lower branch of the isotherm corresponds to the incremental adsorption of N₂ on the adsorbent surface, while the upper branch reflects the desorption process as nitrogen is progressively released. These curves provide insights into the pore structure and adsorption behavior of the activated carbon material. The N₂ adsorption–desorption isotherms corresponding to the AC type H3 are related to having a high affinity for the adsorbent. At very low concentrations, curves do not start at zero but at a positive value on the y-axis corresponding to the amounts adsorbed; this isotherm is encountered when there is chemisorption of the solute.

• Surface area and pore volume: The physical characteristics of the analyzed AC derived from the N₂ adsorption–desorption isotherms, including the total surface area (SBET), total pore volume (Vtot), external surface area (Sext), microporous surface area (Smic), mesoporous surface area (Smes), microporous volume (Vmic), mesoporous volume (Vmes) and average pore diameter (Dp), are summarized in

  Table 5. Characteristics of AC determined from N₂ adsorption isotherms at −196.29 K.

  SBET (m2·g−1)	Vtot (cm3·g−1)	Sext (m2·g−1)	Smic (m2·g−1)	Vmic (cm3·g −1)	Vmes (cm3·g −1)	Dp (nm)
  947.251	0.509	564.438	382.812	0.430	0.079	2.149

  .

Referring to the results in Table 5, the analysis highlights that the activated carbon used has a high total surface area of 947.251 m²·g⁻¹ and a significant microporous volume of 0.430 cm³·g⁻¹, both key factors in its adsorptive efficiency. The microporous structure is especially suited for the adsorption of small molecules like ketoprofen. Additionally, the mesoporous volume (0.079 cm³·g⁻¹) allows for better diffusion and adsorption of larger ketoprofen molecules. The overall pore structure with an average pore diameter of 2.149 nm indicates that the activated carbon is capable of effectively trapping and removing ketoprofen from aqueous solutions, making it ideal for environmental remediation applications.

3.1.3. X-Ray Diffraction Analysis

X-ray diffraction (XRD) analysis reveals the presence of both iron (Fe) and zinc (Zn), with zinc appearing as a surface coating. This protective layer plays a key role in preventing corrosion of the iron wires that make up the grid structure. As shown in Figure 4, the acid treatment effectively removes the zinc layer, confirming that the grid electrode used is composed exclusively of iron.

3.2. Preliminary Tests

To examine the synergistic action of EC and Ads on the removal efficiency of KTP, comparative experiments were performed under the same conditions. According to the Figure 5, after 30 min, the removal efficiency of KTP was evaluated for the process used. The degree of synergy (S), defined based on the kinetic constant from the simultaneous implementation of two processes (k_EC/Ads) and the sum of the individual processes (EC and Ads), was determined using Equation (4) [9,30]:

$$S={\displaystyle \frac{k_{EC/Ads}}{k_{EC}+k_{Ads}}}$$

(4)

Ln [KTP]₀/[KTP]_t versus time was used to determine the apparent rate constant (Figure 6). If the ratio S exceeded one, it indicated synergy between the two processes; otherwise, they either acted additively (S = 1) or competitively (S < 1) [9,30]. The value of S was 1.37, providing clear evidence that employing both processes (EC and Ads) simultaneously resulted in a synergistic effect rather than just an additive one. Therefore, the EC process combined with activated carbon adsorption was recommended.

• The performance of the EC process is significantly impacted by the various electrode materials and electrode combinations used; to this end, two electrode materials were considered: steel and iron grids. The KTP removal efficiency as a function of time and electrode type is shown in Figure 7. The results shows that using an iron grid cathode is more efficient than using a steel cathode. After 40 min of electrolysis, a remarkable elimination in the case of an iron grid cathode was observed, with a yield of 95% compared to 43.13% when using a steel cathode. The superior performance of the iron grid cathode can be explained by the electrochemical properties of iron, which exhibit a higher reduction capacity compared to other metals. This allows iron to produce more hydroxide ions (OH⁻) and hydrogen bubbles (H₂), crucial for efficient pollutant removal. Combined with its large surface area and stability, these properties enhance the overall electrochemical reactions, making iron the optimal choice for water treatment applications. Its non-toxic and recyclable nature further supports its suitability for use in sustainable processes.

• The performances of the three configurations—electrocoagulation alone, electrocoagula-tion with suspended activated carbon (EC/Ads-ACS) and electrocoagulation with an-ode-immobilized activated carbon (EC/Ads-ACT)—were evaluated in terms of KTP re-moval efficiency over different treatment durations (Figure 8).

The following results were generated:

• After 10 min of treatment, the electrocoagulation process alone exhibited a very low removal efficiency (0.179%). In contrast, the addition of activated carbon, either suspended (ACS) or immobilized within the anode (ACT), led to a significant improvement, achieving removal efficiencies of 11% and 10.23%, respectively, with comparable performance at this early treatment stage.
• After 30 min, the differences between the configurations became more pronounced. The EC process reached a removal efficiency of 18.95%, while the EC/Ads-ACS system achieved 32.81%. Notably, the EC/Ads-ACT configuration showed a substantial enhancement, reaching 82.90% removal efficiency.
• At 60 min, all configurations demonstrated improved performance, with EC/Ads-ACT remaining the most effective, achieving a removal efficiency of 96.96%, followed by EC/Ads-ACS (84.15%) and EC alone (59.69%).

The results demonstrate that the efficiency of the electrocoagulation process can be significantly enhanced by incorporating an adsorption mechanism. In particular, the EC/Ads-ACT configuration, in which activated carbon is immobilized within the anode, exhibits superior performance compared to the suspended carbon configuration. This setup optimizes mass transfer phenomena and promotes the localized accumulation of KTP near electroactive zones, resulting in a marked improvement in removal efficiency.

Indeed, the EC/Ads-ACT configuration initially promotes the adsorption of KTP molecules onto the immobilized activated carbon, leading to the formation of a concentration gradient between the grid surface and the bulk solution. This gradient drives a diffusive flux toward the anode, thereby enhancing the transport of the pollutant to electroactive sites and subsequently facilitating its entrapment through the electrocoagulation process. Conversely, when activated carbon is used in suspension and uniformly dispersed throughout the solution, the concentration gradient of KTP between the anode and the solution is significantly reduced due to the uniform diffusion of KTP molecules across the entire volume of the solution. This reduction in the concentration gradient impedes the mass transfer of KTP toward the electrode surface, thereby limiting the efficiency of KTP removal by electrocoagulation.

These observations confirm the presence of a significant synergistic effect between the adsorption and electrocoagulation processes. Overall, this interaction provides a robust basis for the development of advanced treatment strategies, offering enhanced performance for the remediation of complex effluents.

The next preliminary test was conducted to evaluate the efficiency of ketoprofen removal as a function of the activated carbon’s location, either immobilized at the anode or the cathode. As illustrated in Figure 9, the complete removal of KTP (100%) was achieved within 40 min when the activated carbon was placed at the anode. In contrast, when the carbon was positioned at the cathode, the removal efficiency did not exceed 50%, even after 60 min of treatment.

This difference in performance can be explained by the underlying mechanisms of the EC/Ads-ACT configuration. The initial adsorption of KTP onto immobilized activated carbon creates a concentration gradient between the adsorbent surface and the surrounding solution, inducing a diffusive flux of the contaminant toward the electrode. When the activated carbon is located at the anode, the applied electric current promotes the release of metal ions, which subsequently react with hydroxide ions to form metal hydroxides (Equations (5)–(7)), well known for their strong coagulating properties. The synergy between adsorption and electrocoagulation under these conditions enables the rapid and complete removal of KTP.

Conversely, when the activated carbon is immobilized at the cathode, the absence of metal ion release limits the in situ generation of coagulants near the adsorbent surface, thereby significantly reducing the overall removal efficiency. These results highlight the crucial role of anodic reactions in enhancing the performance of the combined electrocoagulation/adsorption system:

$$Anode:{Fe}_{\left(s\right)}{\to Fe}^{2+}+2e^{-}$$

(5)

$$Cathode:{2H}_{2}O+2e^{-}\to H_2+2{OH}^{-}$$

(6)

$${Insolution:Fe}^{2+}+2{OH}^{-}\to {Fe\left(OH\right)}_2$$

(7)

3.3. Modeling and Optimization by CCD

An analysis of regression variance was performed to assess the model’s relevance. The model could be considered significant based on the results of the Fisher test and the p-value < 0.05 test. The results of the analysis of variance (ANOVA) for KTP removal are summarized in

Table 6. A summary of the ANOVA result.

Response	Model	F_Value	p_Value	R2	R2Adj
Ketoprofen removal	quadratic	12,969.48	<0.0001	0.9999	0.9999

. The polynomial quadratic model was used to understand the interaction between the independent and dependent variables, and the predicted regression model for the response in terms of the removal efficiency of KTP (ŷ, %) was as follows:

$$ŷ=46.07+10.43x_3+7.88x_4+7.02x_3{x}_4-0.13x_1^2-0.13x_2^2+6.50x_3^2-0.13x_4^2$$

(8)

According to the regression equation obtained by CCD (Equation (8)), current density (x₃) and activated carbon dose (x₄) had positive effects on the response (b₃ = +10.43 and b₄ = +7.88, respectively). As can be seen in Figure 10, the elimination efficiency of KTP was proportional to the current density and activated carbon dose.

A strong interaction (Figure 11) between x₃ and x₄ with effect +7.02 was found by the experimental design.

The model equation was used to determine the optimal values of the operating pa-rameters to achieve a 100% yield of KTP. Contour plot analysis (Figure 12) allowed us to identify the following optimal conditions: q_AC = 0.63–0.99 g, i = 12.32–14.68 mA·cm⁻², pH = 6.5 and [KTP]₀ = 22.5 mg·L⁻¹. Experimental tests conducted under optimized conditions showed a 97% removal of KTP, which closely matched the predicted value of 100%. Thus, the model was validated.

No significance was recorded for the parameters pH and [KTP]₀ regarding KTP removal (Figure 13 and Figure 14).

4. Comparative Review

The ketoprofen removal efficiency achieved by the combined electrocoagulation and adsorption (EC/Ads) process was compared with those reported in various studies involving different configurations integrating electrocoagulation and adsorption techniques for the treatment of organic pollutants (

Table 7. A comparison of KTP removal with literature studies involving the combination of adsorption and electrocoagulation.

Authors	Pollutant	Method	Experimental Conditions	Reaction Time (min)	Removal Efficiency (%)
Balouchi et al. [11]	Eriochrom black T (Azo dye)	Adsorption	pH = 3, qAC = 0.3 g·L−1	60	75
		Electrocoagulation	pH = 7, i = 20 mA·cm−2	60	80
		EC/Ads	pH = 7, qAC = 0.3 g·L−1, i = 20 mA·cm−2	60	95
Wang et al. [16]	Malachite Green	Adsorption	pH = 5, qAC = 2 g·L−1	30	70
		Electrocoagulation	pH = 5, i = 20 mA·cm−2	30	82
		EC/Ads	pH = 5, qAC = 0.3 g·L−1, i = 20 mA·cm−2	30	97
Yang et al. [18]	Sulfamethoxazole Cephalexin	Adsorption	pH = 6, 0.5 g·L−1	45	35 * 82 **
		EC/filtration	pH = 6, i = 15 mA·cm−2	45	97 * 96 **
		EC/Ads	pH = 6, qAC = 0.5 g·L−1, i = 15 mA·cm−2	45	98 * 100 **
Present Study	Ketoprofen	Adsorption	pH = 6, qAC = 1 g·L−1	30	27
		Electrocoagulation	pH = 6, i = 18.60 mA·cm−2	30	59.69
		EC/Ads	pH = 6, qAC = 1.26–1.98 g·L−1, i = 12.32–14.68 mA·cm−2	30	100

*: Removal Efficiency of sulfamethoxazole and **: Removal Efficiency of Sulfamethoxazole of cephalexin.

). This comparative analysis highlighted that the EC/Ads process, particularly when coupled with activated carbon immobilized on a grid, holds significant potential for use in environmental applications. The process is distinguished by its operational simplicity and effectiveness in removing organic contaminants of pharmaceutical or chemical origin. These qualities make the system highly attractive for wastewater treatment.

5. Conclusions

The performance of the electrocoagulation (EC) process combined with adsorption onto activated carbon (Ads) was investigated, leading to the following conclusions:

• FTIR analysis confirmed the successful adsorption of ketoprofen (KTP) onto activated carbon without any significant modification of the adsorbent surface due to the electrocoagulation process.
• The combined EC/Ads approach demonstrated superior efficiency compared to the individual processes. A synergy factor of 1.37 was calculated, highlighting a significant interactive effect between electrocoagulation and adsorption.
• A comparative study of cathode types revealed that the iron grid cathode outperformed the steel plate cathode in terms of process efficiency.
• The electrocoagulation (EC) process can be significantly improved by integrating an adsorption mechanism, particularly through a configuration in which activated carbon is immobilized within the anode. This combined EC/Ads-ACT approach not only enhances overall pollutant removal efficiency but also allows for the reuse of activated carbon, thereby reducing operational costs and streamlining the treatment process.
• The model derived from the central composite design (CCD) optimization identified the following optimal conditions for complete ketoprofen removal (100%): an activated carbon dose (q_AC) between 0.63 and 0.99 g/500 mL, a current density (i) ranging from 12.32 to 14.68 mA·cm⁻², a pH of 6.5 and an initial KTP concentration of 22.5 mg·L⁻¹.
• The statistical analysis of the model revealed a strong interaction between current density and activated carbon dosage, confirming their combined influence on removal efficiency.

In conclusion, this process is characterized by its operational simplicity and its effectiveness in removing organic contaminants, giving the system strong potential for wastewater treatment. Its flexibility allows for adaptation to various types of pollutants, enhancing its relevance for diverse applications in effluent treatment.