ABSTRACT
Electrocoagulation has several disadvantages, such as electrode passivation, generation of heat due to energy consumption, and significant sludge formation. The constraints restrict its application in the treatment of tannery effluent. Therefore, the introduction of alternating pulse current electrocoagulation (APC-EC) aimed to resolve these concerns. We conducted an empirical examination of the research, specifically focusing on the frequency and stirring speed. This study investigated the impact of different parameters on the effectiveness of removing chemical oxygen demand (COD), the turbidity of trivalent chromium, and the energy consumption of perforated aluminium (Al) electrodes. The implementation of the central composite design (CCD) within the surface response design technique has enhanced multiple operational parameters in the APC-EC process for the treatment of tannery effluent. By employing our advanced mathematical and statistical techniques, we successfully eliminated the highest levels of COD, chromium (III) ion, and turbidity, all while significantly reducing energy usage. As a result, we achieved the optimal conditions for our process. The components that were eliminated most rapidly at 11000 Hz, 576 rpm, and 30 minutes were COD (70.3%), turbidity (96%), and 89.56%. The surface response results provide a description of the frequency dynamics of APC-EC.
HIGHLIGHTS
Innovative alternating pulse current (APC-EC) solution surpasses direct current EC limits in tannery wastewater treatment.
Optimization via central composite design enhances frequency, stirring speed, and reaction time.
Notable chemical oxygen demand and Cr+3 removal lowered energy use with inventive electrode approach.
Research showcases APC-EC's superior efficiency and sustainability.
APC-EC introduces a breakthrough method for efficient tannery effluent treatment.
INTRODUCTION
The leather sector has a prominent place in the world economy. Within the domain of the leather industry, the raw hides obtained from the animal sector undergo a sequence of mechanical and chemical alterations (Costa et al. 2008; Benhadji et al. 2011; Elabbas et al. 2016). In this context, raw hides, which are a perishable byproduct of the cattle sector, undergo a process of conversion into goods that exhibit both durability and economic value. However, within the context of leather manufacture, a wide range of waste is produced as a result of the transformative process, appearing as solid, liquid, and gaseous byproducts throughout the manufacturing process. The resulting byproducts display variation depending on the characteristics of the products, the methods used, and the materials used as inputs. Undoubtedly, the wastewater generated because of numerous manufacturing operations is a significant challenge within the leather sector. The correlation between the amount of wastewater produced and the volume of water used inside a given facility is commonly recognized. The use of different chemical compounds during all stages of processing, including cleaning and preserving untreated animal skins, results in the generation of wastewater with varying compositions in terms of quality and quantity (Feng et al. 2007; Isarain-Chávez et al. 2014; Deghles & Kurt 2016). The characteristics of wastewater generated by the leather industry vary depending on the specific technology applied. The watery effluents that arise from the process of immersion and conditioning consist of many substances, including sodium chloride, hemoglobin, feces, earth matter, epidermal remnants, colloidal particles, and calcium chloride (Cassano et al. 2001; Kumar & Mani 2007; Sekar et al. 2009). The effluents produced as a consequence of the fleshing process include significant amounts of oil-grease, suspended particles, and chemical oxygen demand (COD). The effluent that is produced as a consequence of the unhairing and liming process shows increased concentrations of lime, sulfur, organic compounds, and both dissolved and suspended particulate matter (Bajza & Vrcek 2001; Abreu & Toffoli 2009; Hammami et al. 2009). The lime wash and bating wastewaters demonstrate a substantial concentration of ammonia. The effluents generated during the bating process also include the remaining enzymes used. In the absence of recycling, the degreasing process generates wastewaters that demonstrate significant levels of oil content, alkaline potassium permanganate demand, and COD. The use of sodium sulfide in degreasing processes is widespread, and it is important to highlight that the effluents produced during these processes might include different levels of sulfur (Font et al. 1996). The solution that is formed after the pickling procedure shows a notable level of solutes, which is evident from the existence of dissolved solids (Kanth et al. 2009). The chemical composition of wastewater generated by tanneries varies depending on the exact kind of substrate used in the tanning process. Currently, chromium (Shaalan et al. 2001; Fahim et al. 2006) is the primary agent used in tanneries. The complex nature of this specific wastewater composition presents a potential risk to both human well-being and environmental integrity upon its introduction into ecosystems. The primary cause of this phenomenon may be linked to the potential for chromium (III) to undergo oxidation to chromium (VI) or the presence of azo dyes (Dixit et al. 2015). However, it has been demonstrated in mammalian in vitro investigations that trivalent chromium demonstrates potential toxicity as a result of its function as a competitive inhibitor in many cellular processes (Walsh et al. 1994). In this specific context, the conventional biological treatment method often fails to achieve a satisfactory level of effectiveness due to the harmful characteristics of tannery effluent, which negatively affects the growth and reproduction of bacteria. Furthermore, traditional physicochemical techniques are characterized by a comparatively elevated cost and possess the capacity to produce secondary contaminants. The inclusion of additional chemical constituents is necessitated by the stipulated criteria (Pandey & Thakur 2020).
The restrictions noted above have motivated several businesses to investigate effective alternative treatment methods for the removal of contaminants, with a preference for electrochemical approaches. In contemporary times, electrochemical treatment approaches have reached a state of progress whereby they demonstrate not only cost comparability with other technologies but also higher efficiency and effectiveness (Chen 2004). In the present day, there has been much research conducted on electrochemical technologies in the field of environmental applications, specifically on the treatment of water and wastewater in industrialized countries. One of the procedures used for wastewater treatment is electrocoagulation (EC) due to its inherent benefits, such as its simplicity, dependability, and cost-effectiveness in operation. EC has garnered significant attention owing to its potential applications in the leather industry (Deghles & Kurt 2016; Garcia-Segura et al. 2017). The EC process (Chen 2004; Jiménez et al. 2012) involves several important phenomena. At the interfaces of the electrodes, electrochemical reactions occur. Within the aqueous medium, coagulants are produced. These coagulants then adsorb soluble or colloidal pollutants. Finally, the pollutants are removed through sedimentation and flotation mechanisms. Equations (1) and (2) illustrate the production of ions and oxygen by anodic processes.
Due to the pH value being approximately 4, aluminum (Al) does not remain in cationic form. The cations formed react with water to form complexes of the type ,
, or the poorly soluble hydroxide
.
The metal cations, specifically aluminum ions , undergo sacrificial oxidation in the aqueous medium. This process facilitates the removal of undesired impurities, such as pollutants, through either chemical reactions leading to precipitation or by inducing the fusion of colloidal materials (flocs). Subsequently, these fused materials can be eliminated via electrolytic flotation with the assistance of hydrogen gas
generated during the cathodic reaction.
Nevertheless, in the process of EC utilizing direct current (DC), it is observed that following a specific duration, a non-permeable oxide layer is generated on the anodic surface. This hinders the efficient propagation of electrical currents within the system, resulting in elevated power expenditures and diminished efficacy. This issue is exacerbated when utilizing an aluminum electrode. In a broader context, the augmented utilization of the electrode and electrical energy, along with the escalated generation of sludge, are prominent drawbacks associated with this approach. In the realm of EC, it is customary for power consumption expenditures to surpass 50% when considering the overall expenses of the unit (Mollah et al. 2001). Hence, the enhancement of the electrode processes employed in EC is of utmost importance. While the antecedent electrochemical-based treatability investigations have made significant contributions to this domain, the majority of them solely focus on elucidating the impact of prevailing process conditions on the elimination of a particular contaminant from diverse wastewater sources. To the utmost extent of the authors' understanding, there exists a conspicuous void in the pertinent domain pertaining to the exploration of groundbreaking and economically viable EC setups aimed at mitigating the drawbacks linked to energy usage and waste management of sludge. Based on the available literature, it appears that there is a scarcity of comprehensive publications focused on the investigation of a comparative optimization of tanning wastewater treatment using response surface methodology (RSM). In this context, the current investigation was used to address the gap mentioned previously by focusing on a detailed quantitative analysis of the alternating pulse current EC (APC-EC) process within the experimental domain of the pivotal variables impacting this process, such as frequency, stirring speed, and treatment time, with the aim of achieving minimal energy utilization and maximal removal. RSM was employed for the optimization and analysis of tannery wastewater treatment performance.
Considering the aforementioned evidence, the current investigation introduces a pioneering technique known as APC-EC as a means to address the drawbacks associated with DC-EC. The application of an alternating pulse circuit at the output of the DC source enables the conversion of the DC into a pulsating, alternating current. Utilizing the prevailing supply model at specific temporal intervals, the direction of electric current is altered, thereby inducing a corresponding modification in the functional role of the electrodes. The pulsating current and alternating electrode role in this process serve the dual purpose of preventing anode electrode passivation and allowing ample time for coagulants to fully engage in the coagulation reaction during the off-time intervals between each pulse. Consequently, this results in a decrease in energy consumption (Koparal et al. 2008). The utilization of electrical currents is observed in conjunction with square, sinusoidal, and triangular waveforms during the process of pulsed electrolysis. Nevertheless, it is worth noting that pulsed currents featuring square wave forms, both symmetrically and asymmetrically, are more extensively employed in the realm of electrolysis (Sahay & Kushwaha 2017).
The study aimed to improve tannery wastewater treatment by introducing APC-EC to overcome issues seen in conventional DC-EC. These issues include electrode passivation, high energy consumption, heat generation, and excessive sludge. Experimental analysis focused on factors like frequency, stirring speed, and reaction time. The study aimed to optimize the APC-EC process through statistical methods, achieving maximum removal of COD, and turbidity, while minimizing energy use.
METHODS
Characterization of tannery wastewater
The wastewater produced by a tannery facility located in Damascus, Syria, was collected and afterward subjected to several analyses to determine its properties. These analyses included measuring the pH level, chemical COD, concentration of Cr+3, total dissolved solids (TDS), and turbidity. The specific findings obtained from these analyses are shown in Table 1.
Characteristics of wastewater produced by tannery plant
No . | Parameters . | Value . | Unit . |
---|---|---|---|
1 | pH | 3.85 | – |
2 | COD | 8,970 | mg/L |
3 | ![]() | 190 | mg/L |
4 | TDS | 18,500 | mg/L |
5 | Turbidity | 142 | NTU |
No . | Parameters . | Value . | Unit . |
---|---|---|---|
1 | pH | 3.85 | – |
2 | COD | 8,970 | mg/L |
3 | ![]() | 190 | mg/L |
4 | TDS | 18,500 | mg/L |
5 | Turbidity | 142 | NTU |
Reagents
This research used a range of reagents to determine the COD. Potassium dichromate (K2Cr2O₇) was used as the oxidizing agent because of its strong oxidizing characteristics, which are necessary for the decomposition of the organic components in the sample. Sulfuric acid (H2SO4) and hydrochloric acid (HCl) were used to modify the pH levels and improve the reaction conditions for the oxidation process. Acetone was essential in purifying the electrodes, guaranteeing precise readings by eliminating any possible impurities. The conversion of organic molecules into carbon dioxide (CO2) and water (H2O), which is a crucial step in properly measuring COD levels, was facilitated by the use of silver sulfate (AgSO4) and mercury(II) sulfate (HgSO4) as catalysts. The back-titration procedure required the use of ferrous ammonium sulfate as the titrant to accurately measure the quantity of the oxidizing agent that has undergone a reaction. The selection of ferroin as the indicator is based on its distinct color transition at the titration endpoint, which greatly aids in determining the endpoint of the titration. Pure distilled water was used in all studies to guarantee the integrity of the solutions and the dependability of the outcomes.
Equipment used in the experimental tests
In the above equation, Φ represents the current efficiency expressed as a percentage, whereas and
denote the actual and theoretical dissolving mass of the electrode in grams, respectively.
The comparison facilitates the measurement of the effectiveness of operations such as the elimination of pollutants or the decrease in COD in the field of environmental management. The equipment used includes the HACH Model 2100P pH meter from the United States, the Jenway Model 3,540 conductivity meter from England, the INGCO Model DM7502 digital multimeter from China, and a manually made pulse supply.
Experimental procedures
The EC reactor used in this research was constructed using acrylic material and has the following dimensions: 10 cm length, 15 cm width, and 20 cm height. Two planar sheets with rectangular shapes were used as the cathode and anode electrodes. The dimensions of each sheet were 9.5 cm in width, 10 cm in length, and 0.2 cm in thickness. These sheets were purchased from China and had a purity level of 99.5%. The sheets were perforated with circular holes, each measuring 0.2 cm in diameter, and were made of aluminum. The electrodes were arranged in a vertical configuration, with a gap of 1 centimeter between them. The objects were situated at a vertical distance of 0.5 centimeters from the lowermost part of the cell. The electrodes were connected to the pulser, which was in turn connected to a DC electric generator. The DC was injected into the pulser, where it underwent a conversion process to create an output current. This output current was then delivered to the electrodes. The selection of a duty cycle of 50% in this experimental approach was informed by previous scientific studies (Martiningsih et al. 2017; Zhang et al. 2021). The slurry for the coagulation procedure was created using a magnetic stirrer. During the experimental protocol, the electrodes were exposed to a potential difference, while the ensuing currents were measured using a voltmeter and an ammeter. The aforementioned measurements played a pivotal role in validating the attainment of the intended electrical current and guaranteeing the effective implementation of the procedure. During the first stage, the electrodes underwent abrasion using sandpaper to eliminate any adsorbed contaminants. Subsequently, the electrodes were washed by immersing them in a hydrochloric acid solution with a concentration of 10%. This was followed by treatment with acetone and distilled water, and eventually, the electrodes were dried by desiccation at a temperature of 100 °C. The weights of the plates made of aluminum were measured both before and after each electrochemical experiment.
Costs and efficiencies estimation
An electrolytic cell is linked to a pulse power supply to create an electrolysis system, where the process of electrolysis occurs in pulses. The duration of the current from the on mode to the off mode is referred to as , which represents the pulse duration or pulse width. This is the period of time during which electrolysis is actively taking place. The duration ‘t’ between the transition from the off mode to the on mode is referred to as the intermittent period of electrolysis or pulse interval. The frequency pulse EC technique utilizes a specific power source that generates three distinct parameters: pulse power or current amplitude, pulse width
, and pulse interval
. By adjusting these parameters, one can achieve enhanced decontamination and energy-saving effects.
The variable denotes the theoretical electrode consumption in
. The variable t represents the duration of the treatment in seconds.






Experimental setup for APC-EC: (a) schematic; (b) actual: (1) DC power supply with pulse supply; (2) duty cycle control; (3) frequency control; (4) wire; (5) magnetic stirrer; (6) heat control; (7) speed control; (8) perforated aluminum electrodes; (9) EC cell; (10) tap; (11) tannery wastewater; (12) digital multimeter.
Experimental setup for APC-EC: (a) schematic; (b) actual: (1) DC power supply with pulse supply; (2) duty cycle control; (3) frequency control; (4) wire; (5) magnetic stirrer; (6) heat control; (7) speed control; (8) perforated aluminum electrodes; (9) EC cell; (10) tap; (11) tannery wastewater; (12) digital multimeter.
Response surface methodology
The field of design of experiment (DoE) comprises a wide range of mathematical approaches that are very beneficial for statistical modeling and systematic analysis of a given subject matter. The optimization of intended responses or measures of output is accomplished by using variables or factors (Asaithambi et al. 2016; Ghelich et al. 2019). RSM is a frequently used approach in the field of DoE for the construction of models. It serves as a sequential approach that effectively enhances the design and formulation of innovative items, ultimately increasing their overall performance (Asaithambi et al. 2012; Ghelich et al. 2019). Furthermore, the RSM is widely used as an empirical statistical technique for the purpose of constructing mathematical models, optimizing trials involving several factors, and examining the relationships between the response variable and the explanatory variables (Pavlović et al. 2014). The RSM offers a significant benefit compared to the usual strategy of analyzing one variable at a time due to its capacity to reduce the required number of experimental operations, thereby saving time. The achievement of RSM is facilitated via the simultaneous interaction of factors and the modeling of specific response parameters. As a result, the use of RSM facilitates a more expeditious and methodical investigation of the many factors implicated (Ghelich et al. 2019). The present study used the RSM with the aid of the Design-Expert program, which was created by Stat-Ease Inc. (Pavlović et al. 2014). This facilitated the development of a pragmatic and efficient association between the selected replies and the control variables. The experimental methodologies were devised in line with the concepts of the catalytic chemical decomposition (CCD). The use of RSM was implemented in a series of five consecutive steps described below (Bezerra et al. 2008):
1. The process involves the identification of several elements that have an impact on the intended results, as well as the determination of the minimum and maximum values associated with each of these factors. The pH and conductivity reader measures the pH values that result from electrode dissociation, without the addition of any chemicals to manipulate the pH. The characteristics that were chosen to impact the effectiveness of APC-EC include frequency, stirring speed, and response time. The lower and higher limits of the variables were established based on the results of our first tests conducted on parameters such as COD, turbidity,
removal effectiveness, final pH, consumption of energy, and electrode usage. The characteristics have substantial economic significance within electrochemical processes and, as a result, were included as the variables that are dependent (responses).
- 2. The determination of the DoE and the analysis of the experimental trials are influenced by the design approach. The second step of the RSM involves the careful development of an experimental design model. To achieve this objective, it is possible to use a model that incorporates both first-order and second-order kinetics. As explicated in the scholarly literature, researchers have often shown a preference for the utilization of second-order symmetrical designs, such as the three-level factorial design (Missaoui et al. 2013), the central composite layout (Gengec et al. 2013; Tahir 2019), and the Box–Behnken design (Mehralian et al. 2021). Variability in the selection of experimental sites, the number of levels for factors, and the number of runs is seen across different design models. The selected approach for this research was the CCD, which included lower and higher levels for each variable inside Design-Expert program. The variables were analyzed over five levels, which were designated as -α, −1, 0, +1, and +α. The determination of the α level for each variable is dependent on the number of variables (k), and may be achieved by using Equation (12) (Bezerra et al. 2008).
- 3. The software was used to compute the α levels. Table 2 presents the predetermined values for the experimental design of the three components, namely, frequency, stirring speed, and treatment duration. The program calculated the total number of experiments using a particular formula.where k is the number of variables and cp corresponds to the number of the central points (Bezerra et al. 2008; Pandey & Thakur 2020). The CCD for three experimental variables comprised 20 experiments. Each experiment was conducted in the sequence listed in Table 3 and performed only once. By repeating the experiments six times, the results can achieve better precision due to averaging the inherent variability in the measurements.
- 4. The statistical and mathematical evaluation of experimental data involves the use of polynomial functions to accurately represent and understand the acquired results. A series of 20 tests were conducted based on the specified circumstances outlined in Tables 3 and 4. DoE software recorded the values for the proportion of COD,
, turbidity, final pH, and energy and electrode usage for each experiment. The experimental data were analyzed using the Design-Expert software (commercial package) and fitted to a second-order polynomial model. Subsequently, regression coefficients were derived from the analysis. The response variable (chromium ions removal %, turbidity reduction %, COD removal %, and energy consumption), which is denoted by (y), was used to assess the second-order polynomial equation, which represents the relationship between the chosen variables (x) and their interactions (independent variables multiplication in regression equation and quadratic terms), as described by Equation (14) (Mondal et al. 2012; Assémian et al. 2018). The assessment of model adequacy was conducted using analysis of variance (ANOVA). The regression coefficients
and
were used to ascertain the most suitable polynomial model, while the F-test was utilized to examine the statistical significance. The statistical regression of full quadratic modes is mentioned below:where β0, βi, βii, and βij are the regression coefficients for intercept, linear, quadratic, and interaction terms, respectively.
5. The optimal levels of each variable are achieved to obtain the appropriate levels of replies. In the field of numerical optimization, the selection process involves strategically choosing desired objectives for each component, along with corresponding responses, from the various options available within the software interface. Following this, the computer seeks to identify the optimal configurations that maximize the desirability function (Pavlović et al. 2014). The optimal objective for each variable was selected from the range specified in Tables 3 and 4. There are several response choices including maximizing removal efficiency and minimizing energy consumption (Khataee et al. 2010).
Experimental design levels for frequency, stirring speed, and reaction time
Predicted and actual variable . | |||||||
---|---|---|---|---|---|---|---|
Variable . | Unit . | Factors . | α-low . | Low . | Center . | High . | α-high . |
−1.68 | −1 | 0 | +1 | +1.68 | |||
Frequency | Hz | ![]() | 272.8 | 3,000 | 7,000 | 11,000 | 13,727.2 |
Stirring speed | rpm | ![]() | 127.2 | 400 | 800 | 1,200 | 1,472.7 |
Reaction time | min | ![]() | 9.8 | 15 | 22.5 | 30 | 35.1 |
Predicted and actual variable . | |||||||
---|---|---|---|---|---|---|---|
Variable . | Unit . | Factors . | α-low . | Low . | Center . | High . | α-high . |
−1.68 | −1 | 0 | +1 | +1.68 | |||
Frequency | Hz | ![]() | 272.8 | 3,000 | 7,000 | 11,000 | 13,727.2 |
Stirring speed | rpm | ![]() | 127.2 | 400 | 800 | 1,200 | 1,472.7 |
Reaction time | min | ![]() | 9.8 | 15 | 22.5 | 30 | 35.1 |
Comparison of actual and predicted removal percentages
Run . | Frequency (Hz) . | Stirring speed (rpm) . | Reaction time (min) . | COD removal (%) . | ![]() . | Turbidity removal (%) . | |||
---|---|---|---|---|---|---|---|---|---|
Actual value . | Predicted value . | Actual value . | Predicted value . | Actual value . | Predicted value . | ||||
1 | 7,000 | 800 | 22.5 | 58.6 | 57.6 | 69.1 | 68.5 | 75.5 | 76.0 |
2 | 3,000 | 400 | 30 | 63.2 | 64.4 | 87.0 | 87.3 | 85.0 | 82.1 |
3 | 11,000 | 400 | 30 | 65.0 | 64.4 | 90.8 | 90.5 | 94.0 | 93.2 |
4 | 3,000 | 1,200 | 30 | 65.8 | 64.9 | 95.0 | 92.3 | 93.0 | 92.5 |
5 | 7,000 | 800 | 22.5 | 57.4 | 57.6 | 67.8 | 68.5 | 76.2 | 76.0 |
6 | 13,727.2 | 800 | 22.5 | 50.2 | 50.4 | 71.7 | 72.3 | 80.0 | 80.2 |
7 | 11,000 | 1,200 | 15 | 36.1 | 35.6 | 46.1 | 45.7 | 47.0 | 49.7 |
8 | 7,000 | 1,472.7 | 22.5 | 52.8 | 53.4 | 61.0 | 64.6 | 70.0 | 69.7 |
9 | 272.8 | 800 | 22.5 | 53.6 | 52.4 | 66.0 | 65.2 | 71.0 | 71.4 |
10 | 3,000 | 1,200 | 15 | 36.7 | 38.0 | 40.0 | 40.4 | 50.0 | 50.5 |
11 | 3,000 | 400 | 15 | 34.2 | 35.2 | 24.1 | 27.5 | 32.0 | 34.1 |
12 | 7,000 | 800 | 22.5 | 57.9 | 57.4 | 68.0 | 68.7 | 76.7 | 76.5 |
13 | 7,000 | 800 | 22.5 | 57.1 | 57.1 | 69.8 | 68.5 | 74.8 | 76.0 |
14 | 7,000 | 800 | 22.5 | 57.2 | 57.6 | 65.6 | 68.2 | 75.9 | 76.4 |
15 | 7,000 | 800 | 9.8 | 25.0 | 23.4 | 24.6 | 20.9 | 30.0 | 26.6 |
16 | 7,000 | 800 | 35.1 | 70.0 | 70.7 | 95.0 | 98.6 | 95.0 | 98.8 |
17 | 11,000 | 1,200 | 30 | 63.2 | 62.8 | 81.7 | 78.4 | 90.0 | 87.6 |
18 | 11,000 | 400 | 15 | 33.5 | 35.0 | 47.0 | 49.8 | 49.0 | 49.3 |
19 | 7,000 | 800 | 22.5 | 58.0 | 57.6 | 70.6 | 68.5 | 77.0 | 76.0 |
20 | 7,000 | 127.2 | 22.5 | 54.0 | 52.4 | 67.5 | 63.9 | 60.0 | 60.7 |
Run . | Frequency (Hz) . | Stirring speed (rpm) . | Reaction time (min) . | COD removal (%) . | ![]() . | Turbidity removal (%) . | |||
---|---|---|---|---|---|---|---|---|---|
Actual value . | Predicted value . | Actual value . | Predicted value . | Actual value . | Predicted value . | ||||
1 | 7,000 | 800 | 22.5 | 58.6 | 57.6 | 69.1 | 68.5 | 75.5 | 76.0 |
2 | 3,000 | 400 | 30 | 63.2 | 64.4 | 87.0 | 87.3 | 85.0 | 82.1 |
3 | 11,000 | 400 | 30 | 65.0 | 64.4 | 90.8 | 90.5 | 94.0 | 93.2 |
4 | 3,000 | 1,200 | 30 | 65.8 | 64.9 | 95.0 | 92.3 | 93.0 | 92.5 |
5 | 7,000 | 800 | 22.5 | 57.4 | 57.6 | 67.8 | 68.5 | 76.2 | 76.0 |
6 | 13,727.2 | 800 | 22.5 | 50.2 | 50.4 | 71.7 | 72.3 | 80.0 | 80.2 |
7 | 11,000 | 1,200 | 15 | 36.1 | 35.6 | 46.1 | 45.7 | 47.0 | 49.7 |
8 | 7,000 | 1,472.7 | 22.5 | 52.8 | 53.4 | 61.0 | 64.6 | 70.0 | 69.7 |
9 | 272.8 | 800 | 22.5 | 53.6 | 52.4 | 66.0 | 65.2 | 71.0 | 71.4 |
10 | 3,000 | 1,200 | 15 | 36.7 | 38.0 | 40.0 | 40.4 | 50.0 | 50.5 |
11 | 3,000 | 400 | 15 | 34.2 | 35.2 | 24.1 | 27.5 | 32.0 | 34.1 |
12 | 7,000 | 800 | 22.5 | 57.9 | 57.4 | 68.0 | 68.7 | 76.7 | 76.5 |
13 | 7,000 | 800 | 22.5 | 57.1 | 57.1 | 69.8 | 68.5 | 74.8 | 76.0 |
14 | 7,000 | 800 | 22.5 | 57.2 | 57.6 | 65.6 | 68.2 | 75.9 | 76.4 |
15 | 7,000 | 800 | 9.8 | 25.0 | 23.4 | 24.6 | 20.9 | 30.0 | 26.6 |
16 | 7,000 | 800 | 35.1 | 70.0 | 70.7 | 95.0 | 98.6 | 95.0 | 98.8 |
17 | 11,000 | 1,200 | 30 | 63.2 | 62.8 | 81.7 | 78.4 | 90.0 | 87.6 |
18 | 11,000 | 400 | 15 | 33.5 | 35.0 | 47.0 | 49.8 | 49.0 | 49.3 |
19 | 7,000 | 800 | 22.5 | 58.0 | 57.6 | 70.6 | 68.5 | 77.0 | 76.0 |
20 | 7,000 | 127.2 | 22.5 | 54.0 | 52.4 | 67.5 | 63.9 | 60.0 | 60.7 |
Experimental and predicted results of electrochemical treatment parameters and their effects on process efficiency
Run . | Frequency (Hz) . | Stirring speed (rpm) . | Reaction time (min) . | Final ![]() . | Current density ![]() . | Electrodes consumption (![]() . | Energy consumption ![]() . | ||||
---|---|---|---|---|---|---|---|---|---|---|---|
Actual value . | Predicted value . | Actual value . | Predicted value . | Actual value . | Predicted value . | Actual value . | Predicted value . | ||||
1 | 7,000 | 800 | 22.5 | 5.4 | 5.3 | 21.0 | 20.4 | 0.3 | 0.3 | 2.9 | 3.0 |
2 | 3,000 | 400 | 30 | 5.4 | 5.3 | 24.5 | 24.3 | 0.5 | 0.5 | 5.2 | 5.0 |
3 | 11,000 | 400 | 30 | 5.5 | 5.4 | 14.7 | 16.5 | 0.3 | 0.4 | 3.1 | 3.0 |
4 | 3,000 | 1,200 | 30 | 5.6 | 5.5 | 24.5 | 24.3 | 0.5 | 0.5 | 5.2 | 5.0 |
5 | 7,000 | 800 | 22.5 | 5.3 | 5.3 | 21.0 | 20.4 | 0.3 | 0.3 | 2.9 | 3.0 |
6 | 13,727.2 | 800 | 22.5 | 5.2 | 5.2 | 16.7 | 13.8 | 0.3 | 0.2 | 2.0 | 1.8 |
7 | 11,000 | 1,200 | 15 | 4.9 | 4.9 | 14.7 | 16.5 | 0.2 | 0.1 | 1.5 | 1.6 |
8 | 7,000 | 1,472.7 | 22.5 | 5.3 | 5.3 | 21.0 | 20.4 | 0.3 | 0.3 | 2.9 | 3.0 |
9 | 272.8 | 800 | 22.5 | 5.1 | 5.1 | 25.0 | 26.9 | 0.4 | 0.4 | 4.1 | 4.2 |
10 | 3,000 | 1,200 | 15 | 4.8 | 4.8 | 24.5 | 24.3 | 0.3 | 0.3 | 2.6 | 2.5 |
11 | 3,000 | 400 | 15 | 4.7 | 4.7 | 24.5 | 24.3 | 0.3 | 0.3 | 2.6 | 2.5 |
12 | 7,000 | 800 | 22.5 | 5.3 | 5.3 | 21.0 | 20.4 | 0.3 | 0.3 | 2.9 | 3.0 |
13 | 7,000 | 800 | 22.5 | 5.3 | 5.3 | 21.0 | 20.4 | 0.3 | 0.3 | 2.9 | 3.0 |
14 | 7,000 | 800 | 22.5 | 5.4 | 5.3 | 21.0 | 20.4 | 0.3 | 0.3 | 2.9 | 3.0 |
15 | 7,000 | 800 | 9.8 | 4.5 | 4.5 | 21.0 | 20.4 | 0.1 | 0.1 | 1.3 | 1.3 |
16 | 7,000 | 800 | 35.1 | 5.7 | 5.7 | 21.0 | 20.4 | 0.5 | 0.5 | 4.6 | 4.7 |
17 | 11,000 | 1,200 | 30 | 5.7 | 5.6 | 14.7 | 16.5 | 0.3 | 0.4 | 3.1 | 3.0 |
18 | 11,000 | 400 | 15 | 4.8 | 4.8 | 14.7 | 16.5 | 0.1 | 0.1 | 1.5 | 1.6 |
19 | 7,000 | 800 | 22.5 | 5.3 | 5.3 | 21.0 | 20.4 | 0.3 | 0.3 | 2.9 | 3.0 |
20 | 7,000 | 127.2 | 22.5 | 5.0 | 5.0 | 21.0 | 20.4 | 0.3 | 0.3 | 2.9 | 3.0 |
Run . | Frequency (Hz) . | Stirring speed (rpm) . | Reaction time (min) . | Final ![]() . | Current density ![]() . | Electrodes consumption (![]() . | Energy consumption ![]() . | ||||
---|---|---|---|---|---|---|---|---|---|---|---|
Actual value . | Predicted value . | Actual value . | Predicted value . | Actual value . | Predicted value . | Actual value . | Predicted value . | ||||
1 | 7,000 | 800 | 22.5 | 5.4 | 5.3 | 21.0 | 20.4 | 0.3 | 0.3 | 2.9 | 3.0 |
2 | 3,000 | 400 | 30 | 5.4 | 5.3 | 24.5 | 24.3 | 0.5 | 0.5 | 5.2 | 5.0 |
3 | 11,000 | 400 | 30 | 5.5 | 5.4 | 14.7 | 16.5 | 0.3 | 0.4 | 3.1 | 3.0 |
4 | 3,000 | 1,200 | 30 | 5.6 | 5.5 | 24.5 | 24.3 | 0.5 | 0.5 | 5.2 | 5.0 |
5 | 7,000 | 800 | 22.5 | 5.3 | 5.3 | 21.0 | 20.4 | 0.3 | 0.3 | 2.9 | 3.0 |
6 | 13,727.2 | 800 | 22.5 | 5.2 | 5.2 | 16.7 | 13.8 | 0.3 | 0.2 | 2.0 | 1.8 |
7 | 11,000 | 1,200 | 15 | 4.9 | 4.9 | 14.7 | 16.5 | 0.2 | 0.1 | 1.5 | 1.6 |
8 | 7,000 | 1,472.7 | 22.5 | 5.3 | 5.3 | 21.0 | 20.4 | 0.3 | 0.3 | 2.9 | 3.0 |
9 | 272.8 | 800 | 22.5 | 5.1 | 5.1 | 25.0 | 26.9 | 0.4 | 0.4 | 4.1 | 4.2 |
10 | 3,000 | 1,200 | 15 | 4.8 | 4.8 | 24.5 | 24.3 | 0.3 | 0.3 | 2.6 | 2.5 |
11 | 3,000 | 400 | 15 | 4.7 | 4.7 | 24.5 | 24.3 | 0.3 | 0.3 | 2.6 | 2.5 |
12 | 7,000 | 800 | 22.5 | 5.3 | 5.3 | 21.0 | 20.4 | 0.3 | 0.3 | 2.9 | 3.0 |
13 | 7,000 | 800 | 22.5 | 5.3 | 5.3 | 21.0 | 20.4 | 0.3 | 0.3 | 2.9 | 3.0 |
14 | 7,000 | 800 | 22.5 | 5.4 | 5.3 | 21.0 | 20.4 | 0.3 | 0.3 | 2.9 | 3.0 |
15 | 7,000 | 800 | 9.8 | 4.5 | 4.5 | 21.0 | 20.4 | 0.1 | 0.1 | 1.3 | 1.3 |
16 | 7,000 | 800 | 35.1 | 5.7 | 5.7 | 21.0 | 20.4 | 0.5 | 0.5 | 4.6 | 4.7 |
17 | 11,000 | 1,200 | 30 | 5.7 | 5.6 | 14.7 | 16.5 | 0.3 | 0.4 | 3.1 | 3.0 |
18 | 11,000 | 400 | 15 | 4.8 | 4.8 | 14.7 | 16.5 | 0.1 | 0.1 | 1.5 | 1.6 |
19 | 7,000 | 800 | 22.5 | 5.3 | 5.3 | 21.0 | 20.4 | 0.3 | 0.3 | 2.9 | 3.0 |
20 | 7,000 | 127.2 | 22.5 | 5.0 | 5.0 | 21.0 | 20.4 | 0.3 | 0.3 | 2.9 | 3.0 |
Removal analysis






RESULTS AND DISCUSSION
Removal efficiency of COD, Cr+3, turbidity, energy consumption and electrodes, and final pH
Tables 3 and 4 provide the removal efficiency of COD, Cr+3, turbidity reduction, and the energy consumption using Al–Al perforated electrodes. The results in these tables also include the corresponding projected values obtained by RSM, along with the final pH and current density. Adjusted coefficient of determination (R2) is a modified version of R2 that has been adjusted for the number of predictors in a regression model. It is used to determine how well a regression model fits the data while taking into account the number of predictors in the model. Unlike R2, which can only increase or stay the same as more predictors are added,
can decrease if the addition of a predictor does not improve the model enough to offset the penalty for adding another variable.
Table 3 presents a comprehensive comparison between the actual and expected removal percentages for COD, Cr3+ (chromium), and turbidity in several experimental trials. The results demonstrate a robust association between the observed and projected values, emphasizing the model's precision in predicting treatment effectiveness. Modifications in frequency, agitation velocity, and duration of the reaction had a substantial effect on the rates of removal, demonstrating patterns that align with theoretical predictions. For example, when the frequencies and stirring speeds were raised, there was generally a noticeable improvement in the removal of pollutants. This was evident from the greater percentages of COD and Cr3+ removal, as well as the decreased turbidity. The most significant improvements were reported in experiments with extended reaction durations, particularly in run 16, which achieved a maximum of 70% COD reduction and over 95% removal of chromium ions (Cr3+). Nevertheless, the results also reveal a phenomenon of diminishing returns in certain instances, when an increase in reaction speed or frequency did not lead to significantly improved outcomes. This indicates that optimizing these parameters is essential for maximizing the effectiveness of pollution removal without needless resource consumption.
The comparison between actual and predicted values for various responses: (a) COD removal %, (b) removal %, (c) turbidity reduction %, and (d) energy consumption.
The comparison between actual and predicted values for various responses: (a) COD removal %, (b) removal %, (c) turbidity reduction %, and (d) energy consumption.
Percentage removal: (a) COD, (b) , (c) turbidity, and (d) final pH using a combination of frequency and reaction time (stirring speed: 800 rpm).
Percentage removal: (a) COD, (b) , (c) turbidity, and (d) final pH using a combination of frequency and reaction time (stirring speed: 800 rpm).
Percentage removal: (a) COD, (b) , (c) turbidity, and (d) final pH using a combination of frequency and stirring speed (reaction time: 30 min).
Percentage removal: (a) COD, (b) , (c) turbidity, and (d) final pH using a combination of frequency and stirring speed (reaction time: 30 min).
Surface response plot of the effect of frequency and reaction time on the energy consumption (stirring speed: 800 rpm).
Surface response plot of the effect of frequency and reaction time on the energy consumption (stirring speed: 800 rpm).
Table 4 demonstrates the impact of different electrochemical treatment parameters on the effectiveness of the pollutant removal procedure. The characteristics encompass stirring velocity, frequency, duration of reaction, ultimate pH, current density, electrode usage, and energy expenditure. The stirring speeds ranged between 127.2 and 1,472.7 revolutions per minute (rpm), while the frequency varied from 272.8 to 13,727.2 Hz. Increasing the stirring speeds often enhances the development of flocs by assuring even distribution of ions, which facilitates the aggregation of particles into larger flocs that can be easily removed. The reaction times observed in the studies varied between 9.8 and 35.1 min. Extended reaction durations often result in improved pollutant removal as they provide more chances for ion interaction and the development of flocs. The pH values at the end of the procedure varied between 4.5 and 5.7, which is the ideal range. This range affects the charge and solubility of ions, ultimately affecting the effectiveness of floc production. The variation in current density, ranging from 14.7 to 25.0 mA/cm2, has a direct impact on the production of ions and the creation of flocs. Increasing current densities can expedite the creation of ions, which in turn enhances the formation of flocs. However, this can lead to higher energy consumption without a substantial improvement in pollution removal. The electrode consumption exhibited a range of 0.1–0.5 kg/m3, but the energy consumption had a variation between 1.3 and 5.2 kWh/m3. The current density and reaction time have an impact on these parameters, where greater values are associated with increased current densities and longer reaction times. Efficient floc production, which improves the removal of pollutants while minimizing resource consumption, is achieved by optimal ion replenishment, as indicated by current density and pulse power supply mode, as well as stirring speed. The implementation of the pulse power supply mode reduces electrolysis duration while preserving high current efficiency, leading to decreased energy and electrode usage, enhanced electrolysis efficiency, and low heating of the adapter. Moreover, a higher current density and larger electroactive regions result from perforations in the electrodes, and sharp edges increase the likelihood of distortion. This phenomenon is well recognized within the field of corrosion research. The primary method by which the speciation and solubility of Al are regulated in the aluminum agglomerated coagulant-electrocoagulation (APC-EC) process involves the hydrolysis and agglomeration reactions of trivalent aluminum ions . In a theoretical context, the presence of several aluminum hydroxide species, namely,
(which is preferred for the coagulation process),
,
(also preferred for the coagulation process),
, and
, may be theoretically seen within the pH range of 3–5 (Lu et al. 1999; Bingül et al. 2023). The concentration of aluminum hydroxide
decreased and was transformed into the aluminum hydroxide ion
when the pH decreased below 7 (Lu et al. 1999; Bingül et al. 2023). According to the study findings, the optimal pH for achieving the maximum removal of COD and
was determined to be 5.6 by comparing Tables 3 and 4. Hence, it can be deduced that the observed pH level is optimal for conducting EC due to the predominant formation of aluminum complexes (coagulants) in this pH range (Lu et al. 1999; Bingül et al. 2023).
Figure 3(d) demonstrates that when the frequency rises, the pH of the solution typically increases. This is primarily caused by the heightened generation of hydroxide ions (OH⁻) at the cathode. The natural increase in pH allowed for the most effective elimination of pollutants, eliminating the need for any additional pH changes. The untreated solution maintained an appropriate pH range for efficient elimination of pollutants, indicating that the electrochemical process automatically regulated the pH to favorable levels for treatment.
These particular species have the ability to participate in the oxidation of soluble organic matter due to their elevated oxidative potentials (Güven et al. 2009; Verma Shiv et al. 2013).
Figure 4 demonstrates the collective influence of speed and frequency on both removal efficiencies. The most prominent influence of stirring speed is seen in mass transfer. The findings show that the removal of COD, , and turbidity increases marginally as the mixing speed increased. In the literature, similar findings have been reported (Camcioglu et al. 2017). The augmentation of stirring speed facilitates the removal of gas bubbles that accumulate on the surface of the electrode. This increase in turbulence leads to a reduction in the thickness of the diffusion layer at the electrode surface and subsequently reduces cell resistance (Sridhar et al. 2012). One additional objective of the stirring process is to ensure the even dispersion of iron-based coagulants throughout the reaction. The uniformity of temperature and pH levels inside the reactor is achieved by the stirring action of the electrolyte. Elevated stirring rates, on the other hand, possess the capacity to disintegrate the aggregates formed inside the reactor, resulting in the formation of minuscule aggregates that pose difficulties in their separation from water (Bayar et al. 2011; Parsa et al. 2011).
With the third parameter (stirring speed) set to 30 min, graphs were used to illustrate the combined effects of stirring speed (400–1,200 rpm) and frequency (3,000–11,000 Hz) on the removal efficiencies of COD, Cr+3, and turbidity. These results indicate that to effectively remove when the COD is at its peak, an average frequency of 7,000 Hz and a stirring speed exceeding 685 rpm are required. It has also been demonstrated that while the frequency helps reduce energy consumption, the stirring speed does not have a significant impact on energy usage.
In addition, a decrease in frequency leads to an increase in sludge production. The results indicate that there was an increase in the removal percentages of COD, , and turbidity from 71.6, 89.51, and 89.52% to 72.82, 89.32, and 95.58% respectively, when the stirring speed was raised from 400 to 800 rpm at a frequency of 7,000 Hz. Specifically, the energy consumption for a reaction period of 30 min and a mixing speed of 400 rpm dropped from 4.9 to
and the consumption of electrodes decreased from 0.47 to
as the frequency increased from 3,000 to 11,000 Hz.
Model validation
The model outcome reveals that the R2 and adjusted R2 for COD removal are 0.9951 and 0.9907, respectively. Based on the ANOVA analysis of current study, it can be observed that the interplay between frequency, stirring speed, and reaction time has a significant impact on the parameters of COD,
turbidity, and energy usage.
In the present work, the degree of confidence level is 95%. As seen in Figure 2, the model's predicted values align closely with the experimental data, exhibiting a proximity to the diagonal line across all sites. The ANOVA analysis revealed that the quadratic models provide statistical significance (p-value < 0.05) and may be used to predict the percentage of COD (which is mentioned in the Supplementary material), , and turbidity removal, as well as the energy consumption (Khan et al. 2019; Ano et al. 2020).
Obtained regression correlations
Combination of operating parameters
The percentage removals of COD, , and turbidity are assessed in relation to energy consumption and electrode consumption. Various factors that influence these responses (frequency, reaction time, and stirring velocity) were taken into account, and the impacts of different variables are graphically represented in Figures 3–5 using RSM. The influence of operational settings on predicting the highest percentage removal of COD,
, turbidity, as well as energy consumption and electrode consumption was also examined. The COD,
, and turbidity removal efficiencies exhibited an upward trend with the progressive extension of electrolysis duration and the utilization of low frequency. Simultaneously, the energy consumption and electrode performance experienced a notable surge in response to the prolonged electrolysis period, as visually depicted in Figure 3. In Figures 4 and 5, increasing the stirring speed from 500 to 700 rpm and the frequency from 6,000 to 8,000 Hz led to improvements in pollutant removal, including a higher percentage of COD removal, increased Cr3+ concentration reduction, improved turbidity removal, and reduced energy consumption.
The surface response curve depicted in Figure 5 offers valuable insights into the impact of frequency and reaction time on energy consumption during the electrochemical treatment process, while maintaining a stirring speed of 800 rpm. The results indicate that raising the frequency from 3,000 to 11,000 Hz results in a significant decrease in energy usage, suggesting that higher frequencies improve the energy efficiency of the treatment process. However, there is a clear relationship between reaction time and energy use. Longer reaction times, which can range from 9 to 33 min, lead to higher energy usage. The relationship between frequency and reaction time indicates that higher frequencies consistently reduce energy consumption across different reaction times, highlighting the significance of optimizing frequency to enhance the overall efficiency of the operation.
Optimization with RSM
One of the primary benefits of RSM in the context of CCD is the attainment of optimal conditions for the elimination of pollutants and energy utilization. According to the CCD, the outcomes were optimized by employing the regression equations (Equations (23)–(26)). To optimize the process, the Design-Expert software explores the design space, taking into consideration various constraints and limitations. To achieve the authentic maxima or minima, multiple randomized initial points are selected. Each process variable and response variable must be assigned a predetermined target value (Khan et al. 2019).
During the optimization process, the variables of frequency, stirring speed, and reaction time were carefully chosen within their respective ranges. The objective was to maximize the responses, including the percentage of COD, concentration of , and efficiency of turbidity removal, while simultaneously minimizing energy consumption. Based on the given operating parameters, the optimal value was achieved at a frequency of 11,000 Hz, a stirring speed of 576 rpm, and a time of 30 min. The optimal values of COD removal, Cr3 removal, turbidity reduction are 70, 90, and 96% and for energy and electrode consumption are 3 kW h=m3 and 0:35 kg=m3, respectively.
CONCLUSIONS
This investigation focused on assessing the efficacy of the APC-EC technique employing perforated aluminum electrodes for the elimination of COD, , and turbidity in tannery wastewater. The study also examined the influence of various process parameters, including frequency, stirring speed, and reaction time, on the percentage of removal for COD,
, and turbidity, as well as the energy consumption. This analysis was conducted using RSM with Central Composite Design. The analysis of the experimental results, as determined by ANOVA, revealed that RSM proved to be a practical and suitable approach for optimizing the process variables. The optimal removal efficiencies achieved for COD, Cr3+, and turbidity were 72.70, 90.09, and 95.21%, respectively. In terms of energy and electrode consumption per unit volume, the performance figures were 4.3 kWh/m3 and 0.47 kg/m3, respectively. The obtained results were acquired under these experimental conditions: a frequency of 5,900 Hz, a stirring speed of 830 rpm, and a reaction time of 30 min. To optimize the removal efficiencies of COD, it is imperative to augment the applied frequency, while acknowledging that such augmentation may lead to a concomitant reduction in energy consumption. Present research highlighted that pH level adjustments were unnecessary, as the baseline pH of the tannery effluent was carefully maintained during all experiments due to controlled electrode dissolution. The study also demonstrated that factors like reaction time, frequency, and stirring speed significantly influenced various responses, including the removal efficiencies of turbidity, COD, and Cr3+, as well as pH stability, energy consumption, and electrode material usage. Therefore, the APC-EC technique has considerable promise for practical wastewater treatment applications due to its efficacy, cost-effectiveness, and robustness of the used electrodes. The ANOVA table (Analysis of Variance) indicates a p-value less than 0.05 degree of confident level, which is presented in the Supplementary material.
DATA AVAILABILITY STATEMENT
All relevant data are included in the paper or its Supplementary Information.
CONFLICT OF INTEREST
The authors declare there is no conflict.