Electrocoagulation process for the decolorization of wastewater containing Reactive Red 195 and CFD simulation of the hydrodynamic in a continuous-flow single-channel reactor

Textile industries are known to be among the most water-consuming and polluting industries (Samanta et al. 2019), the World Bank indicates that 17–20% of industrial wastewater comes from textile wet processing (Zheng et al. 2022). The latter generates effluents containing acids, alkalis, dyes, print pigments, hydrogen peroxide, starch, soaps, metals, and surfactants, to name a few (Bidu et al. 2021). Depending on the dye composition and the dying process, 10–50% of the dye initial amount is released in the discharged effluents (Ammayappan et al. 2016), with such a complex composition, constituting thus a major environmental issue. Moreover, reactive dyes are among the most problematic for the environment due to their bright colors, high solubility in water, and low fixation degree (50–90%). Numerous techniques were previously used to remove Reactive Red 195 (RR195) from synthetic wastewater, namely adsorption (Dey & Kumar 2017; Pérez-Calderón et al. 2018; Gündüz & Atabey 2019; Mahanna & Samy 2020; Munagapati et al. 2022), sonoelectrochemical oxidation (Somayajula et al. 2012), chemical coagulation (Thammachai & Charoensuk 2021), electrochemical oxidation process (Umadevi et al. 2022), electro-Fenton (Nazari & Setayesh 2019), and electrocoagulation (EC) (Criado et al. 2020). Compared to the other processes, the EC process guarantees multiple desired outcomes such as chemical dosage reduction and sludge generation reduction (Moussa et al. 2017; Tegladza et al. 2021), and, unlike most conventional processes, the EC process provides in situ delivery and compact equipment (Criado et al. 2020). EC is a wastewater remediation method that has been broadly studied for the decolorization of wastewater containing RR195 and for the removal of a large range of contaminants such as: heavy metals (Doggaz et al. 2019), suspended solids (Ntambwe Kambuyi et al. 2019), oils (Yang et al. 2021), dyes (Criado et al. 2020), Fluorides (Graça et al. 2019). The EC process involves the electrochemical dissolution of the sacrificial metal leading to the release of metal ions, which undergo hydrolysis reactions forming various metallic hydroxides precipitates (coagulants). Simultaneously, at the cathode, electrolytic hydrogen gas bubbles are generated. These bubbles allow a solid–liquid separation via flotation of the coagulated particles (flocs).

The design of the EC process, mainly the operation mode and the reactor geometry, strongly affects its operation and efficiency. The majority of published studies on wastewater treatment using the EC process are characterized by the use of small reactors (volume ranging from 100 mL to 1 L) operating in batch mode. According to Garcia-Segura et al. (2017), most experimental works lean toward stirred batch cells with plate electrodes rather than other reactor geometries that can provide higher efficiency and lower energy consumption. In the last decades, much effort has been put into the design and application of continuous-flow EC reactors in the wastewater treatment and decolorization (Amour et al. 2016; Benazzi et al. 2016). In this context, Ardhan et al. (2015) published an interesting paper based on the design of a continuous-flow reactor using experimental batch data for the decolorization of Reactive Blue 21 by EC. The steady-state conditions could often be predicted using experimental data from the reactor operating in batch mode (Bennajah et al. 2009; Attour et al. 2014; Rodrigues et al. 2020). However, the continuous-flow operation mode leads to flow malfunctions that can reduce the contact time between the coagulants and the pollutants, resulting in low treatment or decolorization efficiency. As such, the total or partial recycling of the solution leaving the reactor may be an interesting outcome as it allows the reuse of the coagulant leaving the reactor.

Additionally, the geometry of the reactor has great influence on bubble path, solid–liquid separation (flotation/settling), floc formation, fluid flow regime, and mixing characteristics (Hakizimana et al. 2017). Hence, for a given reactor geometry, the characteristics of the hydrodynamic as well as their influence on the mechanisms that occur within the reactor must be considered when investigating the reactor performance. Computational Fluid Dynamics (CFD) has been used to assess various aspects of flow through electrochemical reactors (Choudhary & Mathur 2017; McBeath et al. 2020; Villalobos-Lara et al. 2020).

This work investigates the decolorization of Reactive Red 195 using the EC process in a continuous-flow single-channel (CFSC) reactor operating in closed-circuit. The configuration allows a total recycling of the solution at the outlet of the reactor. The performance of the system is evaluated as a function of transfer and transport phenomena that govern the process. This paper is outlined as follows: first, the characterization of the hydrodynamic within the CFSC reactor is performed by CFD simulation using Comsol Multiphysics^®. An experimental Residence Time Distribution (RTD) study is then used to validate the numerical results obtained by CFD simulation. Second, the experimental study of the decolorization of the wastewater containing RR195 is investigated as a function of various parameters such as initial dye concentration, applied current intensity, and electrolysis time.

The experiments were carried out using Reactive Red 195 (RR195) supplied by Meghmani Dyes and Intermediates LLP. Table 1 shows the properties of the RR195 including its chemical class, molecular weight, maximum absorption wavelength, and chemical structure. The textile dye was used without any further purification in order to represent real dyeing conditions.

Table 1

Properties of the Reactive Red 195 (RR195)

Dye .	Chemical structure .	Mw (⁠⁠) .	Chemical class .	λmax (nm) .
Reactive Red 195		1,136.3	Monoazo	542

Dye .	Chemical structure .	Mw (⁠⁠) .	Chemical class .	λmax (nm) .
Reactive Red 195		1,136.3	Monoazo	542

View Large

The preparation of synthetic wastewater consisted of mixing the RR195 with distilled water using a magnetic stirrer. The initial concentration was varied between ⁠, while the initial conductivity was set at used in previous studies (Merzouk et al. 2009) and adjusted using sodium chloride (NaCl) which is the most essential auxiliary added in the reactive dyeing of many cellulosic fibers to enhance dyes’ exhaustion (Burkinshaw & Salihu 2018). The initial pH was maintained at 7 as the EC removal efficiency is proven to be the most efficient between 5 and 9. At this pH range, the prevailing species are Al(OH)₃ which are characterized by their large specific surface area (Aoudj et al. 2010).

The entire closed-circuit consisted of a volume of 9.445 L, 5% of which constituted the storage tank volume. This configuration allowed recycling 100% of the treated solution along with the flocs.

The CFSC reactor (40 cm high, 59.5 cm long, and 5 cm wide) used in this study was previously designed and used by Ntambwe Kambuyi et al. (2021). It is made of transparent glass and is split into 10 rectangular compartments arranged in series. Each compartment (C), with a width of 5 cm, is divided equally into two channels: upflow and downflow channels connected to each other through a passage zone (P) with a height (H_P) of 0.5 cm and recirculation zone (R) with a height (H_R) of 4 cm. The liquid height inside the CFSC reactor while functioning is 34.5 cm. H_P of 0.5 cm allows a maximum flow velocity in the vicinity of the free surface leading to flocs entrainment resulting in a high coagulant specific surface area, while H_R of 4 cm enables flocs settling at the bottom of the CFSC reactor. This configuration allows a vertical serpentine flow, which improves mixing within the reactor. Since the cross-section is different in the flow channels, the passage, and the recirculation zone, the flow velocity is not uniformly distributed throughout the volume of the CFSC reactor for a given volumetric flowrate.

Aluminum electrodes plates (of 50 cm height, 2 cm width, and 0.2 cm thickness) arranged in a monopolar-parallel configuration were used as anodes and cathodes. The experiments were conducted under galvanostatic mode; different current intensities were applied to nine pairs of electrodes placed in the first nine upflow channels (one pair per upflow channel) as shown in Figure 1. A copper bus cable and a wooden bar were used, respectively, to connect the anodes (as well as the cathodes) to each other and to fix an inter-electrode distance (d_int) of 14 mm. The total active electrode surface S_a was equal to 621 cm² (34.5 cm × 2 cm is the surface of each electrode), and the corresponding S_a/V_T ratio for the entire closed-circuit was equal to 6.57 ⁠. A high active electrode surface was chosen in order to avoid the increase in energy consumption. The electrodes were initially rubbed with sandpaper to remove the passive oxide layer, rinsed with tap water, and immersed into diluted acetone to eliminate dye traces from former tests. Finally, they were dipped in 3-M HCl solution to prevent possible electrode passivation. The electrodes were rinsed with distilled water and dried for use.

The conductivity was measured using the portable conductor HI 99300 of HANNA brand. The resolution of the conductivity meter is and the measurement ranges from 0 to with an accuracy of ±2% full scale. Samples were periodically taken at the outlet of the reactor after treatment, centrifuged at 4,000 rpm for 2 min using Hettich ROTOFIX 32A centrifuge. The absorbance was measured before and after treatment using UV-visible spectrophotometer HITACHI, U5100 model at a maximum wavelength of 542 nm.

The mixing flow pattern inside the CFSC reactor was obtained using the RTD method. The experimental distribution function obtained was used to validate the results of the numerical simulation performed through CFD. In order to maintain a constant residence time, the inlet flowrate was adjusted to ⁠. The stimulus–response technique based on pulse injection was employed using of potassium chloride as a tracer. A volume of 1 mL of the tracer was injected during 4 s. The RTD experiments were performed without the electrodes in order to prevent possible solution conductivity variations. The conductivity of the outlet liquid samples was measured for hydrodynamic analysis using the distribution function F(t).

To perform the CFD simulation of the CFSC reactor, COMSOL Multiphysics^® was used and some assumptions related to the operating conditions were defined. First, the flow was assumed to be incompressible and due to the complete dilution of dye in water, the chemical and physical properties of the solution were considered to be similar to those of water (especially the density and dynamic viscosity). Moreover, the CFD simulation was performed without electrodes inside the reactor, assuming that electrochemical reactions do not significantly affect the hydrodynamic behavior within the reactor. Therefore, globally, the hydrodynamic in the reactor is controlled by the liquid flowrate.

As for boundary conditions, a constant normal velocity boundary condition was applied at the inlet of the CFSC reactor (u=–u₀). The turbulent intensity and the turbulent length scale were equal to 0.05 (5%) and 0.025 m, respectively. At the outlet of the CFSC reactor, the boundary condition was set to zero (p = 0). The zero-shear boundary condition was applied to the surface of the upper part (free surface) of the reactor. The no-slip boundary condition was applied for the remaining boundary of the CFSC reactor (the surface of the lower part, the surface of baffles, etc.). Considering the vertical serpentine flow in our case, the force of gravity was taken into account.

The convection–diffusion (Equation (10)) was solved to obtain a numerical result, which was compared to the experimental one for validating the results obtained by CFD.

A constant concentration of the tracer boundary condition was applied at the inlet of the CFSC reactor. For the outlet, the zero diffusive flux boundary condition was applied, which implies that the mass flux was purely convective. Also, the zero-flux boundary condition was applied at the free surface and at the remaining parts of the CFSC reactor.

The CFD simulation is performed using a boundary layer mesh for the discretization of domains. This type of mesh is typically used for fluid flow problems to resolve the thin boundary layers along the no-slip boundaries. First, the solution of the hydrodynamics was evaluated for different mesh numbers until the numerical average velocity was approximately equal to the analytical one calculated using Q/S (Q is the flowrate equal to and S is the area across of the flow channel that is equal to 12.5 cm²). The comparisons are made for the upflow channel 11 and the downflow channel 12 at y = 0.17 m. The results show that the relative error ε values between numerical and analytical average velocity are less than 1% for all the elements used. This proves the accuracy of this type of mesh regardless of the number of elements. However, 204,000 elements are preferred due to the lower value of the relative error (ε = 0.166 and 0.146% for flow channels 11 and 12, respectively) and are used for the hydrodynamic analysis inside the CFSC reactor.

Furthermore, it is observed that the changes in flow directions generate the formation of a jet flow in the following flow channel. The magnitude of this jet flow is higher in the downflow channel at 0.26 m < y < 0.345 m because of the reduced cross-sectional area in the passage zone compared to the recirculation zone. However, regardless of its magnitude, the sudden change in flow direction greatly influences the velocity field in the CFSC reactor.

As indicated in Figure 3, back mixing and recirculation of the flow occur on the opposite side of the flow channel following the passage zone or recirculation zone. In the upflow channel, back mixing and recirculation occur near the recirculation zone at y ranging from 0.04 to 0.07 m and occupy about 30–40% of the cross-sectional area of the upflow channel (Zoom (c)), while they occupy about 60% of the cross-sectional area of the downflow channel (Zoom (a)) at y ranging from 0.26 to 0.34 m, except for the second flow channel where the back mixing and recirculation range from 0.04 to 0.11 m. Finally, the flow tends to homogenize throughout the channel cross-section as it moves away from the passage and recirculation zones as shown in Figure 3 and zoom (b). Moreover, in the recirculation zone located at the bottom of the reactor, a dead zone is observed.

The main conclusion that emerges from the CFD simulation and the RTD study is that the velocity is not uniformly distributed inside the CFSC reactor. In the vicinity of the free surface, the streamwise velocity changes abruptly, and the local velocity of the liquid increases. Therefore, a large vortex is formed on top of the different downflow channels of the CFSC reactor as shown in Figure 3. The CFSC reactor can be used to perform further decolorization experiments, considering the stability of the flow in the reactor.

The amount of electro-dissolved aluminum at the anode is strongly monitored by the applied current intensity and the electrolysis time. The electrochemical reactor of this study operated under current intensities ranging from 30 to 200 mA for investigating the decolorization of synthetic wastewater containing RR195 with an initial concentration of ⁠. These current intensity values are far lower than in conventional EC cells (Merzouk et al. 2009; Kobya et al. 2014; Assémian et al. 2018; Criado et al. 2020). Preliminary experiments were initiated to determine the range of current intensities under which high decolorization is reached, coagulant overdosage is avoided, and the useless high energy consumption is avoided. While applying current intensities ranging from 1 A to 200 mA to the EC reactor, the decolorization temporal evolution remained almost unchanged. Furthermore, at higher current intensities the resulting flocs are able to maintain their larger size due to the larger collision frequency owing to their high concentration (Harif & Adin 2011). All of this is a downside in our case since the flocs act as adsorbent and therefore a high specific surface area is required (Aoudj et al. 2010).

The flowrate was adjusted at resulting in an average inlet velocity of (first downflow channel shown in Figure 1). On the free surface of the CFSC reactor, the flocs appear as a red colored layer which thickness depends largely on the current intensity and the electrolysis time. After about 60 min of operating under current intensities higher than 100 mA, the flocs floating on the free surface create a vertical thin gradient layer with the oldest flocs red on top and the youngest rather white at the bottom of the layer. This layering is a consequence of the decolorization temporal evolution, where the red layer comes from the reaction step as the treatment is still occurring, and the white layer is a result of the steady state during which the treatment remains unchanged regardless of the amount of the coagulant produced.

As might be expected, the flow visualization using the flocs formed during the electrochemical treatment confirms the numerical results shown in Figure 3. Experimentally, the jet flow formed in the passage zone and the increasing velocity carry away a part of the flocs formed at the free surface in the different channels of the CFSC reactor. However, the breakage of the flocs at the free surface only occurs when the flocs layer is thin enough. Conversely, a higher coagulant production leads to a highly stable thick flocs layer on the free surface withstanding the flow.

Figure 5 reveals a similar decolorization efficiency evolution profile for all the higher current intensities. This profile follows two steps:

• - Reaction step: a sharp increase in the decolorization of synthetic wastewater appears. The treatment occurs rapidly owing to the availability of dye molecules. It is influenced by the current intensity; the higher the applied current the sharper the increase.

• - Steady-state step: as the electrolysis time progresses, the rate of decolorization decreases. During this step, the decolorization of synthetic wastewater remains slightly unchanged. The residual amount of dye withstands the treatment despite the availability of the coagulant.

When the electrochemical reactor is operating under a current intensity of 30 mA, the decolorization efficiency evolution remains in the reaction step until the end of the experiment (Figure 5). At an electrolysis time of 58 min, the decolorization efficiency reached 80%. It might be explained by the insufficiency of the coagulant to uptake the dye molecules before that electrolysis time.

Conversely, when applying a current intensity of 50 mA, a decolorization efficiency of about 82% is obtained at a treatment time of approximately 45 min. However, when the reactor is operating under 100 and 150 mA, the electrolysis time yielding a decolorization efficiency of 80% is quasi-unchanged at about 20 min, while 200 mA requires only about 19 min to achieve the same decolorization efficiency as shown in Figure 5.

Figure 6 illustrates the temporal evolution of the specific electrode consumption for the different current intensities ranging from 30 to 200 mA.

As illustrated in Figure 6, the current intensities under 100 mA generate, at first, specific electrode consumptions under or equal to dye removed, while the current intensities above 100 mA generate specific electrode consumptions higher than dye removed even before reaching the steady state. Nonetheless, reaching the steady-state condition leads to a linear relationship between the specific electrode consumption and the electrolysis time. This could be explained by Equation (3), when the steady state is established, the final dye concentration resists to the treatment. Thus, the decolorization efficiency remains almost unchanged. This leads to a specific electrode consumption that only changes as a function of the amount of the coagulant produced that is proportional to the current intensity applied and the electrolysis time (Equation (2)).

The slope of the linear region depends on the applied current intensity (Figure 6). This overconsumption of the electrodes is attributed to the over-production of the coagulant which is increased by raising the electrolysis time.

According to Merzouk et al. (2009), the specific electrode consumption is about dye removed under an applied current intensity of 1.5 A and an initial concentration of of 2-naphthoic acid and 2-naphthol mixture for an 80% decolorization efficiency. Alternatively, Yuksel et al. (2012) reported the lowest electrode consumption for the removal of Reactive Yellow 135 from a real textile effluent at an initial dye concentration of ⁠, being dye removed under a current intensity of 143 mA to reach 40% of the decolorization efficiency. Therefore, in this study, dye removed seems a reasonable limit value of the specific electrode consumption. This should be taken into consideration to preserve the electrodes’ life span.

Table 2 displays the electrode consumption value at different electrolysis times for the different current intensities that yield a decolorization efficiency above 80%. At an electrolysis time of 26 min, increasing the applied current intensity from 100 to 200 mA only increases the decolorization efficiency by 7%, while the specific electrode consumption is raised by 46%.

The results in Table 2 confirm the previously discussed results in Figures 5 and 6. The rise of the applied current intensity from 30 to 100 mA results in an increase of the specific electrode consumption by 44% and a decrease of the electrolysis time by 55%. Moreover, increasing the current intensity from 50 to 100 mA decreases the electrolysis time by 42% and hardly raises the specific electrode consumption by 4%.

Hence, to reach a certain decolorization efficiency, a compromise between time and current intensity must be found. A high-current intensity provides a good decolorization efficiency in a short time, while a low-current intensity produces the same result over a longer time. As discussed, the increase of the current intensity generates many downsides such as overdosage of coagulant, excessive energy consumption, and electrodes’ lifespan reduction; whereas a high electrolysis time allows good contact between the coagulant and dye molecules. However, when designing a reactor, a high time of treatment means larger equipment leading to large investment and operating costs.

The main conclusion that can be drawn from this part is that for an initial dye concentration of ⁠, the CFSC reactor operating in a closed-circuit and under steady-state conditions should be avoided to prevent useless energy consumption and overdosage of the coagulant. Considering that the transition from the reaction step to the steady state of decolorization efficiency of the current intensities (Figure 5) and the specific electrode consumption (Figure 6) occurs at about 26 min for a current intensity of 100 mA. The steady-state condition begins at this time. Hence, it will be referred to as the electrolysis time limit.

During the following experiments, the CFSC reactor operating in a closed-circuit (Figure 1) behaves more like a fed-batch system with the coagulant being added continuously. The feeding rate is monitored only by the current intensity as the electrolysis time is constant.

When the CFSC reactor is operating under a current intensity of 100 mA for an electrolysis time of 26 min, two distinct ranges of concentrations appear (Figure 7). First, when the initial dye concentration is between ⁠, a sharp decrease of the specific electrode consumption occurs while the decolorization efficiency is almost constant at about dye removed. Second, the specific electrode consumption is quasi-constant, and the decolorization decreases almost linearly from 82 to 25% with increasing dye concentration from 10 to ⁠. This could be attributed to the adsorption capacity of the coagulant. At first, the adsorption capacity is not yet exhausted due to the unavailability of the dye molecules, then it slowly stabilizes to reach exhaustion when there are surplus dye molecules in the solution. The coagulant adsorption capacity is exhausted when all the active sites are already occupied.

A robust process must be able to achieve the same decolorization efficiency regardless of the initial dye concentration value. In the following experiments, process performance was investigated, using various current intensities and initial dye concentrations (RR195), while maintaining the electrolysis time at 26 min, in order to estimate the current intensity yielding a decolorization of about 80% for each dye concentration.

The principal aim of these experiments is to create a resilient EC process that is capable of delivering satisfactory results even under critical conditions such as high initial dye concentrations.

Specific electrode consumption was also evaluated for the CFSC reactor operating in a closed-circuit under different current intensities and initial dye concentrations and the same electrolysis time limit defined above as 26 min. As shown in Figure 8(b), an increase in the current intensity leads to a proportionate rise in the specific electrode consumption. For example, under the same current intensity of 500 mA, the respective electrode consumptions are about dye removed and dye removed for initial dye concentrations of 10 and ⁠, respectively. However, at 500 mA the initial concentration of has already reached a decolorization efficiency of 99%. Therefore, the residual dye concentration in the solution is rather low to non-existent. In fact, the rate at which the specific electrode consumption increases with increasing current intensity is inversely proportional to raising the initial concentration.

Similar results were showcased by Chafi et al. (2011), using the EC process for the removal of Acid Orange 7. Increasing the initial dye concentration from 50 to leads to both a decrease of the decolorization efficiency from 90 to 75% and the specific electrode consumption from dye removed to dye removed.

Summing up, in the case of lower initial dye concentration, the decolorization efficiency is already at a maximum value. Therefore, the overdosage electro-dissolved aluminum is considered useless to the process. Conversely, higher initial dye concentration still provides significantly increasing decolorization efficiency with rising current intensity.

This integral study investigated the decolorization of synthetic wastewater containing Reactive Red 195 in a laboratory-scale EC process. A CFSC reactor operating in closed-circuit is chosen to keep the similar hydrodynamics of continuous mode. Besides the transfer phenomena study, this work described transport phenomena inside the reactor using CFD simulation and RTD methodology. Furthermore, the deployed approach takes into consideration mixing, decolorization efficiency, and specific electrode consumption.

CFD simulation results indicate that the velocity is not uniformly distributed in all parts of the CFSC reactor. At the passage zone located in the vicinity of the free surface, the streamwise velocity changes abruptly, and the local velocity of the liquid increases. This results in the formation of a large vortex on top of the different downflow channels of the CFSC reactor. While at the recirculation zone located at the bottom of the reactor, a dead zone is observed

The flow tends to homogenize throughout the channel cross-section as it moves away from the passage and recirculation zones. In general, considering the stability of the flow in the CFSC reactor, it can be used to perform further decolorization experiments.

For an initial dye concentration of ⁠, the electrolysis time required to reach the steady-state conditions depends on the applied current intensity. However, our experimental data demonstrated that the temporal evolution of the decolorization remains unchanged when the current intensity exceeds 100 mA.

To achieve a decolorization efficiency of 80%, a decrease of about 40% in electrolysis time is observed when the electrochemical reactor operates under a current intensity of 100 mA instead of 50 mA. The deviation in the profile of the decolorization efficiency and the specific electrode consumption occurs at about 26 min, the steady-state condition begins at this treatment time, which is referred to as the electrolysis time limit. The electrolysis time limit is considered as an essential parameter for the transposition from closed-circuit to continuous mode.

When raising the initial dye concentration from ⁠, the specific electrode consumption remained slightly unchanged at about dye removed, meanwhile the decolorization efficiency demonstrated a remarkable decrease. Thus, for the same treatment time, the EC process could be enhanced by increasing the current intensity. This exploratory study using the CFSC reactor operating in closed-circuit to perform EC provides a solid base that might be useful for the transposition from batch to continuous mode.

The authors would like to thank Prof El Ibrizy Ouafa for providing linguistic advice and English proofing.

R.L. conceptualized and investigated the study, did software analysis, data curation, and formal analysis, prepared the methodology, wrote the original draft, wrote, reviewed, and edited the article. T.N.K. validated, visualized, wrote, reviewed, and edited the article. Z.Z. collected resources and did data curation. A.A. reviewed and edited the article. B.B. visualized, reviewed, and edited the article. K.D. collected resources and validated the study. A.D. conceptualized the study, did data curation, did project administration, supervised, wrote, reviewed and edited the article.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.