Influence of the main parameters
The effect of the main parameters (i.e., the properties of the electrode materials, the experimental conditions (applied electrode potential, pH, concentration, electric conductivity, residence time, flow rate) and the reactor design) on the electrosorption/desorption kinetics and capacity have been reviewed in detail. The main research articles dealing with the electrosorption of organic compounds in the literature are listed in Table 1.
Nature and surface morphology of the electrode materials
Various electrode materials have been employed for electrosorption technologies, and Fig. 7a shows the occurrence frequency regarding the electrode materials used in the literature (Table 1). These materials are proposed for classification into the following categories (percentage of occurrence in brackets): carbon-based electrodes (85%), metal-based electrodes (9%) and polymer/resin-based electrodes (6%). Under the carbon electrodes, the following subcategories can be found: carbon-based composites (19%), GAC (32%), carbon cloth (19%), carbon fiber (7%), carbon felt (7%), graphite (13%) and graphene (3%).
Metal-based materials have barely been studied in the literature (Fig. 7a), probably due to their usually high cost compared to carbon materials and their lower flexibility in shape, surface area and porosity. The crystallographic orientation of the single-crystal planes (Lust, 1997) and the electronic structure of the metal surface impact the EDL, and therefore the electrosorption efficiency (Schmickler, 1996). It has been further demonstrated that the dipole orientation of water, often used as a solvent, can be impacted by the surface chemistry of metal-based materials (Conway, 1999). In addition, the point of zero charges of metals influences the specific orientation of H2O molecules, which then impacts H2O adsorbability (Trasatti, 1972). Therefore, the adsorbability of solvated ions on metals is also impacted, along with their electrosorption efficiency.
In contrast, it is obvious that carbon represents the majority of electrode materials used in the literature (Fig. 7a). Despite their availability and low prices, this enthusiasm toward carbon materials can be explained by their adequate properties for electrosorption, which are discussed more in detail in sections 184.108.40.206 to 220.127.116.11 of part I. Among their important characteristics, they have a particularly high surface area, high porosity, high electrical conductivity and high electrochemical stability, especially for cathodic reduction (Frackowiak, 2001; Hou et al., 2012). Graphene has notably attracted attention as an emerging two-dimensional material composed of sp2-hybridized carbon atoms arranged in a honeycomb structure, and its characteristics have benefited environmental electrochemistry (Geim & Novoselov, 2007; Le et al., 2015; Mousset et al., 2016a, 2016b; Mousset et al., 2017; Le et al., 2019; Du et al., 2021).
The surface morphology of electrode materials is often inspected using microscope devices as a qualitative approach to roughly evaluate the adsorption ability. Microscopic views obtained via scanning electron microscopy (SEM) are displayed in Figs. 7b and 7c for activated carbon fiber (ACF) (Han et al., 2006b) and single-walled carbon nanotube (SWCNT) films (Yue et al., 2019b), respectively. Fig. 7b depicts the fine structure of ACF, which involves numerous uniform open pores on the surface (Han et al., 2006b), while the SEM image of the prGO/SWCNT film displays a rough surface and corrugated structures, which should provide a large accessible surface area (Fig. 7c) (Yue et al., 2019b).
Specific surface area, porosity, and roughness
A large surface area and an optimal pore size distribution of the electrode are essential requirements for reaching a high electrosorption capacity. In addition, they result in a high capability for charge accumulation at the active sites of the electrode/electrolyte interface (Zaharaddeen et al., 2016). From Fig. 8a, it can be seen that the specific surface area of the electrode used in the literature reached 2000 m2 g-1, while a majority of articles (36%) applied materials with specific surface areas within the range of 0-500 m2 g-1 (Table 1).
One of the most common methods employed to modify the surface area and pore size distribution of carbon materials is their activation via chemical and/or thermal treatment. The electrosorption of 4-chlorophenol was studied by testing AC with different treatments, i.e., nonmodified AC (AC-NM), AC oxidation with nitric acid (AC-NA), AC treated with ammonia (AC-Am), and annealed AC (AC-HT) (Fig. 8b) (Biniak et al., 2013). The results highlighted a decrease in the specific surface area (SBET) estimated by the well-known Brunauer-Emmett-Teller method when the AC was modified. Additionally, the total volume of mesopores (Vme) increased with the chemical treatment (AC-NA and AC-Am) (Fig. 8b), while the electrosorption efficiencies were better with these materials. It is important to note that increasing the surface area of materials does not systematically lead to a higher electrosorption efficiency, unlike many studies intend to show. Therefore, the distribution of pores remains an essential factor to consider, along with the specific surface area. The pore size has previously been shown to induce a change in the EDL. Fig. 8c depicts the EDL within different pore sizes of the carbon material, i.e., higher than 2 nm, between 1 and 2 nm and smaller than 1 nm. It has been emphasized that the existence of an optimal pore size gives an optimal capacitance when the pore size is close to the ion size (Largeot et al., 2008). This feature further highlights the impact of microporosity on the EDL and electrosorption efficiency. Due to these differences in the behavior of EDLs in porous materials, a multiscale model has been proposed according to the pore size for better accuracy of capacitance prediction: a sandwich-type model for micropores (< 2 nm), an electric double-cylinder model for mesopores (2-50 nm)), and a planar model for macropores (> 50 nm) (Fig. 8d) (Huang et al., 2008; Feng et al., 2010; Dai et al., 2020).
A more recent approach for modifying the material properties involves carbonaceous nanostructured materials, such as graphene and/or carbon nanotubes (CNTs). A comparison between the specific surface area and porosity of three graphene-based materials, namely, prGO/SWCNTs, prGO and reduced graphene oxide (rGO), used for MB electrosorption has been performed (Yue et al., 2019b). The specific surface areas of rGO (5.5 m2 g-1) and prGO (29 m2 g-1) were much smaller than that of prGO/SWCNT (94 m2 g-1), and the same trend was obtained for the total pore volume. The prGO/SWCNT electrode material depicted a higher electrosorption capacity (13 g−1). The increase in the specific surface area and total porosity could have explained the higher electrosorption capacity (Yue et al., 2019b).
In addition, many electrode materials used in electrosorption can be considered to have a nonidealized electrode geometry, which then affects the EDL and consequently the electrosorption efficiency (Fan et al., 2014). For example, curvature effects are frequently encountered in cylindrical shapes (e.g., CNTs) and porous materials, which breaks with the linearity of models developed for planar electrode systems (Henstridge et al., 2010; Lian et al., 2016). Moreover, differences have been predicted by using density functional theory (DFT) between concave and convex shapes. Concave induces a lower curvature effect than convex interfaces, leading to higher capacitances in the latter (Lian et al., 2016), and better electrosorption.
Applied electrode potential/current density and electrode polarity
The electrode polarity related to the applied current intensity sign is known to influence the electrosorption capacity of the porous surface, while the electrosorption rate can be increased. According to the reviewed papers (Table 1), the occurrence percentages of electrode polarity configurations tested for electrosorption are as follows: cathode (23%), anode (41%) and both polarities (i.e., cathode and anode) (36%) (Fig. 10a). Most of the time, the electrode polarity can induce electromigration effects that need to be considered according to the charge in the targeted compounds. Since many organic compounds are negatively charged, the reason why more electrosorption studies have been performed with a positively charged anode material could be explained (Fig. 10a). For instance, the electrode polarity was studied using metal wire coated with an epoxy resin-based polymer as the working electrode and platinum wire as the counter electrode for cresol red electrosorption (Wang et al., 2012). The results showed that anodic polarization enhanced the efficiency compared to cathodic polarization. This was explained by the orderly migration of the negatively charged cresol red molecules with the sulfonic group on the benzene ring toward the positively charged wire anode. In contrast, the same charge between the compound and the cathode material limited their interaction (Wang et al., 2012). Similarly, anodic polarization increased the amount of phenol adsorbed on AC, while cathodic polarization decreased it (McGuire et al., 1985). In contrast, the cathodic polarization remarkably increased the thiocyanate anion electrosorption on the AC compared to anodic polarization (Rong & Xien, 2005), ascribed to coulomb interactions. In addition, the cathodic electrosorption of 8-quinolinecarboxylic acid on ACC resulted in higher removal efficiencies (96%) than anodic electrosorption (84%) (López-Bernabeu et al., 2016). These differences in behavior were due to the occurrence of a faradaic reaction at the Ti-Pt anode, while the cloth was the cathode. In this configuration, the anodic oxidation of 8-quinolinecarboxylic acid and its subsequent byproducts could be implemented, which enhanced the removal efficiency of the targeted compound (López-Bernabeu et al., 2016). In contrast, organic compounds were not oxidized when the AC tissue was used as the anode, and only electrosorption could occur.
The faradaic reactions are considered to be competitive when only electrosorption is desired. Their occurrence is directly related to the applied electrode potential value, which is the driving force that permits modification of the adsorption equilibrium by introducing a difference in the potential at the electrode/electrolyte interface. This potential is linked with the applied current density through the Butler-Volmer equation (Lord et al., 2012). These reactions are involved at either anodic potentials that are too high or cathodic potentials that are too low, and some of them have recently been reviewed (Zhang et al., 2018) and schematized (Holubowitch et al., 2017). Water oxidation into oxygen (O2) and water reduction into hydrogen (H2) are the two most famous faradaic equations that can occur in aqueous media. There can be other parasitic reactions according to the electrode material used and/or the composition of the electrolyte. For instance, if a carbon material is used as an anode, it can be easily oxidized (E0 = 0.21 V/standard hydrogen electrode). This can be an issue because most of the porous electrodes employed are based on a carbon material (Fig. 7a), knowing that anodic polarization is the most experimented upon (Fig. 10a). Therefore, caution should be taken in applied operating conditions if porous carbon anodes are implemented to avoid corrosion. Moreover, undesirable nitrogenous and/or chlorinated oxyanions (NO2-, NO3-, ClOH, ClO3-) can be initially generated at a sufficiently high anodic potential in the presence of nitrogenous species and/or chloride ions (Brito et al., 2015; Mousset et al., 2018; Mousset et al., 2020a). When a high O2 overvoltage anode is employed, such as boron-doped diamond (BDD), to cite the most frequently studied anode, perchlorate (ClO4-) can even be produced (Bergmann et al., 2009). Often missing in the literature and especially in reviews dealing with faradaic reactions involved in capacitive deionization (Zhang et al., 2018): the local alkalization at the cathode from O2 reduction and H2O reduction reactions should not be overlooked. Under this condition, magnesium, calcium and/or carbonate are present in solution, which is the case in many actual aqueous effluents (Belarbi et al., 2016; Adnan et al., 2021a), and electroprecipitation phenomena can occur by forming Mg(OH)2 and/or CaCO3 at the cathode surface (Belarbi et al., 2016; Adnan et al., 2021a). These deposits then progressively passivate the cathode and hamper electrolysis.
Type and concentration of the inorganic electrolyte
In most synthetic solution studies, an inorganic supporting electrolyte is added to increase the electrical conductivity of the wastewater to be treated (Koparal et al., 2002; Han et al., 2006b; Gerçel, 2016). Although electrolytes should not be added in the context of wastewater depollution to avoid external contamination, studies in their presence have emphasized the antagonist role of the electrosorption efficiency in organics according to their types and concentrations.
It has been highlighted that the equilibrium electrosorption capacity of phenol on ACF was reduced, whereas the electrosorption rate was enhanced with increased sodium sulfate (Na2SO4) electrolyte concentration from 1 to 100 mM (Han et al., 2006b). In addition, Na2SO4 increased the solution conductivity, which contributed to an increase in ion transport. However, the adsorption sites for phenol on the surface of the adsorbent were occupied by the competitive electrosorption of the Na2SO4 electrolyte (Han et al., 2006b). Similar results have been obtained in other studies involving the electrosorption of aniline on ACF in Na2SO4 electrolyte (Fig. 11a) (Han et al., 2006a), of 2,4-dichlorofenoxyacetic acid on ACC in Na2SO4 electrolyte (Bayram et al., 2018), and of acilan blau dye on AC-perlite mixtures in sodium chloride (NaCl) electrolyte (Han et al., 2006b). This trend was confirmed by the study of 1-adamantanol in the presence of halide (F-, Cl-, Br-)-based electrolytes. This further emphasized the decrease in the organic adsorption when the electrolyte concentrations increased from 5 to 100 mM, regardless of the types of halide anions (Stenina et al., 2001). It was thus reported that organic-inorganic attraction could occur during concurrent adsorption, which the strong hydrogen bonds could explain between the OH- group of the organic compound and the halide anions.
The effect of the electrolyte type on the electrosorption efficiency was more emphasized by Gerçel (2016), who compared two different electrolytes, either sodium nitrate (NaNO3) or Na2SO4, for the electrosorption of burdem orange II textile dye by using AC prepared from a waste material. The highest dye removals of 95% and 88% were achieved with 0.2 M Na2SO4 and 0.2 M NaNO3, respectively. The differences in removal percentages could be attributed to the different hydrated radii and valences of the electrolyte, regardless of the initial solution concentration (Gerçel, 2016).
Current challenges to overcome
A general overview of the main factors and phenomena involved in the electrosorption efficiency is proposed and schematized in Fig. 15. This scheme allows for the identification of which parameters affect each phenomenon that then governs the electrosorption efficiency. The importance of the EDL and mass transport is influenced by most of the main parameters. Therefore, future efforts should be made to enhance both EDL and mass transport. The main output and future improvements in the electrosorption process efficiency are addressed as follows and determine part of the orientation of the PhD thesis:
– Electrode material: The high specific surface area of the material is insufficient to obtain a high electrosorption capacity since the pore size also appears to be a crucial parameter. The presence of mesopores tends to improve efficiency. Therefore, the specific surface area and the pore size distribution of materials should be determined before experiments to ensure a maximum electrosorption of the pollutants. Moreover, the material conductivity and wettability should not be neglected in the mechanisms. Material stability and reusability are other essential criteria for long-term use in practical applications. Existing studies lack life span and regeneration tests, which are representative of real conditions.
– Electrode potential and polarity: An optimal electrode potential is required for optimal electrosorption, while values that are too high or too low lead to faradaic reactions that need to be avoided. The electrode polarity plays a role in the electrosorption efficiency according to the types of targeted compounds, but the porous material can be damaged if it does not support either oxidative or reductive potentials. Carbon-based materials have been widely employed in electrosorption, although they can be easily corroded as anodes. Therefore, caution should be taken in the applied conditions to preserve the lifetime of the electrode material.
Table of contents :
PART I: BIBLIOGRAPHY
1.1. Definition and types of adsorption
1.2. Adsorption of organic pollutants
1.3. Current stage and future prospects
2.1. Electrosorption: background and principle
2.2. Influence of the main parameters
2.3. Reactor design
2.4. Factors influencing the electrodesorption step
2.5. Long-term efficiency of electrosorption
2.7. Current developments and future prospects
PART II: MATERIAL AND METHODS
1.1. Sample collection and characterization
1.2. Preparation of activated carbon
1.3. Preparation of the beads
1.4. Characterization of beads and GAC
1.5. Stirred batch and column sorption experiments
1.6. Batch and fixed-bed column data analysis
1.7. Analytical methods
2.1. Effluents preparation and characterization
2.2. Preparation and characterization of the 3D porous electrode
2.3. Electrochemical setup and procedure
2.4. Electrochemical regeneration of the GAC electrode
2.5. Electrochemical degradation of OMWW
PART III: RESULTS AND DISCUSSIONS
CHAPTER I: COMPOSITE BEADS
I. Performance and dynamic modeling of a continuously operated pomace olive packed-bed for olive mill wastewater treatment and phenol recovery
2. Results and discussion
2.1. Physicochemical properties of the effluent
2.2. Characterization of the adsorbent
2.3. Adsorption efficiency
II. Electrosorption of phenolic compounds from olive mill wastewater: mass transport consideration under transient regime through alginate-activated carbon fixed-bed electrode
3. Results and discussion
3.1. Electrochemical characterization of the 3D porous electrodes
3.2. Electrosorption efficiency
CHAPTER II: GRANULAR ACTIVATED CARBON
I. Granular activated carbon based on pomace olive for olive mill wastewater treatment and phenol recovery
2. Results and discussion
2.1. Physicochemical properties of effluent
2.2. Preparation and characterization of GAC
2.3. Adsorption efficiency
2.3.1. Adsorption equilibrium study
2.3.2. Modeling of adsorption kinetics
2.3.3. Effect of pH
2.3.4. Effect of temperature
2.3.5. Fixed-bed column
II. Electrosorption with bio-sourced granular activated carbon electrode for phenols recovery and combination with electrooxidation for residual olive mill wastewater treatment
2. Results and discussion
2.1. Characterization of GAC
2.1.1. Influence of electrochemical conditioning on GAC characteristics
2.1.2. Electrochemical characterization of the conditioned GAC
2.2.1. Phenol electrosorption
2.2.2. Effect of phenolic compounds properties
2.2.3. OMWW electrosorption
2.3. Electrochemical regeneration of the GAC electrode
2.4. Electrochemical degradation of OMWW