Capacitive deionization for wastewater treatment: Opportunities and challenges

Capacitive deionization for wastewater treatment: Opportunities and challenges

Chemosphere 241 (2020) 125003 Contents lists available at ScienceDirect Chemosphere journal homepage: www.elsevier.com/locate/chemosphere Review C...

2MB Sizes 0 Downloads 12 Views

Chemosphere 241 (2020) 125003

Contents lists available at ScienceDirect

Chemosphere journal homepage: www.elsevier.com/locate/chemosphere

Review

Capacitive deionization for wastewater treatment: Opportunities and challenges Ayelet Kalfa a, b, Barak Shapira a, b, Alexey Shopin a, b, Izaak Cohen a, b, Eran Avraham a, b, *, Doron Aurbach a a b

Department of Chemistry, Bar-Ilan University, Ramat-Gan, 5290002, Israel The Institute of Nanotechnology and Advanced Materials, Bar-Ilan University, Ramat-Gan, 5290002, Israel

h i g h l i g h t s  The utilization of CDI processes for wastewater treatment is reviewed.  Selective separation of heavy metals by typical CDI processes is still a great challenge.  Electro-oxidation of organic pollutants during CDI processes should be promoted much further.  Electro-sorption processes may be also exploited for bacteria cells removal from feed water.

a r t i c l e i n f o

a b s t r a c t

Article history: Received 4 August 2019 Received in revised form 24 September 2019 Accepted 27 September 2019 Available online 27 September 2019

Capacitive deionization (CDI) is an emerging method for removal of charged ionic species from aqueous solutions, based on electrostatic interactions between (mostly) inorganic ions and porous carbon electrodes. Inspection of recent publications related to CDI processes, revealed that the majority of the publications are related to the removal of salt (NaCl) from the water (desalination) or electrosorption processes. However, such a water desalination is only one process in the improvement of the quality water, it is interesting to review the literature in the context of CDI processes for other water treatment processes. Herein wastewater treatments are discussed. In this paper, we critically review the last publications that relate to capacitive deionization with wastewater treatments. Since wastewater treatments may involve broad aspects, we address in this review four specific water treatment processes that are thought to be connected with CDI processes: organic fouling of CDI cells, removal of heavy metals by CDI processes, removal of organic micropollutants with CDI processes and disinfection with CDI processes. We also evaluate herein the status of several research efforts in this area and suggest future directions. © 2019 Elsevier Ltd. All rights reserved.

Handling Editor: Y Yeomin Yoon

Contents 1.

2.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1.1. Wastewater treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1.2. Capacitive deionization (CDI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1.3. Organic fouling in CDI cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 1.4. Removal of heavy metals with CDI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 1.5. Removal of organic micro-pollutants with CDI processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 1.6. Disinfection with CDI processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 Concluding remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

* Corresponding author. Department of Chemistry, Bar-Ilan University, RamatGan, 5290002, Israel. E-mail address: [email protected] (E. Avraham). https://doi.org/10.1016/j.chemosphere.2019.125003 0045-6535/© 2019 Elsevier Ltd. All rights reserved.

2

A. Kalfa et al. / Chemosphere 241 (2020) 125003

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

1. Introduction 1.1. Wastewater treatment Pollution of ground and surface water in many places throughout the world has become a major environmental problem in the last decades. The main cause of water pollution is anthropogenic contaminations, which include industrial waste, sewage and fertilizers. Wastewater in general are all water that come from industrial and municipal uses. Municipal solid waste species come from homes, institutions and small businesses, while industrial waste streams can originate from agricultural sources and factories (e.g. the textile industry). Efficient wastewater reclamation policy not just enables to reduce water shortage but also helps to conserve natural water resources. Recycled wastewater can also be considered as a pseudo sustainable source of water, which does not depend on rainfalls, thus increasing the reliability of water supply (Garcia and Pargament, 2015). Some countries put an ambitious goal to meet up to 60% of total water demand by development and assimilation of wastewater reclamation (municipal and industrial) technologies (EPA, 1998). Wastewater treatments include primary e physical processes, which involve sedimentation of solid waste species moieties in the used water (Removal of suspended soli, 2007). Wastewater is passed through several tanks and filters that separate fresher water from the contaminants. This primary batch of the sludge may contain nearly 50% of suspended solids within wastewater (Liang et al., 2018). The secondary treatments include activated sludge processes, trickling filters or bio-filters, oxidation ditches, and rotating biological contactors (RBC) (Removal of suspended soli, 2007). A combination of two of these processes in series is used in municipal wastewater containing high concentrations of organic materials from industrial sources (Removal of suspended soli, 2007). Oxidation of the organic contaminants in the wastewater is usually promoted by biological treatments, such as “activated sludge processes.” Sludge is mixed with millions of actively growing singlecell microorganisms (mostly bacteria and protozoa) which clean the water from the organic contaminants through their natural metabolic activity. Trickling or bio-filters (EPA, 1998) (Removal of suspended soli, 2007) which are also used at this stage of the general purification process, include basins or towers filled with support media such as stones, plastic shapes, or wooden slats. Tertiary treatments can be applied by demand to specific wastewater constituents, which cannot be removed by secondary treatments (Removal of suspended soli, 2007). In this last step phosphate and nitrates are removed using activated carbon or sand filtration and the water is disinfected with chlorine, chlorine dioxide, sodium hypochlorite and chloramine, ozone or UV radiation (Removal of suspended soli, 2007). It is important to note that there are more advanced wastewater treatments but those abovementioned three steps are the basic and traditional ones. 1.2. Capacitive deionization (CDI) CDI, as a technology for brackish water desalination, has drawn much attention in the last decade, providing promising features such as cost effectiveness, being environmentally friendly and

chemical free process and operate under low pressure (Liang et al., 2018; Porada et al., 2013; Oren, 2008; Ryoo et al., 2003; Seo et al., 2010; Li and Zou, 2011; Zhang et al., 2012; Zhao et al., 2012; Wang et al., 2014; Mayes et al., 2010; Li et al., 2010a). The basic idea of a CDI process is illustrated in Fig. 1a. By applying a potential difference to the CDI cell (which comprises porous, high surface area electrodes), anions and cations are adsorbed on the oppositely charged electrodes (usually a voltage difference up to 1.2 V can be applied between the electrodes). Desalted solution is driven then out from the cell as a product, under the constant feed flow. In the next step, the electrodes are discharged back (at this step, the electrostatic energy can be partially recovered), desorbing the salt out from the interfacial double layer region (and the pores) of the electrodes into the bulk liquid phase to obtain a concentrated solution, which is disposed as waste. In general, CDI electrodes are made of porous carbon materials with high specific surface area, usually commercial activated carbon. However, it is important to mention that major efforts are made in the fabrication of innovative high surface area carbon electrodes, with superior characteristics such high ion transport and extreme ion removal capacity, in a variety of forms such as MOFs-derived carbon electrodes (Shen et al., 2018; Wang et al., 2017a; Yan et al., 2018), hierarchical porous carbon architectures (Zhao et al., 2016, 2017; Zhang et al., 2017), and Graphene-based electrode materials (Wang et al., 2016; Liu et al., 2016a; Duan et al., 2017). There are some architectures of CDI cells, where the prevailing architectures adopt a flow by mode (Liang et al., 2018), where the solution flows between the parallel electrodes. A flow through mode (Liang et al., 2018), where the feed solution flows through the electrode (Fig. 1b) is also relevant and provide some advantages. In typical CDI cells the electrodes’ material is carbon based materials (usually activated carbon) with high electrical conductivity and porous structure providing the high surface area (Oren, 2008). In some cases, ion exchange membranes are incorporated in CDI cells (in the front of the electrodes (Zhao et al., 2013)) in order to prevent co-ion expulsion from the electrodes (in reaction to the cell’s polarization, at the expense of the desirable adsorption of counterions) to the solution and thus to enhance the energy efficiency of the process (termed as membrane CDI (MCDI. Fig. 1b). Because the CDI processes involve mainly electrostatic interactions, with no chemical reactions that may be irreversible, activated carbon in CDI cells can undergo an impressive number of charge/discharge cycles. This of course true if the voltage applied to the cells do not push the electrodes’ potentials to high values which endanger the stability of the system (water splitting, carbon oxidation etc.) CDI, as a water treatment method, was extensively investigated especially for desalination of brackish water. However, removal of ions from water is only one process in the efforts to improve the quality of water. In this review we aim at exploring CDI processes for water treatment beyond just ions removal. At first, we address the consequences of long term exposure of CDI cells to organic matter in the water, which is a key factor in the sustainability of CDI processes in wastewater environment. We also explore to which extent CDI processes can be selective toward removal of heavy metal ions. Next, we try to address electro-oxidation reactions along CDI processes e “parasitic reactions or a strategy”? Finally, we try to understand if CDI processes can be exploited for

A. Kalfa et al. / Chemosphere 241 (2020) 125003

3

Fig. 1. a)- Illustration of the principles of CDI processes for salt removal from water. At charging step, anions and cations are adsorbed on the oppositely charged electrodes, and desalted solution is driven out from the cell as a product. At the discharge step, ions are desorb out from the interfacial double layer region and pores systems of the electrodes into the bulk liquid phase to obtain a concentrated solution, which is disposed as waste. b)- Different CDI cells configurations. A e Flow-by CDI cell, B- Flow-through CDI cell and C- Flowby CDI cell containing ion exchange membranes attached to the electrodes.

disinfection purposes and to which extent the electrostatic interactions are involved with antibacterial activity. 1.3. Organic fouling in CDI cells The inevitable existence of dissolved organic matters (DOMs, in some places expressed as total organic carbon, TOC, or natural organic material, NOM, depending on the detection methods) in the primary or secondary wastewater effluents (mostly nonbiodegradable DOMs) brings up an important question regarding the consequences of long term exposure of CDI cells to DOMs or, in general, the functioning of CDI in real wastewater environment (especially secondary effluents). In the work, reported by Wang et al. (2017b), a typical CDI cell consists of commercial AC electrodes was operated in a batch mode fashion, where the background solution originated from a municipal sewage plant (after a biological treatment). It was shown (taken from a close inspection of the 8th charge - discharge cycle e although it is not directly inferred from the paper), that in comparison to aqueous NaCl (featuring the same conductivity), the performance was lower, both in terms of desalination capacity and desalination rate (i.e. electroadsorption rate). Moreover, after 8 charge-discharge cycles a depletion at the TOC concentration was (Fig. 2) observed, implying on what is proposed therein as a competitive electrostatic adsorption of surface charged DOMs, which, in turn partially blocks the access of inorganic ions into the porous structure of the electrodes. Another important observation was that the CDI cell can be regenerated by washing the electrodes with sodium hydroxide solution. It was assumed (supported by three dimensional fluorescence measurements on the washing water) that the humid-like substance is the main “fouling agent”. It was also hypothesized that

þ 2þ Fig. 2. eIon amounts desorbed from electrode of four ions (NHþ and Mg2þ) 4 , K , Ca in each discharge cycle (Wang et al., 2017b).

the alkaline environment promotes ionization of humic-like substances’ surface groups and thus, in turn, increases the solubility of adsorbed humic like substance in the washing solution. M. Mossad et al. (Mossad and Zou, 2013) Investigated the effect of prolong cycling of typical CDI cells (commercial activated carbon electrodes, single pass mode of operation), where the feed solution contains organic salt (humic acid as the organic matter-denoted therein as organic “folluant”) on CDI performance, applying two humic acid feed concentrations (1 mg/L and 3.1 mg/L TOC). Instead of the conventional discharge step, where the CDI cells are shortcircuited, two consecutive steps of flushing and opposite polarity were considered, denoted therein as a “regeneration step”. It was observed that salt removal efficiency started to decline after about 17 h of operation applying feed solution containing 1 mg/L TOC of

4

A. Kalfa et al. / Chemosphere 241 (2020) 125003

humic acid. When the CDI feed stream contained higher TOC concentration (3.1 mg/L TOC humic acid salt), the TDS removal efficiency decreased gradually from 85% to 70.2% after 30 h of operation. It was believed that humic acid accumulated in the pores of the activated carbon electrodes. Moreover, CDI cells did not respond to desorption by reversing the applied voltage during what was denoted therein as the regeneration cycle. In a subsequent work of Zhang et al. (2013)the effect of long term operation of inland brackish water on a commercial CDI prototype unit (LT Green Energy, Adelaide, Australia) was investigated. As expected, a degradation at the CDI performance (shown in Fig. 4) was observed with a decrease at the TDS removal efficiency from 86% during the first day to 55% after 15 days of continuous operation as a cause of accumulation of DOMs on the electrodes (the nature of the DOMs was not provided in this work). It was found that when the TOC in the feed solution was low enough (in ranges that are not detectable by regular TOC analyzers) no fouling impact was observed. Wang et al. showed that washing fouled CDI cells with sodium hydroxide solution (0.001 M) could lead to their full recovery. Recently, ion intercalation (or insertion) electrodes are attracting much attention as an alternative for the conventional carbonaceous materials (especially for Naþ ions removal) (Srimuk et al., 2016; El-Deen et al., 2015). Intercalation electrodes (sodium manganese oxide (Na4Mn9O18 (NMO)), for instance) may exhibit higher ion removal per one-unit mass or volume of the electrode and better ion transport in comparison to conventional carbonaceous electrode materials. X Liu in his work (Liu et al., 2018) explored the effect of the presence of humic acid in the feed water on prolong charge-discharge cycling of NMO intercalation electrode. It was shown that in the presence of humic acid (200 mg/L in 200 mM NaCl electrolyte), the NMO electrode experienced over 25% capacity loss over 30 cyclic voltammetry (CV) cycles (between 0.15e0.7 V vs. Ag/AgCl reference electrode). Physical analysis (XRD of the NMO electrode after cycling) and detection of manganese ions in the electrolyte after cycling, led to two possible capacity fading mechanism: hindered Naþ ions diffusion to the insertion interface of the NMO electrode, as a cause of humic acid adsorption to NMO surface, and reduction of available binding sites for Naþ ions caused by facilitated manganese dissolution in the presence of humic acid in the electrolyte. These findings are significant when considering the benefits in the replacement of carbonaceous electrode by intercalation electrodes for water treatment. Seemingly, the interactions between activated carbon electrodes and DOMs are complex, involving intensive and extensive

Fig. 4. TDS removal efficiency over 15 days of continuous operation and after cleaning at day 16 for two different feed water originated from two different lakes (Zhang et al., 2013).

parameters such as the internal structure of the electrodes, their surface functionality, the ionic strength and the pH of the feed solution, the applied potentials, etc. For instance, we did not find any differentiation between the impact on positive and the negative electrodes in the CDI cells. Moreover, cleaning or recovering procedures of flushing with caustic solutions deserve more attention. For instance, when dealing with membranes technology, the long term impact of exposing carbonaceous materials to alkaline or acidic solutions should be studied carefully, because aside to their cleaning capability, such solutions can challenge the membranes stability. CDI cells’ performance is strongly correlated with the electrodes’ surface properties. Any variation in the nature of the electrodes’ surface groups (e.g., the ratio between acidic and basic groups) as a cause of exposing the electrodes to an environment which pH is not neutral, might affect pronouncedly the performance of CDI processes (Gao et al., 2016). For instance, the sustainability of carbonaceous electrodes in organic matter-containing solutions is a key point regarding CDI processes for water treatment in general and for wastewater in particular. It is clear that more indepth research about the effect of DOMs on CDI cells and their electrodes (sort and long term exposure) is required in order to enable the full application of CDI processes for treating different types of water streams. Highly important is the development of effective methods for cleaning and recovering CDI cells that were operated with water containing DOMs. 1.4. Removal of heavy metals with CDI

Fig. 3. Product flow rate of CDI unit with time when processing different feed solutions. F1 solution containing NaCl only, F2 containing low TOC concentration 1 mg/L, F3 containing higher TOC concentration 3.1 mg/L, F6 containing ferric ion of feed solution containing 1.8 mg/L TOC (Mossad and Zou, 2013).

The presence of heavy metal ions in wastewater streams poses a major health and environmental concern. Heavy metal cations in water do not degrade naturally and therefore, the disposal of wastewater containing toxic metals should be made carefully under strict regulations. Moreover, heavy metals tend to accumulate in living organisms. Metals that are digested beyond an allowable concentration, may lead to serious health disorders. The most commonly toxic metals in wastewater are arsenic, lead, mercury and cadmium, where less abundant toxic metals are copper, nickel and zinc. It is more likely to find traces of toxic metals in industrial wastewater, where the main industrial sources are metal finishing and plating factories, semiconductor manufacturing, textile manufacturing and landfills. Removal of heavy metals from wastewater can be accomplished by chemical and physico-chemical methods, such as chemical

A. Kalfa et al. / Chemosphere 241 (2020) 125003

precipitation, coagulation, adsorption with activated carbon powder, ion exchange, solvent extraction, floatation and also by membrane filtration (Wan Ngah and Hanafiah, 2008). Seemingly, since CDI treatment methods are mostly based on the electrostatic interactions between charged species and electric charge across the electric double layer, CDI could be the ultimate solution not merely for removal of hazardous metals such as chromium or lead, but also for simple regeneration of these metals. Simple recovery of metals is highly desirable from both environmental and economic considerations. However, we must keep in mind that heavy metals co-exist together with alkali metal ions, where the latter are usually dominant in wastewater streams. Therefore, when using CDI for capturing heavy metals, it is important to make sure that their trapping is not interfered by adsorption of other (non-toxic) metal ions. From viewing the basic models of the electric double layer (The Stern Model (Oldham, 2008)or more advancing model for nanoporous structure (Amphoteric modified Donnan model (Biesheuvel et al., 2014), there is no attribution to specific type of ions (or more precisely ions surrounded by their hydration shells). Although a model for the electric double layer that take into account also the size and valence of ions was recently suggested (Suss, 2017), it is very challenging to design electrodes and CDI processes that display an affinity toward certain types of ions in any kind of water streams. The aim of this section is to review the latest publications on heavy metal ions removal and/or recovery using CDI processes, to understand what can be the origin of a selective affinity toward specific metal ions (if there is relevant data) and to find out if indeed any preference toward adsorption of specific heavy metal ions was ever shown. It is important to distinguish between ion selectivity that is observed after the EDLs reach an equilibrium and ion selectivity that is observed during the charging step, attributed by R. Zhao et al. (2013) in his work, as a ’’time-dependent ion selectivity’’. It was shown that in typical CDI cells (meaning no special surface treatment or membrane insertion) containing feed water with NaCl/CaCl2 mixtures, when the concentration of Naþ ion is five times the Ca2þ ion concentration, right after voltage application, the majority monovalent Naþ cations are preferentially adsorbed in the EDL, and later, they are replaced by the minority, divalent Ca2þ cations. Supported by a theoretical study, it was shown that time dependent ion selectivity is governed by an interplay between transport resistance of ions outside the EDL and the voltagedependent ion adsorption capacity of the EDL. However, in the below mentioned works, although it is not directly inferred, we take the assumption that the ion selectivity is observed after the EDLs reach an equilibrium. In the work of Huang et al. (2016) the preference of CDI processes toward the removal of different ions: cadmium, lead and chromium was studied (with a feed solution concentration of 0.5 mM). It is important to notice that the electrodes in the CDI unit used by them were based on commercial activated carbon cloth (ACC, FM70) what means that no further manipulation or surface treatment were applied to the electrodes. Moreover, adsorption experiments were held at 0 V (between electrodes) in order to distinguish between physical adsorption (or more precisely electroadsorption at the Potential of Zero Charge (PZC (Wang et al., 2017b))) and electro-adsorption under potential application. Interestingly, Pb2þ ions showed high intensity of physical adsorption (nearly 29%). Since the adsorption rates in that work were provided in %, the translation to mg/g or mol/g is difficult). However, when solutions containing Pb2þ, Cd2þ and Cr3þ ions was introduced to the CDI unit, the adsorption of Cd2þ ions was inhibited, with around 46% and 45% adsorption rates for Pb2þ and

5

Cr3þ ions, respectively and only around 13% for Cd2þ ions (compared to adsorption of 31% for Cd2þ ions when being the single ions in the feed stream). This inhibition of Cd2þ ions adsorption was attributed in that work to their relatively lower charge density. This work may shed some light on the adsorption preference of ions related to their effective size and charge. However, since that work did not address the porous structure of the electrodes (no information on pore size distribution and pores shape) and their surface properties. Hence, despite the interesting finding, it was hard to conclude from it about the adsorption priority of Pb2þ over Cd2þ ions, although the former ones exhibited a more preferred adsorption affinity. In the work of Fan et al. (2016), the feasibility of arsenate (H2AsO 4 ) and arsenite (H3AsO3) species removal by CDI processes was studied. When arsenate moieties were presented in high concentration (200 mg/L), a considerable high electro-adsorption capacity was obtained (10.31 mg/g). Surprisingly, when the uncharged form of arsenic - arsenite, was presented in the feed solution, the electro-adsorption capacity was relatively high around 7.57 mg/g. The removal of arsenite was attributed to two consecutive electrochemical processes during potential application: oxidation of arsenite to arsenate at the positive electrode and electro-adsorption of arsenate. However, it was found that the presence of co-existing ions reduces the electro-adsorption capacity of arsenic compounds in the water (with a gradual decrease of the arsenic electroadsorption capacity as the NaCl concentration in the solution was higher) showing no arsenic compounds adsorption preference in the presence of co-existing inorganic ions and NOM (natural organic matter). Li et al. (2010b) investigated the electro-adsorption capacity toward ferric ions in comparison to different cations using graphene nano flakes as the electrode material in CDI processes. It was observed (shown in Fig. 3) that the electro-adsorption capacity of graphene nano-flakes followed the order of Fe2þ>Ca2þ>Mg2þ>Naþ with salt adsorption capacity of 0.62, 0.55, 0.52 and 0.45 mg/g respectively (the feed solutions’ conductivity was 50 mS/cm for all the systems studied therein (Fig. 5)). As suggested by Huang et al. (2016), Li et al. (2010b) also hypothesized that the electroadsorption preference trends are proportional to the ions valence and inversely proportional to ion effective size. However, there was

Fig. 5. eThe effect of ionic species on removal of conductivity with CDI cells employing graphene nano flakes as the electrodes. The applied voltage and the initial solutions conductivity were 2.0 V and 50 mS/cm, respectively. The feed solutions contained NaCl, CaCl2, MgCl2 and FeCl3 as indicated (Li et al., 2010b).

6

A. Kalfa et al. / Chemosphere 241 (2020) 125003

no attribution to electrodes’ structure, and in particular, to the graphene two-dimensional structure in this respect. Liu et al. (2017) adopted another approach of combined removal of alkali metal and lead ions by two different mechanisms. In this work, synthesized 3D graphene electrodes served as both electroadsorbent substance and as a matrix for deposited chelating agents (EDTA) for Pb2þ ions removal. Experiments with asymmetric CDI cells consisting 3D graphene electrode as the positive electrode and EDTA grafted 3D graphene as the negative electrode showed high removal efficiency (about 99%, at pH 6 in a wide range of concentration of Naþ and Pb2þ ions, 5e100 ppm) for Naþ and Pb2þ ions. Moreover, since the adsorption mechanism for Naþ and Pb2þ ions is different, the different regeneration processes (short-circuiting for Naþ ions and acidic elution solution flushing for Pb2þ ions) enables efficient recovery of lead ions, such as the regeneration can take place during different stages. In fact, the approach suggested in this work is a combination of ion exchange and CDI in one system. A schematic illustration of the separation and recovery of heavy metal cations is illustrated in Fig. 6. However, it was interesting to compare the ion exchange capacity toward Pb2þ ions of the synthesized EDTA grafted 3D graphene to commercial ion exchangers. Huang et al. (2016) explored the electrosorption capability of copper ions with typical CDI cells comprising activated carbon electrodes. Since copper ions can undergo electrochemical reduction within the window potentials of water electrolysis, two mechanism for copper ions should be considered; electrodeposition and purely electrostatic interactions, where the former is involved with a charge transfer from the negative electrode to the copper ions. Anyhow, the applied potential between the electrodes dictates which from the two mechanism the dominant is. It was found that at potential difference of higher than 0.8 V between the electrodes, Cu2þ ions removal is involved also with faradaic reactions. It was also demonstrated that with a mixed solution of

50 mg/L Cu2þ ions, with varying concentration of NaCl background solution (1, 10 and 100 mM, at 0.8 V) the electrosorption capacity of Cu2þ ions showed only slight decay (2.21, 2.39 and 2.48 mg/g correspondingly), which is a very impressive finding (the electrosorption capacity for Naþ ions was not provided). The high preferential selectivity for Cu2þ ions over Naþ ions was attributed to the differences at the valence of the ions. Though the main concept beyond CDI separation processes are pure electrostatic interactions, faradaic interactions also can contribute to the total separation process and also can be the dominant mechanism as shown in this work. However, the authors are a somewhat skeptical, given that the high selectivity between copper ions and sodium ions was solely attributed to discrepancy between the ion valences, especially inspecting other works in this regard. When a faradaic reaction is involved with one of the electrodes, the CDI separation process should be carefully inspected. The first implication is asymmetric potential distribution between the electrodes. For instance, the negative electrode (in this case) may undergo only electrodeposition process whereas the electroneutrality is conserved by electrostatic adsorption of anions on the positive electrode. Moreover, choosing appropriate potential is crucial in this regard, since mass transport of copper ions to the electrodes may be suppressed, especially when Cu2þ ions are presented in the water at low concentration. Without derogating the important findings in this work, we recommend on more extensive work on selective copper electrosorption using three electrodes CDI cells, at which the individual potential of the electrode can be fully controlled. In this way, differentiation between faradaic and nonfaradaic reactions can be reached more easily. We can refer to the work of Shapira et al. (2016)as an appropriate example. Utilization of faradaic reactions in conjunction with purely electrostatic interactions for selective removal of ions can be found in a work by Avraham et al. (2011) work. In this proof of concept

Fig. 6. eA Schematic illustration of separation and recovery of heavy metal ions alkali metal ions and chlorides from wastewater by CDI employing EDTA grafted 3D graphene electrodes. 3D macroscopic graphene (3DEGR) is used as the cathode, where EDTA on 3DGR acts as a chelating group to chelate heavy metal ions. Meanwhile, salt ions are adsorbed into the 3DEGR pores via electrostatic adsorption when an external voltage is applied (Liu et al., 2017).

A. Kalfa et al. / Chemosphere 241 (2020) 125003

paper, boron (in the form of boric acid, 480 mg/L, background electrolyte NaCl 1300 mg/L) was introduced to flow through CDI cells comprising activated carbon cloth electrodes (ACC-507-15). It was hypothesized that by the virtue of faradaic reaction on the negative electrode (mostly oxygen reduction) the local pH in the negative electrode’s pores becomes high enough, in a way that boric acid is transferred into the charged form e borate ions, which is further electro-adsorbed on the positive electrode (the removal process of boron is illustrated in Fig. 7). A removal of boron at a level of 30% was observed, like NaCl ions. However, more extensive work is needed in order to explore the feasibility of CDI processes in the context of boron removal from wastewater streams. Although, by definition, ammonia is not regarded as heavy metal, it is worth mentioning Wang et al. work (Kim et al., 2013) in which a new method for nitrogen removal and recovery (in the form of ammonia) was proposed by the coupling of MCDI and ion exchange processes. The rational behind this concept relied on the preferential selectivity for active metals ions (Kþ, Ca2þ and Mg2þ) over NHþ 4 in the course of the MCDI process, in which the feed solution enters the MCDI cells prior to the ions exchange process. The concentration of NHþ 4 (in relation to other ions, in the feed solution) helps to mitigate competition between NHþ 4 and other ions during the ions exchange process. It was demonstrated that the concentration of active metal ions decreased substantially from 8.06 mmol/l to 1.19 mmol/l, where the concentration of NH4þ decreased slightly from 1.44 mmol/l to 0.6 mmol/l during the MCDI process. It is important to mention that the NH4þ ions concentrating process in the first stage, is followed by several chargedischarge cycles in a batch mode fashion, where the effluent in the charging step is collected and separated from the feed solution. It was shown that by coupling MCDI and ions exchange processes in series, a recovery of 65% of ammonium cations can be attained. This work resembles the work of Liu et al. (2017) in a way that

Fig. 7. Illustration of the removal process of boron from water by flow through CDI cells. (1) Boron contaminated water first flow through the negatively polarized electrode, wherea Certain basic pH environment is developed adjacent the negative electrode. (2) The equilibrium B(OH)3 þ 2H2O<> B(OH) 4 þ H3Oþ, adjacent the electrode, is driven toward the formation of borate ions. (3) The borate ions are enforced to flow to the positively polarized and being electro-adsorbed. Note that there are two possibilities for keeping the electroneutrality of the solution: 1. for each borate ion which is electro-adsorbed onto the negative electrode, a hydronium ion is generated by the positive electrode. 2. For each hydroxide ion which is generated by the negative electrode, chloride ion (in this case) is being electro-adsorbed onto the positive electrode (Avraham et al., 2011).

7

combination of CDI process with other existing techniques (membranes based) may result with a considerable selective separation. Removal of anion pollutants with CDI processes. An appropriate content of ions or mineral in drinking water is necessary for human body. However, anions such fluoride, nitrate and bromide, above a certain level, can cause major problems. For instance, fluoride pollution in drinking water recognized to be beneficial as in concentration between 0.5 and 1.0 mg/L. Above this concentration, it can cause permanent bone and joint deformation and dental and skeletal fluorosis. Excess of nitrate in drinking water might lead to methemoglobinemia in infants and alimentary canal cancers with the maximum NO 3 (Yue et al., 2016). World Health Organization (WHO) advises that nitrate concentration in drinking water should be maintained under 10 mg/L (Guidelines and Quality, 1995). It is wide accepted, that electroporation capacity is strongly dependent on ionic charge and hydrated radius. In Yu-Jin K. et al. (Kim et al., 2013), selective removal of nitrate ions was achieved with nitrate-selective composite carbon electrodes (NSCCEs) for use in capacitive deionization processes, in order to selectively remove nitrate ions from a solution containing a mixture of anions. The anion mixture was chloride and nitrate ions, in concentration of 5 mM and 3 mM, respectively. Although chloride ion has better migration capabilities, nitrate anions a adsorption to the composite carbon electrodes were 2 fold higher compared to chloride anion. In the study of Zhaolin C. et al. (Chen et al., 2015), e the effect of two-parameters were investigated: the anion charge and ionic radius on electrosorption of activated carbon in CDI processes. The ions that were selected for evaluation were chloride (Cl), nitrate 2  (NO ) and phosphate (PO4 3). Ex3 ), fluoride (F ), sulfate (SO4 periments were carried out in both single and mixed electrolyte solutions. The authors showed that hydrated radii is a crucial parameter for electrosorption, given same ion valceny. In an experiment of different anion electrosorption capacity the molar saturation ca2  pacity of Cl,NO and PO4 3 were 164.8, 159.2, 138.5, 87.7 3 ,F ,SO4 and 74.3 mmol/g, respectively (Fig. 8). The smaller hydrated radii the higher electrosorption capacity. In the work of Cohen et al. (2018) a selective removal and recovery of bromide from a mixed chloride and bromide solution, in asymmetric CDI cells design (illustration is provided in Fig. 9), was demonstrated. In contrary to most of the abovementioned works,

Fig. 8. Electrosorption capacity of different anions using the activated carbon electrodes at applied voltage of 1.2 V (Chen et al., 2015).

8

A. Kalfa et al. / Chemosphere 241 (2020) 125003

Fig. 9. A schematic illustration of the asymmetric CDI cell design for selective bromide ions recovery (Cohen et al., 2018).

purely electrostatic interactions play a minor role in the selective separation mechanism for the halide ions. The mechanism at which bromide ions are selectively removed from solution involves two consecutive steps: at the first stage, bromide ions are electrooxidized on the positive electrode to form bromine (taking advantage from about 200 mV difference between the standard oxidation potentials of bromide and chloride ions), following physical adsorption of bromine within the porous structure of the activated carbon electrode. A separation factor of 1:80 was demonstrated. Selective separation of heavy metals by CDI processes as described above is still a great challenge in the broad field of water treatment. Although some attempts were made, it seems that most of the works are under proof-of-concept stages. We think that a standardization for the term “selectivity” in CDI processes should be agreed upon first (separation factors concentration thresholds). This will probably facilitate comparison between similar works (now being hardly possible). Moreover, we think that a more emphasis should be given to development of selective ions exchange membranes (employing MCDI configuration). The rationality and practicality in the design of selective ions exchange membranes is much simpler than design and fabrication of selective electrodes through their complex porous structure. 1.5. Removal of organic micro-pollutants with CDI processes Organic micro-pollutants may relate to disinfection by-products (DBPs), pharmaceuticals, pesticides and endocrinal disrupting compounds (EDCs) where the latter are known to interfere badly with the action of hormones in endocrinal systems of living creatures (Snyder and Benotti, 2010). Traditional wastewater treatment technologies are generally not effective in removing these micropollutants, because many of them are hard to separate and are resistive to biological degradation. Membrane filtration technologies (especially nano-filtration and reverse osmosis) may provide a relatively effective barrier for almost all organic matters. The secondary pollution by accumulated micro-organic species in concentrated streams cannot be neglected. However, by applying electro-oxidation methods for organic pollutants removal, a full mineralization of the organic matter can be completed. The oxidation can be direct, i.e. direct exchange of electrons between the organic matter with the electrodes, or by indirect oxidation with the assistance of red-ox mediators. For instance, in advanced oxidation processes (AOP (Oturan and Aaron, 2014)) the catalytic generation of highly reactive non-selective hydroxyl radicals is the main mechanism toward organic matter removal or conversion to less harmful intermediates. As a good option, the source of hydroxyl radicals may be a catalytic reaction between hydrogen peroxide (generated by oxygen electrochemical reduction) and ferric ions (Fenton reaction) (Neyens and Baeyens, 2003). In this section, we aim at inspecting whether using electro-

oxidation reactions simultaneously to purely capacitive interactions can be considered as a good strategy for purifying water steams which are polluted by organic species. In the work of Liu et al. (2016b), the effectiveness of CDI (bench scale CDI cells comprising commercial activated carbon cloths (ACC, FM10, Chemviron, UK) on disinfection by products (DPB) precursors control was evaluated. The premise in this work was that DBP precursors should be eliminated (at least partially) before a disinfection step. DPB precursors models in this work were synthetic humic acid and bromide containing saline water (bromide in water may react with chlorine to form hypobromous acid (HOBr), which may then react with organic matter (for instance humic acid) and generate toxic brominated DBPs (Sivey et al., 2013)). Therefore, the ratio between bromide to organic matter (provided as dissolved organic carbon (DOC)) is an important factor. In the course of CDI treatment processes (applying different potentials between the electrodes: 0.6, 0.9 and 1.2 V), the CDI treatment was able to bring down the ratio by 26% at potential differences of 0.6 V (initial concentrations were 1.78, 9.98 mg/L for bromide and humic acid, respectively), indicating the preferential removal of bromide over the weakly charged humic acid (similar trend was observed at higher potentials). These findings are in agreement with previous works (Wang et al., 2017a; Yan et al., 2018). What is more relevant to the topic discussed in this section, is that at a potential difference of 1.2 V, the reduction of humic content was observed in a way that cannot be associated with electro-sorption mechanism. From the DBP formation potential (FP) tests (which detailed description can be found therein (Liu et al., 2016b)), It is deduced that electrochemical oxidation reactions at the electrode surfaces are responsible for breaking down the humic acid into smaller and more hydrophilic substances. The importance of this work to the topic discussed in this section, is that electro-oxidation of humic acid is likely to occur within the potential window of water electrolysis. However, this finding was considered in this particular work more like as parasitic reaction with adverse implications, rather than a strategy toward reduction of DPB precursors in the water. The authors think that the question whether electro-oxidation of organic matter in the course of CDI processes is mostly a parasitic reaction or rather can be used as a strategy for removing organic pollutants, deserves more extensive research. This brings us to the work of Duan et al. (2015), in which the possibility of simultaneous removal of organic matter and inorganic salt by combining electrochemical oxidation and electrosorption was investigated. Phenol and sodium chloride were chosen as representative of organic pollutants and inorganic salts. These mixtures were examined in typical CDI cells (comprising activated carbon electrodes). The simultaneous removal was believed to be accomplished by applying high potential difference between the electrodes (2 V) with oxygen flow to the cell. Over 90% of phenol, 60% of TOC and 20% of salinity were removed during 300 min of electrolysis time (with a feed solution of 17.1 mM NaCl and 1.1 mM phenol). Moreover, it was shown that the phenol removal rate was accelerated with the increase of the applied voltage and in the presence of oxygen in the water. As expected, two mechanisms for phenol removal were assumed: direct oxidation, in which the electrodes exchange charge with phenol and indirect mechanism, where active red-ox mediators such as hypochlorite and hydrogen peroxide are formed first electrochemically and then react chemically with phenol. Inspection of the intermediate products (benzoquinone, catechol and chloro-phenol moieties) suggested that the removal process involves both direct and indirect oxidation, where the dominant mediator is NaClO (and not H2O2). Indirect oxidation of organic matter may sound as the ultimate

A. Kalfa et al. / Chemosphere 241 (2020) 125003

choice, such as the oxidation reaction does not necessarily take place at the electrode interface. T. Kim in his work (Kim et al., 2016) tried to quantified the hydrogen peroxide generation of CDI cells (occupying commercial activated carbon fibers, two pairs 50 mm diameter each) in a flow mode operation. It was revealed that under conditions of flow rate of 10 ml/min, voltage application of 1.2 V and charge/discharge cycle of 6 min, an accumulation of approximately 0.1 mg mg of H2O2 was detected after 50 charge-discharge cycles. These findings led to the conclusion that the generation of hydrogen peroxide is inevitable on the cathode due to the thermodynamic redox potential under neutral pH conditions and can explain the increase at pH of the effluent during the charging step. However, we do not find any attribution to the role or type of possible functional surface groups that may facilitate oxygen reduction on carbon surface (Srimuk et al., 2017). We can refer to Yeager (1984) work on the role of oxygen functional groups in this respect. The type and origin of functional surface group play a major role in the catalytic activity of the carbon toward oxygen reduction into peroxide. Getting back to the question of whether electro-oxidation reactions in CDI processes are parasitic or can be considered as a desired strategy, the stability of the CDI cells at too high applied voltages may be a major issue. Indeed, a noticeable degradation in the performance was observed in the above described processes even after 5 charge-discharge cycles. However, after flushing cycled cells with a 0.1 M NaOH solution and some cycling at reverse polarity of the CDI cells which were fully symmetrical (i.e., similar positive and negative electrodes produced from the same activated carbon), a full recovery of the cell is obtained. Since CDI is essentially an electrochemical method, the combination of electrochemical reactions, other than purely electrostatic interactions is prompted in CDI processes for water treatment. The challenge is to control and distinguish between desirable and parasitic reactions. Since the porous carbon electrodes, by the virtue of their high surface area, can play as substrates for the insertion of designated catalysts, we think that a combination between electrostatic and catalytic reactions is very promising in this regard. Moreover, long term stability of CDI cells should be developed (quite possible), especially when high potential differences are applied between the electrodes. 1.6. Disinfection with CDI processes In order to limit the effects of organic materials and suspended solids in water streams before their distribution, disinfection processes are usually the final stage in wastewater quality treatment. The addition of chemicals such as ozone (preferentially in combination with ultraviolet radiation) and chlorine in the disinfection stage is widely used. Chlorination is the most abundant technology for disinfection of drinking water and wastewater. The main drawback of this technology is the creation of undesirable by-products such as trichloro-methane and chloro-acetic acid which occurs due to interaction of organic molecules with chlorine. These by-products reduces the effectiveness of disinfection (Boyer and Singer, 2005). Also chlorination can create chloro-amine moieties which are suspected to be carcinogenic (Weng and Blatchley, 2013). CDI cells which contain high surface area electrodes can be utilized in disinfection stages by the incorporation of antibacterial agents into the electrodes via their surface modifications. It may be possible to design and operate bi-functional CDI cells that include composite electrodes possessing both antibacterial and electrosorption capabilities. However, since CDI is a dynamic process in the sense of its varying concentration of ions in the vicinity of the electrodes, and

9

the potential across the double layer region during charging cycle, it is interesting to understand, to which extent, these dynamic properties affects the disinfection efficacy, or, in other words, are the CDI processes beneficial or disadvantageous to disinfection processes? Much like in membrane filtration technologies, fouling of the porous electrodes (by formation of a bio-film) is also a big concern and probably will require a regeneration methodology. In the work of Laxman et al. (2015), brackish ground water that was collected from a well in Oman’s Al Musanaah, was introduced into flow through regular CDI cells (consisting of activated carbon cloth electrodes), in a batch mode fashion, following chargedischarge cycling between 0 and 1.6 V. Concurrently with the conductivity measurements, the antimicrobial properties of the CDI cells were evaluated by estimating living cells mortality in the feed (well water) and at the end of desalination process following flow cytometry measurements. At first, an attention was given to the ratio between dead and vital cells in the water after charge and discharge processes. It was found that the vital cells count decreased by 50% while dead cells count after desalination (at the end of the charging step) increased by 90%. Two possible mechanisms for vital cells count reduction were suggested: the high electric field between the electrodes may lead to cells’ walls rupture, or the electro-sorption of bacterial cells which have a negative charge (due to the presence of phosphate and lipopolysaccharides) together with anions, may lead to dehydration of the cells, as an hypertonic environment is built in the vicinity of the positive electrodes. From the inspection of the cells count (dead and vital) in the water after the regeneration step (the dead cells count was similar to that after the charging step and the vital cells count was similar to that of the feed water), it was deduced that electro-sorption is the primary disinfection mechanism and some of the bacteria cells were just reversibly electroadsorbed to the electrode’s surface. Long term effects of the disinfection properties were also investigated by evaluating bacteria colonies forming (CFU) after 1 week in order to evaluate any possible long term bacterial regrowth. It was shown that vital cells count was reduced by 3-fold from 3  104 CFU/ml to 1  104 CFU/ml in the desalinated water. Moreover, CFU calculation was 2  104 CFU/ml in the water after the discharge step, providing more supporting evidence to that the electro-sorption is the dominant disinfection mechanism. Inspection of Field emission SEM (FESEM) images of the positive electrodes after regeneration step, showed no bacterial presence on them. In that work, it was demonstrated that typical CDI cells might have anti-microbial properties. However, the authors think that this primary work should stimulate more extensive and fundamental works in the field of disinfection properties of regular CDI cells. For instance, the question to which extent the electric field that is generated between the electrodes affects the vitality of the bacteria cells should be rigorously explored. Moreover, whether a sufficient hypertonic environment is built in the vicinity of the electrodes and to which extent is an important yet unresolved question. This question can be well addressed by theoretical modeling supported by relevant experiments (which are still missing). In long term regrowth inhibition, the effect of the high osmotic pressure in salinity water, should also be considered. Microbial fouling (and possible electrodes maintenance procedures) should be evaluated along prolonged desalination cycles. The fact that FESEM imaging showed no bacterial presence is encouraging, but can barely provide any clue for long term cycling and long term exposure of CDI cells to microbial environment. However, this work can certainly serve as a good basis for following works that will closely explore the relationship between typical CDI processes and microbial environments (or vice versa). In the work of El Deen et al. (El-Deen et al., 2016) asymmetrically

10

A. Kalfa et al. / Chemosphere 241 (2020) 125003

functionalized microporous activated graphene electrodes were fabricated for combined disinfection and desalination of brackish water. The “asymmetry” is attributed to the different surface modification of the electrodes: carboxylic groups (COO2) and quaternary amine groups (NRþ 4 ), in which the latter are known for their antibacterial properties (killing microbes upon contact by physical disruption of the anionic microbe cytoplasmic membrane). Feed water containing 104 X CFU/ml of Escherica coli (E. coli) was circulated in the asymmetric CDI cells in a single pass mode of operation where a voltage of 2 V was applied between the electrodes. At the end of the charging step, the disinfection performance reached 98.55% killing rates (corresponding to a 1.9 log reduction). Long term cycling (30 cycles) showed good disinfection capability retention with no detected reduction at the bacteria CFU counting at the effluent. No attribution to the effect of cell voltage or electric field on the antibacterial activity is provided in that work. Moreover, how the presence of E. coli affects the desalination efficiency, remained unclear. In contrary to the work of Laxman et al. (2015), the CDI cells in this particular work, played only as substrates for the anti-bacterial quaternary amine deposition and no attribution is given to any electro-sorption processes. Hence the two abovementioned works provide two different mechanisms. A comparison between the antibacterial efficacies is much complicated (injected charge vs. contact time and surface charge concentration), not mentioning that the contact time of the bacteria cells with the electronic charge or chemical surface charge is completely different. However, simple and important experiments (have not been done yet), where the antibacterial activity is evaluated at different applied potentials between electrodes, may qualitatively point out the contribution of the electric field on the disinfection processes. Similar to the El-Deen et al. work (El-Deen et al., 2016), a composite electrodes made by casting an activated carbon electrode with cationic nano-hybrids of graphene oxide (GO)- quaternized chitosan were fabricated for combined desalination and disinfection of water (denoted therein as capacitive deionization disinfection (CDID) electrodes (Wang et al., 2015). A schematic illustration of the combined deionization and disinfection cell is depicted in Fig. 10). The composite electrodes in the CDI cells consist of three main components, the quaternized chitosan (the antibacterial agent), GO and the activated carbon, Where the GO

component functions as an attractive electric conductive carrier substrate by increasing the grafting density and therefore the antibacterial efficacy. Casting thin layer grafted GO onto activated carbon electrodes, helps to maximize the contact area of the quaternized chitosan with the bacteria cells and enables the combined functioning of both disinfection and desalination. Disinfection and desalination experiments were conducted with feed concentration of 106 CFU/ml E. coli, and varying NaCl concentration 100 and 200 mg/L NaCl in single pass operation. The important findings were: 1 CDI cells with only activated carbon electrodes showed 1 log reduction, which supports Laxman et al. (2015) hypothesis in which reversible electro-sorption is the dominant disinfection mechanism in regular CDI cells. 2. An impressive 6 log reduction was shown for the composite electrodes at the first 20 min and declined to about 2 log reduction after 60 min. With moderate bacteria loading (104 log CFU/ml) with alternating cycles of 4 min bacteria capture and 6 min bacteria desorption, 4 log reduction was maintained for at least 5 h. 3. A slight decrease at the time the cell can achieve 99.9999% of E. coli killing (from 20 min to 16 min) was observed, suggesting that accumulation of Cl-ions may adversely affect the cationicity of the qunatrity functional groups and the antibacterial activity. This comprehensive work, provides more supporting evidence of electro-sorption as the prevailing mechanism in disinfection processes in regular CDI cells. Moreover, it was shown that the contribution of bacteria removal by electrostatic interactions, in comparison to removal by the cytoplasmic membrane distruption (by the quaternized ammonium) is low. However, the idea of utilizing the high surface of the carbonaceous electrodes as attractive conductive carrier for deposition of antibacterial agents was impressively demonstrated. Interestingly, microbial fouling was not observed in the abovementioned works. However, it is important to note that the CDI cells were operated for limited number of cycles. Another point that should be stressed is the fact that the CDI cells were operated under high voltages (1.6 and 2 V). In fact, the electrodes themselves can serve as a mediator for generation of biocides like chlorine or peroxides (Kim et al., 2016). Within this frame of working potentials, parasitic reactions such as oxygen reduction (peroxide pathway (Shapira et al., 2016)]) or chloride oxidation may contribute to the anti-bacterial efficacy and such electrochemical reactions should not be ruled out. 2. Concluding remarks

Fig. 10. Schematic illustration of combined deionization and disinfection with composite quaternized chitosan/GO/Activated carbon electrodes in CDI cells (Wang et al., 2015).

In this critical review, a literature outlook on CDI processes with relation to wastewater treatment processes is provided. There is no questioning that CDI can be harnessed for water treatment processes beyond water desalting. Still, the authors underline the need of much more extensive research and development efforts with this respect. The first important point is the impact of NOM in water streams on CDI cells (more particular, the electrodes), which is still ambiguous and deserves clarification. While it is widely assumed that NOM tends to foul the carbon electrodes very quickly, more quantitative analyses should be performed. It is important to understand the dependence of fouling rates of CDI cells operated in contaminated water streams on the NOM type, concentration, applied voltage and the nature of the electrodes (material, structure, porosity). Regeneration of the electrodes also deserves more attention.

A. Kalfa et al. / Chemosphere 241 (2020) 125003

Regardless the tradeoff that should be seriously considered between the ease of operation and cost effectiveness of a CDI process to addition of consumables such as alkaline washing solutions, the impact of regeneration procedures on the electrodes stability should be explored. The issue of electrodes’ durability in prolonged CDI processes is super important. Although in some researches a preference of heavy metal electro-sorption in CDI processes in comparison to alkali metal ions is demonstrated, it hard yet to see any clear, significant electrosorption preference in regular CDI processes towards selective heavy metal ions adsorption that can pave the way for CDI implementation in industrial heavy metals removal from water streams. However, a great potential of CDI processes in removal of heavy metal ions can be accomplished in processes combining CDI and ions exchange by selective membranes. Electro-oxidation of organic pollutants simultaneous to capacitive de-ionization processes should be promoted much further, since CDI is essentially an electrochemical treatment method that can be extended to useful red-ox processes in addition to electrostatic interactions. However, we suggest that efforts towards improving the long term stability of electrodes exposed to organic pollutant and organic intermediates in the course of such oxidation process, should be prioritized before emphasis on the study of oxidation of organic matter in CDI cells. Electro-sorption processes may be also exploited for bacteria cells removal from feed water. However, it was shown that a judicious utilization of high surface electrodes as substrates for deposited antibacterial agents can promote pronouncedly the effectiveness of CDI processes in disinfection of water sreams. References Avraham, E., Noked, M., Soffer, A., Aurbach, D., 2011. The feasibility of boron removal from water by capacitive deionization. Electrochim. Acta 56, 6312e6317. https://doi.org/10.1016/j.electacta.2011.05.037. Biesheuvel, P.M., Porada, S., Levi, M., Bazant, M.Z., 2014. Attractive forces in microporous carbon electrodes for capacitive deionization. J. Solid State Electrochem. 18, 1365e1376. https://doi.org/10.1007/s10008-014-2383-5. Boyer, T.H., Singer, P.C., 2005. Bench-scale testing of a magnetic ion exchange resin for removal of disinfection by-product precursors. Water Res. 39, 1265e1276. https://doi.org/10.1016/j.watres.2005.01.002. Chen, Z., Zhang, H., Wu, C., Wang, Y., Li, W., 2015. A study of electrosorption selectivity of anions by activated carbon electrodes in capacitive deionization. Desalination 369, 46e50. https://doi.org/10.1016/j.desal.2015.04.022. Cohen, I., Shapira, B., Avraham, E., Soffer, A., Aurbach, D., 2018. Bromide ions specific removal and recovery by electrochemical desalination. Environ. Sci. Technol. 52, 6275e6281. https://doi.org/10.1021/acs.est.8b00282. Duan, F., Li, Y., Cao, H., Wang, Y., Crittenden, J.C., Zhang, Y., 2015. Activated carbon electrodes: electrochemical oxidation coupled with desalination for wastewater treatment. Chemosphere 125, 205e211. https://doi.org/10.1016/ j.chemosphere.2014.12.065. Duan, H., Yan, T., Chen, G., Zhang, J., Shi, L., Zhang, D., 2017. Porous Graphene Frameworks and Their Enhanced, pp. 7465e7468. https://doi.org/10.1039/ c7cc03424e. El-Deen, A.G., Choi, J.H., Kim, C.S., Khalil, K.A., Almajid, A.A., Barakat, N.A.M., 2015. TiO2 nanorod-intercalated reduced graphene oxide as high performance electrode material for membrane capacitive deionization. Desalination 361, 53e64. https://doi.org/10.1016/j.desal.2015.01.033. El-Deen, A.G., Boom, R.M., Kim, H.Y., Duan, H., Chan-Park, M.B., Choi, J.H., 2016. Flexible 3D nanoporous graphene for desalination and bio-decontamination of brackish water via asymmetric capacitive deionization. ACS Appl. Mater. Interfaces 8, 25313e25325. https://doi.org/10.1021/acsami.6b08658. EPA, 1998. How Wastewater Treatment Works, p. 6. Fan, C.S., Tseng, S.C., Li, K.C., Hou, C.H., 2016. Electro-removal of arsenic(III) and arsenic(V) from aqueous solutions by capacitive deionization. J. Hazard Mater. 312, 208e215. https://doi.org/10.1016/j.jhazmat.2016.03.055. Gao, X., Porada, S., Omosebi, A., Liu, K.L., Biesheuvel, P.M., Landon, J., 2016. Complementary surface charge for enhanced capacitive deionization. Water Res. 92, 275e282. https://doi.org/10.1016/j.watres.2016.01.048. Garcia, X., Pargament, D., 2015. Reusing wastewater to cope with water scarcity: economic, social and environmental considerations for decision-making. Resour. Conserv. Recycl. 101, 154e166. https://doi.org/10.1016/ j.resconrec.2015.05.015. Guidelines, W.H.O., Quality, D., 1995. Nitrate and Nitrite in Drinking Water. Nitrate

11

Nitrite Drink. Water. https://doi.org/10.17226/9038. Huang, Z., Lu, L., Cai, Z., Ren, Z.J., 2016. Individual and competitive removal of heavy metals using capacitive deionization. J. Hazard Mater. 302, 323e331. https:// doi.org/10.1016/j.jhazmat.2015.09.064. Kim, Y.J., Kim, J.H., Choi, J.H., 2013. Selective removal of nitrate ions by controlling the applied current in membrane capacitive deionization (MCDI). J. Membr. Sci. 429, 52e57. https://doi.org/10.1016/j.memsci.2012.11.064. Kim, T., Yu, J., Kim, C., Yoon, J., 2016. Hydrogen peroxide generation in flow-mode capacitive deionization. J. Electroanal. Chem. 776, 101e104. https://doi.org/ 10.1016/j.jelechem.2016.07.001. Laxman, K., Myint, M.T.Z., Al Abri, M., Sathe, P., Dobretsov, S., Dutta, J., 2015. Desalination and disinfection of inland brackish ground water in a capacitive deionization cell using nanoporous activated carbon cloth electrodes. Desalination 362, 126e132. https://doi.org/10.1016/j.desal.2015.02.010. Li, H., Zou, L., 2011. Ion-exchange membrane capacitive deionization: a new strategy for brackish water desalination. Desalination 275, 62e66. https://doi.org/ 10.1016/j.desal.2011.02.027. Li, H., Zou, L., Pan, L., Sun, Z., 2010. Novel graphene-like electrodes for capacitive deionization. Environ. Sci. Technol. 44, 8692e8697. https://doi.org/10.1021/ es101888j. Li, H., Zou, L., Pan, L., Sun, Z., 2010. Using graphene nano-flakes as electrodes to remove ferric ions by capacitive deionization. Separ. Purif. Technol. 75, 8e14. https://doi.org/10.1016/j.seppur.2010.07.003. Liang, J., He, D., Wang, D., Tang, W., Tang, L., Zeng, G., Liu, Z., Gong, J., 2018. Various cell architectures of capacitive deionization: recent advances and future trends. Water Res. 150, 225e251. https://doi.org/10.1016/j.watres.2018.11.064. Liu, P., Wang, H., Yan, T., Zhang, J., Shi, L., Zhang, D., 2016. 3D Graphene for Improved Capacitive Deionization y, pp. 5303e5313. https://doi.org/10.1039/c5ta10680j. Liu, D., Wang, X., Xie, Y.F., Tang, H.L., 2016. Effect of capacitive deionization on disinfection by-product precursors. Sci. Total Environ. 568, 19e25. https:// doi.org/10.1016/j.scitotenv.2016.05.219. Liu, P., Yan, T., Zhang, J., Shi, L., Zhang, D., 2017. Separation and recovery of heavy metal ions and salt ions from wastewater by 3D graphene-based asymmetric electrodes: via capacitive deionization. J. Mater. Chem. A. 5, 14748e14757. https://doi.org/10.1039/c7ta03515b. Liu, X., Whitacre, J.F., Mauter, M.S., 2018. Mechanisms of humic acid fouling on capacitive and insertion electrodes for electrochemical desalination. Environ. Sci. Technol. 52, 12633e12641. https://doi.org/10.1021/acs.est.8b03261. Mayes, R.T., Tsouris, C., Kiggans, J.O., Mahurin, S.M., Depaoli, D.W., Dai, S., 2010. Hierarchical ordered mesoporous carbon from phloroglucinol-glyoxal and its application in capacitive deionization of brackish water. J. Mater. Chem. 20, 8674e8678. https://doi.org/10.1039/c0jm01911a. Mossad, M., Zou, L., 2013. Study of fouling and scaling in capacitive deionisation by using dissolved organic and inorganic salts. J. Hazard Mater. 244e245, 387e393. https://doi.org/10.1016/j.jhazmat.2012.11.062. Neyens, E., Baeyens, J., 2003. A review of classic Fenton’s peroxidation as an advanced oxidation technique. J. Hazard Mater. 98, 33e50. https://doi.org/ 10.1016/S0304-3894(02)00282-0. Oldham, K.B., 2008. A Gouy-Chapman-Stern model of the double layer at a (metal)/ (ionic liquid) interface. J. Electroanal. Chem. 613, 131e138. https://doi.org/ 10.1016/j.jelechem.2007.10.017. Oren, Y., 2008. Capacitive deionization (CDI) for desalination and water treatment past, present and future (a review). Desalination 228, 10e29. https://doi.org/ 10.1016/j.desal.2007.08.005. Oturan, M.A., Aaron, J.J., 2014. Advanced oxidation processes in water/wastewater treatment: principles and applications. A review. Crit. Rev. Environ. Sci. Technol. 44, 2577e2641. https://doi.org/10.1080/10643389.2013.829765. Porada, S., Zhao, R., Van Der Wal, A., Presser, V., Biesheuvel, P.M., 2013. Review on the science and technology of water desalination by capacitive deionization. Prog. Mater. Sci. 58, 1388e1442. https://doi.org/10.1016/j.pmatsci.2013.03.005. Removal of suspended solids. Ind. Waste Treat., 2007 53e77. https://doi.org/ 10.1016/b978-012372493-9/50040-5. Ryoo, M.W., Kim, J.H., Seo, G., 2003. Role of titania incorporated on activated carbon cloth for capacitive deionization of NaCl solution. J. Colloid Interface Sci. 264, 414e419. https://doi.org/10.1016/S0021-9797(03)00375-8. Seo, S.J., Jeon, H., Lee, J.K., Kim, G.Y., Park, D., Nojima, H., Lee, J., Moon, S.H., 2010. Investigation on removal of hardness ions by capacitive deionization (CDI) for water softening applications. Water Res. 44, 2267e2275. https://doi.org/ 10.1016/j.watres.2009.10.020. Shapira, B., Avraham, E., Aurbach, D., 2016. Side reactions in capacitive deionization (CDI) processes: the role of oxygen reduction. Electrochim. Acta 220, 285e295. https://doi.org/10.1016/j.electacta.2016.10.127. Shen, J., Li, Y., Wang, C., Luo, R., Li, J., Sun, X., Shen, J., Han, W., Wang, L., 2018. Electrochimica Acta Hollow ZIFs-derived nanoporous carbon for efficient capacitive deionization. Electrochim. Acta 273, 34e42. https://doi.org/10.1016/ j.electacta.2018.04.004. Sivey, J.D., Arey, J.S., Tentscher, P.R., Roberts, A.L., 2013. Reactivity of BrCl, Br2, BrOCl, Br2O, and HOBr toward dimethenamid in solutions of bromide þ aqueous free chlorine. Environ. Sci. Technol. 47, 1330e1338. https://doi.org/10.1021/ es302730h. Snyder, S.A., Benotti, M.J., 2010. Endocrine disruptors and pharmaceuticals: implications for water sustainability. Water Sci. Technol. 61, 145e154. https://doi.org/ 10.2166/wst.2010.791. Srimuk, P., Kaasik, F., Krüner, B., Tolosa, A., Fleischmann, S., J€ ackel, N., Tekeli, M.C., Aslan, M., Suss, M.E., Presser, V., 2016. MXene as a novel intercalation-type

12

A. Kalfa et al. / Chemosphere 241 (2020) 125003

pseudocapacitive cathode and anode for capacitive deionization. J. Mater. Chem. A. 4, 18265e18271. https://doi.org/10.1039/c6ta07833h. €ckel, N., Tolosa, A., Krüner, B., Fleischmann, S., Grobelsek, I., Srimuk, P., Zeiger, M., Ja Aslan, M., Shvartsev, B., Suss, M.E., Presser, V., 2017. Enhanced performance stability of carbon/titania hybrid electrodes during capacitive deionization of oxygen saturated saline water. Electrochim. Acta 224, 314e328. https://doi.org/ 10.1016/j.electacta.2016.12.060. Suss, M.E., 2017. Size-based ion selectivity of micropore electric double layers in capacitive deionization electrodes. J. Electrochem. Soc. 164, E270eE275. https:// doi.org/10.1149/2.1201709jes. Wan Ngah, W.S., Hanafiah, M.A.K.M., 2008. Removal of heavy metal ions from wastewater by chemically modified plant wastes as adsorbents: a review. Bioresour. Technol. 99, 3935e3948. https://doi.org/10.1016/ j.biortech.2007.06.011. Wang, H., Shi, L., Yan, T., Zhang, J., Zhong, Q., Zhang, D., 2014. Design of graphenecoated hollow mesoporous carbon spheres as high performance electrodes for capacitive deionization. J. Mater. Chem. A. 2, 4739e4750. https://doi.org/ 10.1039/c3ta15152b. Wang, Y., El-Deen, A.G., Li, P., Oh, B.H.L., Guo, Z., Khin, M.M., Vikhe, Y.S., Wang, J., Hu, R.G., Boom, R.M., Kline, K.A., Becker, D.L., Duan, H., Chan-Park, M.B., 2015. High-performance capacitive deionization disinfection of water with graphene oxide-graft-quaternized chitosan nanohybrid electrode coating. ACS Nano 9, 10142e10157. https://doi.org/10.1021/acsnano.5b03763. Wang, H., Yan, T., Liu, P., Chen, G., Shi, L., Zhang, J., Zhong, Q., Zhang, D., 2016. Graphene Architectures and Their Application as Deionization Capacitors y, pp. 4908e4919. https://doi.org/10.1039/c5ta10703b. Wang, Z., Yan, T., Chen, G., Shi, L., Zhang, D., 2017. High Salt Removal Capacity of Metal  Organic Gel Derived Porous Carbon for Capacitive Deionization. https:// doi.org/10.1021/acssuschemeng.7b03015. Wang, Z., Gong, H., Zhang, Y., Liang, P., Wang, K., 2017. Nitrogen recovery from lowstrength wastewater by combined membrane capacitive deionization (MCDI) and ion exchange (IE) process. Chem. Eng. J. 316, 1e6. https://doi.org/10.1016/ j.cej.2017.01.082. Weng, S., Blatchley, E.R., 2013. Ultraviolet-induced effects on chloramine and cyanogen chloride formation from chlorination of amino acids. Environ. Sci.

Technol. 47, 4269e4276. https://doi.org/10.1021/es400273w. Yan, T., Liu, J., Lei, H., Shi, L., An, Z., Park, H.S., Zhang, D., 2018. Capacitive deionization of saline water using sandwich-like nitrogen-doped graphene composites via a self-assembling strategy. Environ. Sci. Nano. 5, 2722e2730. https:// doi.org/10.1039/C8EN00629F. Yeager, E., 1984. Electrocatalysts for O2 reduction. Electrochim. Acta 29, 1527e1537. https://doi.org/10.1016/0013-4686(84)85006-9. Yue, M., Wang, R., Cheng, N., Cong, R., Gao, W., Yang, T., 2016. ZnCr2S4: highly effective photocatalyst converting nitrate into N2 without over-reduction under both UV and pure visible light. Sci. Rep. 6, 1e11. https://doi.org/10.1038/ srep30992. Zhang, D., Yan, T., Shi, L., Peng, Z., Wen, X., Zhang, J., 2012. Enhanced capacitive deionization performance of graphene/carbon nanotube composites. J. Mater. Chem. 22, 14696e14704. https://doi.org/10.1039/c2jm31393f. Zhang, W., Mossad, M., Zou, L., 2013. A study of the long-term operation of capacitive deionisation in inland brackish water desalination. Desalination 320, 80e85. https://doi.org/10.1016/j.desal.2013.04.010. Y. Zhang, Y. Zhao, S. Cao, Z. Yin, L. Cheng, L. Wu, Ying Zhang, y Yan Zhao, y shunsheng cao, * , y zhengliang yin, y Li cheng, y and limin Wu * , z , x y z, (2017). doi: 10.1021/acsami.7b08776. Zhao, R., Biesheuvel, P.M., Van Der Wal, A., 2012. Energy consumption and constant current operation in membrane capacitive deionization. Energy Environ. Sci. 5, 9520e9527. https://doi.org/10.1039/c2ee21737f. Zhao, R., Satpradit, O., Rijnaarts, H.H.M., Biesheuvel, P.M., van der Wal, A., 2013. Optimization of salt adsorption rate in membrane capacitive deionization. Water Res. 47, 1941e1952. https://doi.org/10.1016/j.watres.2013.01.025. Zhao, S., Yan, T., Wang, H., Zhang, J., Shi, L., Zhang, D., 2016. Creating 3D Hierarchical Carbon Architectures with Micro - , Meso - , and Macropores via a Simple SelfBlowing Strategy for a Flow- through Deionization Capacitor. https://doi.org/ 10.1021/acsami.6b03704. Zhao, S., Yan, T., Wang, Z., Zhang, J., Shi, L., 2017. RSC Advances Removal of NaCl from Saltwater Solutions Using Micro/Mesoporous Carbon Sheets Derived from Watermelon Peel via Deionization Capacitors y, pp. 4297e4305. https://doi.org/ 10.1039/c6ra27127h.