Progress and outlook for capacitive deionization technology

Progress and outlook for capacitive deionization technology

Available online at ScienceDirect Progress and outlook for capacitive deionization technology James Landon1,2, Xin Gao1, Ayokun...

606KB Sizes 1 Downloads 7 Views

Available online at

ScienceDirect Progress and outlook for capacitive deionization technology James Landon1,2, Xin Gao1, Ayokunle Omosebi1 and Kunlei Liu1,3 Capacitive deionization (CDI), an emerging desalination technology, has received considerable attention in recent years due to porous carbon development, new cell designs, and unique operational modes providing higher performance for capacitance-based salt removal. The energy cost of this separation process has also been rigorously evaluated through a variety of efficiency calculations, and direct comparisons are now being made with more conventional membrane-based separation systems. Currently, the thermodynamic energy efficiency (TEE) of CDI is quite low with values typically below 5%, although there are examples where this efficiency can reach 10% or higher under the proper conditions. As the development of CDI and analogous capacitive processes continue, the TEE should remain a metric for comparison to conventional techniques. In addition, the ability of CDI to perform selective separations for trace compounds, resource recovery, and contaminant removal should be more heavily investigated. Addresses 1 University of Kentucky Center for Applied Energy Research, 2540 Research Park Drive, Lexington, KY 40511, USA 2 Department of Chemical and Materials Engineering, University of Kentucky, 177 F. Paul Anderson Tower, Lexington, KY 40506, USA 3 Department of Mechanical Engineering, University of Kentucky, 151 Ralph G. Anderson Building, Lexington, KY 40506, USA Corresponding authors: Gao, Xin ([email protected]), Liu, Kunlei ([email protected])

Current Opinion in Chemical Engineering 2019, 25:1–8 This review comes from a themed issue on Nanotechnology: waterenergy nexus

For seawater desalination, distillation processes such as multi-stage flash distillation and multiple-effect distillation have traditionally been used. In recent decades, membrane-based processes such as reverse osmosis (RO) have been able to provide similar water quality to distillation at a fraction of the energy cost. This benefit combined with capital cost reductions over time has led to increased use of RO for total dissolved solids (TDS) removal [1]. For brackish water desalination, nanofiltration (NF) as well as RO have been shown to be quite effective. The energy expended for these separations can be within a few times of the thermodynamic minimum [2], making them attractive for a multitude of applications. However, depending on the feed water composition, for instance certain power plant wastewater streams, these membranes may require significant levels of pretreatment to avoid polymer degradation due to chlorine [3] and membrane-fouling due to calcium and magnesiumgenerated inorganic scales [4]. Biological compounds must also be sufficiently treated to diminish growth on and around the membrane surface [5]. Consequently, researchers have been exploring alternative water treatment options that can withstand varying feed water qualities while also providing tailored separation mechanisms at reasonable energy costs. Since pressure-based membrane methods make use of size and Donnan exclusion at an interface, fouling and targeted ion removal (beyond the noted mechanisms) will remain an issue. Therefore, new separation processes are being developed to circumvent these problems.

Edited by Jeff Urban

Background on capacitive deionization 2211-3398/ã 2019 Published by Elsevier Ltd.

Introduction As fresh water sources become further constrained, new and innovative methods are needed for water treatment operations. Suspended solids, organics, microorganisms, and dissolved solids are among the compounds requiring separation.

Capacitive deionization (CDI) is an emerging separation process that generally makes use of symmetric, porous carbon electrode pairs to electrostatically adsorb ions in the electric double layer (EDL) under the influence of an applied potential [6,7]. When this potential is removed, reduced, or reversed, ions are desorbed into a concentrated stream. A schematic of this process where desorption occurs under short-circuit is shown in Figure 1a. Significant efforts have been made in recent decades in the development of customized porous carbon materials for CDI including carbon nanotubes [8], graphene [9], MXene [10], aerogels [11], xerogels [12], mesoporous carbon [13], and activated carbons [14]. While posing unique benefits, so far these materials have generally failed to provide a significant competitive Current Opinion in Chemical Engineering 2019, 25:1–8

2 Nanotechnology: water-energy nexus

Figure 1

Desorption Step

Adsorption Step

(a) Capacitive Deionization (CDI)


(b) Membrane Capacitive Deionization (MCDI)


(c) Inverted Capacitive Deionization (i-CDI)


(d) Faradaic Deionization (FDI)


(e) Flow Capacitive Deionization (FCDI)


Current Opinion in Chemical Engineering

Schematic of (a) capacitive deionization (CDI), (b) membrane capacitive deionization (MCDI), (c) inverted capacitive deionization (i-CDI), (d) Faradaic deionization (FDI), and (e) flow capacitive deionization (FCDI). On the left, adsorption is shown for each process, and on the right, desorption.

advantage to NF and RO in either capital or operating cost due to low salt adsorption capacity and carbon electrode oxidation. In the last ten years, new cell designs and operational modes have been developed that further enhanced the performance, life, and efficiency of CDI systems including flow and fluidized bed capacitive deionization systems [15,16,17]. Membrane capacitive deionization, shown in Figure 1b, incorporates anion exchange membranes (shown in blue) over the anode and cation exchange membranes (shown in red) over the cathode [18]. These ion exchange membranes provide significant Current Opinion in Chemical Engineering 2019, 25:1–8

gains in the charge efficiency of the process, a value defining the amount of salt molecules adsorbed per charge input. Beyond the addition of membranes, chemical functionalization of the carbon surfaces has also taken place, such as in the creation of inverted capacitive deionization (i-CDI), shown in Figure 1c [19]. The chemical functionalization used in i-CDI may be able to provide highly selective separations, given by the proper surface groups. To increase the salt adsorption capacity (SAC) and subsequent utility of these cells further, battery intercalation electrodes have begun to be used in Faradaic deionization (FDI) cells, shown in Figure 1d. While the FDI process shown here

Progress and outlook for capacitive deionization technology Landon et al.

is an oversimplification to many designs, it should be noted that these systems can achieve substantially higher SACs than CDI, MCDI, and i-CDI due to the ability to intercalate salt into the solid phase as opposed to simply on the electrode surface [20]. Additional important enhancements have been made towards making this process continuous through the advent of flow capacitive deionization (FCDI) where charged carbon particles in a slurry are used instead of static electrodes for the desalination process [21–23]. A simplistic version of this process is depicted in Figure 1e. In these capacitive cells, cations and anions are adsorbed at opposing electrodes, differing from the singular interface used for pressure-based separation processes. The ability of these interfaces to be modified and ions adsorbed at opposing sites may decrease the risk for many inorganic fouling issues, although high concentrations may still be found in the waste streams. Efficiency definitions in capacitive deionization

CDI has been touted as an energy efficient, environmentally friendly alternative to commercial water treatment routes. However, direct energy comparisons between CDI and these technologies have been sparse. A variety of energy efficiency calculations for CDI have been defined including charge efficiency, specific energy consumption (SEC), energy-normalized adsorbed salt (ENAS) [24], and thermodynamic energy efficiency (TEE) [25], among others. Classically, the salt adsorption capacity and charge efficiency of a CDI process have been used as a metric for comparison, and indeed high charge efficiencies are needed for any energy-efficient CDI process. The charge efficiency is defined as the equivalent charge of salt removed per total charge used during the adsorption step. A standard equation for the charge efficiency (LÞ is shown in Eq. (1): L ¼ ðF  GÞ=ðMW  Qad Þ


where G is the salt adsorption capacity (mg of salt per g of carbon), F is Faraday’s constant, MW is the molecular weight of the salt, and Qad is the charge density used during adsorption. A high charge efficiency signifies that most of the charge input into the process is used for salt adsorption, that is, a charge efficiency of 100% means that all electronic charge would be used for adsorption. For most CDI systems, the charge efficiency has been found to be significantly lower than 100%. Faradaic inefficiencies due to carbon oxidation and dissolved oxygen reduction primarily contribute to this issue, as shown by Zhang et al. and Omosebi et al. [26,27] but improper surface charge on the carbon electrode can often be the largest contributor, especially over short timescales. Over the last 5–10 years, work by Avraham et al. [28] Gao et al. [29,30] and Lado et al. [31] have shown that improper electrode surfaces can lead to charge efficiencies below 50%. In


order to correlate these charge efficiencies with the potentials used in CDI cells, the potential of zero charge (EPZC) and distributed potentials can be measured in the cell used for the separation. Shown in Figure 2a is a differential capacitance plot (and analogous cyclic voltammogram shown in Figure 2b) displaying the location of the EPZC for a typical pristine microporous carbon material used in a symmetric CDI process. Potentials more positive than the EPZC will result in anion adsorption while potentials more negative will lead to cation adsorption. Figure 2c depicts the resting potential when the cell is short circuited (Eo), and the distributed potentials at the anode (E+) and cathode (E) when a cell potential of 1V is applied during adsorption in a conventional CDI process using an identical carbon electrode pair. An example in Figure 2c depicts that the region between Eo and the EPZC will lead to anion expulsion from the cathode surface during charging, thereby resulting in inefficiencies in the net salt adsorption. As shown in Figure 2d, the charge efficiency will also be affected, as well as other energy calculations mentioned above, as the voltage difference between Eo and EPZC becomes greater. In this example where Faradaic reactions are neglected, considering electrodes with different EPZCs and static Eo, with increases in |Eo-EPZC|, additional anion expulsion will take place at the limiting electrode, lowering the charge efficiency further [30]. The location of the EPZC is subsequently crucial to increasing the general efficiency of a CDI process. Faradaic reactions and long-term stability in capacitive deionization

The carbon electrodes used in these open separation systems are subject to various possible Faradaic reactions, including mainly carbon oxidation at the anode and dissolved oxygen reduction at the cathode. These reactions can affect not only the EPZC and charge efficiency but also the effectiveness and stability of the separation process. Zhang et al. and Gao et al. have reviewed a number of these reactions. Example equations are shown below in Eqs. (2)–(6) with formal potentials shown at a pH of 7 versus NHE. Anode reactions: C þ H2 O ! C  Oad þ 2Hþ þ 2e

E0 ¼  0:207 V ð2Þ

2H2 O ! O2 þ 4Hþ þ 4e

E0 ¼ 0:815 V


Cathode reactions: O2ðdissolvedÞ þ 4Hþ þ 4e ! 2H2 O

E0 ¼ 0:815 V


Current Opinion in Chemical Engineering 2019, 25:1–8

4 Nanotechnology: water-energy nexus

100 (a)

80 60 40 EPZC 20 -0.5



Specific Current / mA g-1

Specific Capacitance / F g-1

Figure 2


50 0 -50 -100



Potential / V vs SCE




Potential / V vs SCE



0.6 E-



0.4 Coion repulsion at cathode driven by EO-EPZC


EPZC -0.5



Potential / V


Charge Efficiency

Ionic Charge / M






0.4 0.0





⎥EO - EPZC⎥ / V Current Opinion in Chemical Engineering

Charge efficiency is shown for the CDI process as a function of the surface charge/EPZC of the electrode and the distributed potential to that electrode (E/E+). Electrochemical characterizations of a microporous carbon electrode are carried out in 9 mM deaerated NaCl solution, (a) electrochemical impedance spectrum at 0.02 Hz, (b) cyclic voltammogram at 0.25 mV s1, (c) ionic charge adsorption for a cell charged at 1 V, and (d) charge efficiency for a charging voltage of 1 V as a function of voltage difference between Eo and EPZC.

O2ðdissolvedÞ þ 2Hþ þ 2e ! H2 O2 E0 ¼ 0:318 V


2Hþ þ 2e ! H2 E0 ¼  0:414V


specifically when considering stabilization of the carbon anode. Ion exchange membranes in the MCDI process can nearly eliminate dissolved oxygen reduction at the cathode, but additional materials development is still needed to mitigate carbon oxidation at the anode.

The driving force for carbon oxidation is >0.2 V when operating at a cell potential of 0.8 V and considering a distributed potential at the anode of at least 0.25 V versus SCE [30]. The formal potential for this reaction is noted as 0.207 versus NHE in a pH 7 electrolyte or 0.207 versus NHE in a pH 0 electrolyte. However, higher formal potential values have also been reported, and indeed more work is needed to identify the carbon oxidation potential in various electrode/electrolyte systems. There is also already a significant driving force, hundreds of millivolts, at the cathode for dissolved oxygen reduction. The limiting of these reactions will be crucial to the further development of these processes,

During long-term cycling, degradation of the carbon anode can result in two issues, loss of chloride (anion) adsorption at the anode (positive electrode) as well as oxidation and loss of the carbon itself with oxidation affecting the surface carbon. This issue has been well detailed in the literature by Cohen et al., Shapira et al., and Gao et al. [32–34]. The use of oxidized carbon anodes in i-CDI, shown in Figure 1, can help to substantially prolong the separation, but electrode choices will still be important [35]. New materials developments such as corrosion resistant (oxidation resistant) anode materials with higher chloride adsorption capacities may help to mitigate this effect and enhance the stability of this process.

Current Opinion in Chemical Engineering 2019, 25:1–8

Progress and outlook for capacitive deionization technology Landon et al.

Energy assessments of capacitive deionization systems

While the charge efficiency mentioned previously is a useful metric, it is difficult to evaluate alongside other separation systems that do not make use of electrons in the separation mechanism and instead use pressure-based methods. A more useful comparison is defining the energy efficiency of the separation itself, not just the efficiency of the adsorption process. Mentioned previously were ENAS and the TEE [24]. ENAS can be evaluated using Eq. (7): ENAS ¼

 Z f

t charge

 ðc 0  c Þd t =ðEin  Eout Þ



where f is the flow rate, tcharge is the charging time, co is the influent concentration, c is the effluent concentration, Ein is the energy input, and Eout is the energy recovered. ENAS can be used to calculate how much salt can be removed per unrecoverable energy input, a highly useful comparison term that can also aid in calculating the cost per unit of water produced or treated. The TEE can be calculated as the Gibbs free energy of separation divided by the SEC, a calculation that will depend on the system operation and amount of energy recovered during a discharge process [25,36]. An example calculation for the SEC is shown in Eq. (8) from Wang et al. for a system where all energy is recovered during the discharge step [25]: SEC ¼

1 nD


t d ;f

V cell ðt Þiðt Þd t


t c;0

Where Vcell(t) is the cell voltage, i(t) is the current, tc,0 is the time at the start of the charging step, td,f is the time at the end of the discharge step, and nD is the volume of the dilute treated water produced. Work in recent years by Hemmatifar et al. [37] Wang et al. [25,38] and Qin et al. [39] has highlighted the importance of the TEE as opposed to salt adsorption capacity and charge efficiency values, and the TEE can be used as the basis for comparison with other separation processes. The Gibbs free energy of separation (e.g. the thermodynamic minimum work), which takes into account the volume of water recovered from the process, can be defined using Eq. (9): DG ¼ 2RT

     c0 c 0  gc D c 0  gc D ln  c D ln g c 0 ð1  g Þ c D ð1  g Þ


where R is the ideal gas constant, T is the absolute temperature, co is the influent concentration, cD is the product water concentration, cB is the brine water concentration (defined by c0, cD, and g), and g is the


water recovery. It should be noted that this equation is meant for a 1:1 electrolyte (such as NaCl), which is conventionally used in the current scientific literature. When evaluating reported CDI process separations, Wang et al. and Hemmatifar et al. show that the CDI TEE only reaches 9%, with the vast majority of CDI and MCDI processes resulting in a TEE of <5%, [25,37] although analysis by Hand et al. has also shown some higher efficiencies are possible for MCDI systems and the importance of limiting resistive losses and lower round-trip efficiencies [15]. Moreno et al. also demonstrate that higher efficiencies of >10% are also possible if higher feed concentrations are used [40]. TEE values of <5% are far below those seen with RO systems and are a function of both the separation mechanism and the energy being recovered in the process. The amount of recoverable charge during the discharge phase will strongly impact the resulting energy cost of these capacitive processes. To date, most reported CDI separations have assumed that a significant portion of the discharge current can be recovered and used for subsequent salt separations. Limited studies have been performed to show the extent to which this charge is practically recoverable, although some recent work by Kang et al. has provided demonstrations of this process with a buck-boost converter [41]. More work is needed to define the extent to which energy recovery is possible and practical through the integration of cells, batteries, and supercapacitors. The integration of solar energy with batteries and electrochemical water treatment process may also pose some unique benefits in remote areas.

Future of capacitive deionization

One of the more interesting developments in recent years has been the progress made with Faradaic deionization (FDI) through the use of battery intercalation materials [42,43,44]. While avoiding detail of the specific adsorption mechanism, which can be found elsewhere, these electrodes are capable of adsorbing salt into the bulk of material, vastly extending their total salt adsorption capacity to >100 mg g1 of electrode in some reported systems [45]. Through this process, TEE values of >10% have been achieved with a recent system achieving a TEE value comparable to many RO systems [20,25,42]. The quality and quantity of water recovered, in addition to the TEE, will need to be emphasized when comparing systems, and recent publications have begun to stress this point. The encouraging FDI results will aid in electrochemical processes being energy competitive with future RO systems, although system size, complexity, and cost, both operating and capital, will need to be further considered. Current Opinion in Chemical Engineering 2019, 25:1–8

6 Nanotechnology: water-energy nexus

Figure 3

(a) +


(b) +

Current Opinion in Chemical Engineering

(a) Ion-selective exchange membranes in an MCDI process for the transport and concentration of particular dissolved solids. (b) Selective adsorption surfaces in a membrane-free CDI process.

Conclusion The energy efficiency values discussed here are generally used to analyze the ability of a process to deionize the bulk of a feed water stream, namely the ability of NaCl to be removed from the influent. Capacitive deionization has made tremendous gains in recent years for these separations through the advent of new carbon materials, cell designs, operational schemes, and adsorption/intercalation mechanisms. While RO may continue to be cost effective for many bulk desalination applications, such as seawater desalination plants, the ability of electrodes and ion exchange membranes to be tailored for selective applications will result in advantageous use cases, in particular for smaller-scale applications where CDI can more easily scale down and where particular contaminants can be more easily removed from the surrounding ionic matrix. Recent work by Cohen et al. on selective bromide removal [46], Suss et al. on selectivity of anion/cation adsorption [47], Su et al. on selective electrochemical interfaces [48], Oyarzun et al. on theory and validation of selectivity [49], and Hawks et al. on selective nitrate removal [50] has already begun to demonstrate ion selectivity using CDI and electrochemical systems. However, more work is warranted to extend the viable use cases of these systems. Shown in Figure 3a is a schematic of an MCDI process that incorporates selective ion exchange membranes (IEM) to allow only particular cations through the membrane. Similar work has been carried out with monovalent selective membranes [51], and as IEMs and alternative membrane layer options become more advanced, highly selective separations will be possible [52]. The use of selective interfaces being modulated by applied voltage/current will enable additional control on the adsorption and separation process Current Opinion in Chemical Engineering 2019, 25:1–8

such as the use of carbon surfaces with highly specific surface functional groups as depicted in Figure 3b. This level of control may find particular use in areas such as rare earth element recovery and harmful contaminant removal, among many possible applications.

Conflict of interest statement Nothing declared.

Acknowledgements This work is supported by the Crosscutting Research, National Energy Technology Laboratory, U.S. Department of Energy (DE-FE0031555) and the U.S.China Clean Energy Research Center, U.S. Department of Energy (DE-PI0000017).

References and recommended reading Papers of particular interest, published within the period of review, have been highlighted as:  of outstanding interest 1.

Van der Bruggen B, Vandecasteele C: Distillation vs. membrane filtration: overview of process evolutions in seawater desalination. Desalination 2002, 143:207-218.


Elimelech M, Phillip WA: The future of seawater desalination: energy, technology, and the environment. Science 2011, 333:712.


Do VT et al.: Degradation of polyamide nanofiltration and reverse osmosis membranes by hypochlorite. Environ Sci Technol 2012, 46:852-859.


Pe´rez-Gonza´lez A et al.: State of the art and review on the treatment technologies of water reverse osmosis concentrates. Water Res 2012, 46:267-283.


Lee KP, Arnot TC, Mattia D: A review of reverse osmosis membrane materials for desalination—development to date and future potential. J Membr Sci 2011, 370:1-22.


Porada S et al.: Review on the science and technology of water desalination by capacitive deionization. Prog Mater Sci 2013, 58:1388-1442.

Progress and outlook for capacitive deionization technology Landon et al.


AlMarzooqi FA et al.: Application of capacitive deionisation in water desalination: a review. Desalination 2014, 342:3-15.


Lee B et al.: Enhanced capacitive deionization by dispersion of CNTs in activated carbon electrode. ACS Sustain Chem Eng 2018, 6:1572-1579.


Liu P et al.: Graphene-based materials for capacitive deionization. J Mater Chem A 2017, 5:13907-13943.

10. Srimuk P et al.: MXene as a novel intercalation-type pseudocapacitive cathode and anode for capacitive deionization. J Mater Chem A 2016, 4:18265-18271. 11. Suss ME et al.: Capacitive desalination with flow-through electrodes. Energy Environ Sci 2012, 5:9511-9519. 12. Landon J et al.: Impact of pore size characteristics on the electrosorption capacity of carbon xerogel electrodes for capacitive deionization. J Electrochem Soc 2012, 159:A1861A1866. 13. Tsouris C et al.: Mesoporous carbon for capacitive deionization of saline water. Environ Sci Technol 2011, 45:10243-10249. 14. Hou C-H, Huang C-Y: A comparative study of electrosorption selectivity of ions by activated carbon electrodes in capacitive deionization. Desalination 2013, 314:124-129. 15. Hand S et al.: Global sensitivity analysis to characterize  operational limits and prioritize performance goals of capacitive deionization technologies. Environ Sci Technol 2019. A comprehensive overview of factors affecting the efficiency of the capacitive deionization process is given. Factors such as ion exchange membranes, flow conditions, and carbon surface charge are evaluated, among other metrics. 16. Moreno D, Hatzell MC: Influence of feed-electrode concentration differences in flow-electrode systems for capacitive deionization. Ind Eng Chem Res 2018, 57:8802-8809. 17. Doornbusch GJ et al.: Fluidized bed electrodes with high carbon loading for water desalination by capacitive deionization. J Mater Chem A 2016, 4:3642-3647. 18. Biesheuvel PM, van der Wal A: Membrane capacitive deionization. J Membr Sci 2010, 346:256-262. 19. Gao X et al.: Surface charge enhanced carbon electrodes for stable and efficient capacitive deionization using inverted adsorption–desorption behavior. Energy Environ Sci 2015, 8:897-909. 20. Choi S et al.: Battery electrode materials with omnivalent cation storage for fast and charge-efficient ion removal of asymmetric capacitive deionization. Adv Funct Mater 2018, 28:1802665. 21. Jeon S-i et al.: Desalination via a new membrane capacitive deionization process utilizing flow-electrodes. Energy Environ Sci 2013, 6:1471-1475. 22. Rommerskirchen A et al.: Modeling continuous flow-electrode capacitive deionization processes with ion-exchange membranes. J Membr Sci 2018, 546:188-196. 23. Hatzell KB et al.: Effect of oxidation of carbon material on suspension electrodes for flow electrode capacitive deionization. Environ Sci Technol 2015, 49:3040-3047. 24. Hemmatifar A et al.: Energy breakdown in capacitive deionization. Water Res 2016, 104:303-311. 25. Wang L, Dykstra JE, Lin S: Energy efficiency of capacitive  deionization. Environ Sci Technol 2019. The energy efficiency of various reported capacitive deionization processes is evaluated and compared to conventional membrane-based systems. Data are compiled from the literature, and examples of higher efficiency capacitive processes are highlighted. 26. Zhang C et al.: Faradaic reactions in capacitive deionization (CDI) – problems and possibilities: a review. Water Res 2018, 128:314-330. 27. Omosebi A et al.: Anion exchange membrane capacitive deionization cells. J Electrochem Soc 2017, 164:E242-E247.


28. Avraham E et al.: Limitation of charge efficiency in capacitive deionization: I. On the behavior of single activated carbon. J Electrochem Soc 2009, 156:P95-P99. 29. Gao X et al.: Enhancement of charge efficiency for a capacitive deionization cell using carbon xerogel with modified potential of zero charge. Electrochem Commun 2014, 39:22-25. 30. Gao X et al.: Capacitive deionization using symmetric carbon  electrode pairs. Environ Sci Water Res Technol 2019, 5:660-671. The direct effect of surface charge on carbon materials is examined for symmetric capacitive deionization systems. Systematic oxidation of the carbon surface is found to play a crucial role in the resulting charge efficiency. 31. Lado JJ et al.: Evaluation of operational parameters for a capacitive deionization reactor employing asymmetric electrodes. Sep Purif Technol 2014, 133:236-245. 32. Cohen I et al.: Long term stability of capacitive de-ionization processes for water desalination: the challenge of positive electrodes corrosion. Electrochim Acta 2013, 106:91-100. 33. Gao X et al.: Dependence of the capacitive deionization performance on potential of zero charge shifting of carbon xerogel electrodes during long-term operation. J Electrochem Soc 2014, 161:E159-E166. 34. Shapira B, Avraham E, Aurbach D: Side reactions in capacitive deionization (CDI) processes: the role of oxygen reduction. Electrochim Acta 2016, 220:285-295. 35. Gao X et al.: Voltage-based stabilization of microporous carbon electrodes for inverted capacitive deionization. J Phys Chem C 2018, 122:1158-1168. 36. Wang L, Lin S: Intrinsic tradeoff between kinetic and energetic efficiencies in membrane capacitive deionization. Water Res 2018, 129:394-401. 37. Hemmatifar A et al.: Thermodynamics of ion separation by electrosorption. Environ Sci Technol 2018, 52:10196-10204. 38. Wang L, Biesheuvel PM, Lin S: Reversible thermodynamic cycle analysis for capacitive deionization with modified Donnan model. J Colloid Interface Sci 2018, 512:522-528. 39. Qin M et al.: Comparison of energy consumption in  desalination by capacitive deionization and reverse osmosis. Desalination 2019, 455:100-114. In this work, a direct experimental comparison between membrane capacitive deionization and reverse osmosis is carried out. Reverse osmosis is found to be more thermodynamically energy efficient than current capacitive deionization systems. 40. Moreno D, Hatzell MC: Efficiency of Carnot and conventional capacitive deionization cycles. J Phys Chem C 2018, 122:22480-22486. 41. Kang J et al.: Direct energy recovery system for membrane capacitive deionization. Desalination 2016, 398:144-150. 42. Kim T, Gorski CA, Logan BE: Low energy desalination using  battery electrode deionization. Environ Sci Technol Lett 2017, 4:444-449. Battery electrode materials are evaluated in a new electrochemical cell for deionization. Significant gains in the salt adsorption capacity and energy efficiency of the process are demonstrated with intercalation electrodes. 43. Porada S et al.: Nickel hexacyanoferrate electrodes for continuous cation intercalation desalination of brackish water. Electrochim Acta 2017, 255:369-378. 44. Biesheuvel PM et al.: Capacitive Deionization — Defining a Class Of Desalination Technologies. . arXiv e-prints 2017. 45. Guo L et al.: A Prussian blue anode for high performance electrochemical deionization promoted by the faradaic mechanism. Nanoscale 2017, 9:13305-13312. 46. Cohen I et al.: Bromide ions specific removal and recovery by electrochemical desalination. Environ Sci Technol 2018, 52:6275-6281. 47. Suss ME: Size-based ion selectivity of micropore electric double layers in capacitive deionization electrodes. J Electrochem Soc 2017, 164:E270-E275. Current Opinion in Chemical Engineering 2019, 25:1–8

8 Nanotechnology: water-energy nexus

48. Su X et al.: Electrochemically-mediated selective capture of heavy metal chromium and arsenic oxyanions from water. Nat Commun 2018, 9:4701. 49. Oyarzun DI et al.: Ion selectivity in capacitive deionization with functionalized electrode: theory and experimental validation. Water Res X 2018, 1:100008. 50. Hawks SA et al.: Using ultramicroporous carbon for the  selective removal of nitrate with capacitive deionization. Environ Sci Technol 2019.

Current Opinion in Chemical Engineering 2019, 25:1–8

Design of carbon materials for adsorption of nitrate is shown. A combination of adsorption site geometry, voltage, kinetics, and feed stream composition are some of the defining factors investigated here for a selective separation. 51. Choi J, Lee H, Hong S: Capacitive deionization (CDI) integrated with monovalent cation selective membrane for producing divalent cation-rich solution. Desalination 2016, 400:38-46. 52. Mao S et al.: Fractionation of mono- and divalent ions by capacitive deionization with nanofiltration membrane. J Colloid Interface Sci 2019, 544:321-328.