Top Banner
Contents lists available at ScienceDirect Separation and Purication Technology journal homepage: www.elsevier.com/locate/seppur Atomic layer deposition of TiO 2 on carbon-nanotube membranes for enhanced capacitive deionization Jianhua Feng a,b , Sen Xiong a , Yong Wang a, a State Key Laboratory of Materials-Oriented Chemical Engineering, Jiangsu National Synergetic Innovation Center for Advanced Materials, and College of Chemical Engineering, Nanjing Tech University, Nanjing 211816, China b College of Material and Chemical Engineering, Chuzhou University, Chuzhou 239000, China ARTICLE INFO Keywords: Carbon nanotubes Atomic layer deposition Surface modication Capacitive deionization Desalination ABSTRACT Capacitive deionization (CDI) is an energy-ecient and environment-benign process to produce fresh water. Carbon nanotubes (CNTs) are promising building blocks in constructing high-performance CDI electrodes. Nevertheless, the strong hydrophobicity of CNTs signicantly impedes their applications in aqueous environ- ment. Herein, hydrophilic CNT-based membrane electrodes are obtained via atomic layer deposition (ALD). We demonstrate, for the rst time, that ALD is an ecient and exible means of enhancing the CDI performance of free-standing CNT membrane electrodes by depositing TiO 2 nanoparticles on the surface of CNTs. The CNT membranes display preferable electrosorption behavior after moderate ALD cycle numbers and stable reusability in the desalination process. 1. Introduction Under the situation of worldwide increasing demands and de- creasing supply of freshwater, extensive attention having being paid on advanced desalination technologies. As a promising alternative to de- salination processes, capacitive deionization (CDI) is an environment- friendly and energy-ecient desalination technique in comparison with other desalination techniques which always suer from the drawbacks such as fouling, water electrolysis and high energy consumption [19]. CDI based on the electrical double-layer capacitor (EDLC) theory is an electrochemical water purication method and capable of reducing the salt concentration of brackish and seawater by electrostatic adsorption of ions on porous electrodes [1012]. The electrosorption behavior relies signicantly on the electrical conductivity, wettability and in- ternal structures of the CDI electrode materials [13]. Owing to their high surface area, good exibility and low electrical resistivity, carbon- based materials such as graphene, activated carbon (AC), carbon aerogels (CAs), carbon nanotubes (CNTs), and their composites have been widely investigated for the application in CDI electrode materials over the past years [1419]. However, the tedious treatments for the synthesis and/or modication to the carbon building blocks, easy ag- gregation of powders, and binder addition often complicate the pre- paration process of the electrodes on one hand [2022], and sometimes it cannot obtain acceptable electrosorption capacity on the other [14]. Therefore, advanced electrode materials with good CDI performances which simultaneously have the advantages of simplifying preparation process, avoiding the aggregation, eliminating the blocking caused by binders and can be directly used as electrodes are urgently needed. Incorporating pristine or modied CNTs with binders, then com- bining them with polymers or other porous carbon-based materials, and nally depositing them onto a current collector is the most common way to fabricate CDI composite electrodes. The use of CNTs is expected to increase the specic surface area and enhance the electrical con- ductivity of composite materials, thus improving CDI performance [2329]. However, these CNTs are commonly existed in the shape of powders and are typically required to be chemically modied in order to have a good dispersion in the nal electrodes, which always requires a great deal of time and energy, and is also a tedious process [30,31]. Furthermore, the release of individual CNTs into water or air is likely to cause safety issues during practical applications [32,33]. If the CNTs can be welded together and utilized as hydrophilic free-standing CNT membrane electrodes, they will be more eective for CDI applications. Compared to CNT-doped composite electrodes, free-standing CNT membrane electrodes with randomly interlaced CNTs in the form of fabrics are a kind of promising carbon-based materials, exhibiting three-dimensionally (3D) interconnected nanoporous networks with higher specic surface areas [34,35]. Owing to their excellent thermal, chemical, mechanical and electronic properties, this type of CNT membranes has been used in diverse applications [3639]. However, the inherent strong hydrophobicity of CNTs dramatically hinders their https://doi.org/10.1016/j.seppur.2018.12.026 Received 1 November 2018; Received in revised form 7 December 2018; Accepted 11 December 2018 Corresponding author. E-mail address: [email protected] (Y. Wang). Separation and Purification Technology 213 (2019) 70–77 Available online 12 December 2018 1383-5866/ © 2018 Elsevier B.V. All rights reserved. T
8

Separation and Purification Technologyfunme.njtech.edu.cn/dfiles/17930/public/home/pdf/...Capacitive deionization (CDI) is an energy-efficient and environment-benign process to produce

Jan 24, 2021

Download

Documents

dariahiddleston
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
  • Contents lists available at ScienceDirect

    Separation and Purification Technology

    journal homepage: www.elsevier.com/locate/seppur

    Atomic layer deposition of TiO2 on carbon-nanotube membranes forenhanced capacitive deionization

    Jianhua Fenga,b, Sen Xionga, Yong Wanga,⁎

    a State Key Laboratory of Materials-Oriented Chemical Engineering, Jiangsu National Synergetic Innovation Center for Advanced Materials, and College of ChemicalEngineering, Nanjing Tech University, Nanjing 211816, Chinab College of Material and Chemical Engineering, Chuzhou University, Chuzhou 239000, China

    A R T I C L E I N F O

    Keywords:Carbon nanotubesAtomic layer depositionSurface modificationCapacitive deionizationDesalination

    A B S T R A C T

    Capacitive deionization (CDI) is an energy-efficient and environment-benign process to produce fresh water.Carbon nanotubes (CNTs) are promising building blocks in constructing high-performance CDI electrodes.Nevertheless, the strong hydrophobicity of CNTs significantly impedes their applications in aqueous environ-ment. Herein, hydrophilic CNT-based membrane electrodes are obtained via atomic layer deposition (ALD). Wedemonstrate, for the first time, that ALD is an efficient and flexible means of enhancing the CDI performance offree-standing CNT membrane electrodes by depositing TiO2 nanoparticles on the surface of CNTs. The CNTmembranes display preferable electrosorption behavior after moderate ALD cycle numbers and stable reusabilityin the desalination process.

    1. Introduction

    Under the situation of worldwide increasing demands and de-creasing supply of freshwater, extensive attention having being paid onadvanced desalination technologies. As a promising alternative to de-salination processes, capacitive deionization (CDI) is an environment-friendly and energy-efficient desalination technique in comparison withother desalination techniques which always suffer from the drawbackssuch as fouling, water electrolysis and high energy consumption [1–9].CDI based on the electrical double-layer capacitor (EDLC) theory is anelectrochemical water purification method and capable of reducing thesalt concentration of brackish and seawater by electrostatic adsorptionof ions on porous electrodes [10–12]. The electrosorption behaviorrelies significantly on the electrical conductivity, wettability and in-ternal structures of the CDI electrode materials [13]. Owing to theirhigh surface area, good flexibility and low electrical resistivity, carbon-based materials such as graphene, activated carbon (AC), carbonaerogels (CAs), carbon nanotubes (CNTs), and their composites havebeen widely investigated for the application in CDI electrode materialsover the past years [14–19]. However, the tedious treatments for thesynthesis and/or modification to the carbon building blocks, easy ag-gregation of powders, and binder addition often complicate the pre-paration process of the electrodes on one hand [20–22], and sometimesit cannot obtain acceptable electrosorption capacity on the other [14].Therefore, advanced electrode materials with good CDI performances

    which simultaneously have the advantages of simplifying preparationprocess, avoiding the aggregation, eliminating the blocking caused bybinders and can be directly used as electrodes are urgently needed.

    Incorporating pristine or modified CNTs with binders, then com-bining them with polymers or other porous carbon-based materials, andfinally depositing them onto a current collector is the most commonway to fabricate CDI composite electrodes. The use of CNTs is expectedto increase the specific surface area and enhance the electrical con-ductivity of composite materials, thus improving CDI performance[23–29]. However, these CNTs are commonly existed in the shape ofpowders and are typically required to be chemically modified in orderto have a good dispersion in the final electrodes, which always requiresa great deal of time and energy, and is also a tedious process [30,31].Furthermore, the release of individual CNTs into water or air is likely tocause safety issues during practical applications [32,33]. If the CNTscan be welded together and utilized as hydrophilic free-standing CNTmembrane electrodes, they will be more effective for CDI applications.Compared to CNT-doped composite electrodes, free-standing CNTmembrane electrodes with randomly interlaced CNTs in the form offabrics are a kind of promising carbon-based materials, exhibitingthree-dimensionally (3D) interconnected nanoporous networks withhigher specific surface areas [34,35]. Owing to their excellent thermal,chemical, mechanical and electronic properties, this type of CNTmembranes has been used in diverse applications [36–39]. However,the inherent strong hydrophobicity of CNTs dramatically hinders their

    https://doi.org/10.1016/j.seppur.2018.12.026Received 1 November 2018; Received in revised form 7 December 2018; Accepted 11 December 2018

    ⁎ Corresponding author.E-mail address: [email protected] (Y. Wang).

    Separation and Purification Technology 213 (2019) 70–77

    Available online 12 December 20181383-5866/ © 2018 Elsevier B.V. All rights reserved.

    T

    http://www.sciencedirect.com/science/journal/13835866https://www.elsevier.com/locate/seppurhttps://doi.org/10.1016/j.seppur.2018.12.026https://doi.org/10.1016/j.seppur.2018.12.026mailto:[email protected]://doi.org/10.1016/j.seppur.2018.12.026http://crossmark.crossref.org/dialog/?doi=10.1016/j.seppur.2018.12.026&domain=pdf

  • applications in the field of CDI as water cannot adequately wet the finepores in the membranes. Creating a hydrophilic interface on the CNTmembranes could markedly reduce the contact resistance between thewater and CNT membranes [40]. However, the interface engineering ofCNT membranes involves complex interactions and dynamics. Hence,there is a strong demand for effective techniques which can realize thehydrophilic modification and functionalization of CNT membraneelectrodes in a simple way.

    Atomic layer deposition (ALD), based on sequential self-limitingreactions of alternately injected gaseous precursors, is a promising thin-film-coating technique which could obtain excellent conformity, highlycontrollable thickness and morphology on the surfaces of various sub-strates. ALD has the capability to deposit a variety of target materials onporous substrates and yields strong chemical bonding between thesubstrate and the deposited materials, thus precisely regulating surfaceproperties, pore sizes and separation/adsorption applications by simplyvarying deposition conditions such as temperatures and cycle numbers[41–46]. Recently, ALD has been successfully employed to depositmetal oxides on carbon-based substrates for increasing charge storagefor supercapacitor and a variety of electrochemical applications[47,48]. Even though CNT membranes have been extensively in-vestigated for CDI applications [28,49,50], there are still some short-comings during the electrode fabrication and application processes.Herein, for the first time, we used ALD of titanium dioxide (TiO2),which is a highly hydrophilic, low-cost and eco-friendly metal oxide[51], onto free-standing CNT membranes to achieve binder-free CNT-TiO2 composite electrodes. Significant enhancement in hydrophilicity,electrochemical behavior and CDI performance were obtained by ALDof TiO2 on CNT membrane electrodes. The TiO2-deposited electrodesalso displayed superior reusability during CDI process. This strategy of“ALD on carbon substrates” opens a new avenue to produce advancedelectrodes for various electrochemical applications in addition to CDIconsidering that a large number of materials can be controllably ALD-deposited on various carbon-based substrates.

    2. Experimental section

    2.1. Materials

    Sheets of multi-walled CNT membranes (Suzhou JiediNanotechnology Co., Ltd) with a thickness of∼8 µm were chosen as thesubstrates in this work. Titanium tetrachloride (TiCl4, 99.99%,Metalorganic Center, Nanjing University) and deionized (DI) water(8–20 µs/cm, Wahaha) were selected for TiO2 deposition. Ultrahighpurity nitrogen (99.999%) and high purity nitrogen (99.9%) were usedas the carrier gas and purge gas in the ALD reactor, respectively.Sodium chloride, hydrochloric acid, anhydrous ethanol and other re-agents were all purchased from commercial sources and used withoutfurther treatment.

    2.2. Fabrication of TiO2-deposited CNT membrane electrodes

    The pristine CNT membranes were cut into pieces with dimensionsof 5 cm×5 cm and pre-treated with 20% hydrochloric acid, thenthoroughly rinsed with DI water, and finally dried at 120 °C as sub-strates. The deposition of TiO2 was performed in a commercialized ALDreactor (Savannah S100, Cambridge NanoTech). The dried CNT mem-branes were put into the chamber of ALD reactor and pretreated at theoperating temperature (100 °C) for 30min in vacuum (∼1 Torr) beforedeposition. The TiCl4 and DI water were stored in stainless cylinders atroom temperature and used as metal and oxygen precursors, respec-tively. The precursors were pulsed into the ALD reactor by the carriergas alternatively. For one TiO2 ALD cycle, the pulse durations of TiCl4and DI water were 0.03 and 0.015 s, respectively. After pulse, eachprecursor was held in the chamber for 5 s, and subsequently thechamber was purged for 20 s with nitrogen after each precursor ex-posure. The “exposure mode” was adopted to ensure uniform deposi-tion on porous CNT membranes and all operations were automaticallycontrolled via fast valves in ALD process. The CNT substrates weredeposited for 5, 10, 20, 40 and 60 cycles with a steady N2 flow rate of

    Fig. 1. Schematic illustration for the capacitive deionization system.

    J. Feng et al. Separation and Purification Technology 213 (2019) 70–77

    71

  • Fig. 2. SEM images of the (a) pristine CNT membrane, the (b) CNT@20TiO2 and (c) CNT@60TiO2 membrane electrode. (a)-(c) have the same magnification and thescale bar is shown in (c).

    Fig. 3. TiO2 mass percentages of CNT electrodes with different ALD cycles.Fig. 4. Water contact angles of TiO2-deposited CNT electrodes with differentALD cycles.

    J. Feng et al. Separation and Purification Technology 213 (2019) 70–77

    72

  • 20 sccm, respectively. The prepared CNT membrane electrodes weredenoted as CNT@xTiO2 (x presents the TiO2 ALD cycle numbers).Silicon wafers with thin native oxide (∼2 nm) and dimensions of2 cm×2 cm were also placed together with CNT membranes inside thereactor under the same ALD condition and used as the substrates toexamine the growth rate of metal oxide.

    2.3. Characterizations

    Scanning electron microscopy (SEM, Hitachi S-4800) was carriedout at the operation voltage of 5 kV to examine the surface morpholo-gies of the membrane electrodes. The weights of each 5 cm×5 cmsized membranes before and after ALD deposition were measured by amicro analytical balance and the corresponding TiO2 mass percentageswere calculated. A contact angle goniometer (Dropmeter A-100, Maist)was used to analyze the surface hydrophilicity of the membranes beforeand after deposition. Each sample was measured on different positionswith 5 µL water at ambient temperature and the average values werereported. The TiO2 film thicknesses deposited on silicon wafers weremeasured by a spectroscopic ellipsometer (Compete EASEM-2000U, J.A. Woollam) with an incident angle of 70°. The electrochemical prop-erties of the electrodes were evaluated by cyclic voltammetry (CV)using a CHI660E (Shanghai Chenhua) at 298 K. The sweep potentialrange was adjusted from −0.5 to 0.8 V in an electrochemical cell withconventional three-electrode system. The ALD-modified CNT mem-brane, platinum wire and Ag/AgCl electrode served as the working

    electrode, counter electrode and reference electrode, respectively. Theelectrochemical properties were measured in a 1M NaCl aqueous so-lution.

    2.4. Evaluation of CDI performances

    The CDI performances of the electrodes were evaluated with acontinuously recycling system including a CDI unit cell, a reservoir, adirect current power source, a conductivity meter and a peristalticpump (Fig. 1). The CDI unit cell is consisted of two parallel membraneelectrodes separated by a non-conductive non-woven spacer and sealedwith rubber gaskets. Before CDI tests, the membrane electrodes werepre-wetted with ethanol and then thoroughly washed with DI water.The electrodes were fabricated in the same manner as the electro-chemical measurements. The effective area of the electrode was12.5 cm2. In each experiment, the total NaCl aqueous solution volumeof 20mL with an initial conductivity of ∼86 µS/cm (NaCl concentra-tion: ∼40mg/L) was supplied to the CDI cell using a peristaltic pumpwith a flow rate of 3.8 mL/min and the solution temperature wasmaintained at 298 K. To confirm the correlation between applied vol-tage and electrosorption capacity, the electrical voltage was increasedfrom 0.8 to 1.6 V with an interval of 0.4 V. During the CDI process, theconductivity was monitored and measured by a conductivity metercontinuously. The relationship between concentration (mg/L) andconductivity (µS/cm) can be obtained by a calibration test prior to CDIexperiment (Fig. S1). The electrosorption capacity (EC, mg/g) is de-fined as the adsorbed ions amounts per gram of the membrane electrodeand can be calculated by the following Eq. (1):

    =

    −EC C C Vm

    ( )o e(1)

    where Co and Ce represent the initial and equilibrated concentrations(mg/L) of NaCl solution, respectively, V (L) is the total volume of theNaCl aqueous solution, and m (g) is the total mass of two membraneelectrodes.

    3. Results and discussion

    3.1. Morphology evolution of the CNT membranes during ALD of TiO2

    Fig. 2 shows the morphologies of pristine CNT, CNT@20TiO2 andCNT@60TiO2 membrane electrodes, respectively. It can be seen fromFig. 2a that the pristine CNT membrane with nanoporous structurearising from intertwined CNTs and larger pores contributed by thematrix of smooth fibers is consisted by a uniform, highly interconnectedCNTs network. As displayed in Fig. 2b and c, the TiO2 coverage on CNTsincreases with more ALD cycles. Below 20 ALD cycles, TiO2 nano-particulates distributed on the CNT surface are very tiny and, therefore,there is no remarkable morphology change compared with the pristineCNT. The ALD reactions can still occur on the CNT surface even thoughthere is barely no active sites. The unobvious morphology change atsmall number of ALD cycles can be attributed to the much smallergrowth rate (growth per cycle, GPC,∼0.49 Å) of TiO2 than that of othermetal oxides, e.g., ZnO. The adjacent TiO2 nanoparticulates are gra-dually approaching each other and thus forming a nearly continuousand intact layer along the CNTs with further increment of ALD cycles,e.g., 60 cycles (Fig. 2c).

    The TiO2 mass percentages of ALD-modified CNT electrodes wereshown in Fig. 3. With low ALD cycles, such as 5 cycles, TiO2 nano-particulates on the CNT surface are in the nucleation stage, so TiO2loading amount is very low (3.1 wt%). After the short nucleationperiod, the TiO2 mass percentages were continually promoted andreached 5.5 wt% and 9.7 wt% for the 10 and 20 cycles deposited CNTelectrodes, respectively. The TiO2 mass percentage of the electrode wasgreatly promoted to 20.7 wt% after 40 ALD cycles. The 60 cycles TiO2deposited CNT electrode exhibits a maximum loading (27.9 wt%). The

    Fig. 5. (a) Cyclic voltammetry curves for the TiO2-deposited CNT electrodes ata scan rate of 50mV/s, (b) CV profiles of the CNT@20TiO2 electrode at dif-ferent scan rates. All the curves were obtained in a 1M NaCl solution.

    J. Feng et al. Separation and Purification Technology 213 (2019) 70–77

    73

  • TiO2 loading on CNT electrodes gradually increases with the enhance-ment of ALD cycles and the loading promotion is relatively slow in theearly stage of ALD. This result confirms the occurrence of surface nu-cleation at the initial stage as described above.

    3.2. Surface hydrophilicity of the TiO2-deposited CNT electrodes

    The water affinity plays a significant role for CNT electrodes used inCDI to ensure that the entire pore volume is participating in the sorp-tion of ions in aqueous environments. Herein, to assess the surfacewettability, the water contact angles (WCAs) of CNT electrodes sub-jected different ALD cycles were measured. As illustrated in Fig. 4, the

    pristine porous CNT electrode showed an initial contact angle of∼114°, implying a strong hydrophobic surface. After 5 cycles TiO2deposition, the WCA slightly decreased to ∼101°, and further de-creased to ∼69° with 20 ALD cycles deposition. An obvious decrease ofthe WCA to ∼37° was achieved after 60 ALD cycles. This result showedthat the TiO2 could apparently enhance the affinity between the waterand the membrane electrode. With the ALD of TiO2, the surface of CNTswas progressively replaced by TiO2. The enhanced wettability is mainlyascribed to the existence of hydrophilic hydroxyl groups on TiO2 sur-face after ALD [52]. The excellent wettability of the TiO2-depositedCNT electrodes offers good compatibility to aqueous solutions, whichwill provide more accessible channels for ions and thus improve the CDIperformance.

    3.3. Electrochemical behavior of the TiO2-deposited CNT electrodes

    Cyclic voltammetry (CV) characterization is often used as an ef-fective measure to explore the electrosorption performance and eval-uate the specific capacitance of the CDI electrode materials [23]. Toevaluate the electrochemical performance of TiO2-deposited CNTmembranes as electrodes, CV measurements in a three-electrode systemwere performed. The CV profiles of TiO2-deposited CNT electrodes withdifferent ALD cycles in the range of −0.5 to 0.8 V are shown in Fig. 5a.The encircled areas of TiO2-deposited CNT electrodes gradually enlargewith the increment of ALD cycles, implying that the electrodes havelarger capacitances. Even though the CNT@40TiO2 (20.7 wt% TiO2)and CNT@60TiO2 (27.9 wt% TiO2) electrodes have larger encircled

    Fig. 6. Desalination curves of the TiO2-deposited CNT electrodes at applied voltages of (a) 0.8 V, (b) 1.2 V, (c) 1.6 V and (d) electrosorption capacities of the pristineand TiO2-deposited CNT electrodes at different electrical voltages.

    Table 1Comparison of the electrosorption capacity of CNT@20TiO2 electrode withother reported in literature.

    Samples Initial NaClconcentration(mg/L)

    Voltage (V) Electrosorptioncapacity (mg/g)

    Ref.

    Graphene/MC 89.5 2.0 0.73 [10]Graphene/AC ∼50 2.0 0.85 [13]Graphene ∼50 2.0 1.85 [14]GHMCSs 68.5 1.6 2.3 [54]CNTs-RGO 100 1.6 0.9 [55]Gr/SnO2 25 1.4 1.49 [56]CNT@20TiO2 ∼40 1.6 5.09 This

    work

    J. Feng et al. Separation and Purification Technology 213 (2019) 70–77

    74

  • areas than other electrodes, however, there are some redox peaksduring the measurements and the oval shape of CV curve which can beattributed to the existence of massive hydrophilic hydroxyl groups onTiO2 surface after ALD [25,53]. On the contrary, the CV curve of CNT@20TiO2 (9.7 wt% TiO2) electrode exhibits a nearly rectangular shape,indicating an ideal electrical double layer (EDL) capacitive behaviorand higher electrochemical performances. The CV curves of CNT@20TiO2 at various scan rates (5–100mV/s) are displayed in Fig. 5b. TheCV curve presents rectangular shape at a low scan rate, which meansthe salty ions of the aqueous solution can quickly and effectivelytransport into the electrode surface. However, the CV curve has anincreased distortion when the scan rate is up to 50mV/s. At a higherscan rate, the sodium and chloride ions have not enough time to ac-cumulate and move into the internal pores of the electrode, andmeanwhile the ohmic resistance increases correspondingly, thus re-sulting in the incomplete EDL formation [53]. From the above analysisresults, we can conclude that the CNT@20TiO2 (9.7 wt% TiO2) elec-trode presents a better EDL capacitive behavior, indicating the prefer-able CDI performance.

    3.4. CDI performance of the TiO2-deposited CNT electrodes

    The desalination curves of pristine and TiO2-deposited CNT elec-trodes in NaCl solution with an initial conductivity of ∼86 µS/cm atdifferent electrical voltages are shown in Fig. 6a-c. The electrical

    potential was set from 0.8 V to 1.6 V with an interval of 0.4 V and allcharging processes were carried out for the time when the conductivitydoesn’t change anymore, indicating the saturation reached. Obviously,it can be seen that the conductivity goes to decrease once the electricalpotential applied. When the electrical potential is set as 0.8 V, the so-lution conductivity exhibits slight decrease on the whole. However,when the applied voltage is increased to 1.2 V or 1.6 V, a dramaticdecrease of the conductivity appears at the initial stage which indicatesquick electrosorption of the salt ions. With the operation continuing,the rate of the adsorption of ions represented in conductivity curvebecomes slowly due to the electrosorption saturation. The electro-sorption capacities of pristine and TiO2-deposited CNT electrodes ob-tained at different electrical voltages are depicted in Fig. 6d. At a higherapplied voltage, the NaCl electrosorption capacity evidently increaseswith operation time, indicating that the higher voltage leads to agreater NaCl removal amount owing to the stronger electrostatic in-teraction. Additionally, at any certain voltage, the electrosorption ca-pacity of CNT@20TiO2 electrode is much higher than that of pristineCNT electrode, demonstrating an enhanced CDI performance. Theelectrosorption capacity of CNT@20TiO2 electrode was calculated ac-cording to Eq. (1) as 3.33 and 5.09mg/g at 1.2 and 1.6 V, respectively,which is much higher than those (1.35 and 3.13mg/g, respectively) ofthe pristine CNT electrode, and exceed those of many other CDI elec-trode materials [10,13,14,54–56]. Meanwhile, it should be worthwhilenoting that, there exist a relatively slight conductivity reduction whichcan be ascribed to the smaller weight of CNT membrane electrode inour work compared with other literature. Table 1 summarizes theelectrosorption capacities of CDI electrode materials, proving that theCNT@20TiO2 electrode is superior than other electrodes. This superiordesalination capacity of the CNT@20TiO2 electrode can be attributed tothe improvement in hydrophilicity and the increase in the number ofadsorption sites for ions under electric field by the participation of ti-tanium atoms [57]. It is noted that the salt adsorption capacities ofCNT@40TiO2 and CNT@60TiO2 are lower than that of CNT@20TiO2under a certain voltage though they possess relatively large CV en-circled areas, which can be ascribed to the decrease of bulk con-ductivity and effective electrosorption sites on the electrodes.

    3.5. Regeneration and reusability of CNT@20TiO2 electrodes

    The regeneration and reusability of the CDI electrode materials isthe most significant factor for practical application. To evaluate theregeneration stability of the CNT@20TiO2 membrane electrode, amultiple electrosorption-desorption experiment was carried out, inwhich the electrosorption voltage was 1.2 V and the desorption voltagewas 0 V. The conductivity variation of NaCl solution during electro-sorption and desorption cycles is shown in Fig. 7a. When the voltagewas applied, the conductivity of the salt solution quickly decreased.After the electrosorption process, the charge voltage was short-circuitedand the adsorbed ions were released from the electrode surface to thebulk solution. No appreciable decrease in the solution conductivity isobserved, indicating that the CNT@20TiO2 electrode has good stabilityin multiple electrosorption-desorption cycles. Fig. 7b displays theelectrosorption capacities of the CNT@20TiO2 electrode during the fourcycling operations which are 3.42, 3.48, 3.31 and 3.20mg/g, respec-tively. It can be seen that there was no obvious decline in removalcapacity after multiple regeneration cycles, indicating that the elec-trode has an excellent and stable reusability for CDI. Therefore, theTiO2-deposited CNT membrane can be considered as a promisingelectrode material for CDI applications.

    4. Conclusions

    In summary, we have demonstrated the successful modification ofCNT membranes via atomic-layer-deposited TiO2 to produce superiorCDI electrodes. The surface coverage of CNTs and TiO2

    Fig. 7. (a) Desalination-regeneration curves and (b) electrosorption capacitiesof the CNT@20TiO2 electrode at 4 cycles.

    J. Feng et al. Separation and Purification Technology 213 (2019) 70–77

    75

  • nanoparticulates loading amounts could be precisely controlled by theALD cycle numbers. The wettability of the TiO2-deposited CNT mem-brane is progressively transformed from strongly hydrophobic to hy-drophilic. Compared to pristine CNT membrane, the functionalizedCNT electrode with moderate ALD cycles presents the preferable elec-trochemical behavior. The CDI performance is dramatically improvedby the CNT@20TiO2 membrane electrode which displays a good re-generation and reusability performances. This “ALD on carbon sub-strates” strategy is expected to extend to fabricate other electrochemicalmaterials for various applications in addition to desalination.

    Conflicts of interest

    The authors declare no competing financial interest.

    Acknowledgements

    Financial supports from the National Basic Research Program ofChina (2015CB655301) and the Jiangsu Natural Science Foundation(BK20150063) are gratefully acknowledged. We also thank the supportfrom the Program of Excellent Innovation Teams of Jiangsu HigherEducation Institutions, and the Project of Priority Academic ProgramDevelopment of Jiangsu Higher Education Institutions (PAPD). Mr.Yimin Guo, a student in Nanjing Foreign Language School, also con-tributed to this work by taking part in preparing membranes and CDItesting.

    Appendix A. Supplementary material

    Supplementary data to this article can be found online at https://doi.org/10.1016/j.seppur.2018.12.026.

    References

    [1] M.A. Shannon, P.W. Bohn, M. Elimelech, J.G. Georgiadis, B.J. Mariñas, A.M. Mayes,Science and technology for water purification in the coming decades, Nature 452(2008) 301–310.

    [2] L. Malaeb, G.M. Ayoub, Reverse osmosis technology for water treatment: state ofthe art review, Desalination 267 (2011) 1–8.

    [3] H. Strathmann, Electrodialysis, a mature technology with a multitude of new ap-plications, Desalination 264 (2010) 268–288.

    [4] S. Bonyadi, T.S. Chung, Flux enhancement in membrane distillation by fabricationof dual layer hydrophilic–hydrophobic hollow fiber membranes, J. Membr. Sci. 306(2007) 134–146.

    [5] M.A. Anderson, A.L. Cudero, J. Palma, Capacitive deionization as an electro-chemical means of saving energy and delivering clean water. Comparison to presentdesalination practices: will it compete? Electrochim. Acta 55 (2010) 3845–3856.

    [6] M.A. Ahmed, S. Tewari, Capacitive deionization: Processes, materials and state ofthe technology, J. Electroanal. Chem. 813 (2018) 178–192.

    [7] M.E. Suss, S. Porada, X. Sun, P.M. Biesheuvel, J. Yoon, V. Presser, Water desali-nation via capacitive deionization: what is it and what can we expect from it?Energy Environ. Sci. 8 (2015) 2296–2319.

    [8] L. Agartan, B. Akuzum, T. Mathis, K. Ergenekon, E. Agar, E.C. Kumbur, Influence ofthermal treatment conditions on capacitive deionization performance and chargeefficiency of carbon electrodes, Sep. Purif. Technol. 202 (2018) 67–75.

    [9] J.E. Dykstra, S. Porada, A. van der Wal, P.M. Biesheuvel, Energy consumption incapacitive deionization - Constant current versus constant voltage operation, WaterRes. 143 (2018) 367–375.

    [10] D. Zhang, X. Wen, L. Shi, T. Yan, J. Zhang, Enhanced capacitive deionization ofgraphene/mesoporous carbon composites, Nanoscale 4 (2012) 5440–5446.

    [11] Y.-J. Kim, J.-H. Choi, Enhanced desalination efficiency in capacitive deionizationwith an ion-selective membrane, Sep. Purif. Technol. 71 (2010) 70–75.

    [12] M. Mossad, L. Zou, Study of fouling and scaling in capacitive deionisation by usingdissolved organic and inorganic salts, J. Hazard. Mater. 244–245 (2013) 387–393.

    [13] H. Li, L. Pan, C. Nie, Y. Liu, Z. Sun, Reduced graphene oxide and activated carboncomposites for capacitive deionization, J. Mater. Chem. 22 (2012) 15556–15561.

    [14] H. Li, T. Lu, L. Pan, Y. Zhang, Z. Sun, Electrosorption behavior of graphene in NaClsolutions, J. Mater. Chem. 19 (2009) 6773–6779.

    [15] P. Xu, J.E. Drewes, D. Heil, G. Wang, Treatment of brackish produced water usingcarbon aerogel-based capacitive deionization technology, Water Res. 42 (2008)2605–2617.

    [16] G. Wang, B. Qian, Q. Dong, J. Yang, Z. Zhao, J. Qiu, Highly mesoporous activatedcarbon electrode for capacitive deionization, Sep. Purif. Technol. 103 (2013)216–221.

    [17] M.A. Tofighy, T. Mohammadi, Salty water desalination using carbon nanotube

    sheets, Desalination 258 (2010) 182–186.[18] C. Santos, J.J. Lado, E. García-Quismondo, I.V. Rodríguez, D. Hospital-Benito,

    J. Palma, M.A. Anderson, J.J. Vilatela, Interconnected metal oxide CNT fibre hybridnetworks for current collector-free asymmetric capacitive deionization, J. Mater.Chem. A 6 (2018) 10898–10908.

    [19] H. Lei, T. Yan, H. Wang, L. Shi, J. Zhang, D. Zhang, Graphene-like carbon na-nosheets prepared by a Fe-catalyzed glucose-blowing method for capacitive deio-nization, J. Mater. Chem. A 3 (2015) 5934–5941.

    [20] Z. Sui, Q. Meng, X. Zhang, R. Ma, B. Cao, Green synthesis of carbon nanotube-graphene hybrid aerogels and their use as versatile agents for water purification, J.Mater. Chem. 22 (2012) 8767–8771.

    [21] S. Chen, Q. Tang, X. Chen, L. Tan, Nitrogen-doped carbon coated LiFePO4/carbonnanotube interconnected nanocomposites for high performance lithium ion bat-teries, New J. Chem. 39 (2015) 9782–9788.

    [22] Z.-Y. Yang, L.-J. Jin, G.-Q. Lu, Q.-Q. Xiao, Y.-X. Zhang, L. Jing, X.-X. Zhang, Y.-M. Yan, K.-N. Sun, Sponge-templated preparation of high surface area graphenewith ultrahigh capacitive deionization performance, Adv. Funct. Mater. 24 (2014)3917–3925.

    [23] M. Moronshing, C. Subramaniam, Scalable approach to highly efficient and rapidcapacitive deionization with CNT-Thread as electrodes, ACS appl. Mater. Interf. 9(2017) 39907–39915.

    [24] J. Cao, Y. Wang, C. Chen, F. Yu, J. Ma, A comparison of graphene hydrogelsmodified with single-walled/multi-walled carbon nanotubes as electrode materialsfor capacitive deionization, J. Colloid Interf. Sci. 518 (2018) 69–75.

    [25] K. Shi, M. Ren, I. Zhitomirsky, Activated carbon-coated carbon nanotubes for en-ergy storage in supercapacitors and capacitive water purification, ACS SustainableChem. Eng. 2 (2014) 1289–1298.

    [26] Q. Dong, G. Wang, T. Wu, S. Peng, J. Qiu, Enhancing capacitive deionization per-formance of electrospun activated carbon nanofibers by coupling with carbon na-notubes, J. Colloid Interf. Sci. 446 (2015) 373–378.

    [27] Y. Wang, L. Zhang, Y. Wu, S. Xu, J. Wang, Polypyrrole/carbon nanotube compositesas cathode material for performance enhancing of capacitive deionization tech-nology, Desalination 354 (2014) 62–67.

    [28] J. Benson, I. Kovalenko, S. Boukhalfa, D. Lashmore, M. Sanghadasa, G. Yushin,Multifunctional CNT-polymer composites for ultra-tough structural supercapacitorsand desalination devices, Adv. Mater. 25 (2013) 6625–6632.

    [29] H. Li, Y. Ma, R. Niu, Improved capacitive deionization performance by couplingTiO2 nanoparticles with carbon nanotubes, Sep. Purif. Technol. 171 (2016) 93–100.

    [30] N.G. Sahoo, S. Rana, J.W. Cho, L. Li, S.H. Chan, Polymer nanocomposites based onfunctionalized carbon nanotubes, Prog. polym. Sci. 35 (2010) 837–867.

    [31] S. Wang, D. Wang, L. Ji, Q. Gong, Y. Zhu, J. Liang, Equilibrium and kinetic studieson the removal of NaCl from aqueous solutions by electrosorption on carbon na-notube electrodes, Sep. Purif. Technol. 58 (2007) 12–16.

    [32] S.K. Manna, S. Sarkar, J. Barr, K. Wise, E.V. Barrera, O. Jejelowo, A.C. Rice-Ficht,G.T. Ramesh, Single-walled carbon nanotube induces oxidative stress and activatesnuclear transcription factor-κB in human keratinocytes, Nano Lett. 5 (2005)1676–1684.

    [33] A. Magrez, S. Kasas, V. Salicio, N. Pasquier, W.S. Jin, M. Celio, S. Catsicas,B. Schwaller, L. Forró, Cellular toxicity of carbon-based nanomaterials, Nano Lett. 6(2006) 1121–1125.

    [34] S. Roy, V. Jain, R. Bajpai, P. Ghosh, A. Pente, B. Singh, D. Misra, Formation ofcarbon nanotube bucky paper and feasibility study for filtration at the nano andmolecular scale, J. Phys. Chem. C 116 (2012) 19025–19031.

    [35] R. Das, M.E. Ali, S.B.A. Hamid, S. Ramakrishna, Z.Z. Chowdhury, Carbon nanotubemembranes for water purification: a bright future in water desalination,Desalination 336 (2014) 97–109.

    [36] U. Vohrer, I. Kolaric, M. Haque, S. Roth, U. Detlaff-Weglikowska, Carbon nanotubesheets for the use as artificial muscles, Carbon 42 (2004) 1159–1164.

    [37] R.H. Baughman, C. Cui, A.A. Zakhidov, Z. Iqbal, J.N. Barisci, G.M. Spinks,G.G. Wallace, A. Mazzoldi, D. De Rossi, A.G. Rinzler, Carbon nanotube actuators,Science 284 (1999) 1340–1344.

    [38] N.A. Prokudina, E.R. Shishchenko, O.-S. Joo, K.-H. Hyung, S.-H. Han, A carbonnanotube film as a radio frequency filter, Carbon 43 (2005) 1815–1819.

    [39] Z. Yang, T. Chen, R. He, G. Guan, H. Li, L. Qiu, H. Peng, Aligned carbon nanotubesheets for the electrodes of organic solar cells, Adv. Mater. 23 (2011) 5436–5439.

    [40] S.B. Darling, Perspective: Interfacial materials at the interface of energy and water,J. Appl. Phys. 124 (2018) 030901.

    [41] H. Chen, S. Wu, X. Jia, S. Xiong, Y. Wang, Atomic layer deposition fabricating ofceramic nanofiltration membranes for efficient separation of dyes from water,AIChE J. 64 (2018) 2670–2678.

    [42] J. Feng, S. Xiong, Z. Wang, Z. Cui, S.-P. Sun, Y. Wang, Atomic layer deposition ofmetal oxides on carbon nanotube fabrics for robust, hydrophilic ultrafiltrationmembranes, J. Membr. Sci. 550 (2018) 246–253.

    [43] S. Xiong, Y. Yang, Z. Zhong, Y. Wang, One-step synthesis of carbon-hybridized ZnOon polymeric foams by atomic layer deposition for efficient absorption of oils fromwater, Ind. Eng. Chem. Res. 57 (2018) 1269–1276.

    [44] H. Wang, M. Wei, Z. Zhong, Y. Wang, Atomic-layer-deposition-enabled thin-filmcomposite membranes of polyimide supported on nanoporous anodized alumina, J.Membr. Sci. 535 (2017) 56–62.

    [45] M. Xie, X. Sun, H. Sun, T. Porcelli, S.M. George, Y. Zhou, J. Lian, Stabilizing anamorphous V2O5/carbon nanotube paper electrode with conformal TiO2 coating byatomic layer deposition for lithium ion batteries, J. Mater. Chem. A 4 (2016)537–544.

    [46] H.C. Yang, R.Z. Waldman, Z. Chen, S.B. Darling, Atomic layer deposition formembrane interface engineering, Nanoscale 10 (2018) 20505–20513.

    [47] X. Sun, M. Xie, J.J. Travis, G. Wang, H. Sun, J. Lian, S.M. George,

    J. Feng et al. Separation and Purification Technology 213 (2019) 70–77

    76

    https://doi.org/10.1016/j.seppur.2018.12.026https://doi.org/10.1016/j.seppur.2018.12.026http://refhub.elsevier.com/S1383-5866(18)33831-0/h0005http://refhub.elsevier.com/S1383-5866(18)33831-0/h0005http://refhub.elsevier.com/S1383-5866(18)33831-0/h0005http://refhub.elsevier.com/S1383-5866(18)33831-0/h0010http://refhub.elsevier.com/S1383-5866(18)33831-0/h0010http://refhub.elsevier.com/S1383-5866(18)33831-0/h0015http://refhub.elsevier.com/S1383-5866(18)33831-0/h0015http://refhub.elsevier.com/S1383-5866(18)33831-0/h0020http://refhub.elsevier.com/S1383-5866(18)33831-0/h0020http://refhub.elsevier.com/S1383-5866(18)33831-0/h0020http://refhub.elsevier.com/S1383-5866(18)33831-0/h0025http://refhub.elsevier.com/S1383-5866(18)33831-0/h0025http://refhub.elsevier.com/S1383-5866(18)33831-0/h0025http://refhub.elsevier.com/S1383-5866(18)33831-0/h0030http://refhub.elsevier.com/S1383-5866(18)33831-0/h0030http://refhub.elsevier.com/S1383-5866(18)33831-0/h0035http://refhub.elsevier.com/S1383-5866(18)33831-0/h0035http://refhub.elsevier.com/S1383-5866(18)33831-0/h0035http://refhub.elsevier.com/S1383-5866(18)33831-0/h0040http://refhub.elsevier.com/S1383-5866(18)33831-0/h0040http://refhub.elsevier.com/S1383-5866(18)33831-0/h0040http://refhub.elsevier.com/S1383-5866(18)33831-0/h0045http://refhub.elsevier.com/S1383-5866(18)33831-0/h0045http://refhub.elsevier.com/S1383-5866(18)33831-0/h0045http://refhub.elsevier.com/S1383-5866(18)33831-0/h0050http://refhub.elsevier.com/S1383-5866(18)33831-0/h0050http://refhub.elsevier.com/S1383-5866(18)33831-0/h0055http://refhub.elsevier.com/S1383-5866(18)33831-0/h0055http://refhub.elsevier.com/S1383-5866(18)33831-0/h0060http://refhub.elsevier.com/S1383-5866(18)33831-0/h0060http://refhub.elsevier.com/S1383-5866(18)33831-0/h0065http://refhub.elsevier.com/S1383-5866(18)33831-0/h0065http://refhub.elsevier.com/S1383-5866(18)33831-0/h0070http://refhub.elsevier.com/S1383-5866(18)33831-0/h0070http://refhub.elsevier.com/S1383-5866(18)33831-0/h0075http://refhub.elsevier.com/S1383-5866(18)33831-0/h0075http://refhub.elsevier.com/S1383-5866(18)33831-0/h0075http://refhub.elsevier.com/S1383-5866(18)33831-0/h0080http://refhub.elsevier.com/S1383-5866(18)33831-0/h0080http://refhub.elsevier.com/S1383-5866(18)33831-0/h0080http://refhub.elsevier.com/S1383-5866(18)33831-0/h0085http://refhub.elsevier.com/S1383-5866(18)33831-0/h0085http://refhub.elsevier.com/S1383-5866(18)33831-0/h0090http://refhub.elsevier.com/S1383-5866(18)33831-0/h0090http://refhub.elsevier.com/S1383-5866(18)33831-0/h0090http://refhub.elsevier.com/S1383-5866(18)33831-0/h0090http://refhub.elsevier.com/S1383-5866(18)33831-0/h0095http://refhub.elsevier.com/S1383-5866(18)33831-0/h0095http://refhub.elsevier.com/S1383-5866(18)33831-0/h0095http://refhub.elsevier.com/S1383-5866(18)33831-0/h0100http://refhub.elsevier.com/S1383-5866(18)33831-0/h0100http://refhub.elsevier.com/S1383-5866(18)33831-0/h0100http://refhub.elsevier.com/S1383-5866(18)33831-0/h0105http://refhub.elsevier.com/S1383-5866(18)33831-0/h0105http://refhub.elsevier.com/S1383-5866(18)33831-0/h0105http://refhub.elsevier.com/S1383-5866(18)33831-0/h0110http://refhub.elsevier.com/S1383-5866(18)33831-0/h0110http://refhub.elsevier.com/S1383-5866(18)33831-0/h0110http://refhub.elsevier.com/S1383-5866(18)33831-0/h0110http://refhub.elsevier.com/S1383-5866(18)33831-0/h0115http://refhub.elsevier.com/S1383-5866(18)33831-0/h0115http://refhub.elsevier.com/S1383-5866(18)33831-0/h0115http://refhub.elsevier.com/S1383-5866(18)33831-0/h0120http://refhub.elsevier.com/S1383-5866(18)33831-0/h0120http://refhub.elsevier.com/S1383-5866(18)33831-0/h0120http://refhub.elsevier.com/S1383-5866(18)33831-0/h0125http://refhub.elsevier.com/S1383-5866(18)33831-0/h0125http://refhub.elsevier.com/S1383-5866(18)33831-0/h0125http://refhub.elsevier.com/S1383-5866(18)33831-0/h0130http://refhub.elsevier.com/S1383-5866(18)33831-0/h0130http://refhub.elsevier.com/S1383-5866(18)33831-0/h0130http://refhub.elsevier.com/S1383-5866(18)33831-0/h0135http://refhub.elsevier.com/S1383-5866(18)33831-0/h0135http://refhub.elsevier.com/S1383-5866(18)33831-0/h0135http://refhub.elsevier.com/S1383-5866(18)33831-0/h0140http://refhub.elsevier.com/S1383-5866(18)33831-0/h0140http://refhub.elsevier.com/S1383-5866(18)33831-0/h0140http://refhub.elsevier.com/S1383-5866(18)33831-0/h0145http://refhub.elsevier.com/S1383-5866(18)33831-0/h0145http://refhub.elsevier.com/S1383-5866(18)33831-0/h0150http://refhub.elsevier.com/S1383-5866(18)33831-0/h0150http://refhub.elsevier.com/S1383-5866(18)33831-0/h0155http://refhub.elsevier.com/S1383-5866(18)33831-0/h0155http://refhub.elsevier.com/S1383-5866(18)33831-0/h0155http://refhub.elsevier.com/S1383-5866(18)33831-0/h0160http://refhub.elsevier.com/S1383-5866(18)33831-0/h0160http://refhub.elsevier.com/S1383-5866(18)33831-0/h0160http://refhub.elsevier.com/S1383-5866(18)33831-0/h0160http://refhub.elsevier.com/S1383-5866(18)33831-0/h0165http://refhub.elsevier.com/S1383-5866(18)33831-0/h0165http://refhub.elsevier.com/S1383-5866(18)33831-0/h0165http://refhub.elsevier.com/S1383-5866(18)33831-0/h0170http://refhub.elsevier.com/S1383-5866(18)33831-0/h0170http://refhub.elsevier.com/S1383-5866(18)33831-0/h0170http://refhub.elsevier.com/S1383-5866(18)33831-0/h0175http://refhub.elsevier.com/S1383-5866(18)33831-0/h0175http://refhub.elsevier.com/S1383-5866(18)33831-0/h0175http://refhub.elsevier.com/S1383-5866(18)33831-0/h0180http://refhub.elsevier.com/S1383-5866(18)33831-0/h0180http://refhub.elsevier.com/S1383-5866(18)33831-0/h0185http://refhub.elsevier.com/S1383-5866(18)33831-0/h0185http://refhub.elsevier.com/S1383-5866(18)33831-0/h0185http://refhub.elsevier.com/S1383-5866(18)33831-0/h0190http://refhub.elsevier.com/S1383-5866(18)33831-0/h0190http://refhub.elsevier.com/S1383-5866(18)33831-0/h0195http://refhub.elsevier.com/S1383-5866(18)33831-0/h0195http://refhub.elsevier.com/S1383-5866(18)33831-0/h0200http://refhub.elsevier.com/S1383-5866(18)33831-0/h0200http://refhub.elsevier.com/S1383-5866(18)33831-0/h0205http://refhub.elsevier.com/S1383-5866(18)33831-0/h0205http://refhub.elsevier.com/S1383-5866(18)33831-0/h0205http://refhub.elsevier.com/S1383-5866(18)33831-0/h0210http://refhub.elsevier.com/S1383-5866(18)33831-0/h0210http://refhub.elsevier.com/S1383-5866(18)33831-0/h0210http://refhub.elsevier.com/S1383-5866(18)33831-0/h0215http://refhub.elsevier.com/S1383-5866(18)33831-0/h0215http://refhub.elsevier.com/S1383-5866(18)33831-0/h0215http://refhub.elsevier.com/S1383-5866(18)33831-0/h0220http://refhub.elsevier.com/S1383-5866(18)33831-0/h0220http://refhub.elsevier.com/S1383-5866(18)33831-0/h0220http://refhub.elsevier.com/S1383-5866(18)33831-0/h0225http://refhub.elsevier.com/S1383-5866(18)33831-0/h0225http://refhub.elsevier.com/S1383-5866(18)33831-0/h0225http://refhub.elsevier.com/S1383-5866(18)33831-0/h0225http://refhub.elsevier.com/S1383-5866(18)33831-0/h0230http://refhub.elsevier.com/S1383-5866(18)33831-0/h0230http://refhub.elsevier.com/S1383-5866(18)33831-0/h0235

  • Pseudocapacitance of amorphous TiO2 thin films anchored to graphene and carbonnanotubes using atomic layer deposition, J. Phys. Chem. C 117 (2013)22497–22508.

    [48] S. Boukhalfa, K. Evanoff, G. Yushin, Atomic layer deposition of vanadium oxide oncarbon nanotubes for high-power supercapacitor electrodes, Energy Environ. Sci. 5(2012) 6872–6879.

    [49] L. Wang, M. Wang, Z.-H. Huang, T. Cui, X. Gui, F. Kang, K. Wang, D. Wu, Capacitivedeionization of NaCl solutions using carbon nanotube sponge electrodes, J. Mater.Chem. 21 (2011) 18295–18299.

    [50] C. Nie, L. Pan, H. Li, T. Chen, T. Lu, Z. Sun, Electrophoretic deposition of carbonnanotubes film electrodes for capacitive deionization, J. Electroanal. Chem. 666(2012) 85–88.

    [51] H. Schäfer, B. Milow, L. Ratke, Synthesis of inorganic aerogels via rapid gelationusing chloride precursors, RSC Adv. 3 (2013) 15263–15272.

    [52] Q. Xu, Y. Yang, J. Yang, X. Wang, Z. Wang, Y. Wang, Plasma activation of porouspolytetrafluoroethylene membranes for superior hydrophilicity and separation

    performances via atomic layer deposition of TiO2, J. Membr. Sci. 443 (2013) 62–68.[53] D. Wang, Z. Geng, B. Li, C. Zhang, High performance electrode materials for electric

    double-layer capacitors based on biomass-derived activated carbons, Electrochim.Acta 173 (2015) 377–384.

    [54] H. Wang, L. Shi, T. Yan, J. Zhang, Q. Zhong, D. Zhang, Design of graphene-coatedhollow mesoporous carbon spheres as high performance electrodes for capacitivedeionization, J. Mater. Chem. A 2 (2014) 4739–4750.

    [55] H. Li, S. Liang, J. Li, L. He, The capacitive deionization behaviour of a carbon na-notube and reduced graphene oxide composite, J. Mater. Chem. A 1 (2013)6335–6341.

    [56] A.G. El-Deen, N.A.M. Barakat, K.A. Khalil, M. Motlak, H.-Y. Kim, Graphene/SnO2nanocomposite as an effective electrode material for saline water desalination usingcapacitive deionization, Ceram. Int. 40 (2014) 14627–14634.

    [57] M.-W. Ryoo, G. Seo, Improvement in capacitive deionization function of activatedcarbon cloth by titania modification, Water Res. 37 (2003) 1527–1534.

    J. Feng et al. Separation and Purification Technology 213 (2019) 70–77

    77

    http://refhub.elsevier.com/S1383-5866(18)33831-0/h0235http://refhub.elsevier.com/S1383-5866(18)33831-0/h0235http://refhub.elsevier.com/S1383-5866(18)33831-0/h0235http://refhub.elsevier.com/S1383-5866(18)33831-0/h0240http://refhub.elsevier.com/S1383-5866(18)33831-0/h0240http://refhub.elsevier.com/S1383-5866(18)33831-0/h0240http://refhub.elsevier.com/S1383-5866(18)33831-0/h0245http://refhub.elsevier.com/S1383-5866(18)33831-0/h0245http://refhub.elsevier.com/S1383-5866(18)33831-0/h0245http://refhub.elsevier.com/S1383-5866(18)33831-0/h0250http://refhub.elsevier.com/S1383-5866(18)33831-0/h0250http://refhub.elsevier.com/S1383-5866(18)33831-0/h0250http://refhub.elsevier.com/S1383-5866(18)33831-0/h0255http://refhub.elsevier.com/S1383-5866(18)33831-0/h0255http://refhub.elsevier.com/S1383-5866(18)33831-0/h0260http://refhub.elsevier.com/S1383-5866(18)33831-0/h0260http://refhub.elsevier.com/S1383-5866(18)33831-0/h0260http://refhub.elsevier.com/S1383-5866(18)33831-0/h0265http://refhub.elsevier.com/S1383-5866(18)33831-0/h0265http://refhub.elsevier.com/S1383-5866(18)33831-0/h0265http://refhub.elsevier.com/S1383-5866(18)33831-0/h0270http://refhub.elsevier.com/S1383-5866(18)33831-0/h0270http://refhub.elsevier.com/S1383-5866(18)33831-0/h0270http://refhub.elsevier.com/S1383-5866(18)33831-0/h0275http://refhub.elsevier.com/S1383-5866(18)33831-0/h0275http://refhub.elsevier.com/S1383-5866(18)33831-0/h0275http://refhub.elsevier.com/S1383-5866(18)33831-0/h0280http://refhub.elsevier.com/S1383-5866(18)33831-0/h0280http://refhub.elsevier.com/S1383-5866(18)33831-0/h0280http://refhub.elsevier.com/S1383-5866(18)33831-0/h0285http://refhub.elsevier.com/S1383-5866(18)33831-0/h0285

    Atomic layer deposition of TiO2 on carbon-nanotube membranes for enhanced capacitive deionizationIntroductionExperimental sectionMaterialsFabrication of TiO2-deposited CNT membrane electrodesCharacterizationsEvaluation of CDI performances

    Results and discussionMorphology evolution of the CNT membranes during ALD of TiO2Surface hydrophilicity of the TiO2-deposited CNT electrodesElectrochemical behavior of the TiO2-deposited CNT electrodesCDI performance of the TiO2-deposited CNT electrodesRegeneration and reusability of CNT@20TiO2 electrodes

    ConclusionsConflicts of interestAcknowledgementsSupplementary materialReferences