Hydroxide Conduction Enhancement of Chitosan Membranes by Functionalized Mxene

materials

Article Hydroxide Conduction Enhancement of Chitosan Membranes by Functionalized MXene

Lina Wang 1,* and Benbing Shi 2,*

1 College of Chemistry and Chemical Engineering, Xi’an University of Science and Technology, Xi’an 710054, China 2 School of Chemical Engineering & Technology, Tianjin University, Tianjin 300072, China * Correspondence: [email protected] (L.W.); [email protected] (B.S.)

Received: 7 September 2018; Accepted: 17 November 2018; Published: 21 November 2018

Abstract: In this study, imidazolium brushes tethered by –NH2-containing ligands were grafted onto the surface of a 2D material, MXene, using precipitation polymerization followed by quaternization. Functionalized MXene was embedded into chitosan matrix to prepare a hybrid alkaline anion exchange membrane. Due to high interfacial compatibility, functionalized MXene was homogeneously dispersed in chitosan matrix, generating continuous ion conduction channels and then greatly enhancing OH− conduction property (up to 172%). The ability and mechanism of OH− conduction in the membrane were elaborated based on systematic tests. The mechanical-thermal stability and swelling resistance of the membrane were evidently augmented. Therefore, it is a promising anion exchange membrane for alkaline fuel cell application.

Keywords: functionalized MXene; chitosan; anion exchange membrane; hydroxide conductivity; alkaline fuel cell

1. Introduction As an efficient, clean, and zero-emission power generation technology, fuel cells can effectively solve energy crises and environmental pollution issues as the focus of global governments. Compared with other types of fuel cells, the ones with alkaline anion exchange membranes have attracted wide attention because of: (1) Low working temperature (60–90 ◦C); (2) high rate of oxygen reduction reaction and catalytic stability, which can decrease the Pt amount or use non-noble metal catalysts such as Ag and Ni instead, thus significantly reducing the cost; (3) low fuel permeability (permeability of methanol: <10−6 m−2 s); and (4) facile water and heat management [1–3]. Alkaline anion exchange membranes are one of the essential components of alkaline fuel cells and are capable of blocking fuel and conducting OH−. Its performance, especially OH− conductivity, directly determines the open circuit voltage and energy output of fuel cells. However, currently available alkaline membranes all have low OH− conductivities owing to a low diffusion coefficient of OH− [4] (movement rate of OH− is 56% of that of H+) and poor channel connectivity inside. Thus, the commercialization of fuel cells has been seriously limited. Alkaline anion exchange membranes are mainly prepared by functionalized polymers, including polyphenyl ether [5], polybenzimidazole [6], and polyetheretherketone [7]. OH− is mainly conducted through conductible cation aggregation zones inside membranes, so the continuity of these zones plays a decisive role. Since traditional polymers have similar functionalized and non-functionalized areas, continuous ion conduction channels cannot form through self-assembly [8,9]. Recently, block polymers have been prepared to significantly increase the OH− conductivity of membranes through self-assembly-generated interconnected ion cluster channels due to entropy-driven microphase separation of hydrophilic and hydrophobic chain segments. Nevertheless, the preparation processes

Materials 2018, 11, 2335; doi:10.3390/ma11112335 www.mdpi.com/journal/materials Materials 2018, 11, 2335 2 of 11 Materials 2018, 11, x FOR PEER REVIEW 2 of 11 forpreparation block polymers processes are for complicated block polymers and the are distribution complicated of and groups the distribution cannot be accurately of groups controlled, cannot be soaccurately it is difficult controlled, and non-universal so it is difficult to generate and non- interconnecteduniversal to generate conduction interconnected channels depending conduction on functionalizedchannels depending polymers on func [10].tionalized polymers [10]. Comparatively,Comparatively, organic–inorganic organic–inorganic hybridization hybridization is is a a convenientconvenient andand effectiveeffective strategystrategy toto elevateelevate thethe ionion conductionconduction performanceperformance ofof membranes.membranes. OnOn oneone hand,hand, thethe surfacesurface ofof functionalizedfunctionalized inorganicinorganic nano-fillernano-filler allowsallows formationformation ofof continuouscontinuous ionion conductionconduction channels.channels. OnOn thethe otherother hand,hand, addingadding functionalizedfunctionalized nano-fille nano-fillerr can can connect connect non-conduction non-conduction dead dead zone zones,s, thereby thereby significantly significantly raising raising the theconductivity conductivity [11,12]. [11, 12In ].addition, In addition, addition addition of inor ofganic inorganic nano-filler nano-filler can markedly can markedly augment augment the thermal the thermalstability stabilityand mechanical and mechanical performance performance of membranes. of membranes. At present, At inorganic present, nano-fillers inorganic nano-fillers are mainly arestructurally mainly structurallyclassified into: classified (1) Zero-dimensional into: (1) Zero-dimensional nanoparticles nanoparticles (SiO2 [13], TiO (SiO2 [14],2 [13 etc.),], TiO (2)2 [one-14], etc.),dimensional (2) one-dimensional nanotubes (halloysit nanotubese (halloysitenanotubes nanotubes[15], carbon [15 ],nanotubes carbon nanotubes [16], etc.), [16 ],and etc.), (3) and two- (3) two-dimensionaldimensional nanolamellar nanolamellar materials materials (graphene (graphene oxide oxide [17], [17 ],hydrotalcite hydrotalcite [18], [18], montmorillonite montmorillonite [[19],19], MoSMoS22 [20],[20], etc.). etc.). Two-dimensional Two-dimensional nano nanolamellarlamellar materials, materials, which which have have high high specific specific surface surface areas areas that thatbenefit benefit the theconstruction construction of long-r of long-rangeange and and wide wide continuous continuous channels, channels, are are thus thus superior superior to otherother materialsmaterials [[21].21]. Nevertheless,Nevertheless, routineroutine two-dimensionaltwo-dimensional nanolamellarnanolamellar materialsmaterials havehave limitedlimited functionsfunctions owingowing toto low contents and uneven uneven distribution distribution of of su surfacerface functional functional groups. groups. Thus, Thus, it itis isimperative imperative to todevelop develop novel novel functionalized functionalized two-dimensional two-dimensional nanolamellar nanolamellar materials materials for for preparing hybrid anionanion exchangeexchange membranes.membranes. InIn thisthis study, we selected a a new new lamellar lamellar material material rich rich in in oxygen-containing oxygen-containing groups, groups, MXene, MXene, to tograft graft imidazolium imidazolium brushes brushes through through precipitation precipitation polymerization polymerization followed followed by quaternization.quaternization. Afterwards,Afterwards, wewe fabricatedfabricated aa hybridhybrid membranemembrane byby usingusing functionalizedfunctionalized MXeneMXene andand chitosanchitosan matrixmatrix withwith excellentexcellent film-forming film-forming ability. ability. Moreover, Moreover, we systematically we systematically studied studied the construction the construction of interfacial of conductioninterfacial channelsconduction and channels the mechanism and the for mechanism ion conduction for byion characterizing conduction theby microstructurecharacterizing andthe testingmicrostructure related properties. and testing related properties.

(a) (b) (c)

2.2. ExperimentalExperimental SectionSection

2.1.2.1. MaterialsMaterials

TitaniumTitanium hydridehydride (TiH(TiH22,, 99%,99%, 325-mesh),325-mesh), titaniumtitanium carbidecarbide (TiC,(TiC, 99%,99%, 325-mesh),325-mesh), aluminumaluminum powderspowders (Al, (Al, 99%, 99%, 325 325 mesh), mesh), and and 1-vinylimidazole 1-vinylimidazo werele were purchased purchased from Alfafrom Aesar Alfa (Shanghai,Aesar (Shanghai, China). HydroﬂuoricChina). Hydrofluoric acid (HF, acid 49%) (HF, was 49%) bought was from bought Aladdin from Reagent Aladdin Inc. Reagent (Shanghai, Inc. (Shanghai, China). Dimethyl China). sulfoxideDimethyl (DMSO,sulfoxide analytic (DMSO, grade), analytic glacial grade), acetic glacial acid acetic (analytic acid grade), (analytic and grade), potassium and hydroxidepotassium (analytichydroxide grade) (analytic were grade) obtained were from obtained Tianjin from Kermel Tianjin Chemical Kermel Chemical Reagent Co.,Reagent Ltd. Co., (Tianjin, Ltd. (Tianjin, China). CSChina). (degree CS of(degree deacetylation: of deacetylation: 95%) was 95%) provided was pr byovided Zhejiang by Golden-ShellZhejiang Golden-Shell Pharmaceutical Pharmaceutical Co., Ltd. (Yuhuan,Co., Ltd. Zhejiang,(Yuhuan, China).Zhejiang, China).

2.2.2.2. ApparatusApparatus AA TEM-100CXTEM-100CX transmissiontransmission electronelectron microscopemicroscope was purchased from JEOL ( (Boston,Boston, MA, USA). AnAn FTIRFTIR ThermoThermo NicoletNicolet IR200IR200 FourierFourier transformtransform infraredinfrared spectroscopespectroscope waswas boughtbought fromfrom ThermoThermo ScientiﬁcScientific (Waltham, MA, MA, USA). USA). A AD8 D8 Advance Advance ECO ECO X-ray X-ray diffractometer diffractometer was obtained was obtained from Bruker from Bruker(Karlsruhe, (Karlsruhe, Germany). Germany). A TGA-50 A TGA-50 thermal thermal gravimetric gravimetric analyzer analyzer was provided was provided by Shimadzu by Shimadzu (Kyoto, (Kyoto,Japan). An Japan). M350 An AX M350 all-purpos AX all-purposee tensile testing tensile machine testing machinewas purchased was purchased from Testometric from Testometric (London, UK). A CHI660B electrochemical workstation was bought from Zhengzhou Shiruisi Instrument

Materials 2018, 11, x FOR PEER REVIEW 3 of 11

Technology Co., Ltd. (Zhengzhou, China). A KJ-1600G muffle furnace was supplied by Zhengzhou Kejia Furnace Co., Ltd. (Zhengzhou, China).

2.3. Methods

(1) Preparation of MXene: First, TiH2, TiC, and Al were completely mixed in a molar ratio of 1:2:1.2. Then, the powders were added in an alundum crucible, compacted, and put into a tube furnace that was heated to 1450 °C at the speed of 10 °C/min and kept constant for 2 h, yielding a solid MAX phase. Finally, the MAX powders (325-mesh) were added into 49% HF solution and completely etched by continuous stirring at 60 °C for 72 h. Subsequently, the powders were washed to neutral by using deionized water and thoroughly dried into MXene lamellar material. (2) Preparation of functionalized MXene (Figure 1): Dried MXene, absolute ethanol, and ammonia water were added into a 250 mL round-bottomed flask, homogeneously distributed by ultrasonication for 4 h, and continuously stirred for 24 h. The mixture was ultrasonicated at 200 W Materials 2018, 11, 2335 3 of 11 for 30 min, combined with 1 mL of Modified polystyrene (MPS), continuously stirred for 32 h, centrifuged with absolute ethanol, washed, and dried into M-MXene. Afterwards, M-MXene (0.6 g) and(London, acetonitrile UK). A(80 CHI660B mL) were electrochemical added into a workstation100 mL one-necked was bought flask fromand dispersed Zhengzhou by Shiruisi6 h of ultrasonication.Instrument Technology After AIBN Co., (0.018 Ltd. (Zhengzhou,g) and 1-vinylimidazole China). A KJ-1600G(0.6 mL) were muffle completely furnace wasdispersed supplied by ultrasonicationby Zhengzhou Kejiain an Furnaceice bath Co.,for 20 Ltd. min, (Zhengzhou, the mixture China). was distilled at 82 °C for 30 min, centrifuged, washed, and dried into MXene-containing imidazole brushes on the surface (VI-MXene). Then VI- MXene2.3. Methods (0.6 g) and chloropropylamine (0.015 mmol) were added into absolute ethanol (150 mL), dispersed(1) Preparation by continuous of MXene: ultrasonication First, TiH for2, 4 TiC, h, and and refluxed Al were at completely 80 °C for 24 mixed h. The in resulting a molar product ratio of was1:2:1.2. centrifuged Then, the powderswith absolute were addedethanol, in washed, an alundum and crucible, dried into compacted, imidazolium and putbrush-functionalized into a tube furnace MXenethat was (QMXene-NH heated to 14502). ◦C at the speed of 10 ◦C/min and kept constant for 2 h, yielding a solid MAX phase.(3) Finally, Preparation the MAX of powdershybrid membrane (325-mesh): wereAppropriate added into amounts 49% HF of solution QMXene-NH and completely2 and CS etched were dissolvedby continuous in glacial stirring acetic at 60 acid◦C solution for 72 h. (2.5%, Subsequently, 60 mL), thecombined powders with were 25% washed glutaraldehyde to neutral bysolution using (0.2deionized mL), and water further and thoroughlystirred for 3 dried h. The into membrane MXene lamellar casting material.solution was thereafter slowly cast onto a levelled(2) Preparation glass plate of and functionalized dried at room MXene temperatur (Figuree1 until): Dried the MXene,solvent absolutecompletely ethanol, volatilized. and ammonia Finally, thewater resulting were added membrane into a 250was mLsubjected round-bottomed to ion exchange flask, homogeneouslyby soaking in 1 M distributed KOH solution by ultrasonication for 24 h and residualfor 4 h, and KOH continuously was removed stirred by soaking for 24 h. in The deionized mixture waswater. ultrasonicated The prepared at 200membrane W for 30 was min, referred combined to aswith CS/QMXene-NH 1 mL of Modified2-X, where polystyrene X is the (MPS), mass continuouslypercentage of stirredfiller material for 32 h,to centrifugedCS (X = 2.5, 5, with 7.5, absoluteand 10). ethanol,(4) The washed, OH− andconductivity dried into across M-MXene. the membrane Afterwards, was M-MXene measured (0.6 in g)a conductivity and acetonitrile cell (80by mL)the alternatingwere added current into a (AC) 100 mL impedance one-necked spectroscopy flask and dispersedmethod. The by membrane 6 h of ultrasonication. resistance was After tested AIBN by two-point(0.018 g) andprobe 1-vinylimidazole alternating current (0.6 impedance mL) were completelyspectroscopy dispersed with an oscillating by ultrasonication voltage of in 20 an mV ice overbath a for frequency 20 min, range the mixture of 106–10 was Hz distilled using an atelec 82trode◦C for system 30 min, connected centrifuged, with a washed,frequency and response dried analyzer.into MXene-containing Recently, a practical imidazole and brushesreproducible on the ex surface situ method (VI-MXene). for measuring Then VI-MXene the true value (0.6 g)of andthe hydroxidechloropropylamine conductivity (0.015 of mmol) AEMs were was added reported. into absolute This method ethanol (150offered mL), excellent dispersed reproducibility by continuous comparedultrasonication with the for methods 4 h, and refluxedcurrently at used 80 ◦ toC measure for 24 h. OH The− resultingconductivity. product We will was try centrifuged this method with of measuringabsolute ethanol, conductivity washed, in andthe following dried into work imidazolium [22]. brush-functionalized MXene (QMXene-NH2).

Figure 1. Schematic diagram for preparation process of QMXene-NH2..

(3) Preparation of hybrid membrane: Appropriate amounts of QMXene-NH2 and CS were dissolved in glacial acetic acid solution (2.5%, 60 mL), combined with 25% glutaraldehyde solution (0.2 mL), and further stirred for 3 h. The membrane casting solution was thereafter slowly cast onto a levelled glass plate and dried at room temperature until the solvent completely volatilized. Finally, the resulting membrane was subjected to ion exchange by soaking in 1 M KOH solution for 24 h and residual KOH was removed by soaking in deionized water. The prepared membrane was referred to as CS/QMXene-NH2-X, where X is the mass percentage of ﬁller material to CS (X = 2.5, 5, 7.5, and 10). (4) The OH− conductivity across the membrane was measured in a conductivity cell by the alternating current (AC) impedance spectroscopy method. The membrane resistance was tested by two-point probe alternating current impedance spectroscopy with an oscillating voltage of 20 mV over a frequency range of 106–10 Hz using an electrode system connected with a frequency response analyzer. Recently, a practical and reproducible ex situ method for measuring the true value of the hydroxide conductivity of AEMs was reported. This method offered excellent reproducibility Materials 2018, 11, 2335 4 of 11 compared with the methods currently used to measure OH− conductivity. We will try this method of measuring conductivity in the following work [22].

3. Results and Discussion

3.1. Experimental Results The addition of functionalized MXene to the chitosan membrane significantly enhanced the thermal stability and mechanical properties of the chitosan membrane. The hybrid membrane became more thermally stable due to a delay of degradation at the second stage (230–320 ◦C), and the tensile strength of the hybrid membrane was significantly raised to 29.2–41.0 MPa. In addition, the functionalized MXene with high specific surface area constructed a continuous ion transport channel inside the membrane, which made the OH− conductivity of the hybrid membrane significantly improved.

3.2. Characterizations of Functionalized MXene For organic–inorganic composite membranes, the material matching was very important, which determined the compatibility between the two phases and the continuity of ion conduction. In this study, the surface of MXene sheet was modified to enhance the interfacial compatibility between inorganic and organic polymers and the defect free composite membrane was prepared. This was a commonly used treatment method for hybrid membranes. The SEM image of MXene was shown in Figure2a. The results of SEM showed that the two phases were compatible. Clearly, MXene had a two-dimensional lamellar structure with high rigidity. After simple dispersion by ultrasonication, mono-lamellar MXene was obtained and its TEM image is exhibited in Figure2b. MXene maintained an intact structure after dispersion and had a light-colored surface. After modification, the color of QMXene-NH2 (Figure2c) deepened owing to imidazolium brushes. Since MXene was etched from Ti3AlC2, its surface contained abundant –OH as well as small amounts of –F and –O–. Therefore, we characterized the chemical structures of MXene before and after surface modification by FTIR spectroscopy (Figure2d). MXene exhibits vibration peaks of –OH at 3448 cm −1 and 1652 cm−1 and a stretching vibration peak of Ti-O-Ti at 500–750 cm−1 [23]. After functionalization, –OH and –O– on the surface of MXene participated in reaction, so the intensities of these peaks plummet. The characteristic peaks of imidazolium were overlapped by the strong ones of MXene. Regardless, −1 there was a stretching vibration peak of N-H (–NH2) in QMXene-NH2 at 3000–3500 cm [24], suggesting that the imidazolium group had been successfully introduced. The TGA curves of MXene and QMXene-NH2 are shown in Figure3a. As an inorganic ceramic material, MXene has high thermal stability and its thermal degradation can be divided into two main stages: (1) Weight loss of only 6% from room temperature to 200 ◦C, mainly originating from water molecules adsorbed on the surface and between lamellas of MXene; and (2) weight loss at above 200 ◦C, mainly owing to the degradation of –OH on the MXene surface [25]. Notably, the weight ◦ loss of QMXene-NH2 from 25 to 200 C exceeded that of MXene, primarily because more free water molecules were adsorbed by hydrophilic imidazolium brushes on the surface of the former, which was accelerated with rising temperature. After functionalization, –OH on the MXene surface was replaced by Si-O-Ti, thus enhancing the thermal stability and reducing the weight loss. Strasser et al. reported that imidazolium cations started to degrade at approximately 300 ◦C[26]. Accordingly, the thermal ◦ degradation of QMXene-NH2 was apparently accelerated at over 300 C. In addition, combined with TGA data, after weight loss the remaining mass was less than 45%, which was within its theoretical range as stated in the reported literature [27]. Therefore, there was no carbonization in the temperature range we tested. Materials 2018, 11, x FOR PEER REVIEW 4 of 11

3. Results and Discussion

3.1. Experimental Results The addition of functionalized MXene to the chitosan membrane significantly enhanced the thermal stability and mechanical properties of the chitosan membrane. The hybrid membrane became more thermally stable due to a delay of degradation at the second stage (230–320 °C), and the tensile strength of the hybrid membrane was significantly raised to 29.2–41.0 MPa. In addition, the functionalized MXene with high specific surface area constructed a continuous ion transport channel inside the membrane, which made the OH− conductivity of the hybrid membrane significantly improved.

3.2. Characterizations of Functionalized MXene For organic–inorganic composite membranes, the material matching was very important, which determined the compatibility between the two phases and the continuity of ion conduction. In this study, the surface of MXene sheet was modified to enhance the interfacial compatibility between inorganic and organic polymers and the defect free composite membrane was prepared. This was a commonly used treatment method for hybrid membranes. The SEM image of MXene was shown in Figure 2a. The results of SEM showed that the two phases were compatible. Clearly, MXene had a two-dimensional lamellar structure with high rigidity. After simple dispersion by ultrasonication, mono-lamellar MXene was obtained and its TEM image is exhibited in Figure 2b. MXene maintained an intact structure after dispersion and had a light-colored surface. After modification, the color of QMXene-NH2 (Figure 2c) deepened owing to imidazolium brushes. Since MXene was etched from Ti3AlC2, its surface contained abundant −OH as well as small amounts of −F and −O−. Therefore, we characterized the chemical structures of MXene before and after surface modification by FTIR spectroscopy (Figure 2d). MXene exhibits vibration peaks of −OH at 3448 cm−1 and 1652 cm−1 and a stretching vibration peak of Ti-O-Ti at 500–750 cm−1 [23]. After functionalization, −OH and −O− on the surface of MXene participated in reaction, so the intensities of these peaks plummet. The Materialscharacteristic 2018, 11 peaks, x FOR ofPEER imidazolium REVIEW were overlapped by the strong ones of MXene. Regardless, 5there of 11 Materialswas a stretching2018, 11, 2335 vibration peak of N-H (−NH2) in QMXene-NH2 at 3000–3500 cm−1 [24], suggesting5 of 11 that theThe imidazolium TGA curves groupof MXene had andbeen QMXene-NH successfully 2introduced. are shown in Figure 3a. As an inorganic ceramic material, MXene has high thermal stability and its thermal degradation can be divided into two main stages: (1) Weight loss of only 6% from room temperature to 200 °C, mainly originating from water molecules adsorbed on the surface and between lamellas of MXene; and (2) weight loss at above 200 °C, mainly owing to the degradation of −OH on the MXene surface [25]. Notably, the weight loss of QMXene-NH2 from 25 to 200 °C exceeded that of MXene, primarily because more free water molecules were adsorbed by hydrophilic imidazolium brushes on the surface of the former, which was accelerated with rising temperature. After functionalization, −OH on the MXene surface was replaced by Si-O-Ti, thus enhancing the thermal stability and reducing the weight loss. Strasser et al. reported that imidazolium cations started to degrade at approximately 300 °C [26]. Accordingly, the thermal degradation of QMXene-NH2 was apparently accelerated at over 300 °C. In addition, combined with TGA data, after weight loss the remaining（d mass） was less than 45%, which was within its theoretical range as stated in the reported literature [27]. Therefore, there was no carbonization in the temperature range we tested. The structure of MXene after surface modification was further characterized by XRD (Figure 3b). Since MXene was etched from Ti3AlC2, there are two characteristic peaks at 8.9° (002 crystal plane) and 38.9° (104 crystal plane) [28]. The interlamellar spacing of MXene was calculated as 0.99 nm by QMXene-NH2 948

Transmittance /a.u. Transmittance 1396 the Bragg equation (d = λ/2sinθ, λ = 1.54 Å). Surface modification of 2θ shifts1652 to 6.8° corresponded to an interlamellar spacing of 1.29 nm. Such increase can3448 mainly be ascribed to weakened hydrogen bonding between lamellas of MXene by ion brushes. In3500 addition, 3000 2500the introduction 2000 1500 1000of an imidazolium cation brush on the surface of the MXene still maintained an intrinsicWavenum belamellarr / cm-1 structure that might facilitateFigure the 2. 2. construction SEMSEM images images of of MXene ofa co MXenentinuous (a); TEM (a); ion TEMimages transport images (b,c) and (channelb, cFTIR) and spectra within FTIR ( spectradthe) of membrane MXene (d) of and MXene asQMXene- reported and in the literature. QMXene-NHNH2. 2.

100 （a）（b） o 96 6.8 MXene o o 300 C QMXene-NH2 200 C O 92 8.9

88 38.9o Weight / % Weight MXene QMXene-NH2 84 / a.u. intensity Relative

80 100 200 300 400 500 600 700 10 20 30 40 50 o Temperature / C 2θ / O

Figure 3. TGA (a) and XRD (b) curves of MXene andand QMXene-NHQMXene-NH22.

3.3. Microstructure,The structure ofPhysical, MXene and after Chemical surface Properties modification of Membrane was further characterized by XRD (Figure3b). ◦ Since MXene was etched from Ti3AlC2, there are two characteristic peaks at 8.9 (002 crystal plane) and 38.9A hybrid◦ (104 crystalalkaline plane) anion [ 28exchange]. The interlamellar membrane spacingwas fabricated of MXene by wasmixing calculated imidazolium as 0.99 brush- nm by thefunctionalized Bragg equation MXene (d =andλ/2sin CS matrixθ, λ = 1.54and Å).its Surfacecross-section modification is presented of 2θ inshifts Figure to 6.8 4. ◦Thecorresponded control CS tomembrane an interlamellar (Figure 4a) spacing had a of flat 1.29 and nm. smooth Such increasecross-section can mainly without be any ascribed defect. to Filling weakened CS matrix hydrogen with bondingQMXene-NH between2 generated lamellas ofmild MXene wrinkles, by ion primarily brushes. In because addition, nano the introductionlamellas interfered of an imidazolium with the cationconfiguration brush on of theCS chain surface segments. of the MXene Unlike still rod-like maintained or spherical an intrinsic materials, lamellar functionalized structure MXene that might did facilitatenot form thecurls construction or large particles of a continuous inside the membrane, ion transport so channel bulge was within barely the visible. membrane QMXene-NH as reported2 was in thecompatible literature. with CS matrix, without obvious interfacial defect. Because QMXene-NH2 was uniformly dispersed in CS matrix, it was considered that it might benefit the construction of continuous OH− 3.3.conduction Microstructure, channels. Physical, and Chemical Properties of Membrane A hybrid alkaline anion exchange membrane was fabricated by mixing imidazolium brush-functionalized MXene and CS matrix and its cross-section is presented in Figure4. The control CS membrane (Figure4a) had a flat and smooth cross-section without any defect. Filling CS matrix with QMXene-NH2 generated mild wrinkles, primarily because nano lamellas interfered with the configuration of CS chain segments. Unlike rod-like or spherical materials, functionalized MXene did

Materials 2018, 11, 2335 6 of 11

not form curls or large particles inside the membrane, so bulge was barely visible. QMXene-NH2 was compatible with CS matrix, without obvious interfacial defect. Because QMXene-NH2 was uniformly dispersed in CS matrix, it was considered that it might beneﬁt the construction of continuous OH− Materialsconduction 2018,, channels.11,, xx FORFOR PEERPEER REVIEWREVIEW 6 of 11

FigureFigure 4.4. SEMSEM imagesimages ofof cross-sectionscross-sections ofofof (((aa))) CSCSCS andandand (((bb))) CS/QMXene-NHCS/QMXene-NHCS/QMXene-NH22-7.5.-7.5.-7.5.

The FTIR spectra of control CS (Chitosan memb membrane)rane) and hybrid membranes are displayed in Figure 55.. TheThe control control CS CS membrane membrane exhibits exhibits characteristic characteristic peaks peaks of of amide amide band band I, band I, band II, andII, and band band III −1 −1 −1 IIIat 1633at 1633 cm cm, 1536−−11,, 15361536 cm cmcm,−− and11,, andand 1390 13901390 cm cmcmrespectively−−11 respectivelyrespectively [29 –[29–31].[29–31].31]. The TheThe peak peakpeak intensities intensitiesintensities of CS/QMXene-NH ofof CS/QMXene-CS/QMXene-2 NHall decrease,22 allall decrease,decrease, mainly mainly becausemainly because two-dimensionalbecause two-dimensionaltwo-dimensional lamellarmaterial lamellarlamellar interfered materialmaterialwith interferedinterfered the IR absorptionswithwith thethe IRIR of absorptionsgroups on CS of chain groups segments. on CS chain In addition, segments. the In interference addition, the was interference further enhanced was further with increasingenhanced with ﬁller −1 increasingincreasingamount, which fillerfiller amount,amount, gradually whichwhich reduced graduallygradually the absorptions. reducedreduced thethe The absorptions.absorptions. broad peaks TheThe of broadbroad all membranes peakspeaks ofof allall at membranesmembranes 3380 cm atcorrespond 3380 cm−− to11 correspond watercorrespond adsorbed toto waterwater inside adsorbed theadsorbed CS membrane insideinside thethe [32 ].CSCS Introducing membranemembrane two-dimensional [32].[32]. IntroducingIntroducing lamellar two-two- dimensionalmaterial interfered lamellar with material the conﬁguration interfered andwith decreased the configuration the motility and of decreased CS chain segments, the motility leading of CS to chainpeak intensitysegments, reduction leading to of peak the hybrid intensity membrane. reduction of the hybrid membrane.

3380 CS Transmittance Transmittance CS/QMXene-NHCS/QMXene-NH22-2.5-2.5 1536 1390 CS/QMXene-NHCS/QMXene-NH22-5-5 CS/QMXene-NHCS/QMXene-NH22-7.5-7.5 CS/QMXene-NHCS/QMXene-NH22-10-10 3500 3000 2500 2000 1500 1000 -1 Wavenumber / cm-1

FigureFigure 5.5. FTIRFTIR spectraspectra ofof controlcontrol CSCS andand CS/QMXene-NHCS/QMXene-NHCS/QMXene-NH22 membranes.membranes.membranes.

For anan ion ion exchange exchange membrane, membrane, the degreethe degree of crystallinity of crystallinity of polymer of chainpolymer segments chain significantly segments significantlyaffected ion conduction.affected ion Generally,conduction. polymer Generally, crystallization polymer crystallization hindered ion hindered conduction ion which, conduction when which,weakened, when promoted weakened, the conductionpromoted the process. conduction The XRD process. patterns The of controlXRD patterns CS and hybridof control membranes CS and ◦ hybridare shown membranes in Figure are6. Theshown crystalline in Figure peaks 6. The of the crystalline control CSpeaks membrane of the control are located CS membrane at 2 θ of 11.3 are ◦ ◦ locatedlocatedand 22.5 atat [2233θ ].ofof In 11.3°11.3° contrast, andand 22.5°22.5° the hybrid [33].[33]. InIn membrane contrast,contrast, thethe has hybridhybrid a characteristic membranemembrane peak hashas of aa QMXene-NHcharacteristiccharacteristic 2peakpeakat 6.8 ofof, accompanied by a decreased degree of crystallinity owing to the spatial interference of QMXene-NH . QMXene-NH22 atat 6.8°,6.8°, accompaniedaccompanied byby aa decreaseddecreased degrdegree of crystallinity owing to the spatial2 It was known that the configuration of CS chain segments affected the crystallinity of polymer chain interferenceinterference ofof QMXene-NHQMXene-NH22.. ItIt waswas knownknown thatthat thethe configurationconfiguration ofof CSCS chainchain segmentssegments affectedaffected thethe crystallinitysegments. However, of polymer the chain addition segments. of two-dimensional However, the lamellar addition materials of two-dimensional just interfered withlamellar the materialsconfiguration just interfered of CS chain with segments, the configuration which reduced of CS the chain crystallinity segments, of thewhich hybrid reduced membrane. the crystallinity So, as the amount of QMXene-NH was elevated, the configuration of CS chain segments was further disrupted, of the hybrid membrane.2 So, as the amount of QMXene-NH22 waswas elevated,elevated, thethe configurationconfiguration ofof CSCS chainlowering segments the degree was offurther crystallinity disrupted, of the lowering hybrid th membranee degree of as crystallinity a result. of the hybrid membrane as a result.

Materials 2018, 11, x FOR PEER REVIEW 7 of 11

o CS 6.8o 11.3 22.5o CS/QMXene-NH2-2.5 CS/QMXene-NH2-5 CS/QMXene-NH2-7.5 CS/QMXene-NH2-10

Materials 2018, 11, 2335 7 of 11 Materials 2018, 11, x FOR PEER REVIEW 7 of 11

o o CS

Relative intensity / a.u. 6.8 11.3 22.5o CS/QMXene-NH2-2.5 CS/QMXene-NH2-5 CS/QMXene-NH2-7.5 10 20 30 CS/QMXene-NH 40 50 2-10 2θ/ o

Figure 6. XRD patterns of control CS and CS/QMXene-NH2 membranes.

The addition of inorganic compounds to the CS introduced the inorganic particles, which Relative intensity / a.u. improved the organic network structure, enhanced mechanical properties, and improved thermal stability. The thermal stability of control CS and hybrid membranes was tested (Figure 7a). 10 20 30 40 50 Containing considerable −NH2 and −OH, CS is highly hydrophilic, so the moisture loss reached 8.3% o at the first stage (25–130 °C). The CS side chain de2gradedθ/ mainly at the second stage (230–320 °C), and the backboneFigureFigure did so 6.6. at XRDXRD the patternspatternsthird stage ofof controlcontrol (320–800 CSCS andand °C) CS/QMXene-NHCS/QMXene-NH [34]. Compared2 membranes. membranes.with control CS membrane, adding QMXene-NH2 reinforced the interfering effects of CS chain segments, which then reduced the hydrophilicityThe addition addition of the of of hybridinorganic inorganic membrane. comp compoundsounds Hence, to the to moisture theCS introduced CS introducedloss at ththee firstinorganic the stage inorganic particles,(25–130 particles,°C) which was lower.whichimproved With improved the a large organic the lamellarorganic network structure, network structure, QMXene-NH structure, enhanced enhanced2 stronglymechanical mechanicalinterfered properties, with properties, theand motility improved and of improved CS thermal chain segments.stability.thermal stability. TheAs athermal result, The thermal thestability hybrid stability of membranecontrol of control CS beca and CSme andhybrid more hybrid thermallymembranes membranes stable was was duetested tested to the(Figure (Figure delay 77a). a).of degradationContaining considerable at the second –NH−NH stage2 andand (230–320 − –OH,OH, CS CS°C). is is Withhighly highly rising hydrophilic, hydrophilic, filling amounts so so the the moisture moisture of QMXene-NH loss loss reached reached2, the 8.3% 8.3% ash ◦ ◦ contentat the firstfirst of the stage hybrid (25–130 membra °C).C). Thene gradually CS side chainincreased degradeddegraded at the mainly third stage at the (320–800 second °C). stage (230–320 °C),C), ◦ and the Additionally, backbone did the sotensile at the properti third stagees of (320–800control CS °C)C) and [34]. [34 hybrid]. Compared membranes with controlwere detected CS membrane, (Figure 7b).adding CS QMXene-NHwas QMXene-NH a semi-crystalline22 reinforcedreinforced polymer. the the interfering interfering Herein, effects th effectse control of ofCS CS chainCS chain membrane segments, segments, had which whicha tensile then then reduced strength reduced the of ◦ 27.5hydrophilicitythe hydrophilicityMPa, a breaking of the of elongationhybrid the hybrid membrane. of membrane. 5.5%, andHence, a Hence, Young’s the moisture the modulus moisture loss of at loss 920.2 the at firstMPa. the firststage After stage (25–130 the (25–130addition °C) was C)of QMXene-NHwaslower. lower. With With a2 large with a large lamellarhigh lamellar mechanical structure, structure, strength QMXene-NH QMXene-NH as an 2inorganic strongly2 strongly interfered lamellar interfered material, with the with motilitythe the tensile motility of CSstrength chain of CS ofsegments.chain the segments.hybrid As membrane a Asresult, a result, the was hybrid the significantly hybrid membrane membrane raised beca to became 29.2–41.0me more more MPa. thermally thermally Nevertheless, stable stable due since due to to thethe the inorganic delay delay of ◦ lamellardegradation material at the had second poor stage toughness, (230–320 the °C). C).breaki Withng risingelongation fillingfilling gradually amountsamounts ofofdropped QMXene-NHQMXene-NH with increasing2,, the ash ◦ fillingcontent amounts. of the hybrid membranemembrane gradually increased at thethe thirdthird stagestage (320–800(320–800 °C).C). Additionally, the tensile properties of control CS and hybrid membranes were detected (Figure

7b). CS was a semi-crystalline polymer. CS Herein, the control CS membrane had a tensile strength of 40（b） 27.5 MPa,90（ a breakinga） elongation of CS/QMXene-NH 5.5%, and 2a-2.5 Young’s modulus of 920.2 MPa. After the addition of CS/QMXene-NH2-5 CS/QMXene-NH2-7.5 230 oC QMXene-NH2 with high mechanical CS/QMXene-NH strength2 -7.5as an inorganic lamellar material, the tensile strength 80 CS/QMXene-NH2-5 CS/QMXene-NH2-2.5 of the hybrid membrane was significantly CS/QMXene-NH raised2-10 to 29.2–41.030 MPa. Nevertheless, since the inorganic lamellar70 material had poor toughness, the breaking elongation gradually dropped with increasing CS filling amounts.60 20 CS/QMXene-NH2-10 Weight /% Weight 50 MPa Stress / o CS 320 C 1040（b） 4090（a） CS/QMXene-NH2-2.5 CS/QMXene-NH2-5 CS/QMXene-NH2-7.5 230 oC CS/QMXene-NH2-7.5 3080 CS/QMXene-NH2-5 CS/QMXene-NH2-2.5 100 200 300 400 500 600 700 0 CS/QMXene-NH2-10 300.00 0.01 0.02 0.03 0.04 0.05 Temperature / oC 70 Strain / % CS 60 20 FigureFigure 7.7. ((aa)) TGATGA curvescurves andand ((b)) stress-strain curves of control CS and CS/QMXene-NH22 membranes.membranes. CS/QMXene-NH2-10 Weight /% Weight 50 MPa Stress / 3.4. IonAdditionally, Exchange Capacity the tensile320 (IEC), oC properties Water Uptake of control Ratio, CS and and10 Swelling hybrid membranesRatio were detected (Figure7b). CS was40 a semi-crystalline polymer. Herein, the control CS membrane had a tensile strength of The IEC of ion exchange membranes can reflect the number of exchangeable OH− inside, 27.5 MPa,30 a breaking elongation of 5.5%, and a Young’s modulus of 920.2 MPa. After the addition of determining100 the OH 200− 300conductivity 400 500 of 600 alkaline 700 membranes0 to a certain extent. The IEC values of 0.00 0.01 0.02 0.03 0.04 0.05 QMXene-NH2 with high mechanicalo strength as an inorganic lamellar material, the tensile strength control CS and hybrid Temperaturemembranes / C are shown in Figure 8a. According to functionalized QMXene- of the hybrid membrane was signiﬁcantly raised to 29.2–41.0 MPa. Nevertheless,Strain / % since the inorganic

lamellarFigure material 7. (a) TGA had curves poor and toughness, (b) stress-strain the breaking curves of elongation control CS and gradually CS/QMXene-NH dropped2 membranes. with increasing ﬁlling amounts. 3.4. Ion Exchange Capacity (IEC), Water Uptake Ratio, and Swelling Ratio

The IEC of ion exchange membranes can reflect the number of exchangeable OH− inside, determining the OH− conductivity of alkaline membranes to a certain extent. The IEC values of control CS and hybrid membranes are shown in Figure 8a. According to functionalized QMXene-

Materials 2018, 11, x FOR PEER REVIEW 8 of 11

NH2 content in CS (CS/QMXene-NH2-2.5, CS/QMXene-NH2-5, CS/QMXene-NH2-7.5, and CS/QMXene-NH2-10), the IEC of the four membranes were 0.43 mmol g−1, 0.46 mmol g−1, 0.49 mmol gMaterials−1, and 20180.51, 11mmol, 2335 g−1 respectively. Rich in −NH2, CS undergoes protonation in the presence of water,8 of 11 thereby producing OH−. The control CS membrane had an IEC of 0.40 mmol g−1, close to that reported by Wan et al. [35]. Given that QMXene-NH2 had abundant imidazolium groups on the surface, filling 3.4. Ion Exchange Capacity (IEC), Water Uptake Ratio, and Swelling Ratio it into CS matrix markedly increased IEC of the hybrid membrane. For instance, IEC reached as high − as 0.49The mmol IEC g of−1 at ion the exchange QMXene-NH membranes2 filling amount can reflect of 7.5%. the number of exchangeable OH inside, − determiningMeanwhile, the OH the conductivitywater uptake of ratios alkaline of membranes control CS to and a certain hybrid extent. membranes The IEC valuesat 20 of°C control were measuredCS and hybrid (Figure membranes 8b). Containing are shown large in Figureamounts8a. of According hydrophilic to functionalized −OH and −NH QMXene-NH2, CS chain segments2 content providedin CS (CS/QMXene-NH a high water uptake2-2.5, CS/QMXene-NH ratio of 115.2% fo2-5,r the CS/QMXene-NH control membrane.2-7.5, Moreover, and CS/QMXene-NH water molecules2-10), −1 −1 −1 −1 exertedthe IEC ofstrong the four solvation membranes effects were on 0.43highly mmol hydrop g , 0.46hilic mmol CS chain g , 0.49segments, mmol g resulting, and 0.51 in mmolapparent g swellingrespectively. of the Rich membrane. in –NH2, CSAt undergoes20 °C, the protonationswelling ratio in of the the presence control of CS water, membrane thereby was producing 31.3%. − −1 QMXene-NHOH . The control2 blocked CS membrane the motility had of an CS IEC chain of 0.40segments mmol through g , close evident to that spatial reported interference, by Wan et al.so [the35]. waterGiven uptake that QMXene-NH ratio and swelling2 had abundant ratio of imidazoliumthe hybrid membrane groups on both the surface, dropped. filling For itexample, into CS matrixwhen −1 7.5%markedly QMXene-NH increased2 IECwas offilled, the hybridthe two membrane. values reduced For instance, to 85.4% IEC and reached 18.4% asrespectively. high as 0.49 The mmol water g permeationat the QMXene-NH was much2 filling higher amount than that of 7.5%. of the most similar membranes as reported [36].

132 36 Membrane Water uptake 0.5（a） 126 （b） Swelling ratio 30 120 0.4

） 114 24 -1 0.3 108 ratio % ratio

18 g mmol g mmol 102 （ 0.2 12 Swellin

Water uptake % 96

IEC IEC

0.1 90 6 84 0.0 0 S .5 C 2-2 2-5 -7.5 2-10 CS -2.5 2-5 7.5 10 e-NH e-NH -NH2 -NH -NH2 e-NH NH2- NH2- Xen Xen ene Xene ene Xen ene- ene- /QM /QM QMX QM QMX /QM MX MX CS CS CS/ CS/ CS/ CS CS/Q CS/Q FigureFigure 8. 8. FillerFiller content-dependent content-dependent IEC IEC values values ( (aa),), water water uptake uptake ratios ratios and and swelling swelling ratios ratios ( (bb)) of of control control

CSCS and CS/ CS/QMXene-NH CS/QMXene-NH2 membranes.2 membranes.

◦ 3.5. HydroxideMeanwhile, Conductivity the water of uptake Membrane ratios of control CS and hybrid membranes at 20 C were measured (Figure8b). Containing large amounts of hydrophilic –OH and –NH 2, CS chain segments provided − a highOH water conductivity uptake ratio was of the 115.2% most for important the control index membrane. for evaluating Moreover, the waterperformance molecules of alkaline exerted membranes,strong solvation determining effects on the highly practical hydrophilic applicability CS chain in preparing segments, alkaline resulting fuel in apparent cells. The swelling conductivity of the of the control CS ◦membrane at 20 °C and 100% RH reached 0.00146 S cm−1 as shown in Figure 9a, membrane. At 20 C, the swelling ratio of the control CS membrane was 31.3%. QMXene-NH2 blocked beingthe motility comparable of CS chainto that segments reported through before evident[37]. Addition spatial interference,of QMXene-NH so the2 constructed water uptake continuous ratio and conduction channels (imidazolium brushes and interfacial water channels) inside the CS membrane, swelling ratio of the hybrid membrane both dropped. For example, when 7.5% QMXene-NH2 was − whichfilled, theincreased two values significantly reduced tothe 85.4% OH andconductivity 18.4% respectively. of the hybrid The membrane. water permeation For instance, was much when higher the fillingthan that amount of the was most 2.5%, similar 5%, membranesand 7.5%, the as conductivity reported [36]. of CS/QMXene-NH2 at 100% RH was raised with addition of QMXene-NH2 as shown in Figure 9a. The conductivity of CS/QMXene-NH2 with addition3.5. Hydroxide of 7.5% Conductivity QMXene-NH of Membrane2 at 100% RH was raised 172% to 0.00398 S cm−1 compared with CS (0.00146 S cm−1 Figure 9a). The addition of 7.5%QMXene-NH2 was enough for alkaline fuel cell OH− conductivity was the most important index for evaluating the performance of alkaline application, because further elevating the content of QMXene-NH2 (10%) induced lamellar membranes, determining the practical applicability in preparing alkaline fuel cells. The conductivity agglomeration and undermined the continuity of interfacial channels, eventually decreasing the OH− of the control CS membrane at 20 ◦C and 100% RH reached 0.00146 S cm−1 as shown in Figure9a, conductivity. With rising temperature, the OH− conductivity gradually increased, mainly because: (1) being comparable to that reported before [37]. Addition of QMXene-NH constructed continuous Such conduction was a thermal control process so the motility of OH2− was promoted as the conduction channels (imidazolium brushes and interfacial water channels) inside the CS membrane, temperature rose; (2) water molecules were more capable of transport, diffusion, and migration due which increased significantly the OH− conductivity of the hybrid membrane. For instance, when to facilitated motility by temperature elevation; and (3) enhancement of the motility of CS chain the filling amount was 2.5%, 5%, and 7.5%, the conductivity of CS/QMXene-NH at 100% RH was segments decreased the energy barrier of OH− conduction [38]. The Arrhenius plots2 of control CS and raised with addition of QMXene-NH2 as shown in Figure9a. The conductivity− of CS/QMXene-NH 2 hybrid membranes are presented in Figure 9b. The activation energy of OH conduction −of1 the control with addition of 7.5% QMXene-NH2 at 100%−1 RH was raised 172% to 0.00398 S cm compared CS membrane at 100%− RH1 was 21.9 kJ mol and the conduction energy barrier plummeted after with CS (0.00146 S cm Figure9a). The addition of 7.5%QMXene-NH 2 was enough for alkaline fuel cell application, because further elevating the content of QMXene-NH2 (10%) induced lamellar agglomeration and undermined the continuity of interfacial channels, eventually decreasing the OH− conductivity. With rising temperature, the OH− conductivity gradually increased, mainly because: Materials 2018, 11, 2335 9 of 11

(1) Such conduction was a thermal control process so the motility of OH− was promoted as the temperature rose; (2) water molecules were more capable of transport, diffusion, and migration due to facilitated motility by temperature elevation; and (3) enhancement of the motility of CS chain segments decreased the energy barrier of OH− conduction [38]. The Arrhenius plots of control CS and hybrid membranes are presented in Figure9b. The activation energy of OH − conduction of the control CS Materials 2018, 11, x FOR PEER REVIEW 9 of 11 membrane at 100% RH was 21.9 kJ mol−1 and the conduction energy barrier plummeted after addition − −1 ofaddition QMXene. of For QMXene. example, For the example, OH conductivity the OH− conductivity of the hybrid of the membrane hybrid membrane dropped to dropped 16.0 kJ mol to 16.0at kJ themol ﬁlling−1 at amountthe filling of amount 7.5%. of 7.5%.

-4.0 0.014 （a）（b） -1 CS -4.5 0.012 CS/QMXene-NH2-2.5 CS/QMXene-NH2-5 0.010 -5.0 )

CS/QMXene-NH2-7.5 1 - CS/QMXene-NH2-10 0.008 -1 CS Ea=21.9 kJ mol

/S cm /S -5.5

σ -1 0.006 CS/QMXene-NH2-2.5 Ea=17.1 kJ mol ln( -1 -6.0 CS/QMXene-NH2-2.5 Ea=16.5 kJ mol 0.004 -1 CS/QMXene-NH2-2.5 Ea=16.0 kJ mol 0.002 -1 -6.5 CS/QMXene-NH2-2.5 Ea=16.7 kJ mol Hydroxiede conductivity / S cm conductivity Hydroxiede 0.000 2.7 2.8 2.9 3.0 3.1 3.2 3.3 3.4 20 30 40 50 60 70 80 90 Temperature / oC 1000/T(K)

Figure 9. (a) Temperature-dependent OH− conductivity and (b) Arrhenius plots of OH− conductivity Figure 9. (a) Temperature-dependent OH− conductivity and (b) Arrhenius plots of OH− conductivity of control CS and CS/QMXene-NH2 membranes. of control CS and CS/QMXene-NH2 membranes. 4. Conclusions 4. Conclusions (1) QMXene-NH2 and CS had high interfacial compatibility and the former could be homogeneously (1) dispersedQMXene-NH in CS2 matrix.and CS had high interfacial compatibility and the former could be homogeneously dispersed in CS− matrix. (2) By constructing continuous OH conduction channels inside the membrane, QMXene-NH2 − (2) increasedBy constructing the number continuous of carrier OH sites conduction allowing ion channels jumping andinside conduction. the membrane, QMXene-NH2 increased the number of carrier sites allowing ion jumping− and conduction. (3) Adding QMXene-NH2 significantly augmented the OH conductivity of the hybrid membrane. 2 − (3) AtAdding 20 ◦C, suchQMXene-NH conductivity significantly of the hybrid augmented membrane the OH (3.98 conductivity mS cm−1) was of the 2.72-fold hybrid that membrane. of the −1 controlAt 20 CS°C, membrane.such conductivity of the hybrid membrane (3.98 mS cm ) was 2.72-fold that of the control CS membrane. (4) Adding QMXene-NH2 boosted both the thermal and mechanical stabilities and reduced the (4) Adding QMXene-NH2 boosted both the thermal and mechanical stabilities and reduced the membrane swelling ratio. membrane swelling ratio. (5) The addition of QMXene-NH2 significantly increased the ion exchange capacity of the hybrid (5) The addition of QMXene-NH2 significantly increased the ion exchange capacity of the hybrid membrane and the water uptake ratio and swelling ratio of the hybrid membrane both dropped. membrane and the water uptake ratio and swelling ratio of the hybrid membrane both dropped. For example, when 7.5% QMXene-NH2 was filled, the two values reduced to 85.4% and For example, when 7.5% QMXene-NH2 was filled, the two values reduced to 85.4% and 18.4% 18.4% respectively. respectively. InIn summary, summary, we we herein herein prepared prepared a novela novel functionalized functionalized nanolamellar nanolamellar material material that that effectively effectively enhancedenhanced the the overall overall performance of of CS CS membranes. membranes. Hence, Hence, it is a it feasibly is a feasibly applicable applicable anion exchange anion exchangemembrane. membrane. Author Contributions: L.W. participated in the design of this study, and carried out the data acquisition, the Author Contributions: L.W. participated in the design of this study, and carried out the data acquisition, the statistical analysis, manuscript preparation and revision. B.S. carried out the statistical analysis and manuscript statistical analysis, manuscript preparation and revision. B.S. carried out the statistical analysis and manuscript revision.revision. All All authors authors read read and and approved approved the the final final manuscript. manuscript. Funding:Funding:This This research research was was funded funded by National by National Natural Natural Science Science Foundation Foundation of China withof China grant with number grant 21506170. number 21506170. Conflicts of Interest: The authors declared that they had no conflicts of interest to this work. Conflicts of Interest: The authors declared that they had no conflicts of interest to this work.

References

Materials 2018, 11, 2335 10 of 11

References

1. Zhou, T.; Shao, R.; Chen, S.; He, X.; Qiao, J.; Zhang, J. A review of radiation-grafted polymer electrolyte membranes for alkaline polymer electrolyte membrane fuel cells. J. Pow. Sources 2015, 293, 946–975. [CrossRef] 2. Liao, X.; Ren, L.; Chen, D.; Liu, X.; Zhang, H. Nanocomposite membranes based on quaternized polysulfone and functionalized montmorillonite for anion-exchange membranes. J. Pow. Sources 2015, 286, 258–263. [CrossRef] 3. Yang, G.; Hao, J.; Cheng, J.; Zhang, N.; He, G.; Zhang, F.; Hao, C. Hydroxide ion transfer in anion exchange membrane: A density functional theory study. Int. J. Hydrog. Energy 2016, 41, 6877–6884. [CrossRef] 4. Lin, C.X.; Zhuo, Y.Z.; Lai, A.N.; Zhang, Q.G.; Zhu, A.M.; Ye, M.L.; Liu, L.Q. Side-chain-type anion exchange membranes bearing pendent imidazolium-functionalized poly(phenylene oxide) for fuel cells. J. Membr. Sci. 2016, 513, 206–216. [CrossRef] 5. Zhu, L.; Zimudi, T.J.; Li, N.; Pan, J.; Lin, B.; Hickner, M.A. Crosslinking of comb-shaped polymer anion exchange membranes via thiol–ene click chemistry. Polym. Chem. 2016, 7, 2464–2475. [CrossRef] 6. Zeng, L.; Zhao, T.S.; An, L.; Zhao, T.G.; Zhao, G. Physicochemical properties of alkaline doped polybenzimidazole membranes for anion exchange membrane fuel cells. J. Membr. Sci. 2015, 493, 340–348. [CrossRef] 7. Yan, X.; Gu, S.; He, G.; Wu, X. Imidazolium-functionalized poly(ether ether ketone) as membrane and electrode ionomer for low-temperature alkaline membrane direct methanol fuel cell. J. Power Sources 2014, 250, 90–97. [CrossRef] 8. Li, N.; Leng, Y.; Hickner, M.A.; Wang, C.Y. Highly stable, anion conductive, comb-shaped copolymers for alkaline fuel cells. J. Am. Chem. Soc. 2013, 135, 10124–10133. [CrossRef][PubMed] 9. Varcoe, J.R.; Atanassov, P.; Dekel, D.R.; Herring, A.M.; Hickner, M.A.; Kohl, P.A.; Kuckernak, A.R.; Mustain, W.E.; Nijmeijrer, K.; Scott, K. Anion-exchange membranes in electrochemical energy systems. Energy Environ. Sci. 2014, 7, 3135–3191. [CrossRef] 10. Price, S.C.; Ren, X.; Jackson, A.C.; Yes, Y.; Elabd, Y.A.; Beyer, F.L. Bicontinuous alkaline fuel cell membranes from strongly self-segregating block copolymers. Macromolecules 2013, 46, 7332–7340. [CrossRef] 11. Shi, B.; Li, Y.; Zhang, H.; Wu, W.; Ding, R.; Dang, J.; Wang, J. Tuning the performance of anion exchange membranes by embedding multifunctional nanotubes into a polymer matrix. J. Membr. Sci. 2016, 498, 242–253. [CrossRef] 12. Chen, H.; Wang, J.; Bai, H.; Sun, J.; Li, Y.; Liu, Y.; Wang, J. Nanohybrid membranes with hydroxide ion transport highways constructed from imidazoliumfunctionalized graphene oxide. RSC Adv. 2015, 5, 88736–88747. [CrossRef] 13. Wu, Y.; Wu, C.; Yu, F.; Xu, T.; Fu, Y. Free-standing anion-exchange PEO–SiO2 hybrid membranes. J. Membr. Sci. 2008, 307, 28–36. [CrossRef] 14. Yang, C.C. Synthesis and characterization of the cross-linked PVA/TiO2 composite polymer membrane for alkaline DMFC. J. Membr. Sci. 2007, 288, 51–60. [CrossRef] 15. Zhang, H.; Shi, B.; Ding, R.; Chen, H.; Wang, Y.; Liu, Y. Composite anion exchange membrane from quaternized polymer spheres with tunable and enhanced hydroxide conduction property. Ind. Eng. Chem. Res. 2016, 55, 9064–9076. [CrossRef] 16. Li, Q.; Liu, L.; Liang, S.; Dong, Q.; Jin, B.; Bai, R. Preparation and characterization of composite membranes with ionic liquid polymer-functionalized multiwalled carbon nanotubes for alkaline fuel cells. RSC Adv. 2013, 3, 13477–13485. [CrossRef] 17. Wan, C.J.; Zhu, L.Q.; Liu, Y.H.; Feu, P. Proton-conducting graphene oxide-coupled neuron transistors for brain-inspired cognitive systems. Adv. Mater. 2016, 28, 3557–3563. [CrossRef][PubMed] 18. Di Vona, M.; Casciola, M.; Donnadio, A.; Noccheti, M.; Pasquini, L.; Narducci, R.; Knauth, P. Anionic conducting composite membranes based on aromatic polymer and layered double hydroxides. Int. J. Hydrog. Energy 2017, 42, 3197–3205. [CrossRef] 19. Liao, G.; Yang, C.; Hu, C.; Pai, Y.L.; Lue, J.S. Novel quaternized polyvinyl alcohol/quaternized chitosan nano-composite as an effective hydroxide-conducting electrolyte. J. Membr. Sci. 2015, 485, 17–29. [CrossRef]

20. Feng, K.; Tang, B.B.; Wu, P.Y. Selective growth of MoS2 for proton exchange membranes with extremely high selectivity. ACS Appl. Mater. Interfaces 2013, 5, 13042–13049. [CrossRef][PubMed] Materials 2018, 11, 2335 11 of 11

21. Zhao, L.; Li, Y.; Zhang, H.; Wu, W. Constructing proton-conductive highways within an ionomer membrane by embedding sulfonated polymer brush modified graphene oxide. J. Power Sources 2015, 286, 445–457. [CrossRef] 22. Noga, Z.; Dekel, D.R. A practical method for measuring the true hydroxide conductivity of anion exchange membranes. Electrochem. Commun. 2018, 88, 109–113. 23. Zhang, J.; Liu, Y.; Lv, Z.; Zhao, T.; Li, P.; Sun, Y.; Wang, Y. Sulfonated Ti3C2Tx to construct proton transfer pathways in polymer electrolyte membrane for enhanced conduction. Sol. State Ion. 2017, 310, 100–111. [CrossRef] 24. Wan, M.; Li, W.; Long, Y.; Tu, Y. Electrochemical determination of tryptophan based on Si-doped nano-TiO2 modified glassy carbon electrode. Anal. Methods 2012, 4, 2860–2865. [CrossRef] 25. Alekseev, S.; Korytko, D.; Iazykov, M.; Khainakov, S.; Lysenko, V. Electrochemical Synthesis of Carbon Fluorooxide Nanoparticles from 3C-SiC Substrates. J. Phys. Chem. C 2015, 119, 20503–20514. [CrossRef] 26. Mashtalir, O.; Naguib, M.; Dyatkin, B.; Gogotsi, Y.; Barsoum, M.W. Kinetics of aluminum extraction from Ti3AlC2 inhydrofluoric acid. Mater. Chem. Phys. 2013, 139, 147–152. [CrossRef] 27. Yahua, L.; Zhang, J.; Zhang, X.; Li, Y.; Wang, Z. Ti3C2Tx Filler Effect on the Proton Conduction Property of Polymer Electrolyte Membrane. ACS Appl. Mater. Interfaces 2016, 8, 20352–20363. 28. Strasser, D.J.; Graziano, B.J.; Knauss, D.M. Base stable poly (diallylpiperidinium hydroxide) multiblock copolymers for anion exchange membranes. J. Mater. Chem. A 2017, 5, 9627–9640. [CrossRef]

29. Wu, X.; Hao, L.; Zhang, J.; Zhang, X.; Wang, Y.; Liu, J. Polymer–Ti3C2Tx composite membranes to overcome the trade–off in solvent resistant nanoﬁltration for alcohol-based system. J. Membr. Sci. 2016, 515, 175–188. [CrossRef] 30. Kanti, P.; Srigowri, K.; Madhuri, J.; Smitha, B. Dehydration of ethanol through blend membranes of chitosan and sodium alginate by pervaporation. Sep. Purif. Technol. 2004, 40, 259–266. [CrossRef]

31. Behera, S.S.; Das, U.; Kumar, A.; Bissoyi, A.; Singh, A.K. Chitosan/TiO2 composite membrane improves proliferation and survival of L929 ﬁbroblast cells: Application in wound dressing and skin regeneration. Int. J. Biol. Macromol. 2017, 98, 329–340. [CrossRef][PubMed] 32. Ma, J.; Sahai, Y. Chitosan biopolymer for fuel cell applications. Carbohydrate Polym. 2013, 92, 955–975. [CrossRef][PubMed] 33. Samuels, R.J. Solid state characterization of the structure of chitosan ﬁlm. J. Polym. Sci. Part B Polym. Phys. 1981, 19, 1081–1105. [CrossRef] 34. Bai, H.; Zhang, H.; He, Y.; Liu, Y.; Zhang, B.; Wang, Y. Enhanced proton conduction of chitosan membrane enabled by halloysite nanotubes bearing sulfonate polyelectrolyte brushes. J. Membr. Sci. 2014, 454, 220–232. [CrossRef] 35. Wan, Y.; Peppley, B.; Creber, K.A.M.; Bui, T.W.; Halliop, E. Preliminary evaluation of an alkaline chitosan-based membrane fuel cell. J. Pow. Sources 2006, 162, 105–113. [CrossRef] 36. Ding, L.; Wei, Y.; Wang, Y.; Chen, H.; Caro, J.; Wang, H. A Two-Dimensional Lamellar Membrane: MXene Nanosheet Stacks. Angew. Chem. Int. Ed. Engl. 2017, 56, 1825–1829. [CrossRef][PubMed] 37. Wan, Y.; Creber, K.A.M.; Peppley, B.; Bui, T.W. Ionic conductivity and related properties of crosslinked chitosan membranes. J. Appl. Polym. Sci. 2003, 89, 306–317. [CrossRef] 38. Chen, C.; Tse, Y.L.S.; Lindberg, G.E.; Knight, C.; Voth, G.A. Hydroxide solvation and transport in anion exchange membranes. J. Am. Chem. Soc. 2016, 138, 991–1000. [CrossRef][PubMed]

© 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).