Direct Assessment of Nanoconfined Water in 2D Ti3C2 Electrode

Subscriber access provided by University of Sussex Library

Article

Direct assessment of nano-confined water in 2D Ti3C2 (MXene) electrode interspaces by a surface acoustic technique Netanel Shpigel, Mikhael D. Levi, Sergey Sigalov, Tyler S. Mathis, Yury Gogotsi, and Doron Aurbach J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.8b04862 • Publication Date (Web): 21 Jun 2018 Downloaded from http://pubs.acs.org on June 21, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 15 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

Direct assessment of nano-confined water in 2D Ti3C2 (MXene) electrode interspaces by a surface acoustic technique Netanel Shpigel1, Mikhael D. Levi1*, Sergey Sigalov1, Tyler S. Mathis2, Yury Gogotsi2, Doron Aurbach1* 1 Department of Chemistry and BINA – BIU center for Nanotechnology and Advanced materials, BarIlan University, Ramat-Gan 5290002, Israel 2 Department of Materials Science and Engineering, and A.J. Drexel Nanomaterials Institute, Drexel University, Philadelphia, PA 19104, USA

Keywords: EQCM, QCM-D, MXenes, Super Capacitors, 2D materieals, Cosmotropic and Chaotropic Ions, Intercalation, Nanoconfined Fluid, Energy Storage ABSTRACT: Although significant progress has been achieved in understanding of ion-exchange mechanisms in the new family of 2D transition metal carbides and nitrides known as MXenes, direct gravimetric assessment of water insertion into the MXene interlayer spaces and mesopores has not been reported so far. Concurrently, the latest research on MXene and Birnessite electrodes shows that nanoconfined water dramatically improves their gravimetric capacity and rate capability. Hence quantification of the amount of confined water in solvated electrodes is becoming an important goal of energy-related research. Using the recently developed and highly sensitive method of in situ hydrodynamic spectroscopy (based on surface-acoustic probing of solvated interfaces) we provide clear evidence that typical cosmotropic cations (Li+, Mg2+ and Al3+) are inserted into the MXene interspaces in their partially hydrated form, in contrast to the insertion of chaotropic cations (Cs+ and TEA+) which effectively dehydrate the MXene. These new findings provide important information about the charge storage mechanisms in layered materials by direct quantification and efficient control (management) over the amount of confined fluid in a variety of solvated battery/supercapacitor electrodes. We believe that the proposed monitoring of water content as a function of the nature of ions can be equally applied to solvated biointerfaces, such as the ion channels of membrane proteins.

INTRODUCTION Two-dimensional (2D) solids containing nanoconfined fluids (water, organic molecules, partially solvated ions) have been recently recognized as promising electrode materials for improved electrochemical energy storage [1,2,3]. 2D transition metal carbides and nitrides (MXenes) are especially attractive in this respect: MXenes possess high electronic conductivity which enables fast, reversible insertion of different cations in organic, aqueous, and acidic environments [4,5,6]. Previous investigations into aqueous based electrolytes revealed that insertion of hydrophilic and hydrophobic cations is accompanied by simultaneous insertion of water molecules [7,8], however, the charge storage mechanism of these materials is still under investigation: Huet et al. examined the MXene Ti3C2 in acidic and neutral electrolytes by in situ electrochemical Raman spectroscopy [9]. Their findings demonstrated a pseudo-capacitive mechanism in the acidic electrolyte and pure adsorption response in neutral electrolyte solutions. The existence of a Faradaic response in acidic medium was also proven by in situ electrochemical X-ray absorption spectroscopy of Ti3C2 [10].

1 ACS Paragon Plus Environment

Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 15

Further studies have revealed that electrochemical or spontaneous intercalation of various cations into Ti3C2 layers in aqueous electrolyte solutions results in the insertion of additional water molecules in accordance with the solvation ability of the respective cation. X-ray diffraction (XRD) [11] and neutron scattering [12] investigations of the spontaneous intercalation of ions resulted in further understanding of the effects that ion intercalation has on Ti3C2, namely the extent of Ti3C2 expansion was correlated to the amount of co-inserted water. A recently reported calorimetric study on the spontaneous intercalation of alkali metal ions (Li+, Na+, K+) into free standing Ti3C2 electrodes suggests there is preferential intercalation of K+ cations due to its smaller hydration radius (compared to that of Li+ and Na+) [13]. Yet, these studies addressed only the non-electrochemical insertion of ions and hence future mechanistic studies of Ti3C2 should be extended to include studying the charging mechanisms of this energy storage material. Recognizing the fact that kinetics of ion transport in 2D layered materials (possessing hierarchical porous structure) strongly depend on the incorporation of nanoconfined solvent [1], which, in turn, is affected by the interlayer ion solvation/desolvation, quantification and efficient control (management) over the amount of confined solvent in 2D layered electrodes become challenging for increasing their energy, power and cycling performance. Indeed, the latest research shows that such solvated electrodes as MXene [14] and Birnessite [15] significantly improve their gravimetric capacity and rate capability with the amount of nanoconfined water contained in their structures. The benefits of electrochemical quartz crystal microbalance with dissipation monitoring (EQCM-D) for in-situ monitoring of intercalation-induced changes in Ti3C2 electrodes have been recently demonstrated using acoustically-thick film electrodes. [16] The assessment and continuous monitoring of the amount of nanoconfined solvent in 2D layered electrodes require application of an alternative EQCM-D approach to a simulataneous monitoring of gravimetric and hydrodynamic responses in acoustically-thin electrode coatings (thickness is much less than the wavelength of sound, hence any viscoelasic effect is negligible [17]). In case of absence of both hydrodynamic and viscoelastic effects linked to the character of the electrode structure, the use of conventional EQCM or AC-gravimetry on single (fundamental) frequency can be sufficient to evaluate the amount of solvent molecules co-adsorbed or co-intercalated into the electrodes together with the related ions. [18,19] Herein using a recently developed method of in situ hydrodynamic spectroscopy [20–22] we first quantify the amount of water confined between the Ti3C2 sheets (interspace) distinguishing it from the water located in the mesoporous space of the hierarchically porous electrode assembly. This approach was further extended to quantification of the amount of nanoconfined water coinserted/co-extracted during intercalation of multiple-charged cations having different hydrophilicity and hydrophobicity, respectively. Whereas hydrophilic (cosmotropic) cations are inserted with a part of their solvation shell, the hydrophobic (chaotropic) cations remain effectively unsolvated showing a clear tendency to push water molecules out of the confined interspaces of the Ti3C2 electrode. Our gravimetric/hydrodynamic study of the electrochemical insertion of different cations into the Ti3C2 electrode coupled with water insertion/extraction contributes to a better understanding of the mechanisms of insertion of multiply charged cations into 2D layered electrodes. [23] RESULTS AND DISCUSSION Gravimetric and hydrodynamic modes of EQCM-D When a quartz crystal (QC) is loaded with a rigid and uniformly spread coating, the width of the resonance peak (∆W/n) does not change, thus the entire rigidly attached coating moves synchronously with the surface of the oscillating QC. In this case, the resonance frequency (∆f/n) decreases proportionally with any added mass (∆mmass) following Sauerbrey equation [24]: ∆⁄ = − ∗ ∆ (1) where Cm is the mass sensitivity constant, and n is the overtone order. Equation (1) shows that mass measurements on different overtone orders are equivalent to measurements at the


Page 3 of 15 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


fundamental frequency (i.e., at n=1) if the related frequency change is normalized by the overtone order. Even though Sauerbrey equation gives a straightforward and convenient way for tracking gravimetric changes, the demands for a rigid and uniform coating are not always practically implemented. Viscoelastic films, for instance, tend to move in a way that is not synchronized with the QC oscillations which is expressed by additional contributions to the recorded frequency change resulting in errors in the gravimetric calculations. With this in mind, we have focused in the present paper on the frequency response of acoustically thin Ti3C2 electrodes (60 nm thick) with a loading mass around 15.9 µg/cm2: very small dissipation factors of this coating in air measured at different overtone orders, and the negligibly small changes of the dissipation factors during the process of ions insertion in contact with electrolyte solutions prove the rigid character of the electrode suitable for both hydrodynamic analysis of the electrode under OCV (open-circuit voltage) conditions and gravimetric analysis of ions and solvent insertions in the electrolyte solutions. Building on this result, we applied the principle of in situ hydrodynamic spectroscopy to quantify the amount of water confined in the porous Ti3C2 electrodes. This method makes use of two structural characteristics of porous hydrated electrodes, namely, the permeability length, ξ (linked to porosity via Kozeni-Carman equation), and thickness of the interfacial porous layer, h. [20–22] The hydrodynamic solid-liquid interactions occurring in a porous electrode are reflected by the resonance frequency (∆f/n) and resonance width (∆W/n) changes which, in turn, depend on both parameters of porous solid, ξ and h, and on the characteristic acoustic characteristic of liquid, called the penetration depth, δ. [20–22] Penetration depth is a natural variable independent parameter of the hydrodynamic problem of solid-liquid interactions characterizing damping (decay) of the shear wave amplitude from the oscillating crystal towards liquid’s interior as a function of liquid’s viscosity/density ratio (ηL/ρL) and shear wave overtone order, n (see SI Eq. S1). When ξ > δ (wide pores) most of liquid in pore interior is not trapped, and hence is moveable with respect to the oscillating crystal. In the first case, the trapped liquid contributes to the frequency change only (the resonance width remains unchanged). In contrast, moveable liquid in the wide pores contributes to both frequency change and resonance width changes. In order to avoid any confusion, we would like to specifically emphasize that both the moveable and non-moveable states of liquid are defined with respect to the oscillating crystal. For the selected thin-layer porous Ti3C2 electrode the condition ξ ≤ δ is an intermediate one between the above mentioned two limiting cases, implying that due to the features of the porous structure of this electrode a significant amount of liquid is trapped in the interspaces of the electrode particles. At the same time true mass of trapped liquid can be assessed only after making correction of the total experimental frequency change, ∆f/n, accounting for the hydrodynamic effect of the moveable liquid in the vertical interparticle pores and horizontal interstack gaps of the Ti3C2 electrode. A full discussion of the porous Ti3C2 electrode structure and its acoustic probing, proof of rigidity of the Ti3C2 coatings in air, and details of hydrodynamic modeling are presented in the Supporting Information (SI) in Fig. S1 and Fig. S2, respectively. Tracking distribution of water in Ti3C2 electrode in a quasi-stationary state (no external polarization) The normalized changes of frequency (∆f/n) and resonance half-bandwidth (∆Γ/n =∆W/2n) for the Ti3C2 electrode fully impregnated with water are shown by closed and open red circles, respectively, in Fig. 1a. For a comparison, the hydrodynamic response of the ideally flat surface of the neat crystal (serving as a reference state) are shown by dotted black straight lines. [25] The downward and upward deviations of the experimental ∆f/n and ∆Γ/n changes, respectively, for the Ti3C2 electrode from the straight lines describing the response of the neat crystal implies that the Ti3C2 electrode has a porous structure containing both moveable and trapped water. As mentioned previously, moveable water contributues both to ∆f/n and ∆Γ/n changes, whereas trapped water results in only negative ∆f/n changes. In order to distinguish between the moveable and trapped water a suitable



Page 4 of 15

hydrodynamic model for fitting the experimental values of ∆f/n and ∆Γ/n changes should be used. As mentioned in the SI and explained in detail in ref. [21], the particular EQCM-D signatures of different types of rough/porous electrode structures can be analyzed by using appropriate hydrodynamic models that are selected on the basis of other surface structure sensitive techniques, e.g. from scanning electron microscopy (SEM) or atomic force microscopy (AFM) images. SEM images of the Ti3C2 elecrtrodes used in this study are shown in Fig. 1b and c. The lower resolution image shows there is an absence of gross features of the external roughness on the scale of the penetration depths of water, the latter ranges from 68 to 140 nm for the 13th and 3rd overtones, respectively. The higher resolution image shows the maximimal lateral size of the platelet-like flakes does not exceed 500 nm. According to the Ti3C2 synthetic procedure used, the individual flakes are expected to be a single or a few layers thick. [7] We distinguish between two types of nano-confined space in Ti3C2 electrodes that were impregnated with water (after removal of air from the pores of the electrode structure by vacuum in a hermetically closed bottle): (I) 2-3 nm size interspaces between the closely packed single/few layer thick sheets in the lamelas, and (II) mesoscopic-size vertical interparticle pores and horizontal interstack gaps (see Fig. S1). Assuming that mesoscopic pores and gaps make up a system of interconnected pores through which water moves during shear oscillation of the coated crystal, we apply the model for uniform porous layers for analyzing the obtained EQCM-D responses. In order to distinguish between the location of water molecules in the interspaces and mesoporous volume, we first fitted the model to the experimental ∆Γ/n changes (red open circles in Fig. 1a, the red solid line represents the best fit) finding two optimal model structural parameters (discussed above): ξ = 28 nm and h = 45 nm (SI Eq. S2a and S2b). In the next step, the evaluated parameters were inserted into SI Eq. S2a to calculate the hydrodynamic contribution of the moveable water to the ∆f/n changes (red solid line in Fig. 1a). It is clearly seen in the same figure that there is an approximate n-independent downward (negative) shift of ∆f/n for the experimental values (red solid symbols and the related thin straight line) from the model curve (red solid line). Such a shift corresponds to inertial loading of the crystal due to the water in the Ti3C2 coating being in an unmoveable (confined) state, i.e. completely trapped in the Ti3C2 interspaces. From the dimensionless frequency change denoted by two arrows (corresponding to the frequency change of 65 Hz) we can easily calculate the ratio between the amount of confined water and Ti3C2 entities. Here we are assuming the chemical composition of the Ti3C2 is Ti3C2(OH)F based on previous studies [11] (molecular mass 204 g/mole). The coating on the quartz crystal had an areal mass density of 0.078 µmole/cm2. Using the Sauerbrey equation for the above-mentioned frequency shift of 65 Hz, the amount of water molecules trapped in the Ti3C2 interspaces was calculated to be 0.0638 µmole/cm2. Thus, after impregnation of water under vacuum, the composition of the Ti3C2 coating was found to be Ti3C2(OH)F (H2O)0.82. It is seen that the molar ratio of the surface functionality –OH to that of the confined water is close to one, suggesting weak hydrogen bonding of water molecules to the –OH surface groups. The ratio of the amount of water trapped in Ti3C2 interspaces and located in the interparticle pores and interlayer gaps (moveable water) can be rougly estimated as 1:3. This follows from application of the Sauerbrey equation to the ∆f/n shift assigned to the moveable water (i.e. located between the straight line for plane surface and the red solid line for the model in Fig. 1a) taking into account that the shift of ∆Γ/n is smaller than the related shift of ∆f/n. As a result the estimated amount of water, 0.186 µmole/cm2, attributed to the mesoporous voids of the electrode is about a triple higher than the amount of water confined in the electrode interspaces. Fig. 1a relates to quasi-stationary EQCMD measurements of Ti3C3 electrode at OCV when the pure water in the cell is replaced first by a 1M solution of LiCl. After ten full fast cycles the solution is replaced by a 1M CsCl solution, and the EQCMD measurements are conducted again at OCV. The normalized bandwidth and frequency change for all three fluids remain practically unchanged. Therefore the amount of water trapped in the interspaces of the electrode also remains essentially constant.


Page 5 of 15 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 1. Hydrodynamic spectroscopy of a thin Ti3C2 electrode after impregnation with pure water (a). Dashed lines correspond to normalized Δf/n and ΔΓ/n changes of an ideally flat surface. The 9

9

2

normalized quantities are expressed as (∆f/n)∙10 /ρ f

o

2

and (∆Γ|n)∙10 /ρ f , for frequency and o

resonance half-width, respectively, where f is fundamental frequency and ρ is liquid’s density. The o

same quantities for the Ti3C2 electrode after consecutive replacement of water with 1M solutions of LiCl and CsCl are shown by solid and open symbols, respectively, as indicated. Red solid lines denote the response obtained by fitting the model of uniform porous layer to the experimental EQCM-D data. The frequency response due to water trapped in the Ti3C2 interspaces is indicated by the two oppositely directed arrows. The HR SEM images of the Ti3C2 electrode of lower and higher resolutions are shown in (b) and (c), respectively. Scale bars are 2µm and 500 nm, respectively.



Page 6 of 15

Dynamic EQCM-D measurements during insertion cations intercalation into Ti3C2 electrodes In this section we report on the number of water molecules that the electrochemically inserted cations can co-insert/co-extract into/from the Ti3C2 interspace, thereby changing the total water content of the electrodes. Since the insertion of all the studied cations results in negligibly small changes of ∆Γ/n (see Figs. 2 and 3 for a series of hydrophilic and hydrophobic cations, respectively) the Sauerbrey equation (Eq. 1) is used to translate the related frequency changes into the effective masses of inserted cations. This is further transformed into the molar ratios of the desolvated cations (calculated from the related masses of the inserted cations) whereas the rest of the frequency change (either positive or negative) is assigned to the number of moles of water accompanying the insertion of the cations (this simple calculation routine was reported previously for nanoporous carbons in full detail [26]).

Figure 2. Timedependent changes of ∆f/n and ∆Γ/n during insertion/extraction of Li+ (a), Mg2+ (b), and Al3+ (c) cations into/from the Ti3C2 electrode (scan rate 50 mV/s) collected at 3rd, 7th, and 11th overtones. The frequency change calculated from the charge was obtained by integration of the related CVs ((d), (e), and (f), respectively) is shown by the dashed lines (Δf/n Faraday) in panels (a), (b) and (c), respectively. The number of moles of water molecules co-inserted (or co-extracted) from the confined Ti3C2 interspace, mw, can be easily calculated by comparison of the slopes of the plots of the experimental change of the amount of the inserted cations and its Faradaic analog (i.e. the population change calculated from the experimental intercalation charge using Faraday law; both quantities are expressed in nmoles/cm2) as a function of the electrode charge density (mC/cm2): , and

, respectively: [26]


Page 7 of 15 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


=

!" !

∗

#$%& #' ( )

(2) *+,- and *.( / denote the molecular masses of unsolvated ion and water, respectively. The Faradaic population change of inserted cations is linked to the Faradaic frequency change, ∆f/nFaraday, via Sauerbrey’s equation.

Figure 3. Time-dependent changes of Δf/n and ΔΓ/n during insertion/extraction of Ba2+(a), Cs+(b), and TEA+ (c) cations into/from the Ti3C2 electrode (scan rate 50 mV/s) collected at 3rd,7th, and 11th overtones. The frequency change calculated from the charge obtained by integration of the related CVs ((d), (e), and (f), respectively) is shown by the dashed lines (Δf/n Faraday) in panels (a), (b) and (c), respectively. A closer inspection of Fig. 2 shows that for the series of hydrophilic cations, Li+, Mg2+, and Al3+, the absolute value of the experimental frequency change, ∆f/nexper, with respect to that of ∆f/nFaraday, increases in the same order, signifying an increase in the number of water molecules accompanying cation insertion. In contrast, for large cations (Ba2+, Cs+ and TEA+), in this order, ∆f/nFaraday appears to be larger by its absolute value than that for ∆f/nexper, implying that the large cations remove water molecules during their insertion into the Ti3C2 interspace (Fig. 3). Moreover, we were able to observe the transfer from the behavior typical for the cosmotropic and chaotropic cations within the entire alkaline-metal cations series: the behavior of the first and last members of this series (i.e. Li+ and Cs+ cations) are shown in Fig. 2a, d and Fig. 3b, e, respectively, whereas the data for Na+ and K+ cations are shown in SI Fig. S3a, c and Fig. S3b, d, respectively. Note that the described approach is valid only for the so-called permselective character of intercalation/adsorption process implying insertion of major counter-ions (e.g. cations during a



Page 8 of 15

negative polarization of the electrode) [18,27]. At a small polarization, e.g. in the vicinity of potential of zero charge of carbon electrodes, or in case of the presence of fixed (localized) charged groups in the electrode structure, the co-ions participate to the charge-compensation process in the electrode in addition to that of the counter-ions, and in this case, the use of electrogravimetric method becomes especially attractive because of the high selectivity of this technique to the nature of inserted ions and solvent molecules. [28] The results from quantifying the number of moles of water accompanying the insertion of different cations into the Ti3C2 interspace (i.e. mw, see Eq. 2) is presented in Fig. 4. The comparison of the slopes of aexper and aFaraday, which visualizes the method of calculation of mw, is shown in Fig. 5, using the intercalation of Ba2+ and Mg2+ cations as an example.

Figure 4. Number of water molecules co-inserted per intercalated hydrophilic cation (positive values) or extracted from the electrode during intercalation of hydrophobic cations (negative values). The quantity mw was obtained from raw experimental frequency changes as compared to the Faradaic frequency change due to insertion of unsolvated cations using Eq. 2. The standard errors were calculated from the measurements of four independent electrode coatings of similar mass.

In case of Al3+ cation (which is subjected to hydrolysis in aqueous solution resulting in appearance of H+ ions), the charge obtained from the related CV (Fig. 2f) is significantly larger than that for the insertion of Li+ and Mg2+ cations due to the insertion of H+ cations into the Ti3C2 interspaces. We have assumed that the charge due to the insertion of Al3+ ions is approximately equal to that for the insertion of Li+ and Mg2+- cations, and in addition that this charge is linked to the ∆f/n quantity obtained from ∆f/nexper, by subtraction of the contribution of H+ cations (for clarity a very detailed calculation of mw for Al3+ cation is included into the SI). As follows from Fig. 5 for two double-valent cations with different unsolvated radii (Mg2+and Ba2+), the insertion of the larger Ba2+ cations is characterized by being practically equal to (hence the Ba2+ cation is inserted into the Ti3C2 electrode in its effectively non-solvated form), whereas for insertion of the smaller Mg2+ cation, a significantly larger experimental slope, , is observed as compared to that of . According to Eq. 2 this implies a positive value of mw


Page 9 of 15 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


meaning that insertion of this cation into the Ti3C2 is accompanied by co-insertion of water molecules.

Figure 5. Changes of the effective amount of Mg2+ and Ba2+ cations as a function of the electrode charge density (circles and squares, respectively). The solid red line (called the Faraday line) was obtained by conversion of the experimental intercalation charge for both cations into the related mass change using the Faraday equation (this slope is designated as aFaraday). The experimentally measured frequency changes for the intercalated Ba2+ and Mg2+ cations were converted into the mass changes using the Sauerbrey equation, and then assuming insertion of unsolvated cations the mass changes were transformed into the related molar changes (see dashed black lines with the different slopes, aexper). The population changes expressed by these lines were averaged over the entire range of overtone orders, n. The differences in the slopes of the experimental lines and the Faraday line, which reflects the co-insertion of water molecules during cation intercalation, were introduced into Eq. 2 to quantify the number of water molecules per each cation co-inserted into the electrode (see Fig. 4). Intercalation starts at a charge density equal to zero and results in change of ∆f/n exper which appears to be larger by its absolute value than that of ∆f/n Faraday, see Fig. 2a-c). When ∆f/n exper = ∆f/n Faraday (see Fig. 3a) the cation intercalates in its effectively desolvated form, meaning no water is gained or lost by the electrode. When the absolute value of the ∆f/n exper is smaller than ∆f/n Faraday (this case is not shown in Fig. 5) the intercalation of the hydrophobic cation is accompanied by dehydration of the electrode, see Fig. 3b-c).

Similar calculations have been performed for gravimetric analysis of intercalation of hydrophilic and hydrophobic cations of alkaline and alkaline-earth series, Al3+ and TEA+ cations. As shown in Fig. 5, insertion of the hydrophilic cations increases the content of water molecules in the Ti3C2 interspaces whereas insertion of the hydrophobic cations decreases the water content. The “dehydration” effect of large cations significantly increases when coming from K+ to Cs+, and more gradually, when coming further to TEA+. This tendency is in good agreement with an earlier prediction that intercalation of large ions results in removal of water molecules from the interspces



Page 10 of 15

implied from the studies of the actuation and nanoscale viscoelastic properties of MXene electrodes. [29,30]As an example, we have estimated that the insertion of Li+ cations increases the initial water content in the Ti3C2 interspace by 20% whereas a similar insertion of Cs+ cations dehydrates the electrode by 41%.

DISCUSSION The number of water molecules either co-inserted or, alternatively, extracted by ions inserted during the electrochemical cycling of Ti3C2 electrodes depends strongly on the ion-water interactions, which is a function of the different ion’s charge per size ratio. The existence of such a relationship obtained experimentally by EQCM-D technique can be perfectly explained by a combination of two concepts implying that: (I) the cosmotropic or chaotropic nature of small and large cations, respectively, determines their behavior in bulk aqueous solutions, [31] and (II) that the essentially confined space in the Ti3C2 electrode (van-der-Waals interspaces) containing water molecules prior to when the cations are electrochemically inserted into the electrode modifies the former bulk effect. Table 1 collects the data listing the unhydrated and hydrated radii of cations and the hydration numbers for several different cations used in our study. The last column indicates the viscosity Bcoefficient used in the Jones-Dole equation [32]: 0/02 = 1 + 567/8 + 96 + 6 8 + … (3) where η and η0 are viscosities of solution and pure solvent, respectively, c is the electrolyte solution concentration, and A, B and C the related coefficients.

Table 1. Radii of hydrated and unhydrated cations (rhydr and runhydr, respectively expressed in Å), hydration numbers, n, and viscosity B-coefficients (dm3/mol) from the Jone-Dole equation [31] Cation

runhydr

rhydr,

n

B

Al3+

0.53

3.77

20.4

0.744

Mg2+

0.72

2.99

10.0

0.385

Li+

0.69

2.41

5.2

0.146

Ba2+

1.36

2.54

5.3

0.216

Cs+

1.70

2.19

2.1

-0.047

TEA+

3.37

3.45

1.1

0.385

The Jone-Dole coefficient B characterizes the strength of ion-water interaction, with positive values representing structure-making cations (cosmotropes) and negative values for the structure-breaking cations (chaotropes). It is implied that small-size multivalent cations modify the hydrogen bonds existing in the water in close vicinity to the cosmotropic cations by the formation of hydration shells; these cations are said to possess hydrophilic properties. The enthalpic gain from the watercation interactions exceeds the entropic loss due to the rearrangement of the water molecules. [32] Due to the decrease of the strength of the cation-water interaction in the sequence: Al3+>Mg2+>Ba2+≥Li+, the viscosity coefficient B decreases. For the large monovalent Cs+ cation with only weak hydrophobic ability, B is small and negative, implying that this cation perturbs the water structure slightly so that the reorganization of hydrogen bonds easily occurs. In contrast, larger-size monovalent structure-breaking cations such as tetraalkyl ammonium cations (with increasing hydrophobicity for the cations with longer chains) profoundly disturbs the bulk water structure in close proximity to the ions. Strengthening, and reorganization of the hydrogen bonds between the


Page 11 of 15 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


water molecules in the form of the iceberg-like structure takes place in the bulk of the electrolyte solution, which results in the increase of B. Surely this process is accompanied by a large entropy decrease [32]. Moreover, large hydrophobic cations when located at a short enough distance knock out the water molecules significantly decreasing the water density (and hence content) down to that of water vapor. [32] The above scenario seems to describe the behavior of the six selected cations when electrochemically inserted into the Ti3C2 electrode, taking into account the effect of the confined space. In fact, as follows from our EQCM-D study, three typically cosmotropic cations (Al3+, Mg2+ and Li+) are characterized by a smaller number of the water molecules linked to the cations when coinserted into the Ti3C2 electrode (see Fig. 4) compared to the bulk hydration number for the same cations (see Table 1). Thus, the partial dehydration of the cosmotropic cations (enhancing with the increase of cation’s charge-to-size ratio) is a direct consequence of the confined effect of Ti3C2 interspaces. The effect of the larger size of Ba2+ cation is counter-balanced by the effect of its higher charge so that its insertion only slightly perturbs the water structure in the Ti3C2 electrodes. A gradual transition from the cosmotropic to the chaotropic behavior was observed along the entire series of alkaline-metal cations (see Fig. 4). For both hydrophobic Cs+ and TEA+ cations, an excluded-volume region is formed in close proximity to the cations in which the water density is smaller than that in the bulk water. [32] Our EQCM-D studies clearly confirm this conclusion by calculation of the average number of water molecules replaced by the inserted K+, Cs+ and TEA+ cations. The excluded volume effect exists already for the hydrophobic cations in bulk solution. However, the confined character of the hydrophobic cations inserted into the Ti3C2 electrode makes the excluded-volume effect the dominant one whereas the iceberg-like water structure typical for the strongly hydrophobic cations in bulk solutions [32] is not formed because of the confined space of the Ti3C2 electrode. Although our study relates to heavily exfoliated single/a few layer thick Ti3C2 particles the main results are in agreement with that obtained in a recent calorimetric study of ions exchange in multilayered clay-like Ti3C2. [13] The conlusion about a weak interaction of water with the Ti3C2 interspace is in accord with our result of the existence of almost 1:1 ratio of the number of –OH functionalities and the number of inserted water molecules. The evidence obtained in the present EQCM-D study that hydrophilic cations bring water molecules into the Ti3C2 interspaces is in accord with the conclusion about the expansion of the interplanalar Ti3C2 spacing caused by the co-inserted water. [11] CONCLUDING REMARKS Using multiharmonic EQCM-D we have applied a new method of in situ hydrodynamic spectroscopy to quantify the amount of water in the Ti3C2 electrode after its impregnation with water. The hydrodynamic modeling of the frequency and resonance width changes allows to distinguish between the water trapped in the Ti3C2 interspaces, and moveable water located in the vertical interparticle pores and horizontal interstack gaps of the Ti3C2 electrode. In order to understand the effect of the inserted cations on the amount of water in the Ti3C2 interspaces eight hydrophilic and hydrophobic cations with different charge-to-size ratio have been selected. Four typically hydrophilic cations (Na+, Li+, Mg2+, Al3+) co-insert water molecules in larger amount as the charge-to-size ratio of the cation increases. In contrast, hydrophobic K+, Cs+ and TEA+ cations knock out the water molecules due to the volume exclusion effect. The gravimetric/hydrodynamic method of analysis that we developed for the Ti3C2 electrode to throw light on the nature of ion-solvent interactions in confined interspaces of Ti3C2 electrode is equally applicable for the characterization of super-capacitive nanoporous carbons and solvated battery and redox-type super-capacitive electrodes. EXPERIMENTAL SECTION



Page 12 of 15

Electrode preparation.Ti3C2 was synthesized using HCl+LiF as an etchant. The full synthesis method was reported elsewhere [16]. The prepared Ti3C2 was mixed in double deionized water (4mg in 20ml) and sprayed on a quartz sensor placed on hot plate at 40°C, using an airbrush supplied with compressed nitrogen. The coated crystal (after removal of air from the pores of the electrode by vacuum) was impregnated with either double-distilled water or a 1M electrolyte solutions of the chloride salts the examined cations (all salts mentioned were purchased from Sigma-Aldrich). EQCM-D characterization: Multiharmonic quartz-crystal measurements were performed by QSense E1 module (QCM-D from Biolinve Scientific) using 5 MHz Au-coated QCM sensors (q-sense Scientific AB, Swedenusing). Frequency and dissipation responses were simultaneously collected from 3rd to 13th overtones. The coated QC (served as a working electrode) was assembled in a home designed 3 electrode cell using Pt as a counter electrode and Ag/AgCl/KCl (sat.) as a reference electrode. electrochemical measurements were conducted by BioLogic VSP-300 with EC-Lab software. Complementary morphological characterizations: HR-SEM imaging was performed using Magellan XHR 400L FE-SEM (FEI Company). Ti3C2 electrode thickness was measured (relative to an uncoated fraction of Au electrode) by Fast Scan Bio atomic force microscope (Bruker AXS) using contact mode with a single FASTSCAN-A silicon probe (force constant of 18 N/m, Bruker).

ASSOCIATED CONTENT Supporting Information Supporting Information contains discussion of gravimetric and non-gravimetric EQCM, and basics of in situ hydrodynamic spectroscopy. AUTHOR INFORMATION Corresponding Authors [email protected] [email protected] Notes The aurhors declare no competing fincancial interests. ACKNOWLEDGEMENTS The authors acknowledge funding from the Binational Science Foundation (BSF) grant No. 2014083/2016 and to the Israel Ministry of Science Technology and Space (Grant 66032) for their financial support. N.S. thanks the Israel Ministry of Science Technology and Space for its support. Synthesis of Ti3C2 MXene was supported by the Fluid Interface Reactions, Structures and Transport (FIRST) Center, an Energy Frontier Research Center funded by the U.S. Department of Energy, Office of Science, Basic Energy Sciences.

REFERENCES (1)

Augustyn, V.; Gogotsi, Y. 2D Materials with Nanoconfined Fluids for Electrochemical Energy Storage. Joule 2017, 1 (3), 443–452.

(2)

Mitchell, J. B.; Lo, W. C.; Genc, A.; LeBeau, J.; Augustyn, V. Transition from Battery to Pseudocapacitor Behavior via Structural Water in Tungsten Oxide. Chem. Mater. 2017, 29 (9), 3928–3937.

(3)

Sai Gautam, G.; Canepa, P.; Richards, W. D.; Malik, R.; Ceder, G. Role of Structural H 2 O in Intercalation Electrodes: The Case of Mg in Nanocrystalline Xerogel-V 2 O 5. Nano Lett. 2016, 16 (4), 2426–2431.


Page 13 of 15 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(4)

Wang, X.; Kajiyama, S.; Iinuma, H.; Hosono, E.; Oro, S.; Moriguchi, I.; Okubo, M.; Yamada, A. Pseudocapacitance of MXene Nanosheets for High-Power Sodium-Ion Hybrid Capacitors. Nat. Commun. 2015, 6, 6544.

(5)

Gao, Y.; Wang, L.; Li, Z.; Zhang, Y.; Xing, B.; Zhang, C.; Zhou, A. Electrochemical Performance of Ti3C2 Supercapacitors in KOH Electrolyte. J. Adv. Ceram. 2015, 4 (2), 130–134.

(6)

Luo, J.; Zhang, W.; Yuan, H.; Jin, C.; Zhang, L.; Huang, H.; Liang, C.; Xia, Y.; Zhang, J.; Gan, Y.; Tao, X. Pillared Structure Design of MXene with Ultralarge Interlayer Spacing for High-Performance Lithium-Ion Capacitors. ACS Nano 2017, 11 (3), 2459–2469.

(7)

Ghidiu, M.; Lukatskaya, M. R.; Zhao, M.-Q.; Gogotsi, Y.; Barsoum, M. W. Conductive Two-Dimensional Titanium Carbide “clay” with High Volumetric Capacitance. Nature 2014, 516 (7529), 78–81.

(8)

Anasori, B.; Lukatskaya, M. R.; Gogotsi, Y. 2D Metal Carbides and Nitrides (MXenes) for Energy Storage. Nat. Rev. Mater. 2017, 2 (2), 16098.

(9)

Hu, M.; Li, Z.; Hu, T.; Zhu, S.; Zhang, C.; Wang, X. High-Capacitance Mechanism for Ti3C2Tx MXene by in Situ Electrochemical Raman Spectroscopy Investigation. ACS Nano 2016, 10 (12), 11344–11350.

(10)

Lukatskaya, M. R.; Bak, S.-M.; Yu, X.; Yang, X.-Q.; Barsoum, M. W.; Gogotsi, Y. Probing the Mechanism of High Capacitance in 2D Titanium Carbide Using In Situ X-Ray Absorption Spectroscopy. Adv. Energy Mater. 2015, 5 (15), 1500589.

(11)

Ghidiu, M.; Halim, J.; Kota, S.; Bish, D.; Gogotsi, Y.; Barsoum, M. W. Ion-Exchange and Cation Solvation Reactions in Ti3C2 MXene. Chem. Mater. 2016, 28 (10), 3507–3514.

(12)

Osti, N. C.; Naguib, M.; Ostadhossein, A.; Xie, Y.; Kent, P. R. C.; Dyatkin, B.; Rother, G.; Heller, W. T.; Van Duin, A. C. T.; Gogotsi, Y.; Mamontov, E. Effect of Metal Ion Intercalation on the Structure of MXene and Water Dynamics on Its Internal Surfaces. ACS Appl. Mater. Interfaces 2016, 8 (14), 8859–8863.

(13)

Sharma, G.; Muthuswamy, E.; Naguib, M.; Gogotsi, Y.; Navrotsky, A.; Wu, D. Calorimetric Study of Alkali Metal Ion (K+, Na+, Li+) Exchange in a Clay-Like MXene. J. Phys. Chem. C 2017, 121 (28), 15145–15153.

(14)

Hu, M.; Hu, T.; Li, Z.; Yang, Y.; Cheng, R.; Yang, J.; Cui, C.; Wang, X. Surface Functional Groups and Interlayer Water Determine the Electrochemical Capacitance of Ti 3 C 2 T

(15)

X

MXene. ACS Nano 2018, 12 (4), 3578–3586.

Nam, K. W.; Kim, S.; Lee, S.; Salama, M.; Shterenberg, I.; Gofer, Y.; Kim, J. S.; Yang, E.; Park, C. S.; Kim, J. S.; Lee, S.S.; Chang, W.S.; Doo S.G.; Jo Y.N.; Jung, Y.; Aurbach, D.; Choi J.W. The High Performance of Crystal Water Containing Manganese Birnessite Cathodes for Magnesium Batteries. Nano Lett 2015, 15 (6), 4071–4079.

(16)

Shpigel, N.; Lukatskaya, M. R.; Sigalov, S.; Ren, C. E.; Nayak, P.; Levi, M. D.; Daikhin, L.; Aurbach, D.; Gogotsi, Y. In Situ Monitoring of Gravimetric and Viscoelastic Changes in 2D Intercalation Electrodes. ACS Energy Lett. 2017, 2 (6), 1407–1415.

(17) (18)

Johannsmann, D. The Quartz Crystal Microbalance in Soft Matter Research; 2015. Levi, M. D.; Salitra, G.; Levy, N.; Aurbach, D.; Maier, J. Application of a Quartz-Crystal Microbalance to Measure Ionic Fluxes in Microporous Carbons for Energy Storage. Nat. Mater. 2009, 8 (11), 872–875.

(19)

Arias, C. R.; Debiemme-Chouvy, C.; Gabrielli, C.; Laberty-Robert, C.; Pailleret, A.; Perrot, H.; Sel, O. New Insights into Pseudocapacitive Charge-Storage Mechanisms in Li-Birnessite Type MnO2 Monitored by Fast Quartz Crystal Microbalance Methods. J. Phys. Chem. C 2014, 118, 26551–26559.

(20)

Shpigel, N.; Levi, M. D.; Sigalov, S.; Girshevitz, O.; Aurbach, D.; Daikhin, L.; Pikma, P.; Marandi, M.; Jänes, A.; Lust, E.; Jäckel, N.; Presser, V. In Situ Hydrodynamic Spectroscopy for Structure Characterization of Porous Energy Storage Electrodes. Nat. Mater. 2016, 15 (5), 570–575.

(21)

Levi, M. D.; Shpigel, N.; Sigalov, S.; Dargel, V.; Daikhin, L.; Aurbach, D. In Situ Porous Structure Characterization of Electrodes for Energy Storage and Conversion by EQCM-D: A Review. Electrochim. Acta 2017, 232, 271–284.

(22)

Sigalov, S.; Shpigel, N.; Levi, M. D.; Feldberg, M.; Daikhin, L.; Aurbach, D. Electrochemical Quartz Crystal Microbalance with Dissipation Real-Time Hydrodynamic Spectroscopy of Porous Solids in Contact with Liquids. Anal. Chem. 2016, 88 (20), 10151–10157.

(23)

Muckley, E. S.; Naguib, M.; Wang, H.-W.; Vlcek, L.; Osti, N. C.; Sacci, R. L.; Sang, X.; Unocic, R. R.; Xie, Y.; Tyagi, M.;



Page 14 of 15

Mamontov, E.; K.L, Page.; Kent, P.; Nanda, J.; Ivanov I.N. Multimodality of Structural, Electrical, and Gravimetric Responses of Intercalated MXenes to Water. ACS Nano 2017, acsnano.7b05264. (24)

Sauerbrey, G. Use of a Quartz Vibrator for Weighing Thin Films on a Microbalance. Zeitschrift für Phys. 1959, 155 (2), 206–222.

(25)

Keiji Kanazawa, K.; Gordon, J. G. The Oscillation Frequency of a Quartz Resonator in Contact with Liquid. Anal. Chim. Acta 1985, 175 (C), 99–105.

(26)

Levi, M. D.; Sigalov, S.; Salitra, G.; Elazari, R.; Aurbach, D. Assessing the Solvation Numbers of Electrolytic Ions Confined in Carbon Nanopores under Dynamic Charging Conditions. J. Phys. Chem. Lett. 2011, 2 (2), 120–124.

(27)

Griffin, J. M.; Forse, A. C.; Tsai, W.-Y.; Taberna, P.-L.; Simon, P.; Grey, C. P. In Situ NMR and Electrochemical Quartz Crystal Microbalance Techniques Reveal the Structure of the Electrical Double Layer in Supercapacitors. Nat. Mater. 2015, 14 (8), 812–819.

(28)

Escobar-Teran, F.; Arnau, A.; Garcia, J. V.; Jiménez, Y.; Perrot, H.; Sel, O. Gravimetric and Dynamic Deconvolution of Global EQCM Response of Carbon Nanotube Based Electrodes by Ac-Electrogravimetry. Electrochem. commun. 2016, 70, 73–77.

(29)

Come, J.; Black, J. M.; Lukatskaya, M. R.; Naguib, M.; Beidaghi, M.; Rondinone, A. J.; Kalinin, S. V.; Wesolowski, D. J.; Gogotsi, Y.; Balke, N. Controlling the Actuation Properties of MXene Paper Electrodes upon Cation Intercalation. Nano Energy 2015, 17, 27–35.

(30)

Come, J.; Xie, Y.; Naguib, M.; Jesse, S.; Kalinin, S. V.; Gogotsi, Y.; Kent, P. R. C.; Balke, N. Nanoscale Elastic Changes in 2D Ti3C2Tx (MXene) Pseudocapacitive Electrodes. Adv. Energy Mater. 2016, 6 (9).

(31)

Zhao, H. Are Ionic Liquids Kosmotropic or Chaotropic? An Evaluation of Available Thermodynamic Parameters for Quantifying the Ion Kosmotropicity of Ionic Liquids. Journal of Chemical Technology and Biotechnology. 2006, pp 877– 891.

(32)

Jones, G.; Dole, M. THE VISCOSITY OF AQUEOUS SOLUTIONS OF STRONG ELECTROLYTES WITH SPECIAL REFERENCE TO BARIUM CHLORIDE. J. Am. Chem. Soc. 1929, 51 (10), 2950–2964.


Page 15 of 15 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


47x26mm (600 x 600 DPI)

ACS Paragon Plus Environment

Direct Assessment of Nanoconfined Water in 2D Ti3C2 Electrode

Recommend Documents