Effects of Applied Potential and Water Intercalation on the Surface

Dec 9, 2016 - A.J. Drexel Nanomaterials Institute and Department of Materials Science and Engineering, Drexel University, Philadelphia, Pennsylvania ...
2 downloads 0 Views 2MB Size
Download PDF
Article pubs.acs.org/JPCC

Effects of Applied Potential and Water Intercalation on the Surface Chemistry of Ti2C and Mo2C MXenes Kurt D. Fredrickson,†,‡ Babak Anasori,§ Zhi Wei Seh,∥ Yury Gogotsi,§ and Aleksandra Vojvodic*,⊥ †

Department of Chemical Engineering, Stanford University, Stanford, California 94305, United States SUNCAT Center for Interface Science and Catalysis, SLAC National Accelerator Laboratory, Menlo Park, California 94025, United States § A.J. Drexel Nanomaterials Institute and Department of Materials Science and Engineering, Drexel University, Philadelphia, Pennsylvania 19104, United States ∥ Institute of Materials Research and Engineering, Agency for Science, Technology and Research (A*STAR), 2 Fusionopolis Way, Innovis, Singapore 138634 ⊥ Department of Chemical and Biomolecular Engineering, University of Pennsylvania, Philadelphia, Pennsylvania 19104, United States ‡

S Supporting Information *

ABSTRACT: Two-dimensional transition metal carbides and nitrides, also known as MXenes, represent an attractive class of materials for a multitude of electrochemical and other applications. While single sheets of MXenes have been widely studied theoretically, there have been much fewer studies on layered bulk MXenes, which are more representative of multi- or few-layer MXenes used in actual applications. Herein, we investigate the structural and electronic effects of water intercalation, multiple functional groups and applied potential on layered bulk Ti2C and Mo2C MXenes using density functional theory. The out-of plane lattice parameter, c, was found to vary significantly with the functional group, and is greatly increased upon intercalation of water. Experimental results confirm the change in lattice constant due to addition or removal of intercalated water. Under zero applied potential, both Ti2C and Mo2C were found to be functionalized by one monolayer of O; bare MXenes were never found to be stable, regardless of the applied potential. Applying a potential changed the adsorbate coverage, changing the systems from O covered to H covered at negative potentials and, in some cases, giving rise to a metal−insulator transition. Understanding of the effects of surface functionalization and water intercalation of MXenes provides a better insight of their use for catalytic and electronic applications.

■

INTRODUCTION Two-dimensional materials have attracted great interest in recent years for their exciting physical and chemical properties. One of these intriguing new families of two-dimensional materials are the transition metal carbides and nitrides, MXenes.1,2 MXenes are produced by etching their ternary layered carbide or nitride precursors,3−5 which consist of an early transition metal M, a post-transition metal or semiconductor A (such as Al, Si, or Ga), and X, which is either C or N. These ternary carbides and nitrides are typically composed of MnXn‑1 layers, interleaved by layers of A, where n = 2, 3, or 4. The selective removal of the A atom from the precursor results in a MXene,6 two-dimensional sheets of a transition metal carbide or nitride weakly bound via secondary bonds, such as van der Waals forces3 and hydrogen bonds.7 MXenes are being explored for a wide variety of applications in many different fields,2 including both Li-ion3 and multivalent-ion batteries,3,8,9 supercapacitors,10 hydrogen storage,11 photocatalysis,12 electrocatalysis,1,13−15 and nanoparticle supports for catalysis.16−18 Bare MXenes are very reactive, and in any real-life application © XXXX American Chemical Society

are functionalized by either O, OH, or F groups, although F groups have been shown to be replaced by OH groups when the MXene is stored in water or washed in a high-pH solution (a base).19,20 In fact, the electronic properties of MXenes vary drastically with both the type of functional groups present on the surface, and their adsorption site;21,22 for example, Ti2CO2 is predicted to be semiconducting, even though Ti2C with most other functional groups is metallic.23,24 While there are several theoretical studies on single sheets of MXenes,22,23,25−27 there are only a few performed on layered bulk MXenes,7,28 which are more representative of multi- or few-layer MXenes used in actual electrochemical applications. Functionalization with OH groups, along with water intercalation, was shown to have large effects on the c- lattice parameters (c-LP) of V2C and Nb2C.19 While many studies have shown the effects of functionalization on the electronic Received: September 8, 2016 Revised: November 1, 2016

A

DOI: 10.1021/acs.jpcc.6b09109 J. Phys. Chem. C XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry C properties of MXenes, few have considered the possibilities of multiple functional groups present at once; e.g., functionalized with both O and H groups simultaneously. Also, while previous studies have shown the effects of applied anodic potential on the oxidation of MXenes,10 there have been few systematic studies of stability or thermodynamic effects on differently functionalized MXenes. We address these critical gaps in knowledge by investigating layered bulk Ti2C and Mo2C MXene structures, including the effects of water intercalation, multiple functional groups, and applied potential. We choose to study Ti2C and Mo2C because of their different bulk carbide chemistries (TiC and β-Mo2C), and difference in crystal structure. Both Ti2C and Mo2C MXenes have been already synthesized, thus offering a possibility of experimental verification of the computationally reported predictions. Prior theoretical calculations on MAX phases show similarities between a MAX phase and its corresponding bulk MX phase (e.g., Ti3SiC2 and TiC),29,30 and this may have important consequences for the functionalization of these materials.

Figure 1. Side and top view of the atomic structure of bare Ti2C or Mo2C MXene. On the right, the possible adsorption sites for the different functional groups (allowed H, OH, O, H2 and H2O) on MxCx−1. The black diamond indicates the (1 × 1) unit cell.

electrolyte solutions. We consider four adsorption sites: a bridging site (on top of M-M bond), an ontop site (on top of M atom), a face centered cubic ( fcc) site (on top of hollow formed between surface M atoms, with no atom present in the second layer) and a hexagonal close packed (hcp) site (on top of hollow formed between surface M atoms, with C present in the second layer). There are three bridge sites per (1 × 1) cell of M2C, and one site each for the ontop, fcc and hcp sites. In order to reduce the large number of possibilities for functionalization, we assume that functionalized groups are placed in equal amounts on each side of both M2C planes/ surfaces in the structure, in the same adsorption site; for example, we do not consider an H in a fcc site on one side of the MXene sheet and an H in a hcp site on the other side, or unequal amounts of functional groups on each sheet. The details of calculating the free energy of adsorption are given in the Supporting Information.

■

THEORETICAL METHODS All calculations are done using the periodic plane-wave basis set implementation of density functional theory (DFT) within the PWscf program of Quantum Espresso31 via the ASE simulation package.32 We use the Bayesian error estimation functional with van der Waals correction (BEEF-vdW) for exchangecorrelation.33 We use Vanderbilt ultrasoft pseudopotentials34 and the valence configuration 2s22p2 for C, 3s23p64s23d1 for Ti, 4s24p65s14d5 for Mo, 2s22p4 for O, and 1s1 for H. For structures containing Ti, we use a plane wave energy cutoff of 600 eV, and for structures containing Mo we use a plane wave energy cutoff of 700 eV, in order to converge the total energies to less than 0.002 meV per cell. For sampling in k-space, we use Monkhorst−Pack35 k-point meshes. We use a (6 × 6 × 6) mesh for bulk TiC, a (4 × 4 × 1) mesh for the (√2 × √2) TiC(111) surface slab, a (5 × 4 × 5) mesh for bulk β-Mo2C, and a (5 × 4 × 1) mesh for the β-Mo2C(0001) surface; in addition, we use (8 × 8 × 1) meshes for the MXene ionic and cell relaxations and (16 × 16 × 2) meshes for density of states (DOS) calculations. The TiC(111) and β-Mo2C(0001) surface calculations consist of 4.5 unit cells of TiC and β-Mo2C, respectively, in order to recover the bulk electronic properties at the center of the slab; the TiC slabs are terminated by Ti on either side, and the β-Mo2C slabs are terminated by Mo on either side. The DOS are calculated with a 0.1 eV Fermi−Dirac smearing. All atoms and lattice parameters are relaxed until the forces are smaller than 0.01 eV/Å. Single sheet calculations are augmented with ∼12 Å of vacuum. Our simulation cell is one bulk cell of M2C with M = {Mo,Ti}, which consists of two layers of M2C stacked on top of one another. Previous experimental reports of TixCx−1Tx4 and MoxCx−1Tx36 show that these MXenes keep the same parent structure as their associated MAX phase, so we do not allow MxCx−1Tx planes to shift relative to one another; in this manner, we keep the same space group and lattice group as is observed in experiment, which simplifies the calculations and more easily allows comparison with experiment. The space group is P3̅m1, in which the MXene planes exhibit AB stacking (Figure 1a). We study adsorption of O, H, H2, H2O, and OH in the adsorption sites indicated in Figure 1b. We consider these adsorbates because, to date, MXenes are synthesized by wet etching in aqueous solutions and moreover, for electrochemical applications, MXenes are often submerged in water or aqueous

■

EXPERIMENTAL METHODS To synthesize Ti2C, 2 g of Ti2AlC13 were added, over ≈120 s, to 20 mL of 10% aqueous hydrofluoric acid (HF) and held for 18 h at room temperature. To make Mo2C, 2g of Mo2Ga2C powders36 were added, over ≈120 s, to 20 mL of 48−51% aqueous HF solution and held at 55 °C for 6.6 days.36 Both solutions were stirred with a magnetic Teflon coated bar. All mixtures were washed 5 times by adding distilled water, shaking for 1 min, centrifuging at 3500 rpm for 120 s for each cycle and finally decanted. After the last centrifuge, the pH of the supernatant was >6. Lastly, the solid residue, after the last decanting, was mixed with ∼10 mL of distilled water and filtered on a porous membrane (3501 Coated PP, Celgard, USA). In order to dry the synthesized MXene powder, they were heated in a vacuum oven at 140 °C for 18 h. X-ray diffraction (XRD) was done on a Rigaku Smart Lab (Tokyo, Japan) diffractometer using Cu Kα radiation (40 kV and 44 mA) and step scan 0.02°, 3−65° 2 theta range and step time of 1 s.

■

RESULTS AND DISCUSSION First, we calculate the properties of bulk bare Ti2C and Mo2C without functional groups. We find that the in-plane lattice parameters (a-LPs) to be 3.04 and 3.03 Å for Ti2C and Mo2C, respectively, and the out-of plane c-LPs of the structures to be 10.24 and 10.14 Å for Ti2C and Mo2C, respectively. Our calculations show that the a-LP has only a weak dependence on functionalization, with changes less than 0.05 Å due to the exchange of functional group. However, the c-LP has a very strong dependence on functionalization, as discussed in more detail below. Since the exact functionalization of MXenes is B

DOI: 10.1021/acs.jpcc.6b09109 J. Phys. Chem. C XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry C

Figure 2. Calculated Pourbaix diagrams for (a) Ti2C and (b) Mo2C MXenes. Only the lowest free energies at a given potential are labeled by the functionalization.

adsorbed hydrogen (H*) is the dominant functional group. As the applied potential becomes more negative, two monolayers of H* (Ti2CH4 at ca. −0.7 V and Mo2CH4 at ca. −0.4 V) become more stable; for Ti2C, the most stable sites are the fcc and hcp sites, and for Mo2C the most stable sites are the ontop and hcp sites. Finally, for Mo2C three monolayers of H* (Mo2CH6 at ∼ −1.0 V) become stable at even more negative potentials, which consist of the ontop, fcc, and hcp sites. We also find that the bare MXene is never energetically favorable, indicating that bare (non functionalized) MXenes should not exist under normal experimental conditions in aqueous solution or in open air. A previous DFT study investigated the terminations of various MXenes with respect to the H chemical potential μH, showing that at any value of μH the surface is covered with O.27 This is in contrast with our results, which show that the surface is covered with multiple monolayers of H at negative potentials. Although we cannot make a direct comparison with previous results (coverage as a function of μH instead of applied potential), we note that ref 27 did not consider the functionalization of Mo2C and Ti2C with H. We now focus our discussion on the most stable functionalizations at different points of the Pourbaix diagram; that is, Ti2CH2, Ti2CH4, Ti2CO2, Mo2CH4, Mo2CH6, and Mo2CO2. We compare the cohesive energy EC depending on the functionalization of the MXenes, which is defined as

strongly dependent on the synthesis conditions and sample history,7,20,37 we try a variety of adsorbates, including multiple simultaneous monolayers of individual adsorbates and mixed coadsorbates. For example, we consider the coexistence of both OH adsorbed on one site and H adsorbed on another, which corresponds to one layer of H and one layer of OH coverage on different sites. Our findings show that only single layers of O, OH, H2O and H2 were stable, except for the system with one layer of H in addition to a monolayer of O, OH, H2O, and H2. However, two layers of H were stable on Ti2C, and two and three layers of H were stable on Mo2C. Our identified adsorption site for O and OH are in agreement with previous theoretical results.23,25,26,38 However, H2 and H2O were never the most stable functionalizations for Ti2C or Mo2C at any applied potential, and are not referred to again in the text. Although we did not explicitly try adsorbing O2, previous theoretical calculations show that O2 spontaneously dissociates on the bare Ti2C surface and adsorbs on the fcc sites,39 in agreement with our calculations. A special note must be made for H2; H2 is unstable on bare Ti2C, and dissociates into two coadsorbed H atoms. On Mo2C, H2 is stable, but the two dissociated H atoms are lower in energy than a molecularly adsorbed H2. For the remainder of the text, we refer to the functionalized MXenes as M2XT2, with T being the functional group(s) on either side of the MXene. For example, Ti2CH2 refers to Ti2C with a monolayer of H on each side, Ti2CH4 refers to Ti2C with two monolayers of H on each side, and Ti2CH2(OH)2 refers to Ti2C with both a monolayer of H and a monolayer of OH on each side. Understanding of the functionalization of MXenes with respect to the applied potential is important for electrochemical applications. In Figure 2, we show Pourbaix diagrams as a function of applied potential for Ti2C and Mo2C; note that only the most energetically favorable coverage at a given potential is labeled in the diagram. At zero potential, both MXenes prefer to be O covered; this can be understood due to their electronic and stoichiometric mismatch compared to their parent bulk carbides (Supporting Information). For both Ti2C and Mo2C, we find that, at potentials larger than ca. −0.5 V, the coverage is one layer of O (which corresponds to one O per (1 × 1) surface cell), and potentials more negative than ca. −0.5 V,

EC = 2ESheet − E Bulk

(1)

where EBulk is the energy of a layered bulk structure of the MXene, and ESheet is the energy of a single sheet of the MXene; EC for different functional groups is given in Table 1. We find that the cohesion energies of the functionalized MXenes are much smaller than that of the bare MXenes, indicating that the bare MXenes are bound much more strongly that the functionalized MXenes, which are only bound via van der Waals forces. Next, we compare our calculated change in the c-LP with respect to the different functional groups of the MXenes (Figure 3) with the experimental values taken from ref 4 for Ti2C and ref 36, for Mo2C. We find that the functionalization of the system has huge effects on the c-LP, ranging from ∼10 Å for bare Mo2C to ∼24 Å for Mo2CH2(OH)2. There are few C

DOI: 10.1021/acs.jpcc.6b09109 J. Phys. Chem. C XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry C

for a total of two water molecules per computational unit cell. In other words, there is one layer of H2O between the MXene sheets. In Figure 3, the new c-LPs of the structures with intercalated water (square) are compared to the c-LPs without the intercalated water (circles). We find that the inclusion of water increases the c-LP in all cases, independent of the choice of the functional groups. Interestingly, the Pourbaix diagrams with the inclusion of one layer of water do not change the relative ordering of the systems with different functional groups. This indicates that the change in free energy of functionalization is only very weakly affected by the intercalation of water. A few of the functional groups become unstable due to interaction with the intercalated water and the system relaxes to one of the other stable configurations; these unstable systems do not have the diamond data point in Figure 3. We note that this is a proof-of-concept model of water intercalation; a more intricate model would involve more complex water structures and intercalation of water in a larger cell than the (1 × 1) to accommodate hydrogen-bonding effects in the water layers and the possibility of submonolayer water intercalation, which is beyond the scope of this paper. Upon intercalation of water, the calculated lattice constants cLP for Mo2CH4, Mo2CH6 and MoC2O2 are all in good agreement with previous experimental values, leading us to conclude that the measured c-LP of Mo2C corresponds to a system that is intercalated with water, in contrast to the measured value for Ti2C, which does not contain water. This raises the question of why one MXene in experiment contains intercalated water, and the other does not, and whether this is synthesis dependent or an inherent property of different MXenes. Previous experimental studies have shown that intercalation of water in MXenes is greatly dependent on the etching condition; e.g., different c-LPs were reported for Ti3C2 made by etching Ti3AlC2 in hydrofluoric acid (HF) at different etching conditions (different temperatures and etching time).4 Also, Ti3C2 manufactured by etching Ti3AlC2 in lithium fluoride (LiF) and hydrochloric acid (HCl) has very large cLPs (as high as 40 Å), which is attributed to the large amount of intercalated water and ions shown by swelling of the compound after being stored in water; Ti3C2 could be dehydrated and rehydrated, similar to clay.41 A decrease of cLP was found after dehydration and an increase after rehydration.40 Finally, prior reports show that heating Nb2C at 673 K in vacuum leads to a reduction of c-LP by 6.49 Å,19 which was attributed to the complete removal of intercalated water; inelastic neutron scattering measurements show signals of bound water and OH* groups that vanish after heating in vacuum. Interestingly, this same heating procedure at 773 K on Ti3C2 leads to only a reduction of c-LP by 0.58 Å. This was due to the fact that the starting Ti3C2 before vacuum annealing had no intercalated water and vacuum annealing only changed the functional groups.19 What these prior results show is that there are very different c-LPs of MXenes depending on water intercalation and functional groups; therefore, it may not make sense to declare one specific c-LP for a given MXene. Rather, the out-of plane lattice parameter c-LP of the MXene must be understood as a LP at a given state (hydrated or dehydrated, number of intercalated water layers, different functional groups present). Based on our calculations, however, we find that the intercalated water layers do not significantly change the overall ordering of the Pourbaix diagrams and result in fairly uniform shift of all states.

Table 1. Calculated Cohesive Energy of the Ti2C and Mo2C MXenes with Different Functional Groups functionalized MXene

cohesive energy (eV)

bare Ti2C Ti2CH2 Ti2CH4 Ti2CO2 bare Mo2C Mo2CH4 Mo2CH6 Mo2CO2

3.14 0.23 0.24 0.19 2.61 0.24 0.25 0.24

Figure 3. Calculated out-of plane lattice parameter c-LP for (a) Ti2C and (b) Mo2C with different functional groups (circle), and same functional group together with one monolayer of intercalated water (squares). The green circles correspond to the functionalized MXene systems with the lowest free energy at any point on the Pourbaix diagram. The unstable systems with intercalated water have been omitted from the figures. Experimental values are shown as red diamonds. For Ti2C, E1 is adapted from ref 4; E2 and E3 are from this work. For Mo2C, E1 is adapted from ref 36, while E2 and E3 are from this work.

theoretical studies of bulk MXenes; we find a c-LP of 21.78 Å in comparison with 19.47 Å (using a Perdew−Burke−Ernzerhof (PBE) exchange-correlation functional), and 18.49 Å (using a PBE functional with van der Waals interactions).19 In other words, the c-LP value is sensitive to the choice of exchangecorrelation functional, especially with respect to van der Waals interactions. First, we examine the c-LP’s for the functionalized MXenes with no intercalated atoms or molecules between the layers (circles in Figure 3). The c-LP for Ti2CH2, Ti2CH4, and Ti2CO2 are in very good agreement with prior experiment.4 However, the experimental value for Mo2C is much larger than that of either Mo2CH4, Mo2CH6, or Mo2CO2 (the lowest free energy functional groups at some point on the Pourbaix diagram).36 To understand this discrepancy, we intercalate H2O in the functionalized MXene systems, as prior experimental measurements show swelling of MXenes due to presence of water.40 To model this, we intercalate one layer of water, which means there is 1 H2O molecule intercalated per 2 metal atoms, D

DOI: 10.1021/acs.jpcc.6b09109 J. Phys. Chem. C XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry C

molecular intercalation and changes of c-LP. However, after vacuum drying, both dry MXenes show similar (0002) peak around ∼8.8°, which corresponds to one layer of water between the MXene layers. Based on these results, the experimental c-LP of Mo2C using a model of one intercalated layer of water is found to be between 20.1 to 21.0 Å, in decent agreement with the calculated 23.55 Å for Mo2CO2, and the c-LP of Ti2C with one intercalated layer water is 20.1 Å, in good agreement 20.61 Å for Ti2CO2. In general, similarities in the Ti2C and Mo2C results confirm the computational finding that c-LP of MXenes largely depend on the synthesis conditions. The difficulty in getting a sample with no intercalated water can be supported up by the DFT calculations; the free energy of water-intercalated structures in the Pourbaix diagram (i.e., Mo2CO2, Mo2CH4, Mo2CH6, Ti2CO2, Ti2CH2, and Ti2CH4) are always lower than those of the systems without intercalated water. Although we did not obtain any samples with no intercalated water, prior reports show that other MXenes can be dried sufficiently that no water remains in between the MXene sheets.7,43 To further understand the effect of drying the MXene after synthesis, we also performed XRD measurements immediately after synthesis and washing of the Ti2C powder (Figure 4b). A very strong (0002) peak at ∼5.3° is observed (c-LP of 32.1 Å) for the after-washed sample, corresponding to three layers of water between Ti2C sheets. As discussed here, the c-LP of Ti2C was reduced to 26.5 Å by keeping the MXene powder at room temperature for about 2 h (wet Ti2C in Figure 4a,b), changing it to two intercalated water layers. To our knowledge, the spontaneous disappearance of this third intercalated water layer has not been shown in MXenes before. Besides the structural and chemical changes, we also find very interesting electronic structure changes of the functionalized MXenes under applied potential. In Figure 5, we compare the electronic structure of MXenes to their bulk and surface metalcarbide counterparts. The DOS of bulk TiC and bulk β-Mo2C are given in Figure 5a,b, the DOS of TiC(111) and βMo2C(0001) surfaces are given in Figure 5c,d, respectively. These should be compared to the DOS of the bare MXene structures given in Figure 5e,f. We find two main bands near the Fermi level; the fully occupied C-localized band at ca. −4 eV for Ti2C and ca. −6 eV for Mo2C, and a much wider band that is partially filled. When it comes to the DOS of the functionalized MXenes (see Figure 5g,h, for Ti2CH2 and Mo2CH2, respectively), we find that they are quite similar to that of the bare MXenes, except for the change in Fermi level due to the charge transfer from the MXene to the adsorbed H. The direction of this charge transfer is unusual, in that these systems, H acts as an acceptor, rather than a donor. Figure 5i shows the DOS of Ti2CO2, which is remarkably different than that of the bare Ti2C, due to the strong hybridization between the adsorbed O and both the Ti and C of the MXene. Although most of the functionalized MXenes are computationally found to be metals, this O-functionalized Ti2C is a small gap insulator, consistent with prior theoretical calculations of single-sheet Ti2CO2.23,44 On the other hand the O-functionalized Mo2C (Mo2CO2) shown in Figure 5j, is superficially similar to that of Ti2CO2, except for the position of the Fermi level; Mo2CO2 also does not become insulating. One important finding is the change in the electronic structure for Ti2C with O and H as functional groups, corresponding to Ti2CO2 and Ti2CH2, respectively, at an applied potential ca. −0.6 eV. We find that Ti2CO2 is insulating,

To achieve a better understanding of the effect of MXene synthesis and water intercalation, we synthesized both Ti2C and Mo2C MXenes with a similar etching method, using HF. For both Ti2C and Mo2C, we perform X-ray diffraction (XRD) measurements on two different samples: the “wet” sample, for which we perform the XRD ∼ 2 h after the last step of washing and filtration, and the “dry” sample, for which the XRD is performed after 18 h at 140 °C under vacuum (Figure 4). Both

Figure 4. XRD patterns of (a) the “wet” and “dry” layered Mo2C and Ti2C samples, and (b) the “wet” and “dry” Ti2C compared with XRD pattern of the Ti2C powder immediately (∼5 min) after washing and filtration labeled as “after wash”.

wet Ti2C and Mo2C MXenes had a peak at ∼6.5°, which, assuming a water diameter of 2.8 Å, would correspond to two layers of water intercalated between each MXene layer. The wet Mo2C XRD pattern has also a second peak (c-LP of 21.0 Å), possibly corresponding to one layer of intercalated water, consistent with prior experimental results for Mo2C.42 The appearance of the second peak in only “wet” Mo2C and not “wet” Ti2C shows that the MXene chemistry can also affect E

DOI: 10.1021/acs.jpcc.6b09109 J. Phys. Chem. C XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry C

Figure 5. Calculated DOS of (a) bulk TiC, (b) bulk β-Mo2C, (c) clean TiC(111) surface, (d) clean Mo2C(0001) surface, (e) bare Ti2C MXene, (f) bare Mo2C MXene, and functionalized MXenes: (g) Ti2CH2, (h) Mo2CH2, (i) Ti2CO2, (j) Mo2CO2, (k) Ti2C(OH)2, and (l) Mo2C(OH)2. Ti2CO2 appears to be metallic due to Fermi−Dirac smearing. F

DOI: 10.1021/acs.jpcc.6b09109 J. Phys. Chem. C XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry C

Figure 6. Calculated DOS of (a) Mo2CO2 MXene and (b) Mo2CO2 MXene with intercalated water.

whereas Ti2CH2 is metallic; we find a gap of 0.39 eV, which is in agreement with previous GGA exchange correlation calculations and should be compared to the hybrid functional calculations which show an even larger gap of 0.88 eV.38 In other words, this shows that Ti2C has a metal−insulator transition that can be induced by an applied voltage, but the transition itself is mediated by adsorbates. This raises the question of what other phase transitions can be induced in MXenes by different functional groups, and what external means can be used to control functionalization. As a representative example, we show the changes in Mo2CO2 electronic structure upon intercalation of water (Figure 6a and 6b). We find that, while the DOS of water broadens when it is intercalated in Mo2CO2, the DOS of the MXene is essentially identical. This indicates that any change in electronic structure is limited to that of intercalated water, with minimal effect on the MXene itself. This explains why the relative positioning of the functionalized MXenes in the Pourbaix diagrams with and without intercalated water are essentially the same; electronically, the MXenes are indifferent to the presence of water, and any change in the MXene itself is limited to a larger c-LP. Hence, the metal−insulator transition is robust and is not affected by the intercalation of water. At the same time, this transition and change in the oxidation state of titanium upon change of the surface termination may contribute to capacitive energy storage by MXene in aqueous electrolytes.45

transitions on a wide variety of MXenes, which can have farreaching implications on their physical and chemical properties for electronic and electrochemical applications.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jpcc.6b09109. Additional information on the construction of the Pourbaix diagrams and on adsorption energies (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Aleksandra Vojvodic: 0000-0002-5584-6711 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported by the Laboratory-Directed Research and Development program funded through the SLAC National Accelerator Laboratory. We thank Jens K. Nørskov for many insightful discussions.

■

■

CONCLUSION In conclusion, we used DFT to calculate the atomic structure and electronic properties of Ti2C and Mo2C MXenes with different functional groups relevant in aqueous media. We calculated the changes in free energy of functionalization, and found that there were different stable sites for H, O, OH, H2, and H2O. The c-LP varies significantly with the functionalization of the MXene, and the c-LP is greatly increased upon intercalation of water. Due to the large variation in the c-LPs reported in experimental results, especially on the effects of hydration of these materials, leads us to propose a question on what the measured c-LP for a given MXene is. We find that the answer is highly dependent on the exact hydration conditions, because the c-LP is very sensitive to these conditions. Under zero applied potential, both sets of MXenes are found to be functionalized by one layer of O; bare MXenes are never stable, regardless of the applied potential. An adsorbate-mediated metal−insulator transition was discovered for Ti2C; applying a voltage of ca. −0.6 V changes the system from insulating Ti2CO2 to metallic Ti2CH2. This work paves the way for further explorations of other adsorbate-mediated phase

REFERENCES

(1) Naguib, M.; Mochalin, V. N.; Barsoum, M. W.; Gogotsi, Y. MXenes: A New Family of 2D Materials. Adv. Mater. 2014, 26, 992− 1004. (2) Lei, J.-C.; Zhang, X.; Zhou, Z. Recent Advances in MXene: Preparation, Properties, and Applications. Front. Phys. 2015, 10, 276− 286. (3) Naguib, M.; Kurtoglu, M.; Presser, V.; Lu, J.; Niu, J.; Heon, M.; Hultman, L.; Gogotsi, Y.; Barsoum, M. W. Two-Dimensional Nanocrystals Produced by Exfoliation of Ti3AlC2. Adv. Mater. 2011, 23, 4248−4253. (4) Naguib, M.; Mashtalir, O.; Carle, J.; Presser, V.; Lu, J.; Hultman, L.; Gogotsi, Y.; Barsoum, M. W. Two-Dimensional Transition Metal Carbides. ACS Nano 2012, 6, 1322−1331. (5) Naguib, M.; Halim, J.; Lu, J.; Cook, K. M.; Hultman, L.; Gogotsi, Y.; Barsoum, M. W. New Two-Dimensional Niobium and Vanadium Carbides as Promising Materials for Li-Ion Batteries. J. Am. Chem. Soc. 2013, 135, 15966−15969. (6) Naguib, M.; Gogotsi, Y. Synthesis of Two-Dimensional Materials by Selective Extraction. Acc. Chem. Res. 2015, 48, 128−135. (7) Wang, H.-W. W.; Naguib, M.; Page, K.; Wesolowski, D. J.; Gogotsi, Y. Resolving the Structure of Ti3C2Tx MXenes through Multilevel Structural Modeling of the Atomic Pair Distribution Function. Chem. Mater. 2016, 28, 349−359. G

DOI: 10.1021/acs.jpcc.6b09109 J. Phys. Chem. C XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry C (8) Er, D.; Li, J.; Naguib, M.; Gogotsi, Y.; Shenoy, V. B. Ti3C2 MXene as a High Capacity Electrode Material for Metal (Li, Na, K, Ca) Ion Batteries. ACS Appl. Mater. Interfaces 2014, 6, 11173−11179. (9) Gogotsi, Y. What Nano Can Do for Energy Storage. ACS Nano 2014, 8, 5369−5371. (10) Lukatskaya, M. R.; Mashtalir, O.; Ren, C. E.; Dall’Agnese, Y.; Rozier, P.; Taberna, P. L.; Naguib, M.; Simon, P.; Barsoum, M. W.; Gogotsi, Y. Cation Intercalation and High Volumetric Capacitance of Two-Dimensional Titanium Carbide. Science 2013, 341 (6153), 1502− 1505. (11) Hu, Q.; Sun, D.; Wu, Q.; Wang, H.; Wang, L.; Liu, B.; Zhou, A.; He, J. MXene: A New Family of Promising Hydrogen Storage Medium. J. Phys. Chem. A 2013, 117, 14253−14260. (12) Guo, Z.; Zhou, J.; Zhu, L.; Sun, Z. MXene: A Promising Photocatalyst for Water Splitting. J. Mater. Chem. A 2016, 4, 11446− 11452. (13) Seh, Z. W.; Fredrickson, K. D.; Anasori, B.; Kibsgaard, J.; Strickler, A. L.; Lukatskaya, M. R.; Gogotsi, Y.; Jaramillo, T. F.; Vojvodic, A. Two-Dimensional Transition Metal Carbides (MXene) as an Efficient Electrocatalyst for Hydrogen Evolution. ACS Energy Lett. 2016, 1, 589−594. (14) Mashtalir, O.; Naguib, M.; Mochalin, V. N.; Dall’Agnese, Y.; Heon, M.; Barsoum, M. W.; Gogotsi, Y. Intercalation and Delamination of Layered Carbides and Carbonitrides. Nat. Commun. 2013, 4, 1716−1722. (15) Ma, T. Y.; Cao, J. L.; Jaroniec, M.; Qiao, S. Z. Interacting Carbon Nitride and Titanium Carbide Nanosheets for HighPerformance Oxygen Evolution. Angew. Chem., Int. Ed. 2016, 55, 1138−1142. (16) Gao, Y.; Wang, L.; Li, Z.; Zhou, A.; Hu, Q.; Cao, X. Preparation of MXene-Cu2O Nanocomposite and Effect on Thermal Decomposition of Ammonium Perchlorate. Solid State Sci. 2014, 35, 62−65. (17) Xie, X.; Chen, S.; Ding, W.; Nie, Y.; Wei, Z. An Extraordinarily Stable Catalyst: Pt NPs Supported on Two-Dimensional Ti3C2X2 (X = OH, F) Nanosheets for Oxygen Reduction Reaction. Chem. Commun. 2013, 49, 10112−10114. (18) Li, X.; Fan, G.; Zeng, C. Synthesis of Ruthenium Nanoparticles Deposited on Graphene-like Transition Metal Carbide as an Effective Catalyst for the Hydrolysis of Sodium Borohydride. Int. J. Hydrogen Energy 2014, 39, 14927−14934. (19) Xie, Y.; Naguib, M.; Mochalin, V. N.; Barsoum, M. W.; Gogotsi, Y.; Yu, X.; Nam, K. W.; Yang, X. Q.; Kolesnikov, A. I.; Kent, P. R. C. Role of Surface Structure on Li-Ion Energy Storage Capacity of TwoDimensional Transition-Metal Carbides. J. Am. Chem. Soc. 2014, 136, 6385−6394. (20) Halim, J.; Cook, K. M.; Naguib, M.; Eklund, P.; Gogotsi, Y.; Rosen, J.; Barsoum, M. W. X-Ray Photoelectron Spectroscopy of Select Multi-Layered Transition Metal Carbides (MXenes). Appl. Surf. Sci. 2016, 362, 406−417. (21) Tang, Q.; Zhou, Z.; Shen, P. Are MXenes Promising Anode Materials for Li Ion Batteries? Computational Studies on Electronic Properties and Li Storage Capability of Ti3C2 and Ti3C2X2 (X = F, OH) Monolayer. J. Am. Chem. Soc. 2012, 134 (40), 16909−16916. (22) Enyashin, A. N.; Ivanovskii, A. L. Structural and Electronic Properties and Stability of MXenes Ti2C and Ti3C2 Functionalized by Methoxy Groups. J. Phys. Chem. C 2013, 117 (26), 13637−13643. (23) Khazaei, M.; Arai, M.; Sasaki, T.; Chung, C.-Y.; Venkataramanan, N. S.; Estili, M.; Sakka, Y.; Kawazoe, Y. Novel Electronic and Magnetic Properties of Two-Dimensional Transition Metal Carbides and Nitrides. Adv. Funct. Mater. 2013, 23 (17), 2185− 2192. (24) Lai, S.; Jeon, J.; Jang, S. K.; Xu, J.; Choi, Y. J.; Park, J.-H.; Hwang, E.; Lee, S. Surface Group Modification and Carrier Transport Properties of Layered Transition Metal Carbides (Ti2CTx, T: -OH, -F and -O). Nanoscale 2015, 7, 19390−19396. (25) Khazaei, M.; Arai, M.; Sasaki, T.; Ranjbar, A.; Liang, Y.; Yunoki, S. OH-Terminated Two-Dimensional Transition Metal Carbides and Nitrides as Ultralow Work Function Materials. Phys. Rev. B: Condens. Matter Mater. Phys. 2015, 92, 075411−075420.

(26) Khazaei, M.; Arai, M.; Sasaki, T.; Estili, M.; Sakka, Y. TwoDimensional Molybdenum Carbides: Potential Thermoelectric Materials of the MXene Family. Phys. Chem. Chem. Phys. 2014, 16, 7841−7849. (27) Ashton, M.; Mathew, K.; Hennig, R. G.; Sinnott, S. B. Predicted Surface Composition and Thermodynamic Stability of MXenes in Solution. J. Phys. Chem. C 2016, 120, 3550−3556. (28) 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.; et al. Effect of Metal Ion Intercalation on the Structure of MXene and Water Dynamics on Its Internal Surfaces. ACS Appl. Mater. Interfaces 2016, 8, 8859−8863. (29) Ivanovskii, A. L.; Medvedeva, N. I. Electronic Structure of Hexagonal Ti3AlC2 and Ti3AlN2. Mendeleev Commun. 1999, 9, 36−38. (30) Medvedeva, N. I.; Novikov, D. L.; Ivanovsky, A. L.; Kuznetsov, M. V.; Freeman, A. J. Electronic Properties of Ti3SiC2-Based Solid Solutions. Phys. Rev. B: Condens. Matter Mater. Phys. 1998, 58, 16042− 16050. (31) Giannozzi, P.; Baroni, S.; Bonini, N.; Calandra, M.; Car, R.; Cavazzoni, C.; Ceresoli, D.; Chiarotti, G. L.; Cococcioni, M.; Dabo, I.; et al. QUANTUM ESPRESSO: A Modular and Open-Source Software Project for Quantum Simulations of Materials. J. Phys.: Condens. Matter 2009, 21, 395502−395520. (32) Bahn, S. R. S. R.; Jacobsen, K. W. K. W. An Object-Oriented Scripting Interface to a Legacy Electronic Structure Code. Comput. Sci. Eng. 2002, 4, 56−66. (33) Wellendorff, J.; Lundgaard, K. T.; Møgelhøj, A.; Petzold, V.; Landis, D. D.; Nørskov, J. K.; Bligaard, T.; Jacobsen, K. W. Density Functionals for Surface Science: Exchange-Correlation Model Development with Bayesian Error Estimation. Phys. Rev. B: Condens. Matter Mater. Phys. 2012, 85, 235149−235171. (34) Vanderbilt, D. Soft Self-Consistent Pseudopotentials in a Generalized Eigenvalue Formalism. Phys. Rev. B: Condens. Matter Mater. Phys. 1990, 41, 7892−7895. (35) Monkhorst, H. J.; Pack, J. D. Special Points for Brillouin-Zone Integrations. Phys. Rev. B 1976, 13, 5188−5192. (36) Meshkian, R.; Näslund, L.-Å.; Halim, J.; Lu, J.; Barsoum, M. W.; Rosen, J. Synthesis of Two-Dimensional Molybdenum Carbide, Mo2C, from the Gallium Based Atomic Laminate Mo2Ga2C. Scr. Mater. 2015, 108, 147−150. (37) Hope, M. A.; Forse, A. C.; Griffith, K. J.; Lukatskaya, M. R.; Ghidiu, M.; Gogotsi, Y.; Grey, C. P. NMR Reveals the Surface Functionalisation of Ti3C2 MXene. Phys. Chem. Chem. Phys. 2016, 18, 5099−5102. (38) Xie, Y.; Kent, P. R. C. Hybrid Density Functional Study of Structural and Electronic Properties of Functionalized Tin+1Xn (X = C, N) Monolayers. Phys. Rev. B: Condens. Matter Mater. Phys. 2013, 87, 235441−235450. (39) Gan, L.-Y.; Huang, D.; Schwingenschlö gl, U. Oxygen Adsorption and Dissociation during the Oxidation of Monolayer Ti2C. J. Mater. Chem. A 2013, 1 (43), 13672. (40) 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, 78−81. (41) 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, 3507−3514. (42) Halim, J.; Kota, S.; Lukatskaya, M. R.; Naguib, M.; Zhao, M. Q.; Moon, E. J.; Pitock, J.; Nanda, J.; May, S. J.; Gogotsi, Y.; et al. Synthesis and Characterization of 2D Molybdenum Carbide (MXene). Adv. Funct. Mater. 2016, 26, 3118−3127. (43) Mashtalir, O.; Lukatskaya, M. R.; Kolesnikov, A. I.; RaymundoPiñero, E.; Naguib, M.; Barsoum, M. W.; Gogotsi, Y. The Effect of Hydrazine Intercalation on the Structure and Capacitance of 2D Titanium Carbide (MXene). Nanoscale 2016, 8, 9128−9133. (44) Li, X.; Dai, Y.; Ma, Y.; Liu, Q.; Huang, B. Intriguing Electronic Properties of Two-Dimensional MoS2/TM2CO2 (TM = Ti, Zr, or Hf) Hetero-Bilayers: Type-II Semiconductors with Tunable Band Gaps. Nanotechnology 2015, 26, 135703−135709. H

DOI: 10.1021/acs.jpcc.6b09109 J. Phys. Chem. C XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry C (45) 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, 1500589−1500592.

I

DOI: 10.1021/acs.jpcc.6b09109 J. Phys. Chem. C XXXX, XXX, XXX−XXX