Adsorption of Li2O2 , Na2O2 and NaO2 on TiC(111) Surface for Metal

Subscriber access provided by University of Sussex Library

C: Energy Conversion and Storage; Energy and Charge Transport 2

2

2

2

2

Adsorption of LiO , NaO and NaO on TiC(111) Surface for Metal-Air Rechargeable Batteries: A Theoretical Study Keren Raz, Polina Tereshchuk, Diana Golodnitsky, and Amir Natan J. Phys. Chem. C, Just Accepted Manuscript • DOI: 10.1021/acs.jpcc.8b01983 • Publication Date (Web): 02 Jul 2018 Downloaded from http://pubs.acs.org on July 3, 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

The Journal of Physical Chemistry

Adsorption of Li2O2, Na2O2 and NaO2 on TiC(111) Surface for Metal-Air Rechargeable Batteries: A Theoretical Study †

Keren Raz,

‡

Polina Tereshchuk,

†School ‡Department ¶The

Diana Golodnitsky,

†

and Amir Natan

∗,‡,¶

of Chemistry, Tel Aviv University, Israel

of Physical Electronics, Tel Aviv University, Israel 69978

Sackler Center for Computational Molecular and Materials Science, Tel-Aviv University, Tel-Aviv 69978, Israel

E-mail: [email protected]

Abstract We analyze, with Density Functional Theory (DFT) calculations, the adsorption energies of Li2 O2 , Na2 O2 and NaO2 on clean and oxygen passivated TiC (111) surfaces. We show, that after deposition of two molecular layers of alkali metal oxides, the initial state of the TiC surface becomes unimportant for the adsorption energy and that all adsorption energies approach their native crystal values. The structure of the adsorbed molecular layers is analyzed and compared to their native oxide crystal structure. Finally, we discuss the similarities and dierences of Li peroxide and Na oxides adsorption at the electrode surface. Introduction

her the discharge products are formed at the cathode surface, or in the solution, and the prevalence of unwanted side reactions. 11,12 The performance and rechargeability of metal-air cells strongly depends on the positive electrode material, where oxygen reduction and evolution reactions take place. A suitable cathode material for an aprotic alkali metal/air cell should have sucient electronic conductivity; low density; high stability, over the operating voltage of the cathode, towards nucleophilic attack by LiO2 and O2 2− ; low cost; and non-toxicity. 11 Carbon has been the material of choice for the porous cathode. It is known that carbon oxidizes above 4V versus Li. However, more importantly, carbon raises problems that impede its use in Li/O2 cells. Carbon decomposes during oxidation of Li2 O2 on charging above 3V as a result of the attack by intermediates of Li2 O2 oxidation and it actively promotes electrolyte decomposition on

Rechargeable metal-air batteries are widely considered to be the next generation highenergy-density electrochemical storage devices. The intense interest in Li-air and Na-air batteries stems from their high theoretical specic energy, which exceeds that of lithium-ion batteries. 111 The high specic energy densities of metal-air batteries result from the use of alkali metals as anodes and ambient-air oxygen, as cathode materials. The discharge process in Li-air batteries involves an electrochemical reaction between Li+ ions and O2 to form Li2 O2 onto the cathode surface. During charging, oxidation of lithium peroxide, followed by generation of oxygen gas occurs. The optimization of charge and discharge processes involves the careful design of the cathode material, the electrolytes, solvents, and mediators. The selection of solvents may aect whet-

ACS Paragon Plus Environment

1

The Journal of Physical Chemistry 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

gated by Wang et al . 21,22 using density functional theory. In this work, we performed ab-initio DFT simulations of nLi2 O2 , nNa2 O2 and nNaO2 (n = 1−6) molecular adsorption on the pristine and oxidized TiC(111) surfaces. We investigated structural and adsorption energy trends as a function of molecular density on both pristine and oxygen passivated TiC(111) surfaces. Additional properties such as the work function change and charge transfer are also analyzed. We show the following main ndings:

discharge and charge, rendering it unsuitable for aprotic Li/O2 cells. 13,14 Titanium carbide can overcome some of the disadvantages of carbon. 14 Recently, nanocrystalline TiC has been shown to be an ecient gas diusion cathode. 14 TiC has good metallic conductivity. Furthermore, the formation of a passivating monolayer of oxygen on the TiC surface was reported to be vital to the system's cycleability. This is because the oxygen passivated surface becomes less reactive in comparison to carbon. 1416 While the Li-air battery has the highest theoretical energy density, 811,17 the low availability of lithium might lead to future depletion. In contrast to lithium, there are abundant sodium sources in both the earth's crust (2.3%) and in the oceans (1.1%). 18 Moreover, the production of sodium is cheaper than that of lithium. Na-air battery systems have a lower theoretical specic energy density compared to Li-air battery systems (1605 or 1108 Wh kg-1 considering Na2 O2 or NaO2 as discharge products, respectively). However, Na-air batteries also demonstrate lower charge/discharge overpotential, which may result in better durability. 3 Therefore, Na-O2 battery oers an interesting alternative to the Li-O2 battery. Even though sodium and lithium share many physicochemical properties, the chemistry of the Liair and Na-air cells is not the same. While sodium forms stable sodium superoxide, lithium superoxide is thermodynamically unstable. 18 It is expected that both sodium peroxide and superoxide would be formed under dierent physicochemical conditions, however kinetic factors, temperature and oxygen pressure, the type of support and catalysts may stabilize a certain phase over the other. 19,20 A key process of lithium-air and sodium-air systems is the possible adsorption of reaction products and intermediates at the cathode surface. The surface adsorption energy and product growth (e.g., Na2 O2 ) at the surface, can be compared to the energetics of product formation in solution, where both kinetics and Li+ and Na+ energetics can aect the nal result. Recently, the adsorption of Li2 O2 clusters on the TiC(111) surface was theoretically investi-

• The dierent Alkali Metal Oxides (AMOs) show similar trends in adsorption energies. • The oxygen passivation layer reduces signicantly the adsorption energy relative to the clean surface. • The rst two AMO layers, adsorbed on the clean surface, act also as a passivation layer and reduce signicantly the adsorption energy of the next layer AMO molecules. Finally, the AMOs' adsorption energies at both the oxygen passivated and the clean TiC(111) surfaces approach the values of the respective AMOs' binding energies on their native crystal surface.

Theoretical

Approach

and

Computational Details We performed spin-polarized density functional theory (DFT) calculations with the generalized gradient approximation (GGA) 23 functional as proposed by Perdew, Burke, and Ernzerhof (PBE). 24 We also analyze, in the supporting information (SI), the eects of Hubbard correction (PBE+U) and Van der Waals (VdW) correction. We show that both corrections do not signicantly change our conclusions. We therefore use the PBE functional throughout all the calculations that are shown here. We used projected augmented wave (PAW) pseudopotentials, 25,26 as implemented in


2

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


the Vienna Ab-initio Simulations Package (VASP). 27,28 We used a Monkhorst-Pack k-point scheme 29 with k-point meshes of 11 × 11 × 11, 11 × 11 × 1 and gamma point for the bulk, surface and gas-phase molecules, respectively, and a cuto energy of 500 eV throughout. The structures were considered as optimized when the atomic forces were smaller than 0.02 eV per Å and a total energy convergence of 10−6 eV was achieved. To model the TiC, Li2 O2 and Na2 O2 bulks and surfaces we used the AFLOW 30 package and the ICSD database. 31 The TiC(111) surface was constructed by applying the repeated slab model with 8 layers, 2 × 2 hexagonal surface unit cell and 21 Å vacuum region, which was found to be sucient. Thicker slabs of TiC were shown to give similar results and are discussed in the SI. We found that the Ti-terminated TiC(111) surface is more stable than the C-terminated TiC(111) surface, which is in agreement with previous results. 21,32,33 Thus, we use the Titerminated TiC(111) surface throughout. Our calculated surface energy for the Ti-terminated TiC(111) surface, 201.8 meV/Å2 , is slightly lower than the result reported by Wang et al. (208 meV/Å2 ). 21 The oxidized TiC(111) surface was modeled by covering the Ti-terminated TiC(111) surface by oxygen atoms at the positions of the next imaginary layer of carbon atoms layer. We used the Zur and McGil algorithm 34 to nd common surface cells for the TiC surface and the dierent AMO native crystals (001) surfaces. The 2 × 2 TiC hexagonal surface cell, used also by Wang et al., 21 was found to have a good enough t to all AMO crystals as shown in the results section. In this cell, a single crystalline layer accommodates two molecules for Li2 O2 and NaO2 but only 1.5 molecules for the Na2 O2 crystal. We have performed, for Na2 O2 , additional calculations with a rectangular 4 × 2 TiC(111) surface cell which has a twice larger area and can accommodate 3 Na2 O2 molecules in a single crystalline layer. Those additional calculations completely agree with the hexagonal cell calculations and are shown in the SI. In all of the calculations we placed the mole-

cules on one side of the slab and employed dipole correction in order to obtain accurate total energies. We allowed the molecules to relax along with four slab layers, while the remaining bottom layers were frozen. We simulated, as an initial guess, dierent nLi2 O2 , nNa2 O2 , and nNaO2 (n = 16) structures on the pristine and oxidized TiC(111) surfaces by applying the following procedure: (i) Ab-initio MD simulations with the Nosé thermostat and slowly lowering the temperature from 300 K (and 500 K) to 0 K for 30 ps. The initial structural models were built by putting the optimized gas-phase AMO molecules at random positions at about 3 − 4 Å above the pristine and oxidized TiC(111) surfaces. (ii) We also took the lowest energy nLi2 O2 structure and used it as a starting point for geometrical relaxation of the respective nNa2 O2 system, and vice-versa. This was helpful in some particular cases to nd lower energy structures that were missed by the MD procedure.

Results

TiC, Li2O2, Na2O2 and NaO2 bulks, common cells and gas-phase molecules Titanium carbide (TiC) bulk has a face centered cubic (fcc) structure with the Oh5 space group symmetry. Our calculated TiC equilibrium lattice constant, a = 4.337 Å, is close to the thermal expansion experimental value (4.318 Å) 35 and recent PBE calculations (4.333 Å). 22 Li2 O2 and Na2 O2 bulks crystallize in hexagonal P 63/mmc and P 62m space groups, respectively. The lattice parameters we obtained are 3.158 Å and 7.686 Å for Li2 O2 and 6.195 Å and 4.472 Å for Na2 O2 , in good agreement with the experimental values 36 and previous theoretical GGA results. 22,37,38 NaO2 can crystallize in a pyrite structure in P a3 space group (between 196 and 223 K), and in F m3m space group (above 223 K), which corresponds to a pyrite structure with a disorder of O2 orientation, as obtained by powder and single crystal X-ray diraction methods. 39 Our calcula-


3


Page 4 of 15

0

ted structure corresponds to the ordered pyrite structure in P a3 space group with a calculated lattice parameter of 5.526 Å, which is in a good agreement with other PBE calculations (5.509 Å) 40 and experiment. 39 The lattice parameters calculated by us along with the corresponding literature values can be found in the SI. For the common cells created, Li2 O2 (001) on TiC(111), Na2 O2 (001) on TiC(111), and NaO2 (001) on TiC(111), we calculated a mi2Ω ) ∗ 100%, as dest factor η = (1 − Ω+A 22 ned by Wang et al., and an area ratio α= A , where Ω is the surface area of TiC(111) Ω 2 (32.597 Å ) and A is the original surface area of Li2 O2 (001) (34.543 Å2 ), Na2 O2 (001) (33.237 Å2 ) and NaO2 (001) (30.532 Å2 ). We found a compression and a relatively small η of 2.90% and 0.97% and α of 1.06 and 1.02 for the Li2 O2 (001)/TiC(111) and Na2 O2 (001)/TiC(111), respectively, and an expansion with η of −3.24% and α of 0.94 for the NaO2 (001)/TiC(111) cell. To obtain the lowest energy structures of the molecules in the gas-phase, we selected reasonable geometries, such as linear, square, triangle and rhombus for Li2 O2 and Na2 O2 molecules, and linear and triangle for the NaO2 molecule to be optimized. We found that the lowest energy structures for the Li2 O2 and Na2 O2 molecules are the planar rhombus, in which two O and two Li (Na) atoms are at opposite corners with Li−O (Na−O) bond lengths of 1.74 Å (2.08 Å) and O−O bond lengths of 1.59 Å (1.60 Å). The gas-phase structure of NaO2 corresponds to the isosceles triangle with Na−O and O−O bond lengths of 2.13 Å and 1.37 Å, respectively. The structures are in a good agreement with the results reported by Lau et al. 41 and Arcelus et al. 42 A gure that shows the lowest energy geometry for the gas-phase molecules is presented in the SI.

-2 NaO2 Na2O2

-4 Li2O2

-6 Li2O2/TiC(111)

-8

Li2O2/TiC(111)-O Na2O2/TiC(111) Na2O2/TiC(111)-O

-10

NaO2/TiC(111) NaO2/TiC(111)-O

-12

0.03

0.06

0.09

0.12

0.15

0.18

0.21

0.24

Figure 1: The adsorption energy (Eq. 1) of nLi2 O2 , nNa2 O2 and nNaO2 molecules on TiC(111) and TiC(111)-O surfaces vs molecular density. Triangles correspond to the AMOs' self-binding energies. The dashed vertical lines correspond to the molecular densities that lead to a single molecular layer, i.e., 2 molecules of Li2 O2 and NaO2 and 1.5 molecules of Na2 O2 in the TiC surface cell.

n*mol/TiC

n Ead1 = (Etot

molecule TiC − nEtot − Etot )/n (1)

where n is the number of molecules in the syn*mol/TiC TiC molecule correstem and Etot and Etot , Etot spond to the total energies of the lowest energy structures for: nM2 O2 and nMO2 molecules on the non-oxidized and oxidized TiC(111) slabs, a single gas-phase molecule, and the TiC slabs, respectively. In order to evaluate the contribution of the recently adsorbed molecule, we calculated also: n*mol/TiC

(n-1)*mol/TiC

n Ead2 = Etot − Etot n−1 n . = nEad1 − (n − 1)Ead1

molecule − Etot (2)

n−1 n Here Ead1 and Ead1 are the adsorption energies of the n and n − 1 molecules on the TiC(111) surfaces as dened in Eq. 1 and we 0 dene Ead1 = 0. We calculated the adsorption energies of systems with n = 1 to n = 6 molecules corresponding to molecular surface densities of ∼ 0.03[1/Å2 ] (3 × 1014 [1/cm2 ]) to ∼ 0.18[1/Å2 ]

Adsorption energies We calculated the adsorption energy per molecule as:


4

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


(1.8 × 1015 [1/cm2 ]). Those results are shown in Figure 1, for comparison, we show also the AMO's self-binding energy, which we dened as the energy gained on average when an AMO molecule is adsorbed at the respective AMO crystal surface. We found similar trends in the adsorption energies (Ead1 ) of nLi2 O2 , nNa2 O2 and nNaO2 molecules on the clean TiC(111) surface. For example, Ead1 decreases with increasing coverage of molecules on the clean TiC(111) surface, from −9.73 eV to −5.84 eV for 1 Li2 O2 and 6 Li2 O2 , respectively. This trend can be explained by the following argument: The larger value of Ead1 at n=1,2 is due to the direct binding of the molecules with the highly reactive Ti surface atoms. When increasing the number of molecules in the system, n Ead1 tends to decrease as the additional molecules are adsorbed on top of the rst molecular layer and are not in the direct contact with the TiC surface. Since the interaction between molecular layers is weaker than the interaction of the molecules with the clean and reactive TiC surface, a smaller adsorption energy is obtained. On the oxidized surface, the Ead1 values follow a dierent trend and are in the range of −1.91 eV to −3.02 eV for nLi2 O2 , from −2.91 eV to −3.27 eV for nNa2 O2 and −1.75 eV to −1.86 eV for nNaO2 systems. This behavior demonstrates that the oxygen layer eectively passivated the TiC surface. The AMO molecules now interact with the oxygen layer and not with the reactive Ti atoms of the pristine surface, leading to signicantly lower adsorption energies for the rst AMO layer. Equation 2 gives the adsorption energy of the recently added molecule; these data are shown in Figure 2. The adsorption of one and two molecules to the pristine TiC(111) surface yields larger energies because of direct binding with the reactive TiC(111) surface, e.g. Ead2 = −9.73 eV for 1 Li2 O2 , while the adsorption of the next molecules require less energy, e.g. −6.68 eV for 3 Li2 O2 , as the molecules are above the rst bilayer. Finally, at n = 5 and 6 the adsorption energy gain approaches a plateau, namely, −2.56 eV for 6 Li2 O2 , and tends to AMO's self binding energy. This is to be expected as

a result of the increasing role of the intermolecular binding. Another important observation from Figure 2 is that after two molecular layers (density of 0.12 [1/Å2 ]), the adsorption energy for additional molecules is almost the same for the clean and oxidized surfaces for all the three AMO species. 0 -4 Li2O2/TiC(111)

-8

Li2O2/TiC(111)-O

-12 0.03

0.06

0.09

0.12

0.15

0.18

0.21

0.24

0 -4 Na2O2/TiC(111)

-8

Na2O2/TiC(111)-O

-12 0.03

0.06

0.09

0.12

0.15

0.18

0.21

0.24

0 -4 NaO2/TiC(111)

-8

NaO2/TiC(111)-O

-12 0.03

0.06

0.09

0.12

0.15

0.18

0.21

0.24

Figure 2: The adsorption energy (Eq. 2) of nLi2 O2 , nNa2 O2 and nNaO2 on TiC(111) and TiC(111)-O surfaces per molecular density. Triangles correspond to the AMOs' self-binding energies.

Geometry of adsorbed AMO layers Figures 3, 4 and 5 present the geometry of the lowest energy structures of nM2 O2 (M = Li, Na) and nMO2 (M = Na) molecules on the pristine and oxidized TiC(111) surfaces. We also present in table 1 the following structural parameters: the molecular surface density, ρmol , calculated as the number of molecules per surface area, the minimal Ti−O, M−O and M−M bond lengths, and the molecular layer thickness, D, calculated as the vertical distance between the M atoms nearest and farthest from the surface. The lowest energy structures of the adsorbed AMO molecules at n = 1, 2 (with a molecular density of ∼ 0.03 [1/Å2 ] and 0.06 [1/Å2 ] respectively) on the pristine TiC(111) surface form O and M atomic layers. In this structure


5

The Journal of Physical Chemistry 2Li2 O2 /TiC(111)

4Li2 O2 /TiC(111)

3Li2 O2 /TiC(111)

5Li2 O2 /TiC(111)

6Li2 O2 /TiC(111)

side view

top view

1Li2 O2 /TiC(111)

3Li2 O2 /TiC(111)−O

4Li2 O2 /TiC(111)−O

5Li2 O2 /TiC(111)−O

6Li2 O2 /TiC(111)−O

top view

1Li2 O2 /TiC(111)−O 2Li2 O2 /TiC(111)−O

side view

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 6 of 15

Figure 3: Lowest energy congurations of nLi2 O2 (n = 16) molecules on the non-oxidized and oxidized TiC(111). Blue and brown balls correspond to Ti and C atoms, while red and green balls are O and Li atoms, respectively. Ball sizes are drawn according to atomic radii. the O atoms are at the hcp hollow positions of Ti atoms, taking the positions of the missing C atoms. The Li/Na atoms are on the positions of the missing Ti atoms in the next atomic layer. The M−O bond lengths are close to the corresponding crystal bonds (see Table 1). The formation of the layered structures due to the creation of O−Ti bonds and the complete break of the molecular O−O bonds can be explained by the strong binding of the O atoms with the reactive Ti-terminated surface. Full molecular (2 molecules,0.06 [1/Å2 ]) coverage of the TiC(111) by 2 M2 O2 and 2 MO2 leads to the expansion of the rst O layer relative to a partial coverage, for example the Ti−O bond lengths of 2 Li2 O2 /TiC(111) increase by 0.18 Å compared with 1 Li2 O2 /TiC(111). Our nding is in a good agreement with the structures obtained by Wang et. al. 22 for 1 Li2 O2 molecule on the TiC(111) surface. At larger molecular density (0.09 [1/Å2 ] and 0.12 [1/Å2 ], which correspond to n = 3 and 4) we found that molecules resemble a layered

structure arrangement with M/O/M/O molecular layers on the TiC(111) surface, in which the lowest O atoms stay at the hcp hollow positions on the TiC(111) surface, similarly to the case of n = 1, 2, however, the M and O atoms of the next molecular layers are slightly displaced from their ideal atomic hcp and fcc hollow positions. The displacements are more signicant for the nNaO2 molecules compared with the nLi2 O2 and the nNa2 O2 molecules because of the dierent stoichiometry. An addition of the next two molecules (n = 5 and 6 with the density of 0.15 [1/Å2 ] and 0.18 [1/Å2 ]) leads to large atomic displacements of the M/O/M layers, except for the rst O layer bound to the TiC(111) surface. The M−O, O−O and M−M bond lengths of the distorted structures are shorter compared with the ordered structures. To summarize this part, it is possible to say that initially the AMO atoms prefer to follow the atomic positions dictated by the TiC crystal. As more molecules are adsorbed, the


6

Page 7 of 15

2Na2 O2 /TiC(111)

3Na 2 O 2 /TiC(111)

4Na 2 O 2 /TiC(111)

5Na 2 O 2 /TiC(111)

6Na 2 O 2 /TiC(111)

side view

top view

1Na 2 O 2 /TiC(111)

4Na 2 O 2 /TiC(111)−O 5Na 2 O 2 /TiC(111)−O

6Na 2 O 2 /TiC(111)−O

top view

1Na 2 O 2 /TiC(111)−O 2Na2 O2 /TiC(111)−O 3Na 2 O 2 /TiC(111)−O

side view

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 4: Lowest energy congurations of nNa2 O2 (n = 16) molecules on the non-oxidized and oxidized TiC(111). Blue, brown, red and yellow balls correspond to Ti, C, O and Na atoms, respectively. Ball sizes are drawn according to atomic radii.


7

The Journal of Physical Chemistry 2NaO 2/TiC(111)

3NaO 2 /TiC(111)

6NaO2 /TiC(111)

5NaO2 /TiC(111)

4NaO 2/TiC(111)

side view

top view

1NaO 2/TiC(111)

3NaO 2 /TiC(111)−O

4NaO2 /TiC(111)−O

5NaO2 /TiC(111)−O

6NaO2 /TiC(111)−O

top view

1NaO 2/TiC(111)−O 2NaO 2/TiC(111)−O

side view

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 8 of 15

Figure 5: Lowest energy congurations of nNaO2 (n = 16) molecules on the non-oxidized and oxidized TiC(111), where blue, brown, red and yellow balls present Ti, C, O and Na atoms, respectively. Ball sizes are drawn according to atomic radii. structure changes in the direction of the native AMO crystal but more layers are needed to fully reach that. The thickness, D, of the AMO molecular layers, is another parameter which can be compared with the respective D of the corresponding compressed AMO crystal structure. Figure 6 shows the layers thickness of the lowest energy structures as a function of molecular density. We found that for nLi2 O2 , nNa2 O2 , and nNaO2 molecules on the pristine TiC(111) surface D is near zero at n = 1 and 2 showing that in the rst layer all metal atoms are at about the same height, while at n = 3 and 4 D in most cases change only slightly with respect to the native crystal, namely by 13−19%, except for 4 Na2 O2 and 3 NaO2 systems which change by 38% and 68%, respectively. For the larger systems (at n = 5 and 6) the dierences in D between the adsorbed AMOs and the crystal structures are higher, for example 46% for 6 Li2 O2 .

8 Li 2 O2 /TiC(111)

6

Li 2 O2 /TiC(111)-O

4 2 0

0.03

0.06

8

Na 2 O2 /TiC(111)

6

Na 2 O2 /TiC(111)-O

0.09

0.12

0.15

0.18

0.21

0.09

0.12

0.15

0.18

0.21

0.09

0.12

0.15

0.18

0.21

4 2 0

0.03

0.06

8 NaO 2 /TiC(111)

6

NaO 2 /TiC(111)-O

4 2 0

0.03

0.06

Figure 6: The layers thickness for the lowest energy structures of nLi2 O2 , nNa2 O2 and nNaO2 molecules on TiC(111) and TiC(111)-O surfaces vs molecular density. Triangles correspond to the thickness of the crystals in the original cell.


8

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


Table 1: Adsorption energy, Ead , and structural properties, such as minimal Ti−O, min min M−O and M−M bond lengths, dmin Ti-O , dM-O and dM-M , respectively, of the nM2 O2 (M = Li, Na; n = 16) and nNaO2 (n = 16) molecules with molecular density, ρmol , and thickness, D, on TiC(111) and TiC(111)-O surfaces. The data for crystal structure and molecules in gas-phase are also provided for comparison. The adsorption energies are given in eV, while the bond lengths and thickness are in Å, and ρmol is in [1/Å2 ]. n

ρmol

Ead

1 2 3 4 5 6 crys. mol.

0.031 0.061 0.092 0.123 0.153 0.184

−9.73 −9.65 −8.66 −7.80 −6.50 −5.84

1 2 3 4 5 6 crys. mol.

0.031 0.061 0.092 0.123 0.153 0.184

−10.53 −10.00 −8.83 −7.37 −6.47 −5.87

1 2 3 4 5 6 crys. mol.

0.031 0.061 0.092 0.123 0.153 0.184

−10.87 −10.28 −7.82 −6.52 −5.66 −5.02

min min dmin Ti-O dM-O dM-M nLi2 O2 /TiC(111) 2.05 1.90 3.06 2.23 1.93 3.07 2.06 1.80 2.46 2.03 1.89 2.36 2.04 1.81 2.31 2.05 1.83 2.30 1.98 2.65 1.74 3.09 nNa2 O2 /TiC(111) 2.01 2.31 3.46 2.14 2.31 3.07 2.05 2.19 2.76 2.09 2.08 2.26 2.05 2.14 2.79 2.05 2.11 2.58 2.31 3.05 2.08 3.83 nNaO2 /TiC(111) 1.99 2.27 6.13 2.02 2.27 3.54 2.04 2.21 3.26 2.03 2.21 3.11 2.03 2.25 3.07 2.03 2.22 3.12 2.44 3.91 2.13 −

D 0.00 0.00 2.20 1.56 3.72 5.62 1.92 0.19 0.00 2.59 3.10 7.67 8.34 2.24

− 0.13 0.89 3.11 3.44 5.88 2.76

Ead

min min dmin Ti-O dM-O dM-M nLi2 O2 /TiC(111)-O −1.91 2.03 1.90 3.02 −2.40 2.02 1.90 2.26 −2.63 2.03 1.93 2.43 −3.02 2.04 1.79 2.67 −2.63 2.00 1.84 2.18 −2.78 2.00 1.84 2.25 1.98 2.65 1.74 3.09 nNa2 O2 /TiC(111)-O −2.91 2.01 2.29 3.43 −3.18 2.03 2.20 3.11 −3.25 2.06 2.21 2.97 −3.19 2.06 2.20 2.81 −3.27 1.92 2.15 2.75 −2.98 2.05 2.15 2.55 2.31 3.05 2.08 3.83 nNaO2 /TiC(111)-O −1.75 2.00 2.32 6.13 −1.81 2.00 2.31 3.36 −1.63 2.00 2.30 3.51 −1.82 2.02 2.28 3.08 −1.78 2.02 2.27 3.22 −1.86 2.02 2.25 3.06 2.44 3.91 2.13 −


9

D 0.52 2.25 2.52 2.75 4.59 6.02 1.92 0.15 3.11 3.28 5.95 6.21 8.60 2.24

− 0.20 3.83 3.61 6.21 6.26 2.76


The eect of surface oxidation on the geometry of the adsorbed AMO layers

Page 10 of 15

This is to be expected as none of the grown crystals is polar and so it is logical to assume that the work function will just oscillate as more layers are grown. Possible insights for the work function change can be found by an analysis of the Bader charge on every atom for the 1 Li2 O2 , 1 Na2 O2 and 1 NaO2 in gas-phase and on the TiC(111) and TiC(111)-O surfaces. Although the total charge is neutral, the atoms possess positive and negative charges (see the SI for details), which is in accordance to their electro-negativity. A Bader charge analysis of the bulk TiC system shows that the Ti atoms are positively charged (+1.50 e), while the C atoms are negatively charged (−1.50 e). Moreover, the charge distribution at the surface layers of the TiC(111) surface is dierent than the bulk value and is calculated to be +1.02 e per atom for the top Ti layer and −1.69 e per atom for the carbon layer beneath it. Upon O layer passivation of the TiC(111), the oxygen atoms receive a negative charge of −0.99 e per atom, at that, the Ti surface atoms become more positively charged, i.e. 1.74 e per atom (vs. +1.02 e per atom in the pristine surface). Upon the AMO adsorption on the pristine surface and O and M layers formation on the TiC(111) surface, the AMO layers, in total, gain negative charge of 1.16 e, 1.33 e and 1.62 e for 1 Li2 O2 , 1 Na2 O2 and 1 NaO2 , respectively. The charge transfer would support an increase in the work function, however, the work function is decreased. This happens because the positive metal atoms are above the negative oxygen atoms, so that the dipole eect is stronger than the charge transfer eect. When the AMOs are adsorbed on the oxygen passivated TiC(111) surface, there is a change of sign in the charge transfer and the AMOs give charge to the surface, i.e. we obtained a positive total charge on the AMOs, +0.79 e, +0.88 e and +0.57 e, for 1 Li2 O2 , 1 Na2 O2 and 1 NaO2 , respectively. Here the charge transfer is expected to reduce the work function, while the dipole is now expected to increase the work function, since the AMO metal is now below the AMO oxygen layer, still the net eect is that of work function reduction.

The presence of an oxide layer on the TiC(111) surface, (TiC(111)-O), strongly aects the molecular structures which dier signicantly from the structures on the pristine surface. Despite the similar trends, such as breaking molecular O−O bonds and structural reassembling, less ordered structures are already formed on the surface at n = 1, as a result of no direct binding with the reactive Ti atoms on the surface. The layer thickness parameter tends to increase compared with the value of D on the pristine TiC(111) surface for all systems as a result of higher structural distortions. This trend can be observed already at n = 1 and 2. One example is D of 0.00 Å and 2.25 Å for 2 Li2 O2 on the pristine and oxidized TiC(111) surfaces, respectively. For higher molecular density the dierences in D still exist but show an oscillating pattern.

Work function change The adsorption of AMO molecules aects also the surface work function (Φ). The work function of the pristine TiC(111) surface that was calculated by us is 4.57 eV (close to experimental value of 4.7 eV 43 ), the addition of the oxygen layer results in an increase of the work function of the TiC(111) to 5.20 eV. Bader charge analysis 44,45 (details are found in the SI) reveals that the oxygen acquires a negative charge of 0.99 e upon adsorption. This is consistent with the increase in the work function. By contrast, when the AMO adsorb on both the pristine and oxidized TiC(111) surfaces, the surface work function decreases (∆Φ < 0). The work function changes are presented in Figure 7, and the electrostatic potentials of these systems, before and after adsorption, which yield the surface dipole, can be found in the SI. Examination of the changes in the work function shows that in all systems, after the adsorption of the rst AMO molecule, there is no clear trend with increasing the AMO coverage.


10

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

The Journal of Physical Chemistry 4

Li2O2/TiC(111)

2

Li2O2/TiC(111)-O

already from the rst molecule for all the AMO species that we checked. The molecules we have checked have the option to adsorb at the surface or to self-assemble in solution and form a cluster. AMOs' Cluster formation energies are not necessarily equal to the AMOs' self-binding energy and can be larger at the beginning and smaller at later stages where there are dierent facets for the formed clusters. 19 As discussed before 12 the way Li2 O2 is crystallized, surface adsorption vs. solution based self-assembly, is strongly aected by the solvent properties. However, the absolute surface adsorption energy may have some eect as well. It is therefore interesting for future work to map and compare those energies for dierent electrode surfaces and check their possible relationship to cell performance. All the AMOs show similar trends for the adsorption energy, we can therefore expect that the behavior of surface adsorption and clustering of sodium products (Na2 O2 and NaO2 ) will be similar to that of Li2 O2 . Another important observation is that the rst layer of adsorbed AMOs functions as a passivation layer. A possible conclusion is that the initial surface reactivity might not be an important parameter, however, checking this might require to test more surfaces and have a more detailed comparison to experiment. Finally, we have analyzed in this work just the products (Li2 O2 ,Na2 O2 ,NaO2 ), this was already sucient to get interesting comparison between trends of sodium and lithium, however, to get a full understanding of those systems, the behavior of additional intermediates, side products and also electrolyte molecules should be investigated.

0 -2 -4 0.03

0.06

0.09

0.12

4

0.15

0.18

Na2O2/TiC(111)

2

Na2O2/TiC(111)-O

0 -2 -4 0.03

0.06

0.09

0.12

0.15

4

0.18

NaO2/TiC(111)

2

NaO2/TiC(111)-O

0 -2 -4 0.03

0.06

0.09

0.12

0.15

0.18

Figure 7: The change in the work function for nLi2 O2 , nNa2 O2 and nNaO2 on TiC(111) and TiC(111)-O surfaces.

Summary and Outlook In this work we have calculated the adsorption energies of varying coverage of Li2 O2 , Na2 O2 and NaO2 molecules on clean and oxygen passivated TiC(111) surfaces. We showed that all the dierent AMO molecules exhibit a similar behavior. In addition, the adsorption on a clean surface is initially much more favorable energetically than on an oxygen passivated surface. Furthermore, we showed that after the deposition of two molecular layers, the adsorption energies at the clean and oxygen passivated surface approach one another (see Fig. 2) and in fact the eect of the surface preparation becomes almost unimportant. This is mainly because the newly adsorbed molecules are now lying on the previous AMO molecular layers and not on the TiC itself, clean or oxygen passivated. To verify this we have compared the adsorption energies to the crystal AMO binding energy for the respective crystals of Li2 O2 ,Na2 O2 and NaO2 and got very close values. It should be noted that for the oxygen passivated surface, which is more realistic experimentally, the adsorption energies are close to that of the AMO crystals self binding energies

Acknowledgement This work was suppor-

ted by the Planing & Budgeting Committee of the Council of High Education and the Prime Minister Oce of Israel, in the framework of the INREP project.


11


Supporting Information Avai-

Page 12 of 15

(8) Lu, J.; Li, L.; Park, J.-B.; Sun, Y.-K.; Wu, F.; Amine, K. Aprotic and Aqueous Li−O2 Batteries. Chem. Rev. 2014, 114, 56115640.

lable Additional information is found in the Supporting Information: (1) Bulk lattice constants calculations; (2) Gas phase molecules calculations; (3) DFT level of theory and model size: eects on adsorption energy trends; (4) Density of States (DOS) graphs; (5) Electrostatic potential graphs; (6) Bader charge.

(9) Chen, Y.; Freunberger, S. A.; Peng, Z.; Fontaine, O.; Bruce, P. G. Charging a Li−O2 Battery Using a Redox Mediator. Nature Chemistry 2013, 5, 489494. (10) Wen, Z.; Shen, C.; Lu, Y. Air Electrode for the Lithium-Air Batteries: Materials and Structure Designs. ChemPlusChem 2015, 80, 270287.

References (1) Littauer, E. L.; Tsai, K. C. Corrosion of Lithium in Aqueous Electrolyte. J. Electrochem. Soc. 1997, 124, 850855.

(11) Bruce, P. G.; Freunberger, S. A.; Hardwick, L. J.; Tarascon, J.-M. Li−O2 and Li−S Batteries with High Energy Storage. Nature Materials 2012, 11, 19.

(2) Abraham, K. M.; Jiang, Z. A Polymer Electrolyte-Based Rechargeable Lithium/Oxygen Battery. J. Electrochem. Soc. 1996, 143, 15.

(12) Johnson, L.; Li, C.; Liu, Z.; Chen, Y.; Freunberger, S. A.; Ashok, P. C.; Praveen, B. B.; Dholakia, K.; Tarascon, J.-M.; Bruce, P. G. The Role of LiO2 Solubility in O2 Reduction in Aprotic Solvents and its Consequences for Li−O2 Batteries. Nature Chemistry 2014, 6, 10911099.

(3) Peled, E.; Golodnitsky, D.; Mazor, H.; Goor, M.; Avshalomov, S. Parameter Analysis of a Practical Lithium- and SodiumAir Electric Vehicle Battery. Journal of Power Sources 2011, 196, 68356840.

(13) Gallant, B. M. Chemical and Morphological Changes of LiO2 Battery Electrodes Upon Cycling. J. Phys. Chem. C 2012, 116, 2080020805.

(4) Chakkaravarthy, C.; Udupa, H. V. K. On the Suppression of Self Discharge of the Zink Electrodes of Zink Air Cells And Other Related Battery Systems. Journal of Power Sources 1983, 10, 197200.

(14) Thotiyl, M. M. O.; Freunberger, S. A.; Peng, Z.; Chen, Y.; Liu, Z.; Bruce, P. G. A Stable Cathode for the Aprotic Li−O2 Battery. Nature Materials 2013, 12, 10501056.

(5) Yang, S.; Knickle, H. Design and Analysis of Aluminum/Air Battery System for Electric Vehicles. Journal of Power Sources 2002, 112, 162173.

(15) Kundu, D.; Black, R.; Berg, E. J.; Nazar, L. F. A Highly Active Nanostructured Metallic Oxide Cathode for Aprotic Li−O2 Batteries. Energy Environ. Sci. 2015, 8, 12921298.

(6) Ma, Y.; Li, N.; Li, D.; Zhang, M.; Huang, X. Performance of Mg-14 Li-1 Al0.1 Ce as Anode for Mg-Air Battery. Journal of Power Sources 2011, 196, 2346 2350.

(16) Kozmenkova, A. Y.; Kataev, E. Y.; Belova, A. I.; Amati, M.; Gregoratti, L.; Velasco-Vélez, J.; Knop-Gericke, A.; Senkovsky, B.; Vyalikh, D. V.; Itkis, D. M. et al. Tuning Surface Chemistry of TiC Electrodes for LithiumAir Batteries. Chem. Mater. 2016, 28, 82488255.

(7) Ito, A.; Zhao, L.; Okada, S.; Yamaki, J. I. Synthesis of Nano-Fe3 O4 -Loaded Tubular Carbon Nanobers and Their Application as Negative Electrodes for Fe/Air Batteries. Journal of Power Sources 2011, 196, 81548159.


12

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


(17) Zu, C.-X.; Li, H. Thermodynamic Analysis on Energy Densities of Batteries. Energy Environ. Sci. 2011, 4, 26142624.

(26) Kresse, G.; Joubert, J. From Ultrasoft Pseudopotentials to the Projector Augmented-Wave Method. Phys. Rev. B 1999, 59, 17581775.

(18) Sun, Q.; Yang, Y.; Fu, Z.-W. Electrochemical Properties of Room Temperature Sodium-Air Batteries with Nonaqueous Electrolyte. Electrochem. Commun. 2012, 16, 2225.

(27) Kresse, G.; Furthmüller, J. Ecient Iterative Schemes for Ab Initio Total-Energy Calculations Using a Plane-Wave Basis Set. Phys. Rev. B 1996, 54, 1116911186.

(19) Kang, S.; Mo, Y.; Ong, S. P.; Ceder, G. Nanoscale Stabilization of Sodium Oxides: Implications for Na−O2 Batteries. Nano Lett. 2014, 14, 10161020.

(28) Hafner, J. Ab-initio Simulations of Materials Using VASP: Density-Functional Theory and Beyond. Journal of Computational Chemistry 2008, 29, 20442078.

(20) Lutz, L.; Yin, W.; Grimaud, A.; Corte, D. A. D.; Tang, M.; Johnson, L.; Azaceta, E.; Sarou-Kanian, V.; Naylor, A. J.; Hamad, S. et al. High Capacity Na−O2 Batteries: Key Parameters for SolutionMediated Discharge. J. Phys. Chem. C 2016, 120, 2006820076.

(29) Monkhorst, H. J.; Pack, J. D. Special Points for Brillouin-Zone Integrations. Phys. Rev. B 1976, 13, 51885192. (30) Curtarolo, S.; Setyawan, W.; Hart, G. L. W.; Jahnatek, M.; Chepulskii, R. V.; Taylor, R. H.; Wang, S.; Xue, J.; Yang, K.; Levy, O. et al. AFLOW: An Automatic Framework for High-Throughput Materials Discovery. Comput. Mat. Sci. 2012, 58, 218226.

(21) Wang, Z.; Sun, J.; Cheng, Y.; Niu, C. Adsorption and Deposition of Li2 O2 on TiC(111) Surface. J. Phys. Chem. Lett. 2014, 5, 39193923.

(31) ICSD

crystallographic

database.

https://icsd.fiz-karlsruhe.de/ search/basic.xhtml, Accessed: Jan 2017.

(22) Wang, Z.; Chen, X.; Cheng, Y.; Niu, C. Adsorption and Deposition of Li2 O2 on the Pristine and Oxidized TiC Surface by First-Principles Calculation. J. Phys. Chem. C 2015, 119, 2568425695.

(32) Zaima, S.; Shibata, Y.; Adachi, H.; Oshima, C.; Otani, S.; Aono, A.; Ishizawa, Y. Atomic Chemical Composition and Reactivity of the TiC(111) Surface. Surf. Sci 1985, 157, 380392.

(23) Perdew, J. P.; Chevary, J. A.; Vosko, S. H.; Jackson, K. A.; Pederson, M. R.; Singh, D. J.; Fiolhais, C. Atoms, Molecules, Solids, and Surfaces: Applications of the Generalized Gradient Approximation for Exchange and Correlation PW91 Reference. Phys. Rev. B 1992, 46, 6671 6687.

(33) Vojvodic, A.; Ruberto, C.; Lundqvist, B. I. Trends in Atomic Adsorption on Titanium Carbide and Nitride. Surf. Sci. 2006, 600, 36193623. (34) Zur, A.; McGill, T. C. Lattice Match: An Application to Heteroepitaxy. J. Appl. Phys. 1984, 55, 378386.

(24) Perdew, J. P.; Burke, K.; Ernzerhof, M. Generalized Gradient Approximation Made Simple. Phys. Rev. Lett. 1996, 77, 38653868.

(35) Elliott, R. O.; Kempter, C. P. Thermal Expansion of Some Transition Metal Carbides. J. Phys. Chem. 1958, 62, 630631.

(25) Blochl, P. Projector Augmented-Wave Method. Phys. Rev. B 1994, 50, 17953 17979.


13


(36) Cota, L. G.; de la Mora, P. On the Structure of Lithium Peroxide, Li2 O2 . Acta Cryst. 2005, B61, 133136. (37) Radin, M. D.; Rodriguez, J. F.; Tian, F.; Siegel, D. J. Lithium Peroxide Surfaces Are Metallic, While Lithium Oxide Surfaces Are Not. J. Am. Chem. Soc. 2012, 134, 10931103. (38) Araujo, R. B.; Chakraborty, S.; Ahuja, R. Unveiling the Charge Migration Mechanism in Na2 O2 : Implications for SodiumAir Batteries. Phys. Chem. Chem. Phys. 2015, 17, 82038209. (39) Carter, G. F.; Templeton, D. H. Polymorphism of Sodium Superoxide. J. Am. Chem. Soc. 1953, 75, 52475249. (40) Arcelus, O.; Li, C.; Rojo, T.; Carrasco, J. Electronic Structure of Sodium Superoxide Bulk, (100) Surface, and Clusters Using Hybrid Density Functional: Relevance for Na−O2 Batteries. Phys. Chem. Lett. 2015, 6, 20272031. (41) Lau, K. C.; Assary, R. S.; Redfern, P.; Greeley, J.; Curtiss, L. A. Electronic Structure of Lithium Peroxide Clusters and Relevance to LithiumAir Batteries. J. Phys. Chem. C 2012, 116, 23890 23896. (42) Arcelus, O.; Suaud, N.; Katcho, N. A.; Carrasco, J. Insight from First Principles into the Stability and Magnetism of Alkali-Metal Superoxide Nanoclusters. J. Chem. Phys. 2017, 146, 184301. (43) Oshima, C.; Aono, M.; Zaima, S.; Shibata, Y.; Kawai, S. The Surface Properties of TiC(001) and TiC(111) surfaces. Journal of the Less-Common Metals 1981, 82, 6974. (44) Bader, R. F. W. Quantum Theory

Atoms in Molecules:

a

; Clarendon Press, 1994.

(45) Henkelman, G.; Arnaldsson, A.; Jónsson, H. A Fast and Robust Algorithm for Bader Decomposition of Charge Density. Comput. Mater. Sci. 2006, 36, 354360. ACS Paragon Plus Environment

14

Page 14 of 15

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


Graphical TOC Entry

Density Functional Theory inverstigation of the adsorption of nLi2 O2 , nNa2 O2 and nNaO2 molecules (n = 16) on TiC(111) surface is presented. We demonstrate that all the Alkali Metal Oxide (AMO) molecules show similar trends in their surface adsorption energy versus the surface coverage density. We also show that the rst AMO layer acts as a passivation layer on the pristine TiC(111) surface.


15

Adsorption of Li2O2 , Na2O2 and NaO2 on TiC(111) Surface for Metal

Recommend Documents