Ab Initio Approach for Prediction of Oxide Surface ... - ACS Publications

Subscriber access provided by ECU Libraries

Letter

An Ab initio Approach for Prediction of Oxide Surface Structure, Stoichiometry, and Electrocatalytic Activity in Aqueous Solution Xi Rong, and Alexie M Kolpak J. Phys. Chem. Lett., Just Accepted Manuscript • Publication Date (Web): 22 Apr 2015 Downloaded from http://pubs.acs.org on April 23, 2015

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 free 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 accessible to all readers and 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.

The Journal of Physical Chemistry Letters 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 17

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 Letters

An Ab initio Approach for Prediction of Oxide Surface Structure, Stoichiometry, and Electrocatalytic Activity in Aqueous Solution Xi Rong and Alexie M. Kolpak* Department of Mechanical Engineering, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139, United States Corresponding Author *[email protected]

ACS Paragon Plus Environment

1


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 17

ABSTRACT The design of efficient, stable, and inexpensive catalysts for oxygen evolution and reduction is crucial for the development of electrochemical energy conversion devices such as fuel cells and metal-air batteries. Currently, such design is limited by challenges in atomic-scale experimental characterization and computational modeling of solid-liquid interfaces. Here we begin to address these issues by developing a general, first-principles and electrochemical-principles based framework for prediction of catalyst surface structure, stoichiometry, and stability as a function of pH, electrode potential, and aqueous cation concentration. We demonstrate the approach by determining the surface phase diagram of LaMnO3, which has been studied for oxygen evolution and reduction, and computing the reaction overpotentials on the relevant surface phases. Our results illustrate the critical role of solvated cation species in governing catalyst surface structure and stoichiometry, and thereby catalytic activity, in aqueous solution. TOC GRAPHICS

KEYWORDS: electrocatalysis, oxygen evolution and reduction reactions, perovskite oxides, density functional theory, surface structure


2

Page 3 of 17

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 design of efficient and inexpensive catalysts for the oxygen evolution reaction (OER) and the oxygen reduction reaction (ORR) is crucial for development of electrochemical energy conversion devices such as fuel cells and metal-air batteries. First-principles computational studies have made significant contributions to such efforts, identifying fundamental correlations between catalytic activity and electronic structure for transition metal1-2 and, more recently, perovksite oxide3-5 catalysts. Despite the insights elucidated in these studies, a comprehensive picture that can describe experimental observations of OER and ORR activity in oxides6-10 remains elusive. A key challenge is the complexity of the oxide-solvent interface. Even in atmosphere, surface structure/composition of oxides is highly sensitive to conditions11-15, and surface stoichiometry exhibits significant influence on oxide activity in various gas phase reactions16-18. Experiments suggest that surface structure and stoichiometry are also important for understanding catalyst stability and activity in liquid environments such as alkaline cells. For example, Ba0.5Sr0.5Co0.8Fe0.2O3-δ, a promising high-efficiency OER catalyst7, is inactive for ORR6; this behavior has been attributed to surface amorphization due to cation leaching under OER conditions9,

19

. However, few in situ atomic-scale surface characterization studies have

been performed to investigate phenomena in oxide-solvent systems, as the solvent makes such techniques extremely challenging. Computationally, approaches have been developed to implicitly include pH and electrode potential (U) in first-principles calculations1, 5, Pourbaix diagrams have been used with such approaches to predict stability of bulk oxides and ideal oxide surface orientations, as well as the presence of –O or –OH adsorbates20-24. However, high computational costs have limited ab initio studies of non-ideal surface stoichiometries. To our knowledge, the effects of surface cation exchange with solvated species have not previously been


3


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 4 of 17

taken into account in determining the atomic structure, stoichiometry, stability, or catalytic activity of surface-solvent interfaces. In this work, we combine first-principles density functional theory (DFT) calculations and electrochemical principles to develop a new approach for prediction of surface structure and stoichiometry in aqueous solution. We derive a general formalism for incorporating pH- and Udependent ion exchange at the catalyst surface-solvent interface, and we use this approach to determine the surface structure/composition phase diagram of LaMnO3 (LMO), a perovskite oxide of interest for aqueous OER/ORR, as a function of pH and U. Our results demonstrate that the surface of LMO is highly sensitive to the environment, with substantially different surface phases present under OER and ORR conditions. Further, computing the free energies of reaction intermediates on these phases, we illustrate the important effects of surface structure and stoichiometry on OER/ORR activity. A variety of oxide surface reconstructions due to surface-solvent ion exchange are possible. In alkaline solution at high U (OER conditions), surface oxidation can occur via formation of OER/ORR intermediates from H2O (i.e., adsorption of O or OH) or introduction of cation vacancies by cation leaching and subsequent solvation. Conversely, at low U (ORR conditions), the surface may be reduced via formation of oxygen vacancies or adsorption of metal ad-atoms. More complex reconstructions including kinks, steps, and changes in surface orientation or termination can also occur. As free energy differences are independent of mechanism, thermodynamically such processes can be considered as series of adsorption, vacancy formation, solvation, and/or precipitation steps.


4

Page 5 of 17

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. Schematic of surface reconstruction via surface-solvent ion exchange. As an example, Fig. 1 illustrates a surface reconstruction due to formation of A vacancies, where A is any atomic species in the oxide. First, a neutral atom of A leaves the surface, with free energy change , where and are the free energies of the reference and reconstructed surfaces, respectively, and is the chemical potential of A. Second, A reacts with protons, electrons, and/or water molecules to form a solvated ion via , where the identity of depends on pH and U. The free energy of this step is ∑ " " , where is the chemical potential of , " and " are the number and chemical potential of species i, and i = H2O, H+, and e−. The total free energy of formation is . As is thus independent of , and can be determined relative to ° ( at standard temperature and pressure). is then computed directly from DFT by using VASP25 with PAW pseudopotentials26 and the RPBE-GGA27 functional as detailed in the Supplemntal Information (SI). is computed using the standard hydrogen electrode (SHE)28 approach, in which protons and electrons are assumed to be in equilibrium with H2(g) at electrode potential $%& 0, ion concentration ()* +,-. =1,


5


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 17

pH=0, and T=25°C. Under operating conditions, as shifted from the SHE Gibbs free ° energy, %& , by finite values of $%& , ()* +,- . , and pH is then given by:

° %& 0$%& 1 2.3 567 56 8 ()* +,- .

(1)

More complex surface reconstruction can similarly be constructed as a series of process involving loss and adsorption of atoms from a reference surface and solvents to form a new surface, which leads to: 9 ∑ ° : ∑ ; <

(2)

Here is the number of A atoms exchanged. A detailed derivation of Eqs. 1 and 2 is provided in SI. As the generality of the derivation suggests, our formalism can be readily employed for other solid-aqueous interfaces. Using this approach, we predict the structure and stoichiometry of the LMO surface as a function of pH and U at room temperature. We consider [001] surfaces, which are observed experimentally3,

29

, and investigate reconstructions derived from ideal MnO2 and LaO(001)

terminations with various combinations and concentrations of adsorbates and vacancies. The ideal MnO2-terminated surface (A0) is taken as the reference. Several A0-derived surfaces are illustrated in Fig. 2.


6

Page 7 of 17

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 2. Optimized surface structures for representative stoichiometries. Figure 3 shows computed free energies for several reconstructions relative to A0 as a function of USHE at pH=12; the stability boundaries for different aqueous Mn species are also shown. Importantly, the figure shows that the slope of changes at each of these boundaries. Furthermore, the slope within a region is different for each surface phase. As suggested from Eq. 2, this behavior is related to the difference in the oxidation state of the Mn on a given surface and in the stable aqueous species. Thus, the relative stability of various surface phases is dramatically affected by the U and pH-dependent identities and formation free energies of aqueous species formed by all atoms of the bulk oxide. In addition to emphasizing the effects of aqueous species, Fig. 3 illustrates the range of accessible surface compositions. For pH=12, A0 is stable over a relatively small range of USHE. At more negative USHE, the stable surface has lost ¼ of the Mn and 1 ML of the O relative to A0,


7


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 17

leading to exposure of the subsurface layer; nominally, this phase is reduced by five additional electron per three remaining surface Mn. At more positive USHE, the surface becomes increasingly oxidized, transitioning from A0 to A1 via adsorption of –OH at all surface Mn sites, then from A1 to B1 via loss of ¼ of the –OH adsorbates along with ¼ of the surface Mn, as seen in Fig. 2.

Figure 3. Surface reconstruction free energies vs USHE at pH=12. MnOx(aq) stability ranges are calculated from the experimental data as shown by Fig. S1 (SI). The essential features of the surface evolution over the entire pH-U phase space are similar to those at pH=12. However, the full phase diagram, shown in Fig. 4, illustrates additional key points. First, La-rich surfaces are only stable for USHE

Ab Initio Approach for Prediction of Oxide Surface ... - ACS Publications

Recommend Documents