An ab Initio Investigation of Proton Stability at BaZrO3 Interfaces

Subscriber access provided by NATIONAL SUN YAT SEN UNIV

Article

An ab initio investigation of proton stability at BaZrO3 interfaces Tania Tauer, Ryan P. O'Hayre, and J. Will Medlin Chem. Mater., Just Accepted Manuscript • DOI: 10.1021/cm500035e • Publication Date (Web): 19 Aug 2014 Downloaded from http://pubs.acs.org on August 25, 2014

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.

Chemistry of Materials 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 13

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

Chemistry of Materials

An ab initio investigation of proton stability at BaZrO3 interfaces Tania Tauer1,3, Ryan O’Hayre2,3, J. Will Medlin1,3* 1

Department of Chemical and Biological Engineering, University of Colorado Boulder, Campus Box 596, Boulder, CO 80303, Colorado, USA

2

Department of Metallurgical and Materials Engineering, Colorado School of Mines, 1500 Illinois St., Golden, CO 80401

3

Renewable Energy Materials Research Science and Engineering Center, Colorado School of Mines, 920 15th Street, Hill Hall 313, Golden, CO 80401

ABSTRACT: Growing evidence that proton chemistry at the perovskite interface influences both proton conduction and catalyst activity has motivated more thorough examinations of proton behavior in these interfacial environments. This study utilizes density functional theory to examine proton stability at two prominent perovskite interfaces, the perovskite surface and perovskite-metal heterointerface, to identify opportunities to screen for perovskites with enhanced functionality. An analysis of the perovskite surface revealed fluctuations in proton stability as a function of the depth below the surface that leads to a barrier for proton mobility. The addition of a metal heterointerface can act to decrease this barrier by stabilizing protons near the surface. Finally, an electronic structure parameter, the p-band center, was identified as a useful predictor for proton adsorption energies in uniform perovskite structures, such as the perovskite surface and bulk, where detailed analyses reveal how local characteristics alter proton stability.

1. INTRODUCTION Proton-conducting perovskite oxides are rapidly gaining interest for a number of energy conversion applications including fuel cells, membrane reactors, and electrolyzers. Protons encounter several varieties of interfaces within these perovskite system, such as the grain boundary (GB) during proton conduction,1-5 the material surface during catalysis,6-8 and the triple phase boundary (TPB) within electrochemical devices.9-11 Proton chemistry at and near these regions greatly influences the behavior of devices for which these perovskites are being implemented; yet many questions remain about the proton behavior within these complex environments. For example, it is well known that proton conductivity in many perovskites, such as BaZrO3, is orders of magnitude lower across the grain boundary (GB) interfaces than within the material bulk.12 While experimental5,13-16 and computational17-20 studies have identified potential sources for slow proton conduction across this region, there is still only limited understanding of proton stability and mobility at the GB interfaces. Additionally, recent data show reasonable catalytic activities for several electrochemical reactions, such as the oxygen reduction reaction and oxygen evolution reaction, on various perovskite surfaces.21-24 While initial studies have begun to reveal important parameters that influence the catalytic properties of perovskites,23-26 they still lack a detailed understanding of the interaction of the proton with the catalyst surface. Finally, little work has been done to assess proton chemistry at the TPB, despite its importance to the performance of electrochemical devices, such as proton conducting ceramic fuel cells and electrolyzers.27,28 The importance of and dearth of knowledge about proton chemistry at perovskite interfaces motivates further detailed studies of these interfacial environments.

In this study, density functional theory (DFT) is used to gain an atomistic level understanding of proton behavior at BaZrO3 (BZY) interfaces to identify opportunities and strategies to design and screen for perovskites with enhanced functionality. First, we present an analysis of proton stability in the near surface environments of the (001) BaO and ZrO2 surface terminations of BZY. An examination of the density of states of the oxygen atoms in this interfacial region provides evidence to explain stability trends of protonic defects in layers near the perovskite surface. We show that changes in the oxygen p-band center are linearly related to the stability of protons at the surface and in the near interfacial region. To verify the generality of these results we then test the p-band center hydration predictor model within bulk yttrium-doped barium cerate. Finally, we assess how the introduction of metal nanoclusters to the perovskite surface changes proton stability in the near-heterointerfacial region of BaZrO3 as a model of the triple-phase boundary. Palladium, platinum, nickel, and silver clusters are arranged at the two (001) surfaces to form BaZrO3-metal interfaces. A comparison of proton stability between the BaZrO3-vacuum and BaZrO3-metal interfaces indicates that different metals have different stabilization effects on the two barium zirconate surfaces tested. Trends across the metals help in identifying factors that control proton stabilization at heterointerfaces. 2. COMPUTATIONAL METHODS All DFT calculations were carried out using the Vienna Ab-Initio Simulation Package (VASP).29-32 Plane waves with an energy cutoff of 500 eV were constructed using projector augmented wave (PAW) potentials.33 The generalized gradient approximation (GGA) was applied using the PW91 exchange correlation

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

functional. A 2 x 2 x 1 Monkhorst-Pack sampling of the Brillouin zone was utilized for all surface calculations.34 A 4 x 4 BaZrO3 unit cell was used throughout. 2.1 BaZrO3-vacuum Interfacial Study The proton stability studies at the BaZrO3-vacuum interface were conducted on an 18-layer asymmetrical slab at both the BaO and ZrO2 (001) surface terminations. The (001) surfaces were chosen because they have been determined to be the most stable surface terminations of BaZrO3.35,36 The thickness of the slab was set at 18 layers, or approximately 38 Å. This was determined following a size screening that revealed energy fluctuations due to the placement of yttrium dopants and oxygen vacancies at various depths from the perovskite surface subsided to less than 0.01 eV after 6 layers. This distance was then tripled to mitigate surface effects from the bottom of the slab interfering with the targeted area of study at the top interface. An asymmetric slab was chosen to allow for the formation of a stoichiometric unit cell in which electroneutrality could be maintained. The bottom 9 layers of the slab were fixed to simulate the perovskite bulk, while top 9 layers were allowed to relax. A vacuum gap of 10 Å was included to prevent any spurious effect between cells. To maintain electroneutrality, the unhydrated unit cell contained two yttrium dopants and one oxygen vacancy. The three defects were arranged in the bottom, fixed layers of the unit cell. Each hydrated unit cell contained two yttrium ions and two protons, yielding an effective doping concentration of about 5%. The dopant atoms were arranged in the same configuration as the unhydrated cell. Additionally, one proton was positioned in the bottom, fixed portion of the cell, while the remaining proton was situated in the top nine relaxed layers. This arrangement ensured minimal interactions between the proton of interest and the remaining defects within the cell. Thus, any variations in energy were able to be attributed to the proximity of the top proton to the perovskite surface rather than the proton’s interaction with other defects. (When a proton is referred to in this work, it refers to this upper proton). In each hydrated cell, the protons were configured 1 Å away from an oxygen atom to model OH bonding, consistent with previous perovskite studies.37,38 The upper protons were bonded to oxygen ions at different depths from the perovskite surface, varying from surface oxygen (i=1) to bulk oxygen (i=8) (refer to Figure 1a and 1b). The orientation of the proton with respect to the perovskite surface was also considered. In the layers where the proton was able to rotate perpendicularly to the cell surface, the proton was arranged in an upward, side, or downward configuration to investigate whether orientation influenced proton stability (Figure 1c). In the layers where the proton was able to rotate parallel to the material surface, the proton was positioned toward one of its 4 next nearest neighbor oxygen ions, forming a stabilizing OH—O bond that has been characterized by perovskite studies.39,40 Proton stability was assessed by calculating the hydration enthalpy of the slab (i.e. the energy change due to the absorption of water) for each proton configuration: ∆

(Eq.1)

where denotes the energy of the hydrated cell with the top proton positioned in layer i from the surface, represents the energy of the unhydrated cell, and is the energy of a single water molecule, calculated as -14.4 eV.

Page 2 of 13

While it is well documented that structural relaxations occur in the proximity of dopant atoms and oxygen vacancies,39,41,42 relaxations were not allowed in the fixed region of the slabs in either the hydrated or unhydrated states. Enabling relaxation around defects in this region (which is retained in all calculations) will not alter the overall stabilization trend, but may systematically alter the absolute values of the hydration enthalpies. As discussed below, hydration enthalpies were shifted relative to the bulk values in order to facilitate direct comparison between the two systems. An electronic density of states (DOS) analysis was conducted to probe the origins of the proton stability trends revealed from this hydration study. The center of the valence band describes the weighted average of the electrons within the band, i.e. the first moment normalized by the zeroth moment. The p-band centers were calculated for each oxygen in the unhydrated cell to identify how the average p-band electron energies within the atoms change between different perovskite layers in the near-interfacial region. The p-band centers of the interfacial oxygens were calculated relative to the bulk oxygen p-band center, -1.2 eV. This was done by setting the p-band centers of the oxygen ions in layer eight to the bulk value and adjusting the remaining values accordingly. a)

b)

c)

Figure 1. Surface structures of the (a) BaO and (b) ZrO2 termination constructed of oxygen (red), barium (green) and zirconium (beige) ions. The circled oxygen ions indicate those oxygens with which the protons were bonded to during the proton stability studies. The tested proton configurations (upward, parallel, downward) are depicted in (c). 2.2 BaCeO3 Bulk Study To determine whether the trend between p-band center and hydration enthalpy that was found in the BaZrO3 interfacial study could also be applied to different perovskite systems, we carried out a similar analysis for bulk yttrium-doped BaCeO3, a system we have previously studied in detail.43 Three dopant concentrations were examined: 50%, 25%, and 12.5%. Unit cells were constructed such that two yttrium atoms were incorporated into each cell to obtain the desired dopant concentration. Therefore, the 50%-doped cell contained 4 chemical formula units of BaCeO3, the 25%-doped cell included 8 chemical formula units, and the 12.5%-doped cell comprised 16 chemical formula units. The yttrium atoms were arranged in the most stable configurations determined from previous bulk BaCeO3 studies.43 For this case, the hydration enthalpy was not calculated, but rather proton adsorption strength was probed. While this would

2


Page 3 of 13

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


represent the process of hydrogen absorption into the material, we have chosen to use the term “adsorption” specifically to indicate that the proton binding (the hydrogen becomes protonated once bonded to the oxygen), or proton adsorption, to oxygen atoms both at the surface and in the bulk is the process under investigation. The proton adsorption strengths were calculated such that: ∆

(Eq. 2)

denotes the cell with a single hydrogen atom, represents the BCY cell with no hydrogen atoms within it, and is the energy of a single hydrogen. As discussed below, we opted to consider H adsorption rather than hydration because our aim was simply to verify that basic trends with respect to pband center in bulk as well as surface models; using H adsorption avoids the complication of introducing a new oxygen vacancy that affects the distribution of electrons around the remaining oxygen ions. The p-band center analysis for the bulk material was performed on the unprotonated BCY unit cell to identify how the proximity to yttrium dopants influenced the distribution of electrons within the oxygen atoms in the system. (If an oxygen vacancy were incorporated, then this analysis would not allow us to comment on the source of the variation in electron distribution.) 2.3 BaZrO3-Metal Interfacial Study In the final part of this study, material hydration was probed at the BaZrO3-metal heterointerface. Due to the increase in computational resources required to complete the BaZrO3-metal interfacial simulations, the unit cells for these calculations were simplified. A smaller unit cell was constructed with 8 layers, such that the bottom 4 layers were fixed and the top 4 layers were allowed to relax. To simulate the complex BaZrO3-metal heterointerface, a 10 atom metal cluster was optimized atop each of the two perovskite surfaces. A 10 atom cluster was chosen because it forms a stable particle (a cleaved section of a full shell “magic number” 13 cluster). The width of the 10 atom cluster was also advantageous because it could be arranged on the 4 x 4 BaZrO3 unit cell such that there were gaps between neighboring metal clusters. This ensured that the minimum amount of perovskite lattice deformation occurred at the heterointerface. It is important to note that BaZrO3-metal interfacial structure is currently unknown experimentally. While a 10 atom metal cluster does not likely represent the accurate 2D interface that occurs in situ, it allows us to explore the effect of the metal interaction on the stability of protons in the environment directly under the metal cluster. Four metals with varying electronic properties were investigated: platinum, palladium, silver, and nickel. First, the Pd metal clusters were geometrically optimized to provide a model for how to arrange the remaining metal atoms. This was done first by examining several different initial configurations of Pd atoms on the two perovskite surfaces. Cluster stability was assessed by calculating the clustering energy:

(Eq. 3)

where denotes the total energy of the metal cluster on the perovskite surface, represents the energy of the bare perovskite slab, and n describes the number of metal atoms in the cluster. First, a layer of 7 Pd atoms was arranged on the two BaZrO3 surfaces to optimize the bottom layer of metal atoms. It was noted that when the metal was allowed to relax the arrangements geometrically optimized to a more stable 5 atom bottom layer with 2 Pd atoms above. We thus probed two 10 atom

cluster configurations: a 7/3 configuration (7 metal atoms on bottom, three on top) and a 5/4/1 configuration (5 atoms on bottom, 4 in the middle, 1 on top). Both schemes were tested, and the 5/4/1 arrangement proved to be the most energetically favorable on both surfaces. This 5/4/1 arrangement was then used to geometrically optimize the remaining 4 metal clusters to control for cluster configuration between the different BaZrO3metal heterointerfaces. Next, unhydrated cells were organized by positioning two yttrium dopants and one oxygen vacancy in the bottom four fixed layers of the perovskite. The defect positions were arranged identically for all BaO interfaces and all ZrO2 interfaces. The hydrated cells were then produced by positioning the yttrium dopants in the same locations as in the unhydrated cells, filling the oxygen vacancy, and adding two protons to the system. One proton was arranged in the bottom fixed layers of the perovskite slab (positioned in the same place for all BaO interfaces and all ZrO2 interface), while the second proton was located in various positions in the first three perovskite layers near the BaZrO3-metal heterointerface. The defect arrangements for both unhydrated and hydrated cells were also replicated at identical BaO-vacuum and ZrO2-vacuum interfaces. This provided a direct comparison between the vacuum and metal interfaces to determine the effect of the metal on proton stability. Proton stability was assessed by comparing changes in the hydration enthalpies between the BaZrO3-metal systems and the control, vacuum configurations: ∆!"#$$% &'(/*+, &'(/$-*+,

(Eq. 4)

∆%./"0 &'(%/*+, &"'123%/$-*+,

(Eq. 5)

∆4 ∆%./"0 ∆5

(Eq. 6)

A positive value of ∆4 represents a reduction of proton stability in the ith layer as a result of the introduction of the metal clusters. A negative ∆4 value denotes that the introduction of the metal to the BaZrO3 surface leads to an increase in proton stability in the ith layer.

To investigate the origins of the proton stability trends revealed from this hydration study an electronic density of states (DOS) analysis was completed. These calculations were carried out in a similar manner to the BaZrO3-vacuum interfacial study described above. Since the slab was not thick enough to distinguish a “bulk” oxygen, the p-band centers of the interfacial oxygens were calculated relative to a reference point. This reference atom (an oxygen ion) was chosen in the middle of the BaZrO3 slab and was used to normalize each of the p-band centers for the different systems. 3. RESULTS AND DISCUSSION 3.1 Proton Stability at the BaZrO3-Vacuum Interface Figure 2 depicts the hydration enthalpies for each proton configuration at both the BaO- and ZrO2- vacuum interfaces. Since the quantitative values of the hydration enthalpies were dependent on the location of the defects in the bottom fixed layers of the perovskite slab, the values are presented relative to the bulk hydration enthalpy, which was calculated as -0.5 eV. Here we have assumed that the bulk hydration enthalpy has been reached by the eighth layer from the perovskite surface, as the enthalpy change from layer seven to layer eight changes by less then 0.06 eV, approximately the error of the simulation.

3



As the proton is moved away from the BaO- and ZrO2 surfaces, it experiences a destabilizing effect when bonded to an oxygen atom in the first subsurface layer. Proton stability then increases slightly in the second subsurface layer (Layer 3) and decreases slightly in Layer 4. After layer 4 the perturbation in proton stability subsides and adsorption energies continues to stabilize toward the bulk adsorption energy, -0.5eV, as the proton is moved away from the perovskite surface. Other than at the ZrO2 surface, the data suggests that orientation plays a small role in the stability of the protonic defect.

2

ZrO2 up ZrO2 side ZrO2 down BaO up BaO down BaO side

1.5 1 0.5 0 0

1

2

3

4

5

6

7

8

9

10

-0.5 -1

Figure 3 depicts the p-band centers for each oxygen atom that was probed at both the BaO-vacuum and ZrO2-vacuum interfaces. The results of the p-band analysis revealed a similar, yet opposite, fluctuation in the p-band centers compared to the hydration enthalpy trend displayed in Figure 2. That is, p-band center trend analysis revealed high p-band centers for surface oxygens, followed by low p-band center values for oxygens in the first subsurface layers, and then a slight increase in p-band centers in the subsequent layer before leveling out into consistent p-band values toward the material bulk.

0 P-Band center relative to the bulk

Figure 2 reveals that a similar trend in proton stability occurs at both the (001) terminations. First, a stabilization of 0.1-1 eV occurs when the proton is bonded to an oxygen at the perovskite surface. At the BaO surface, the orientation of the surface proton (whether the proton is pointed up away from the surface or to the side perpendicular to the surface) has no impact on the stability of the proton. On the other hand, proton orientation does influence the stability of the proton at the ZrO2 surface. The results indicate that the proton experiences a significant stabilization, -0.8 to -1eV, when oriented parallel and upward at the perovskite surface, while the downward orientation experiences a less favorable but still stabilizing hydration enthalpy of -0.1 eV. An analysis of the relaxed structure reveals that the sideways proton reorients to point slightly upward, suggesting that the upward configuration is more favorable at the ZrO2 surface as opposed to down orientation.

-0.2 -0.4 -0.6 -0.8 -1 -1.2 -1.4 -1.6 -1.8 0

1

2

3

4 5 Layer

6

7

8

9

Figure 3. P-band centers plotted relative to the bulk p-band center of oxygen atoms at various distances from the (001) BaO () and ZrO2 () perovskite surfaces. To determine whether the fluctuations in p-band center directly correlate with hydration enthalpy, the data was plotted against each other in Figure 4. Specifically, the p-band center of each oxygen was plotted against the average value of the hydration enthalpies for the given oxygen. The resulting trend indicates that hydration enthalpy decreases, or becomes more favorable, as the p-band center of the oxygen increases.

-1.5 -2

Layer

Figure 2. Hydration enthalpy of protons bonded to oxygen ions near the BaO and ZrO2 surfaces relative to the perovskite bulk. Oxygen location ranges from surface oxygens (Layer 1) to bulk oxygens (Layer 8). Three proton orientations were tested throughout the cell: up toward the surface, sideways parallel to the surface, and down away from the surface.

To understand proton stability variations in the near surface environment of the BaZrO3 crystal, an analysis of the electron density of states was conducted on oxygen atoms at various depths within the slab. (An exemplary density of states plot can be found in Figure S1 in the Supporting Information). Specifically, the p-band structure was explored, as the s-band poorly describes the valence electrons available for bonding. As the p-band center, or the weighted average of the p-band, shifts to higher energy values, more electrons become available for bonding. In general, an increase in the valence band center typically leads to a greater adsorption strength between the atom and its adsorbate.

0 Hydration enthalpy relative to bulk (eV)

Relative hydration enthalpy (eV)

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 13

-0.2 -0.4 -0.6 -0.8 -1 -1.2 -1.4 -2

-1.5 -1 -0.5 P-band center relative to bulk P-band center

0

4


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. Hydration enthalpy vs. p-band center of oxygen atoms near the (001) BaO () and ZrO2 () surfaces relative to perovskite bulk.

The trend in hydration enthalpy versus p-band center is attributable to the effects of the oxygen valence electron energy on the stability of a proton to which it is bound. To make this explicit, we also calculated the relationship between p-band center and the proton adsorption energy on the BZY surfaces and found a strong linear correlation (see Supporting Information). Because these calculations involve the adsorption energy of a proton (rather than a proton-hydroxide pair), electroneutrality is not maintained; nevertheless, the results demonstrate the trends in proton stability can be accurately predicted using the oxygen pband center as a metric. The variation in electronic structure that alters proton stability at the perovskite surfaces is likely caused by band bending at the perovskite-vacuum interface, which induces a space charge effect. Upon cleavage of the symmetric bulk structure, electronic band bending must occur to form a surface state in equilibrium with the vacuum potential. This bending causes variations in the distribution of electrons within the atoms in the near interfacial environment. These data suggest that the band bending phenomenon that occurs at the BaZrO3 surface causes the electrons in the oxygen p-band to shift to higher energy levels at the surface, followed by lower energy levels in the first subsurface layer. This fluctuation persists within the first four layers of the interfacial region and then subsides. Electron distribution shifts to higher energy levels make more electrons available for bonding, thus increasing the stability of protons in this layer and increasing hydration favorability. On the other hand, less electrons are available for binding when the p-band distribution shifts to lower energy levels, thus destabilizing the protons and decreasing the favorability of hydration in these interfacial layers. The results from this surface interfacial study have multiple implications. First, the instability of protons in the first subsurface layer could provide potential insight into the reduced grain boundary conductivities seen experimentally. Recent examination of the perovskite GB region has revealed that a positive grain boundary core charge likely forms within these materials, creating a space charge layer with a diminished proton concentration.5,15,16 This study demonstrates proton instability in the first subsurface layer of the perovskite interface, creating a potential activation barrier for proton mobility to/from the surface. While the grain boundary interface is much more complex than the perovskitevacuum surface, our previous studies have shown that studying the simplified perovskite surface can yield important fundamental trends that may also translate to the more complicated GB region.20 Here we see that proton conduction in the GB interface may result from a combination between proton instability near perovskite interfaces as well as a positive grain boundary core charge determined from experiments. Furthermore, these results demonstrate a distinct difference in proton stability between the perovskite surface and the material bulk, likely due to disparate electron distributions in the oxygen atoms in these two environments. Previous studies have shown a correlation between bulk oxygen p-band centers and ORR catalytic activity.25 We show here that bulk oxygen p-band centers are likely not the most accurate descriptors for catalytic activity because oxygen electronic properties vary at the perovskite surface. While it is more difficult to evaluate the surface oxygens of perovskite oxides because the surface structures are not

precisely known, analyzing the most stable surfaces as determined by DFT may provide a more accurate descriptor of ORR activity. Finally, the relationship between p-band center and hydration enthalpy discovered in this study provides a potentially significant model to screen for perovskites with enhanced hydration properties. The immensely large number of combinations of perovskite oxides and dopant atoms is expensive and time consuming to explore experimentally. This p-band center model establishes a simple parameter that can be computed in silico to identify perovskite-dopant combinations that may exhibit enhanced hydration properties.

3.2 P-Band Analysis within Bulk Yttrium-Doped Barium Cerate The linear correlation between oxygen p-band center and hydration enthalpy revealed in the BaZrO3-vacuum interfacial study has potential implications toward understanding why hydration within perovskite oxides varies between systems. To determine whether this analysis can be extended to other environments, such as the bulk perovskite microenvironment, we carried out a similar DOS analysis on the bulk yttrium-doped barium cerate (BCY) system. The results of this study, depicted in Figure 5, reveal that the linear correlation between oxygen p-band center and hydration enthalpy found at the BaZrO3 surface also exists within the BCY bulk. While this trend does hold within each of the dopant concentrations tested, it is important to note that adsorption energies should not be compared between each of the doping levels, as electroneutrality was not maintained within the unit cells. Although this testing platform can allow us to make comparisons and observations between unit cells of the same size (eg. 50%, 25%, and 12.5%), it does not afford intercellular comparisons and thus cannot be used to predict optimal doping levels for enhanced adsorption energies. It is simply a convenient system to demonstrate how p-band center affects the energetics of proton adsorption on perovskite O atoms.

-2 Adsorption energy (eV)

Page 5 of 13

-2.5 -3 -3.5 -4 -4.5 -5 -5.5 -2

-1.5 P-Band center

-1

Figure 5. P-band center vs. proton adsorption energy of oxygen atoms in 50% (), 25% (), and 12.5% () yttrium-doped barium cerate (BCY).

5


Chemistry of Materials To gain a better understanding of the relationship between the pband center and proton adsorption energy within each of the BaCeO3 systems, we first take a closer look at the 50%-doped sample. Figure 6 depicts the unit cell and p-band center plot for 50%-doped BCY. In this unit cell, every oxygen ion is bonded with one cerium and one yttrium ion, thus allowing for a structural analysis irrespective of yttrium coordination. While 50%-doped BCY is not synthesized experimentally because B-site doping is limited to dopant concentrations less than about 25%, this system is computationally convenient to study due to its small unit cell.1 Since BaCeO3 is comprised of an orthorhombic crystalline lattice, there are variations between Ce-O bond lengths in different axes of the system. The experimental lattice constants for BaCeO3 are A=8.78 Å, B=6.23 Å, and C= 6.22 Å,44 so the CeO (or Y-O) bond lengths along the vertical axis of the pictured unit cell are longer than the bond lengths along the horizontal axis. The p-band vs. adsorption energy plot in Figure 7 distinguishes between the two groups of oxygens, where the “vertical” oxygens are those that reside in the vertical O-Ce-O axis and the “horizontal” oxygens are those that exist in the horizontal O-Ce-O axis. (a)

Figure 6. (a) The unit cell of 50% yttrium-doped barium cerate (BCY50), comprised of oxygen (red), barium (green), cerium (beige), and yttrium (blue) ions. (b) P-band center vs. adsorption energy of oxygens within BCY50 configured along the vertical () and horizontal () axes.

The results indicate a discrepancy between the p-band centers of the horizontal and vertical oxygen atoms. The oxygens situated on the horizontal axes, where the Ce/Y-O bond lengths are shorter, have a p-band center shifted higher toward the Fermi level than those oxygens positioned on the vertical axes. The higher p-band centers appear to contribute to more favorable proton adsorption on the horizontal oxygen ions. This behavior has been discussed previously. Kreuer suggests that the deviation from the cubic lattice structure within orthorhombic perovskites causes different binding energies between disparate lattice sites. These different binding sites lead to preferential proton diffusion in one direction with restricted mobility between different paths, leading to decreased proton mobility within the material.39,45 The results presented here provide clear electronic evidence to support this explanation. A closer investigation of the BCY25 (25% doped) results can provide further insight into how dopant coordination influences the electronic structure and hydration favorability of oxygen ions. Figure 7a depicts the BCY25 unit cell, where the two yttrium atoms have been substituted into two neighboring Ce ion lattice sites. This configuration results in two vertical oxygen ions bonded to the two yttrium dopant ions, while the remaining six vertical oxygens are coordinated with two cerium ions. Figure 7b displays the plot of p-band centers versus adsorption energies between each of the oxygen ions within the system. The results show that the two oxygens coordinated only to yttrium ions have the highest p-band centers as well as the most favorable adsorption energies within the system. The remaining vertical oxygen ions follow the same trend as in the BCY50 systems, showing lower p-band centers and less favorable adsorption energies compared to the horizontal oxygens.

(a)

(b)

-2 Adsorption energy (eV)

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 13

-2.2 -2.4 -2.6 (b)

-2.8 -3 -1.5

-1.4

-1.3 -1.2 P-Band center

-1.1

-1

6


Page 7 of 13

3.3 Proton Stability at the BaZrO3-Metal Heterointerface

-2.8 Adsorption energy (eV)

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


-2.9 -3 -3.1 -3.2 -3.3 -3.4 -3.5 -3.6 -1.8

-1.6

-1.4 -1.2 P-Band center

-1

Figure 7. (a) The unit cell of 25% yttrium-doped barium cerate (BCY25), comprised of oxygen (red), barium (green), cerium (beige), and yttrium (blue) ions. (b) P-band center vs. adsorption energy of oxygens within BCY25 configured along the vertical () and horizontal () axes. Horizontal oxygens with dashed outlines are coordinated to one yttrium. Vertical oxygens with dashed outlines are coordinated to two yttrium ions.

Building on the insights provided by the analysis of the BaZrO3vacuum interface and BaCeO3 bulk, we now consider the more complex BaZrO3-metal heterointerface. An example of a metal cluster at the BaO surface is depicted in Figure 8a, which also includes the three proton configurations probed in this study. The hydration enthalpies resulting from these proton arrangements near the BaO-vacuum and BaO-metal interfaces are displayed in Figure 8b, while the resulting proton stability changes due to the metal heterointerface are plotted in Figure 8c. The results show an improvement in proton stability due to the introduction of platinum, palladium, and silver metal nanoparticles at the BaO surface. The most pronounced enhancement in stability occurred in the first subsurface layer (Layer 2) for these metals. Palladium showed the greatest increase in proton stability near the heterointerfaces, while stability owing to the remaining metals varied in different interfacial layers with nickel leading to a destabilization of protons in the near perovskite surface. (a)

The horizontal oxygens can also be assessed in relation to the yttrium dopants. Eight horizontal oxygen ions are bonded to one cerium and one yttrium ion, while the remaining eight are bonded only to cerium ions. The results show that the “horizontal” oxygens coordinated with the yttrium ion have slightly higher pband centers, as well as more favorable adsorption energies, than those oxygens bonded only to cerium ions. It has long been theorized that yttrium dopants increase the basicity of the oxygen ions in the surrounding area, influencing the hydration favorability of those oxygens. The data presented here supports this hypothesis, revealing that yttrium coordination shifts the oxygen’s electron distribution toward the Fermi level, causing the oxygen to become more basic. This increases the number of electrons available for binding, thus increasing the favorability of proton adsorption. Additionally, this data supports a recent study by Yamazaki et. al that provides evidence for proton trapping within doped perovskites.46 The group found that proton-dopant association limits macroscopic proton transport within yttrium-doped barium zirconate. Protons bonded to yttrium-coordinated oxygen ions must overcome an additional association energy barrier of 29 kJ/mol (0.30 eV) to escape the dopant environment. This phenomenon can arise due to an electronic or geometric mismatch between the dopant and substituted ion.1 Our data provides evidence that protons are more stable on dopant-coordinated oxygen ions due to a shift in these oxygens’ p-band centers. This enhanced adsorption strength likely contributes to proton trapping that occurs within this and other perovskites. To determine the depth of proton trapping a simple analysis could be completed in a larger unit cell with a low dopant concentration. Overall, these results yield a valuable technique to identify dopant ions that may mitigate proton trapping effects. Researchers can screen for dopant ions that cause a minimal shift in the p-band centers of dopant-coordinated oxygen ions. Identifying perovskite-dopant systems that reduce proton preference for particular lattice sites should yield perovskites with enhanced macroscopic proton conductivities.

(b)

7



Hydration enthalpy (eV)

0.5 0 0

1

2

3

4

-0.5 -1 BaO-vac BaO-Ag BaO-Pd BaO-Pt BaO-Ni

-1.5 -2 Layer

(c) 0.5

0.0 0

1

2

3

4

(b)

-0.5

0 0

-1.5

-2.0

Layer

BaO-Ag BaO-Pd BaO-Pt BaO-Ni

Figure 8. (a) The unit cell of BaO-Pd heterointerface comprised of oxygen (red), barium (green), zirconium (light blue), palladium (dark blue), and hydrogen (white) ions. (b) Hydration enthalpy plot of protons bonded to oxygen ions in the first three layers near the perovskite interfaces: BaO-vacuum (), BaO-Ag (), BaONi (), BaO-Pd (), BaO-Pt (). (c) The change in hydration enthalpy between the BaO-vacuum and the BaO-metal interfaces. Symbols correspond to same metal heterointerfaces as given in (b).

An example of a metal cluster at the ZrO2 surface, as well as the hydration and stability results at the ZrO2-metal heterointerfaces, is depicted in Figure 9. Unlike at the BaO-metal heterointerfaces, the introduction of metal nanoparticles at the ZrO2 surface decreases proton stability in nearly all cases tested. The only stabilization that occurred was in the first subsurface layers near the palladium and platinum heterointerfaces.


-1.0

1

3

4

-1 ZrO2-vac ZrO2-Ag ZrO2-Ni ZrO2-Pt ZrO2-Pd

-1.5

Layer

(c) 1.50

1.00

ZrO-Ag ZrO-Ni ZrO-Pt ZrO-Pd

0.50

0.00 0 -0.50

(a)

2

-0.5

-2

Change in hyration enthalpy (eV)

Change in hydration enthalpy (eV)

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 13

1

2

3

4

Layer

Figure 9. (a) The unit cell of ZrO2-Pd heterointerface comprised of oxygen (red), barium (green), zirconium (light blue), palladium (dark blue), and hydrogen (white) ions. (b) Hydration enthalpy plot of protons bonded to oxygen ions in the first three layers near the perovskite interfaces: ZrO2-vacuum (), ZrO2-Ag (), ZrO2Ni (), ZrO2-Pd (), ZrO2-Pt (). (c) The change in hydration enthalpy between the ZrO2-vacuum and the ZrO2-metal interfaces.

8


Page 9 of 13

Symbols correspond to same metal heterointerfaces as given in (b). To understand why the addition of metal to the perovskite interface led to variable changes in proton stability in the near heterointerfacial environment, an electronic analysis of the p-band centers of oxygen ions near the heterointerface was completed. Figure 10 depicts the p-band centers of oxygens near the BaOand ZrO2- interfaces and their corresponding hydration enthalpies. (An exemplary density of states plot for an oxygen at the BaZrO3Pd heterointerface can be found in Figure S4 in the Supporting Information). At the ZrO2 surface, the p-band center acts as a more accurate predictor of hydration enthalpy. The p-band centers are lower near the ZrO2-metal heterointerfaces compared to the ZrO2-vacuum surface, which leads to a decrease in hydration enthalpy near these interfaces. This trend follows in accordance to the results from the perovskite-vacuum interfacial study. However, the p-band center does not as accurately predict proton stabilization at the BaO interfaces.

(a) 1.5 BaO-Ag BaO-Pd BaO-Ni BaO-Pt BaO-Vac


0.5

-0.5

-1.5

-2.5 -2.5

-2.0

-1.5 -1.0 -0.5 P-Band Center

0.0

0.5

(b) 0


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


ZrO-Ag ZrO-Pd ZrO-Ni ZrO-Pt ZrO-Vac

-0.5

-1

Upon a closer investigation into the geometrically optimized hydrated structures, it was found that the protons near the BaO surface were stabilized by forming either an O-H--M bond (in Layer 1) or O-H--O bond (in Layer 2) near the metal heterointerface when compared to the perovskite-vacuum surface. The final configurations of protons in Layers 1 and 2 at the BaOPd surface are depicted in Figure 11. We hypothesize that the addition of the metal nanocluster on the uniform perovskite structure alters how neighboring ions stabilize the proton. For example, it is well documented for previous computational studies that the perovskite lattice relaxes around a protonic defect, allowing the proton to form a weak hydrogen bond with a neighboring oxygen ions, forming an O-H--O bond. Furthermore, a comparison between the O-H—M bond lengths within different perovskite-metal systems revealed a correlation between bond length and hydration enthalpy, such that shorter stabilizing bonds led to more favorable hydration enthalpies. For example, the shortest bond distance occurred between the proton and palladium, a metal with a high affinity for hydrogen, which also produced the strongest proton stabilization. The p-band center model developed previously would not adequately predict this H-M stabilizing effect, as the model only relies on the electron distribution within the oxygen ions. Thus, it appears the p-band center model cannot be used as a predictor for hydration enthalpy in systems that introduce additional interactions in the proton binding site. In other words, the existence of an additional stabilizing interaction with the proton at this interface cannot be accounted for by a p-band center metric that considers only a single proton-oxygen interaction. Despite the complexity of identifying the underlying cause of proton stability at the BaZrO3-metal heterointerface, the results still present several potential implications. First, the stabilization of the proton in the first subsurface layer of the perovskite-metal heterointerface may decrease the energy barrier for diffusion in the interfacial region. Above we have identified an energy barrier of about 0.5 eV for diffusion toward the perovskite interface and a barrier ranging from 0.1-1 eV for proton transport away from this interface. If the introduction of a metal at the barium zirconate surface decreases this barrier it could facilitate proton transport both from and toward the TPB. However, an overstabilization of the proton in this first subsurface layer may cause a decrease in proton mobility if the proton is strongly enough adsorbed to become “trapped” at the interfacial site.

-1.5

-2 -4

-3

-2

-1

0

P-Band Center

Figure 10. (a) Plot of p-band centers of oxygen atoms and their correlating hydration enthalpies at BaO-metal heterointerfaces: BaO-Ag (), BaO-Ni (), BaO-Pd (), BaO-Pt (). (b) Plot of p-band centers vs. hydration enthalpies at ZrO2-metal heterointerfaces: ZrO2-Ag (), ZrO2-Ni (), ZrO2-Pd (), ZrO2Pt ().

a)

9



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 10 of 13

with higher p-band centers yield more favorable hydration enthalpies. This relationship between electron distribution and proton stability explains the fluctuations in hydration enthalpy in different layers of the BaO- and ZrO2- near interfacial environments. Furthermore, this investigation revealed several important insights into proton chemistry at the perovskite surface. The instability of protons in the first subsurface layers of the BaOterminated surface, as well as throughout the ZrO2 near surface environment, reveals a potential barrier for proton mobility to the perovskite interface. This could lead to reduced proton conduction across perovskite GB interfaces previously attributed to a positive grain boundary core charge.

b)

Figure 11. The geometrically optimized configurations of protons in a) Layer 1 and b) Layer 2 at the BaO-Pd heterointerface (only first three perovskite layers shown). The proton in the first layer forms a O-H--Pd bond while the proton in the second layer forms a O-H--O bond to stabilize the proton within the lattice.

Additionally, these results may help explain an enhancement in proton conductivity observed upon the incorporation of Pd metal nanoparticles into the bulk of BZY. According to a study completed by Tong et. al, the Pd nanoparticles introduced into the perovskite during synthesis resided at the grain boundary of the perovskite oxide and caused an increase in both GB and overall conductivity.47 The group hypothesized that a band bending effect at the heterointerface increased proton concentration within the perovskite at the perovskite-metal junction. As it is widely accepted that proton concentration at the perovskite GB is low due to a positive GB core charge, the incorporation of the Pd nanoparticles into the GB may have increased the proton concentration within the region, thus increasing proton conduction. The computational results presented above suggest the effect may additionally be caused by an enhancement in proton mobility across this region due to a decrease in the energetic barrier in the first subsurface layer near the perovskite interface.

The relationship between p-band center and proton adsorption energy was examined in bulk yttrium-doped barium cerate to substantiate whether the p-band center model could be used to understand hydration trends between different perovskite systems. The results not only yielded the same correlations between the pband center and proton adsorption strength, they also led to an enhanced understanding of the implications of structural properties on material hydration. First, a study of 50%-doped BCY confirmed that divergence from the cubic perovskite crystalline structure produces two types of oxygen lattice sites. The disparate electron distributions between these sites leads to an enhanced proton stability on one type of site, likely leading to restricted proton mobility and decreased overall proton conductivities within these perovskites. Furthermore, an analysis of the 25%-doped BCY sample revealed that dopant-coordinated oxygen ions have higher p-band centers and, thus, increased proton adsorption energies. The results from this study demonstrate how the p-band center can be used to understand variations in proton stability within disparate perovskite structures. Finally, an examination of various BaZrO3-metal heterointerfaces revealed that the introduction of various types of metals at the perovskite surface may enhance proton stability near the perovskite interface. We found that the p-band center model developed previously in this study does not accurately predict hydration enthalpies at the perovskite-metal interface due to the presence of different types of binding sites for the proton. Despite the complications in understanding the proton stability trends observed here, the evidence for proton stabilization at the perovskite-metal heterointerfaces supports the hypothesis that the incorporation of metal nanoparticles into the perovskite can change the mobility of protons in the near heterointerface environment. By exploiting this principle, nanoparticles may be able to be incorporated into regions of low proton conductivities to boost proton mobility and enhance proton conduction within perovskites.

AUTHOR INFORMATION Corresponding Author *Email: [email protected]

Phone: (303) 492-2418 Funding Sources 4. CONCLUSION The goal of this work was to investigate proton chemistry at perovskite interfaces using ab initio techniques. An electronic analysis at the BaZrO3-vacuum surface revealed that oxygen ions

National Science Foundation funds the Renewable Energy Materials Research Science and Engineering Center (Award DMR-0820518).

10


Page 11 of 13

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


ACKNOWLEDGMENT

16.

The authors acknowledge support from the National Science Foundation for the funding of the Renewable Energy Materials Research Science and Engineering Center (Award DMR0820518). Computational resources were provided by NSF-MRI Grant CNS-0821794, MRI-Consortium: Acquisition of a Supercomputer by the Front Range Computing Consortium (FRCC), with additional support from the University of Colorado and NSF sponsorship of the National Center for Atmospheric Research.

17.

Example density of states plot of an oxygen at the barium zirconate BaO (001) surface (Figure S.1). Adsorption energy of protons at the BaZrO3-vacuum interface (Figure S.2). Proton adsorption strengths versus p-band centers of oxygen ions at the BaZrO3-vacuum interface (Figure S.3). Exemplary density of states plot of an oxygen at the BaZrO3-Pd heterointerface (Figure S.4).

REFERENCES

3.

4. 5. 6. 7.

8. 9. 10. 11. 12.

13. 14.

15.

20.

22.

Supporting Information

2.

19.

21.

ASSOCIATED CONTENT

1.

18.

Kreuer, K. D. Annu. Rev. Mater. Res. 2003, 33, 333. Babilo, P.; Uda, T.; Haile, S. M. J.Mater. Res. 2007, 22, 1322. Azad, A. K.; Savaniu, C.; Tao, S.; Duval, S.; Holtappels, P.; Ibberson, R. M.; Irvine, J. T. S. J. Mater. Chem. 2008, 18, 3414. Iguchi, F.; Chen, C. T.; Yugami, H.; Kim, S. J. Mater. Chem. 2011, 21, 16517. Iguchi, F.; Sata, N.; Yugami, H. J. Mater. Chem. 2010, 20, 6265. Lee, Y. L.; Kleis, J.; Rossmeisl, J.; Shao-Horn, Y.; Morgan, D. Energy Environ. Sci. 2011, 4, 3966. Suntivich, J.; Gasteiger, H. A.; Yabuuchi, N,; Nakanishi, H.; Goodenough, J. B; Shao-Horn, Y. Nature Chem. 2011, 3, 546. Suntivich, J.; Perry, E. E.; Gasteiger, H. A.; ShaoHorn, Y. Electrocatalysis 2013, 4, 49. Maguire, E.; Gharbage, B.; Marques, F. M. B.; Labrincha, J. A. Solid State Ionics 2000, 127, 329. Bertrand, G. L.; Caboche, G.; Dufour, L. C. Solid State Ionics 2000, 129, 219. O'Hayre, R.; Barnett, D. M.; Prinz, F. B. J. Electrochem. Soc. 2005, 152, A439. Kjolseth, C.; Fjeld, H.; Prytz, O.; Dahl, P. I.; Estournes, C.; Haugsrud, R.; Norby, T. Solid State Ionics 2010, 181, 268. Iguchi, F.; Yamada, T.; Sata, N.; Tsurui, T.; Yugami, H. Solid State Ionics 2006, 177, 2381. Shirpour, M.; Rahmati, B.; Sigle, W.; van Aken, P.A.; Merkle, R.; Maier, J. J. Phys. Chem. C 2012, 116, 2453. Chen, C. T.; Danel, C. E.; Kim, S. J. Mater. Chem. 2011, 21, 5435.

23. 24.

25.

26.

27. 28. 29. 30. 31. 32. 33.

34. 35. 36. 37.

38. 39. 40. 41.

Shirpour, M.; Merkle, R.; Maier, J. Solid State Ionics 2012, 216, 1. Helgee, E. E.; Lindman, A.; Wahnstrom, G. Fuel Cells 2013, 13, 19. Lee, H. B.; Prinz, F. B.; Cai, W. Acta Mater. 2010, 58, 2197. Nyman, B. J.; Helgee, E. E.; Wahnstrom, G. Appl. Phys. Lett. 100 (2012). Tauer, T.; O'Hayre, R.; Medlin, J. W. J. Mater. Chem. A 2013, 1, 2840. Bockris, J. O.; Otagawa, T. J. Electrochem. Soc. 1984, 131, 290.. Wang, Y.; Cheng, H.-P. J. Phys. Chem. C 2013, 117, 2106. Choi, Y.; Mebane, D. S.; Lin, M. C.; Liu, M. Chem.Mater. 2007, 19, 1690. May, K. J.; Carlton, C. E.; Stoerzinger, K. A.; Risch, M.; Suntivich, J.; Lee, Y.-L.; Grimaud, A.; Shao-Horn, Y. J. Phys. Chem. Lett. 2012, 3, 3264. Lee, Y.-L.; Kleis, J.; Rossmeisl, J.; Shao-Horn, Y.; Morgan, D. Energy Environ. Sci. 2011, 4, 3966. Stoerzinger, K. A.; Risch, M.; Suntivich, J.; Lue, W. M.; Zhou, J.; Biegalski, M. D.; Christen, H. M.; Ariando; Venkatesandeh, T.; Shao-Horn, Y. Energy Environ. Sci. 2013, 6, 1582. Fehribach, J. D.; O'Hayre, R. SIAM J. Appl. Math. 2009, 70, 510. Malagoli, M.; Liu, M. L.; Park, H. C.; Bongiorno, A. Phys. Chem. Chem. Phys. 2013, 15, 12525. Kresse, G.; Furthmuller, J. Comput. Mater. Sci. 1996, 6, 15. Kresse, G.; Furthmuller, J. Phys. Rev. B 1996, 54, 11169. Kresse, G.; Hafner, J. Phys. Rev. B 1994, 49, 14251. Kresse, G.; Joubert, D. Phys. Rev. B 1999, 59, 1758. Perdew, J. P.; Chevary, J. A.; Vosko, S. H.; Jackson, K. A.; Pederson, M. R.; Singh, D. J.; Fiolhais, C. Phys. Rev. B 1992, 46, 6671. Monkhorst, H. J.; Pack, J. D. Phys. Rev. B 1976, 13, 5188. Heifets, E.; Ho, J.; Merinov, B. Phys. Rev. B 2007, 75, 155431. Ho, J.; Heifets, E.; Merinov, B. Surf. Sci. 2007, 601, 490. Cordero, F.; Craciun, F.; Deganello, F.; La Parola, V.; Roncari, E.; Sanson. A. Phys. Rev. B 2008, 78, 054108. Munch, W.; Seifert, G.; Kreuer, K. D.; Maier, J. Solid State Ionics 1996, 86-8, 647. Kreuer, K. D. Solid State Ionics 1999, 125, 285. Munch, W.; Kreuer, K. D.; Seifertli, G.; Maier, J. Solid State Ionics 1999, 125, 39. Longo, A.; Giannici, F.; Balerna, A.; Ingrao, C.; Deganello, F.; Martorana, A. Chem. Mater. 2006, 18, 5782.

11



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

42. 43. 44.

45.

46. 47.

Page 12 of 13

Munch, W.; Kreuer, K. D.; Seifert, G.; Maier, J. Solid State Ionics 2000, 136, 183. Tauer, T.; O'Hayre, R.; Medlin, J. W. Solid State Ionics 2011, 204, 27. Takeuchi, K.; Loong, C.-K.; Richardson, J. W.; Guan, J.; Dorris, S. E.; Balachandran, U. Solid State Ionics 2000, 138, 63. Kreuer, K. D.; Muench, W.; Ise, M.; He, T.; Fuchs, A.; Traub, U.; Maier, J. Ber. Bunsen Ges. Phys. Chem 1997, 101, 1344. Yamazaki, Y.; Babilo, P.; Haile, S. M. Chem. Mater. 2008, 20, 6352. Tong, J. H.; Subramaniyan, A.; Guthrey, H.; Clark, D.; Gorman, B. P.; O'Hayre, R. Solid State Ionics 2012, 211, 26.

12


Page 13 of 13

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


For Table of Contents

13


An ab Initio Investigation of Proton Stability at BaZrO3 Interfaces

Recommend Documents