First-Principles Design of Hydrogen Dissociation Catalysts Based on

May 8, 2014 - First-Principles Design of Hydrogen Dissociation Catalysts Based on Isoelectronic Metal Solid Solutions. Dong-Hwa Seo†, Hyeyoung Shinâ...
1 downloads 10 Views 385KB Size
Download PDF
Letter pubs.acs.org/JPCL

First-Principles Design of Hydrogen Dissociation Catalysts Based on Isoelectronic Metal Solid Solutions Dong-Hwa Seo,†,⊥ Hyeyoung Shin,‡ Kisuk Kang,† Hyungjun Kim,*,‡ and Sang Soo Han*,§ †

Department of Materials Science and Engineering, Research Institute of Advanced Materials (RIAM), Seoul National University, 1 Gwanak-ro, Gwanak-gu, Seoul 151-742, Republic of Korea ‡ Graduate School of Energy, Environment, Water, and Sustainability (EEWS), Korea Advanced Institute of Science and Technology (KAIST), Yuseong-gu, Daejeon 305-701, Republic of Korea § Center for Computational Science, Korea Institute of Science and Technology (KIST), Seoul 136-791, Republic of Korea S Supporting Information *

ABSTRACT: We report an innovative route for designing novel functional alloys based on first-principles calculations, which is an isoelectronic solid solution (ISS) of two metal elements to create new characteristics that are not native to the constituent elements. Neither Rh nor Ag exhibits hydrogen storage properties, whereas the Rh50Ag50 ISS exhibits properties similar to Pd; furthermore, Au cannot dissociate H2, and Ir has a higher energy barrier for the H2 dissociation reaction than Pt, whereas the Ir50Au50 ISS can dissociate H2 in a similar way to Pt. In the periodic table, Pd is located between Rh and Ag, and Pt is located between Ir and Au, leading to similar atomic and electronic structures between the pure metals (Pd and Pt) and the ISS alloys (Rh50Ag50 and Ir50Au50). From a practical perspective, the Ir−Au ISS would be more cost-effective to use than pure Pt, and could exhibit catalytic activity equivalent to Pt. Therefore, the Ir50Au50 ISS alloy can be a potential catalyst candidate for the replacement of Pt. SECTION: Surfaces, Interfaces, Porous Materials, and Catalysis

T

solution phases, several synthesis techniques, such as quenching to yield a metastable state or nanoscale fabrication to stabilize the nonequilibrium phases under ambient conditions, have been developed to prepare solid-solution phases at room temperature. For example, Rh−Ag and Pd−Pt, which are known to be immiscible pairs, were successfully synthesized as solid-solution structured alloy nanoparticles.14,15 Therefore, we expect that fabrication methods for a variety of solid-solution phases will become readily available. Similar atomic and electronic structures could lead to similar chemical properties. When the atomic structures of a pure metal and a solid-solution alloy composed of two metallic elements neighboring the pure metal are same, the electronic structure of the pure metal would be similar to one of the solidsolution alloy phases due to the free electrons of metals. Based on this hypothesis, one can design new functional materials that are cost-effective to use. In particular, one can develop a solidsolution structured alloy of two metal elements to create new characteristics that are not native to the constituent elements, which cannot be obtained by the classical alloying method. In this study, we have investigated the electronic structures and chemical properties of solid-solution phases of 50:50 Rh/ Ag and 50:50 Ir/Au and compared them with pure Pd and Pt

he design of novel functional materials such as catalysts has been a long-standing goal in the field of computational materials science.1−3 The successful development of an electronic structure calculation method using density functional theory (DFT) and its accurate predictive ability with relatively affordable computational costs are bringing this goal to fruition.4−6 Currently, in silico screening of materials has been widely pursued for applications involving lithium ion batteries,7,8 fuel-cell catalysts,9,10 gas storage,11 and others.12 Most of the procedures employed in computational materials design are based on combinatorics.13 Within a predefined search domain, we perform a computational test for nearly every possible combination. In particular, for the development of metallic catalysts, combinatorial materials screening has been widely employed to search for nonprecious (or at least lessprecious) or better performing metallic alloys.10 A solid-solution alloy in which metallic elements are homogeneously mixed at the atomic level can expand the tunability of the chemical and physical properties of the metallic systems beyond classical alloying techniques. Because the constituent elements are completely intermingled with each other at the atomistic level, variations in composition and/or combination allow us to continuously tailor the material’s properties. However, thermodynamically, only certain element combinations allow the formation of a solid-solution alloy, whereas others favor the formation of segregated phases. Although this problem has limited the utilization of solid© 2014 American Chemical Society

Received: March 10, 2014 Accepted: May 8, 2014 Published: May 8, 2014 1819

dx.doi.org/10.1021/jz500496e | J. Phys. Chem. Lett. 2014, 5, 1819−1824

The Journal of Physical Chemistry Letters

Letter

Figure 1. Simulated X-ray diffraction patterns for several Rh50Ag50 structures, where the patterns were simulated with λ = 0.55277 Å. The values in parentheses for each atomic structure indicate formation energies of the structure relative to pure Rh and Ag metals.

system with the PBE functional and compared the values with the reported RPBE (revised PBE) results (Ir: −2.9 eV, Pt: −2.4 eV, and Au: −3.5 eV),23 which indicated reasonable agreement. If two materials with the same atomic structure have similar electronic structures, they might exhibit similar chemical and physical properties. Therefore, we need to elucidate the atomic structures of the Rh50Ag50 and Ir50Au50 solid solution alloys along with their electronic structures. An experimental study found that the solid-solution structured Rh50Ag50 alloy displays a diffraction pattern consistent with single face-centered cubic (FCC) pure Rh and Ag phases, and its lattice parameter is between the values for Rh and Ag, indicating that the XRD pattern of the solid-solution structured Rh50Ag50 is similar to the XRD pattern of pure Pd.14 Figure 1 shows several atomic structures of the Rh50Ag50 phase and compares each of the Xray diffraction (XRD) patterns simulated with λ = 0.55277 Å, as used in the previous experiment.14 We also calculated the formation energies relative to pure Rh and Ag for each atomic structure. Of these structures, the FCC structure is most favorable, and the XRD pattern of this structure is most similar to the XRD pattern of Pd. Although the Rh and Ag atoms in the NaCl structure are mixed better than in the FCC structure, the NaCl structure is thermodynamically less favorable than the FCC structure, and the XRD pattern of the NaCl structure does not match well with that of Pd. Based on these results, the FCC structure shown in Figure 1 is a plausible structure of the solid-solution phase of Rh50Ag50. From the cluster expansion result shown in Figure S1 of the Supporting Information (SI), we find that the random structure is more stable than the layered one shown in Figure 1. An electron density of states (DOS) of the random structured Rh50Ag50 looks like a mixture of the DOS of pure Rh and Ag, which is unlike the DOS of Pd or the layered structure Rh50Ag50, shown in Figure S2. Based on an experiment14 claiming that the synthesized Rh50Ag50 solidsolution phase would exhibit an electronic structure similar to pure Pd, a random structure of the Rh50Ag50 alloy can be ruled out. The formation energy of the FCC structure is positive relative to the pure elements by 0.331 eV per f.u., which indicated that Rh and Ag intrinsically form a segregated domain

metals, respectively. In the periodic table, Pd is located between Rh and Ag, and Pt is located between Ir and Au. Therefore, the 50:50 Rh/Ag and 50:50 Ir/Au alloys have the same total number of electrons as pure Pd and Pt metals, respectively. We found that not only the electronic structures but also the chemical activities of the solid-solution phases were comparable to their isoelectronic pure metallic analogues. Specifically, the Rh50Ag50 alloy exhibits a hydrogen storage property similar to that of Pd, even though Rh and Ag have no hydrogen storage ability, and the Ir50Au50 alloy exhibits a hydrogen dissociation ability on the surface similar to that of Pt. Indeed, the Rh−Ag case has already been experimentally demonstrated,14 which supports the predictive ability of our calculations. Therefore, we propose that the synthesis of a solid-solution structured Ir50Au50 alloy could provide a potential replacement for the precious Pt catalyst.16 We performed density functional theory (DFT) calculations within a plane wave basis set using the Vienna Ab-initio Software Package (VASP)17 to investigate the atomic and electronic structures of the pure metals and solid-solution alloys, using the project augmented wave pseudopotential method18,19 considering scalar relativistic effects. We employed the Perdew−Burke−Ernzerhof (PBE)20 exchange-correlation functional. All of the calculated structures were fully relaxed under periodic boundary conditions. In modeling the surface slabs, we used optimized lattice parameters for bulk systems. With the lattice parameters, we built the slabs with four layers and a 15 Å vacuum. Then we relaxed the top two layers of the slabs while the bottom two layers remained fixed. We used a kinetic cutoff energy of 520 eV and a 10 × 10 × 10 k-point mesh and considered a spin-polarization effect. In addition, to evaluate the energy barriers for the H2 dissociation reaction on metal surfaces, we used the nudged elastic band (NEB)21 method with an additional 11 images to interpolate between the initial and final states. The present calculation with the PBE functional shows that the lattice parameters of Pd and Pt are 3.96 and 3.98 Å, respectively, which are similar to the experimental values (3.89 Å for Pd and 3.92 Å for Pt).22 We also calculated d-band centers relative to the Fermi energies (Ir: −2.3 eV, Pt: −2.2 eV, and Au: −3.4 eV) for each metallic 1820

dx.doi.org/10.1021/jz500496e | J. Phys. Chem. Lett. 2014, 5, 1819−1824

The Journal of Physical Chemistry Letters

Letter

Figure 2. Electronic structures of pure elements and their ISS analogues: (a,b) density of states where the Fermi energy is located at zero energy, and (c,d) charge density difference between pure elements and their ISS analogues along the [101] direction in their atomic structures. Here panels a and c correspond to the pair Pd and Rh50Ag50, and b and d correspond to the pair Pt and Ir50Au50. In a and c, black and gray atoms indicate Rh and Ag, respectively, and in b and d, orange and yellow atoms indicate Ir and Au, respectively.

and Rh50Ag50 systems, a similar characteristic is observed (Figures 2c and S2). This similarity between the electronic structures of the pure metals (Pd, Pt) and their ISS phases composed of neighboring elements can be explained by considering a free electron model of the metal. The electronic structure of these FCC metals near the Fermi level is primarily determined by the d-electrons (Figure S5). In comparison to the d-electrons, the contribution of the s-electrons near the Fermi level is less than 1%. The valence electrons experience a Coulombic attraction from the “nuclei + core-electrons” located at every FCC lattice site. The same lattice type (FCC) with comparable lattice parameters for the neighboring metals (i.e., (1) Rh: 3.842 Å, Ag: 4.165 Å vs Pd: 3.955 Å; (2) Ir: 3.876 Å, Au: 4.174 Å vs Pt: 3.976 Å) leads to lattice parameters for the ISS phases (i.e., Rh50Ag50: 4.005 Å, Ir50Au50: 4.019 Å) that are nearly identical to their isoelectronic pure metal analogues, resulting in similar locations of the Coulomb attraction centers. Although the extent of the Coulombic potential is either slightly larger (for the right neighboring element of Ag or Au) or slightly smaller (for the left neighboring element of Rh or Ir) than the pure metal case, the valence electron should experience a similar extent of Coulombic field in both the ISS phase and its isoelectronic pure metal analogue because (1) the solid-solution phase yields a good-mixture of these heterogeneous Coulomb attracting centers and (2) the screening of the nuclear charge is quite substantial for the transition metal elements, such that the small perturbation of the nuclei charge has an even smaller effect on the dynamics of the valence electrons. In particular, when the total number of electrons is preserved (i.e., isoelectronic case), the Fermi energy should be located at a very similar position

structure in the bulk phase. However, the nanosizing of such an intrinsically immiscible alloy can lead to a homogeneous solidsolution structure that is readily stable near room temperature, as shown in previous experiments.14 For the Ir50Au50 system, we performed a similar procedure and found that it has the same FCC structure as Rh50Ag50, as shown in Figure S3. After determining the atomic structures of the Rh50Ag50 and Ir50Au50 solid−solution alloys, we investigated the DOS of each system to compare their electronic structures with their isoelectronic pure metal analogues (Pd and Pt) (Figure 2a,b). We found that the electronic structure of Pd or Pt is very similar to the electronic structure of the isoelectronic solid solution (ISS) phase of Rh50Ag50 or Ir50Au50 near the Fermi energy, respectively.24 In addition, we also calculated line profiles of the charge density difference25 for pure Pd and Pt and their isoelectronic analogues (Rh50Ag50 and Ir50Au50) along the [101] direction in their structures (Figure 2c,d). For Ir50Au50 (Figure 2d), the (0,0,0) and (1,0,1) sites are occupied by the Ir atom, and the (0.5,0,0.5) site is occupied by the Au atom. Because the number of electrons in the Pt atom is higher than in the Ir atom and lower than in the Au atom, the charge density difference at (0,0,0) and (1,0,1) is positive, and the charge density difference at (0.5,0,0.5) is negative. However, the charge density differences at a certain distance (approximately 1 Å) between the two atoms for the Pt and Ir50Au50 are nearly zero, which indicates that the numbers of valence electrons in the Pt and Ir50Au50 systems are similar in the bonding regime. We note that the homogeneous phase, either Ir or Au, does not exhibit such similar charge density distributions in the bonding regime (Figure S4). For the Pd 1821

dx.doi.org/10.1021/jz500496e | J. Phys. Chem. Lett. 2014, 5, 1819−1824

The Journal of Physical Chemistry Letters

Letter

within the band structure. Therefore, we find that the ISS phase of Rh50Ag50 or Ir50Au50 has an electronic structure similar to Pd or Pt metal near the Fermi energy. Assuming that the catalytic activities of the metals fundamentally originate from their electronic structure, one can easily hypothesize that an ISS with a similar electronic structure would possess similar physics and chemistry. To explore this idea, we calculated the ability of the Rh50Ag50 ISS to store hydrogen and the ability of the Ir50Au50 ISS to perform as a hydrogen dissociation catalyst because Pd and Pt are wellknown hydrogen storage and hydrogen dissociation metals, respectively. The Pd metal stores hydrogen via the formation of metal hydride (i.e., PdHx), where the hydrogen atoms are absorbed into the interstitial sites of the FCC lattice at either the octahedral sites (degeneracy is 4 within a unit cell) or the tetrahedral sites (degeneracy is 8 within a unit cell). In Table 1,

Rh50Ag50 and Pd during hydrogen storage is quite similar (Pd: 4.7% and Rh50Ag50: 7.5%). Based on this result, we expect that the stability of the ISS alloys would be good. We further investigated H2 dissociation on top of the Ir50Au50 ISS surface to explore its potential as a hydrogen activating catalyst. We considered the (111) surface of the Ir50Au50 ISS, and calculated an energy profile for the H2 dissociation reaction on the Ir50Au50 (111) surface using the NEB theory. Then, the energy profile on the Ir50Au50 ISS was compared to the energy profiles on the (111) surfaces of pure Pt, Ir, and Au, as shown in Figure 3. On the (111) surface of the Ir50Au50 ISS, both Ir

Table 1. Formation Enthalpy (ΔHf) for Hydrogen Absorption in Pd, Rh, Ag, and Rh50Ag50 Calculated by DFT Calculationsa ΔHf

ΔHf

ΔHf

ΔHf

Pd

Rh

Ag

Rh50Ag50 Rh50Ag50-HO 8 Rh50Ag50-HT −191 Rh50Ag50160 2HT

Pd−HO Pd−HT Pd-2HT

−132 −208 162

Rh−HO Rh−HT

90 177

Ag−HO Ag−HT

728 422

Figure 3. Calculated energy along the minimum-energy reaction path for H2 dissociation on the (111) surface of Ir, Pt, Au, and Ir50Au50. The calculations were done for a supercell geometry with one H2 molecule per √3 × √3 surface unit cell on one side of metal slabs consisting of four atomic layers. Here the Ir50Au50 alloy shows similar catalytic behavior to Pt, although pure Ir and Au metals do not.

a Here HO and HT indicate the octahedral and tetrahedral sites in the FCC lattice, respectively. The formation enthalpy is shown in meV per formula unit.

and Au atoms coexist. Therefore, we considered several adsorption sites (11 cases) for the two H atoms on the surface to identify the most favorable site. The most preferred site is where each H atom’s position is between two adjacent Ir atoms, as shown in Figure S7. Along this reaction path, the calculated energy barrier (Ea) over the Ir50Au50 (111) surface is nearly zero, which is comparable to 0.03 eV over the Pt (111) surface.27 In addition, the dissociation reaction energy (ΔHrxn) is −0.84 eV/2H, which is very close to −0.87 eV/2H over the Pt (111) surface.27 We also note that pure Au metal does not catalyze the H2 dissociation reaction, which is confirmed in our calculations. The Ea values are 0.46 and 1.09 eV over the Ir (111) surface and the Au (111) surface, respectively, and ΔHrxn for the Au metal is positive (endothermic reaction), although ΔHrxn for the Ir metal is negative (exothermic reaction), as previously found in an experiment.30 The catalytic activity of the Ir50Au50 alloy can be well explained by the d-band theory.29,31−35 The d-band model developed by Nørskov and co-workers has been used successfully to explain trends in the reactivity of transitionmetal and alloy surfaces. We investigated the DOS for H atomically chemisorbed on the (111) surfaces of Ir, Pt, Au, and Ir50Au50 ISS alloy in Figure 4. Here, the Ir50Au50 alloy shows a similar DOS feature to Pt. The H 1s-metal d bonding resonances for both of the Ir50Au50 and Pt are at energies between −5 and −10 eV for Pt, and the H 1s-metal d antibonding resonances for the metal systems are above the Fermi energies. For the Au case, the H 1s-metal d bonding resonances are located at similar energies to ones for Ir50Au50 and Pt; however, some antibonding resonances can be found just below the Fermi energy, which leads to weak interaction of H atoms on the Au surfaces shown in Figure 3. For the Ir case,

we show the calculated enthalpies of formation (ΔHf) using the Pd + x/2 H2 → PdHx reaction for the cases when the hydrogen atoms occupy all of the octahedral sites (Pd−HO), half of the tetrahedral sites (Pd−HT), and all of the tetrahedral sites (Pd− 2HT). For Pd−HO and Pd−HT, the ΔHf values are calculated to be negative (favorable), which is consistent with previous theoretical results.26 The experimental neutron scattering result supports octahedral occupation,26 i.e., Pd−HO, although ΔHf of Pd−HT is more negative than ΔHf of Pd−HO in our calculations. This slight discrepancy between theory and experiment is due to the lack of zero-point energy corrections, which are considerable in these systems,26 and/or imperfection in the exchange-correlation functional. However, it is clear that DFT can predict the hydrogen storage ability of Pd. For the Rh50Ag50 ISS, as shown in Table 1, we found that hydrogen is absorbed into half the tetrahedral sites with a substantially negative ΔHf (i.e., −191 meV/f.u., compared with −132 meV/f.u. of Pd−HO or −208 meV/f.u. of Pd−HT), where the hydrogen atom bonds to two Rh and two Ag atoms. This result is even more interesting because the ΔHf values for the formation of RhHx or AgHx are highly positive (see Table 1), indicating that the Rh or Ag metal alone cannot exothermically form a metal hydride, which is consistent with previous experiments.14 Indeed, Kubota and his co-workers demonstrated that nanoparticles of a 50:50 solid solution of Rh−Ag with an FCC lattice exhibit hydrogen storage ability.14 This result provides a good example to support that ISSs with similar electronic and structural properties will inherit the chemical properties of their pure metal analogues. In addition, we find no phase transformation in the Rh50Ag50 alloy during its interaction with hydrogen, and the volume expansion of 1822

dx.doi.org/10.1021/jz500496e | J. Phys. Chem. Lett. 2014, 5, 1819−1824

The Journal of Physical Chemistry Letters

Letter

Figure 4. Density of one-electron states (DOS) (solid line) for H atomically chemisorbed on the (111) surface of Ir, Pt, Au, and IrAu. For comparison, the DOS (dashed line) for the four clean metal surfaces are shown. As indicated by the gray-shading, only states below the Fermi energy (which is the energy zero in all cases) are filled.

■

the antibonding resonances are above the Fermi energy; however, the bonding resonances are found at higher energies (−3 to −7 eV) than for Ir50Au50 and Pt. In addition, we calculated d-band centers for Pt and Ir50Au50 ISS alloy in Figure S8 and found that the two systems show very similar d-band centers (−2.2 eV for Pt and −2.1 eV for the Ir50Au50 alloy), which definitely verifies their similar catalytic activity. Due to the similar electronic structures of Pt and Ir50Au50, they can exhibit similar H2 dissociation behaviors. Moreover, in Figure S9, we clearly see that the dissociated H atoms occupy similar sites on the (111) surfaces of Pt and Ir50Au50 with a similar M− H bond distance. According to previous experiments36,37 regarding thin Au films grown on Ir (111), hydrogen atoms can exist at the interfaces between Ir and Au. This phenomenon definitely indicates that the catalytic effects of ISS alloys could be maintained if Ir and Au atoms are mixed well in the crystals. Therefore, the catalytic effects of the Ir−Au ISS alloy could be maintained in the composition range of Ir25Au75 to Ir75Au25 (see Figure S10). Additionally, we computed the projected DOS of the surface monolayer of Ir25Au75 on Ir(111) and found it to be similar to the projected DOS of the Ir50Au50 ISS alloy surface monolayer and Pt surface monolayer (see Figure S11). In conclusion, the hypothesis that an ISS with an electronic structure comparable to its pure metal analogue would exhibit similar physics or chemistry is true, at least for the cases involving the Rh−Ag and Pd pair and the Ir−Au and Pt pair. This result provides a new design principle for the development of functional metallic systems with better catalytic activity and/ or at a reduced cost. We emphasize that the design principle based on an ISS is distinct from the design principle based on the classical alloying method. The latter is typically pursued to enhance or amplify the desirable characteristics of the constituent elements, while the former is employed to create new characteristics that are not native to the constituent elements. From a practical perspective,16 the Ir−Au ISS would be more cost-effective to use than pure Pt while displaying similar catalytic activity. Therefore, the Ir50Au50 ISS alloy can be a potential catalyst candidate for the replacement of Pt. We expect that a design strategy based on the ISS principle can provide innovative insight into the development of new functional materials.

ASSOCIATED CONTENT

S Supporting Information *

Cluster expansion method for the RhAg structures, atomic structures for the IrAu system, electronic structures of Pd/ RhAg and Pt/IrAu, and chemisorption sites of two H atoms on the Ir50Au50 and Pt (111) surfaces. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected] (H.K.). *E-mail: [email protected] (S.S.H.). Present Address ⊥

(D.-H.S.) Department of Materials Science and Engineering, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139, United States.

Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS S.S.H. is grateful for the financial support from the Korea Institute of Science and Technology (Grant No. 2E24320 & 2E24630). H.K. acknowledges the support by the Global Frontier R&D Program (2013-073298) on Center for Hybrid Interface Materials (HIM) funded by the Ministry of Science, ICT & Future Planning. This work was supported by the Supercomputing Center/Korea Institute of Science and Technology Information with supercomputing resources including technical support (KSC-2012-C3-049).

■

REFERENCES

(1) Yu, W.; Porosoff, M. D.; Chen, J. G. Review of Pt-Based Bimetallic Catalysis: From Model Surfaces to Supported Catalysts. Chem. Rev. 2012, 112, 5780−5817. (2) Viswanathan, V.; Hanse, H. A.; Rossmeisl, J.; Nørskov, J. K. Unifying the 2e− and 4e− Reduction of Oxygen on Metal Surfaces. J. Phys. Chem. Lett. 2012, 3, 2948−2951. (3) Jacobsen, C. J. H.; Dahl, S.; Clausen, B. S.; Bahn, S.; Logadottir, A.; Nørskov, J. K. Catalyst Design by Interpolation in the Periodic Table: Bimetallic Ammonia Synthesis Catalysts. J. Am. Chem. Soc. 2001, 123, 8404−8405.

1823

dx.doi.org/10.1021/jz500496e | J. Phys. Chem. Lett. 2014, 5, 1819−1824

The Journal of Physical Chemistry Letters

Letter

(4) Shan, B.; Wang, L.; Yang, S.; Hyun, J.; Kapur, N.; Zhao, Y.; Nicholas, J. B.; Cho, K. First-Principles-Based Embedded Atom Method for PdAu Nanoparticles. Phys. Rev. B 2009, 80, 035404(1)− 035404(8). (5) Sha, Y.; Yu, T. H.; Liu, Y.; Merinov, B.; Shirvanian, P.; Goddard, W. A., III. Oxygen Hydration Mechanism for the Oxygen Reduction Reaction at Pt and Pd Fuel Cell Catalysts. J. Phys. Chem. Lett. 2011, 2, 572−576. (6) Pan, M.; Brush, A. J.; Pozun, Z. D.; Ham, H. C.; Yu, W.-Y.; Henkelman, G.; Hwang, G. S.; Mullins, C. B. Model Studies of Heterogeneous Catalytic Hydrogenation Reactions with Gold. Chem. Soc. Rev. 2013, 42, 5002−5013. (7) Ceder, G.; Hautier, G.; Jain, A.; Ong, S. P. Recharging Lithium Battery Research with First-Principles Methods. MRS Bull. 2011, 36, 185−191. (8) Ceder, G. Opportunities and Challenges for First-Principles Materials Design and Applications to Li Battery Materials. MRS Bull. 2010, 35, 693−701. (9) Greeley, J.; Jaramillo, T. F.; Bonde, J.; Chorkendorff, I.; Nørskov, J. K. Computational High-Throughput Screening of Electrocatalytic Materials for Hydrogen Evolution. Nat. Mater. 2006, 5, 909−913. (10) Studt, F.; Abild-Pedersen, F.; Bligaard, T.; Sørensen, R. Z.; Christensen, C. H.; Nørskov, J. K. Identification of Non-Precious Metal Alloy Catalysts for Selective Hydrogenation of Acetylene. Science 2008, 320, 1320−1322. (11) Lin, L.-C.; Berger, A. H.; Martin, R. L.; Kim, J.; Swisher, J. A.; Jariwala, K.; Rycroft, C. H.; Bhown, A. S.; Deem, M. W.; Haranczyk, M.; Smit, B. In Silico Screening of Carbon-Capture Materials. Nat. Mater. 2012, 11, 633−641. (12) Curtarolo, S.; Hart, G. L. W.; Nardelli, M. B.; Mingo, N.; Sanvito, S.; Levy, O. The High-Throughput Highway to Computational Materials Design. Nat. Mater. 2013, 12, 191−201. (13) Hart, G. L. W.; Curtarolo, S.; Massalski, T. B.; Levy, O. Comprehensive Search for New Phases and Compounds in Binary Alloy Systems Based on Platinum-Group Metals, Using a Computational First-Principles Approach. Phys. Rev. X 2013, 3, 041035(1)− 041035(33). (14) Kusada, K.; Yamauchi, M.; Kobayashi, H.; Kitagawa, H.; Kubota, Y. Hydrogen-Storage Properties of Solid−Solution Alloys of Immiscible Neighboring Elements with Pd. J. Am. Chem. Soc. 2010, 132, 15896−15898. (15) Kobayashi, H.; Yamauchi, M.; Kitagawa, H.; Kubota, Y.; Kato, K.; Takata, M. Atomic-Level Pd−Pt Alloying and Largely Enhanced Hydrogen-Storage Capacity in Bimetallic Nanoparticles Reconstructed from Core/Shell Structure by a Process of Hydrogen Absorption/ Desorption. J. Am. Chem. Soc. 2010, 132, 5576−5577. (16) From BASF website. http://apps.catalysts.basf.com/apps/ eibprices/mp/ (accessed June 27, 2013). The prices of Ir, Pt, and Au are $925, $1318, and $1239 per troy ounce (oz t), respectively. Here 1 oz t is equivalent to 0.0311034768 kg. (17) Blöchl, P. E. Projected Augmented-Wave Method. Phys. Rev. B 1994, 50, 17953−17979. (18) Kresse, G.; Joubert, D. From Ultrasoft Psedopotentials to the Projector Augmented-Wave Method. Phys. Rev. B 1999, 59, 1758− 1775. (19) Kresse, G.; Furthmuller, J. Efficiency of Ab-Initio Total Energy Calculations for Metals and Semiconductors Using a Plane-Wave Basis Set. Comput. Mater. Sci. 1996, 6, 15−50. (20) Perdew, J. P.; Burke, K.; Ernzerhof, M. Generalized Gradient Approximation Made Simple. Phys. Rev. Lett. 1996, 77, 3865−3868. (21) Sheppard, D.; Terrell, R.; Henkelman, G. Optimization Methods for Finding Minimum Energy Paths. J. Chem. Phys. 2008, 128, 134106(1)−134106(10). (22) Kittel, C. Introduction to Solid State Physics, 7th ed.; John Wiley & Sons, Inc.: New York, 1976; p 23. (23) Gajdoš, M.; Eichler, A.; Hafner, J. CO Adsorption on ClosePacked Transition and Noble Metal Surfaces: Trends from Ab Initio Calculations. J. Phys.: Condens. Matter 2004, 16, 1141−1164.

(24) To quatitatively compare the similarity of electronic structures, paticularly near Fermi energy (EF), we defined the following measure:

∫

ΔDOS2 − 1 = {

[DOS2(E) − DOS1(E)]2 g (E ; σ ) dE}1/2

g (E ; σ ) =

2 2 1 e−(E − EF) /2σ σ 2π

which compares two densities of states (DOS1 and DOS2) with higher weights near EF within range of σ. Within 3 eV near EF, ΔDOSIr−Pt(σ) = 1.83, ΔDOSAu−Pt(σ) = 2.80, and ΔDOSIr50Au50−Pt(σ) = 1.06, which clearly shows the similarity between DOS of suggesting Ir50Au50 alloy and DOS of Pt near EF. We further find that ΔDOSRh−Pd(σ) = 2.43, ΔDOSAg−Pd(σ) = 5.16, and ΔDOSRh50Ag50−Pd(σ) = 2.08, showing the similarity between DOS of Rh50Ag50 alloy and DOS of Pd near EF. (25) The charge density difference is calculated by the charge density of the pure metals (Pt or Pd) minus that of their ISSs (Ir50Au50 or Rh50Ag50). (26) Caputo, R.; Alavi, A. Where Do the H Atoms Reside in PdHx Systems? Mol. Phys. 2003, 101, 1781−1787. (27) These results are similar to the previous reports (refs 28 and 29). According to ref 28, the energy barrier and enthalpy for H2 dissociation over Pt (111) is 0.03 and −1.10 eV, respectively. (28) Lee, K.; Kim, Y.-H.; Sun, Y. Y.; West, D.; Zhao, Y.; Chen, Z.; Zhang, S. B. Hole-Mediated Hydrogen Spillover Mechanism in Metal−Organic Frameworks. Phys. Rev. Lett. 2010, 104, 236101(1)− 236101(4). (29) Hammer, B.; Nørskov, J. K. Why Gold is the Noblest of All the Metals. Nature 1995, 376, 238−240. (30) Hagedorn, C. J.; Weiss, M. J.; Weiberg, W. H. Dissociative Chemisorption of Hydrogen on Ir(111): Evidence for Terminal Site Adsorption. Phys. Rev. B 1999, 60, R14016−R14018. (31) Hammer, B.; Nørskov, J. K. Electronic Factors Determining the Reactivity of Metal Surfaces. Surf. Sci. 1995, 343, 211−220. (32) Mavrikakis, M.; Hammer, B.; Nørskov, J. K. Effect of Strain on the Reactivity of Metal Surfaces. Phys. Rev. Lett. 1998, 81, 2819−2822. (33) Greeley, J.; Nørskov, J. K.; Mavrikakis, M. Electronic Structure and Catalysis on Metal Surfaces. Annu. Rev. Phys. Chem. 2002, 53, 319−348. (34) Nørskov, J. K.; Rossmeisl, J.; Logadottir, A.; Lindqvist, L.; Kitchin, J. R.; Bligaard, T.; Jónsson, H. Origin of the Overpotential for Oxygen Reduction at a Fuel-Cell Cathode. J. Phys. Chem. B 2004, 108, 17886−17892. (35) Nørskov, J. K.; Bligaard, T.; Rossmeisl, J.; Christensen, C. H. Towards the Computational Design of Solid Catalysts. Nat. Chem. 2009, 1, 37−46. (36) Ferrin, P. A.; Kandoi, S.; Zhang, J.; Adzic, R.; Mavrikakis, M. Molecular and Atomic Hydrogen Interactions with Au-Ir Near-Surface Alloys. J. Phys. Chem. C 2009, 113, 1411−1417. (37) Okada, M.; Nakamura, M.; Moritani, K.; Kasai, T. Dissociative Adsorption of Hydrogen on Thin Au Films Grown on Ir{111}. Surf. Sci. 2003, 523, 218−230.

1824

dx.doi.org/10.1021/jz500496e | J. Phys. Chem. Lett. 2014, 5, 1819−1824