Journal of Alloys and Compounds 580 (2013) S6–S9
Contents lists available at SciVerse ScienceDirect
Journal of Alloys and Compounds journal homepage: www.elsevier.com/locate/jalcom
Water co-adsorption and electric ﬁeld effects on borohydride structures on Os(1 1 1) by ﬁrst-principles calculations Mary Clare Sison Escaño a,⇑, Ryan Lacdao Arevalo b, Elod Gyenge c, Hideaki Kasai b a
Graduate School of Engineering, University of Fukui, 3-9-1 Bunkyo, Fukui 910-8507, Japan Department of Precision Science & Technology and Applied Physics, Osaka University, 2-1 Yamadaoka, Suita, Osaka 565-0871, Japan c Department of Chemical and Biological Engineering, The University of British Columbia, Vancouver, BC, Canada V6T 1Z3 b
a r t i c l e
i n f o
Article history: Available online 17 February 2013 Keywords: Borohydride Osmium Water monomer Electric ﬁeld Density functional theory Hydrogen generation
a b s t r a c t Periodic density functional theory calculations are performed to investigate the nature of the BH4ad and its interaction with H2Oad in the presence of homogenous electric ﬁeld. We observed a signiﬁcant charge polarity of BH4ad on Os(1 1 1) and such property could explain the electrostatic interaction with water monomer (Had) with its HOH plane parallel to the surface. This interaction changes the BHad molecular structure to BH3ad + Had. In the presence of homogenous electric ﬁeld, the water co-adsorption effect is reduced due to the stabilization of H2Oad on the surface and the deviation of the O–H bond from the plane, decreasing the electrostatic interaction between BH4ad and H2Oad. These fundamental ﬁndings imply accessible control of borohydride structures on an electrode surface, which could be relevant for direct borohydride fuel cell (DBFC) and reversible hydrogen storage/release applications. Ó 2013 Elsevier B.V. All rights reserved.
1. Introduction Borohydrides are group of compounds with high hydrogen content (i.e. 10.6 wt% in NaBH4 form) [1,2]. Thus, signiﬁcant number of research has been devoted to investigate their applicability for hydrogen generation. Furthermore, because of high theoretical energy density, the direct electricity generation from borohydride ion has also been extensively investigated [2–4]. When used as a hydrogen source, borohydride undergoes hydrolysis reaction (1): BH 4 + 4H2O = BðOHÞ4 + 4H2, where the hydrogen is then supplied to a fuel cell. On the other hand, as an anodic fuel for direct borohydride fuel cell (DBFC), borohydride is oxidized through the following reaction (2): BH (E° = 4 + 8OH = BO2 + 6H2O + 8e 1.24 VSHE). In the latter method of power generation, the main issue is the co-occurrence of (1) and (2) reactions on the anode. The presence of atomic hydrogen on the anode leads to a ‘‘mixed’’ potential of reactions (2) and (3) H2 + 2OH = 2H2O + 2e (E° = 0.828 V) or to H2 evolution, which reduces the coulombic efﬁciency. Achieving borohydride oxidation at more negative potentials with no hydrogen evolution is an outstanding challenge since typical catalyst known for high activity for borohydride oxidation (e.g. Pt, Ni, Pd) are also highly active for hydrogen evolution. On the other hand, in the former method (hydrogen storage and release), the hydrolysis reaction is exothermic and occurs even without a catalyst at pH < 9. To prevent H2 formation upon standing, ⇑ Corresponding author. Tel.: +81 776 27 9802; fax: +81 776 27 9742. E-mail address: [email protected]
(M.C.S. Escaño). 0925-8388/$ - see front matter Ó 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jallcom.2013.02.018
the borohydride solution is typically maintained as a strongly alkaline solution. Now, H2 is signiﬁcantly produced only when these solutions comes in contact with catalyst. A number of noble and non-noble transition metals/alloys have been identiﬁed to be catalytically active towards hydrolysis reaction: Ru, Pt, Pt–Ru, Pt–Pd, Raney Ni and Co [5–10]. One of the main targets of this method is a rapid (but carefully controlled) H2 generation at ambient temperatures (and even down to 0 °C) without mechanical compression, addition of water, acidity or heat. In some reports, the H2 yield can be manipulated using applied voltage [11,12]. Given the above scenarios, whether the borohydride act as energy carrier or as a storage medium, the hydrolysis reaction (1) on a catalyst under relevant electrode potentials is an integral process, which should be well understood. Because of the complexity of the catalytic hydrolysis reaction of borohydride, which also involves solid-phase dissolution and liquid-phase transfer of the reactant and by-product, the fundamental understanding of the reactions occurring on the catalyst surface itself is often inaccessible by fundamental electrochemical studies. It is the aim of the present paper to investigate the borohydride and water co-adsorption on a catalyst surface under the inﬂuence of the electric ﬁelds. Here, we focus on the changes in the B–H bond as a result of borohydride interaction with the catalyst, water monomer and applied voltage. Based on the previous work on the adsorption of borohydride on 4d and 5d metals , it was found that the dissociative adsorption of BH4ad on Pt(1 1 1), Ir(1 1 1) and Pd(1 1 1) yielding BHad and 3Hads on surface, to some extent, corroborates with the hydrogen evolution experimentally observed on these metals [3,14–16]. Here, the
M.C.S. Escaño et al. / Journal of Alloys and Compounds 580 (2013) S6–S9
BH4ad represents the surface bound adsorbate which results from the oxidative adsorption of borohydride anion (BH 4 ): BH4 + (meta site) ? BH4ad + e. This initial electrochemical step has been conﬁrmed in electrochemical experiments on both noble and nonnoble metals [3,6,11,14,15,17]. Interestingly however, an equally stable borohydride is observed in Os(1 1 1), albeit in a molecular structure but with a B–H bond elongated with respect to the gas phase. This implies possible vulnerability of the molecule structure on surface to changes in its environment. Experimentally, the tendency of Os to catalyze hydrogen generation is seen to depend on many factors such as the solvent, catalyst structure and applied potential . Aside from the fact that Os is four to ﬁve times cheaper than Pt, Os-based catalysts exhibit a strong potential for a more controlled hydrolysis reactions, hence, the appropriateness of Os for the above-mentioned fundamental studies.
2. Computational method The fundamentals of the BH4ad interaction with Os and with co-adsorbed H2Oad are studied using spin-polarized density functional theory (DFT) [18,19], implemented in the Vienna ab-initio Simulation Package (VASP) [20–23]. The generalized gradient approximation (GGA) of Perdew, Burke and Ernzerhof (PBE)  is used for the exchange–correlation functional and the projector augmented wave (PAW) method [25,26] is employed to describe the ion-valence electron interactions. A supercell with a 4-layer fcc(1 1 1) metal slab and 15 Å of vacuum is used to simulate the metal surface. The slab is optimized by relaxing the two topmost layers. A 3 3 surface unit cell is employed corresponding to an adsorbate coverage of 1/ 9 ML with respect to the number of surface atoms. This coverage corresponds to small concentration of borohydride typical in experiments [2–4]. Brillouin-zone integrations are performed on a grid of (4 4 1) Monkhorst–Pack k points with a smearing of Methfessel–Paxton method . A plane-wave expansion with a cutoff of 400 eV is used throughout the calculations. The most stable conﬁguration of BH4ad on Os(1 1 1) is determined by ﬁrst placing the center of mass (boron) on four major adsorption sites (fcc-hollow, hcp-hollow, bridge and top) at 1.80 Å vertical distance from the surface. This is denoted as ‘‘down’’ conﬁguration – a trigonal BH4ad where the 3H’s spans toward the surface. The ‘‘up’’ conﬁguration, on the other hand is when the 3H’s point away from the surface (see inset of Fig. 1). Then, an in-plane rotation of the molecule in increments of 60° explores the potential minimum. A total of 16 conﬁgurations are considered. This is an important potential energy scanning method to carefully locate the most stable structure on the metal surface. The conjugate gradient minimization method  was used for the structural relaxation of the adsorbate and the top-two layers of the metal slab. Optimization is terminated when the Hellman–Feynman forces acting on each atom dropped below 0.01 eV Å3. The calculated equilibrium lattice constant for Os is 3.85 Å in excellent agreement with previous DFT calculations . The binding energy, Eb of the adsorbate on the metal surface is deﬁned as: Eb = ET EM Eg, where ET is the total energy of the adsorbate/metal system, EM is the total energy of the bare metal slab and Eg is the energy of the gas phase adsorbate. The activation barriers are obtained using nudged elastic band method (NEB). The changes particularly in the structure of the BH4ad due to interaction with adsorbed H2Oad under
the inﬂuence of homogenous electric ﬁeld are discussed. The most stable conﬁguration of BH4ad + H2Oad adlayer on Os(1 1 1) is obtained after calculating the total energies of all the possible symmetric conﬁgurations on the surface. The geometry and the details of the interaction will be discussed in Section 3. The uniform electric ﬁeld is applied to BH4ad + H2Oad adlayer on Os(1 1 1) system as well as on BH4ad/ Os(1 1 1) by introducing a dipole sheet in the middle of the vacuum region which then polarizes the slab [30,31]. The application of the electric ﬁeld is expounded in the next sections.
3. Results and discussion 3.1. Atomic and electronic structure of BH4ad/Os(1 1 1) We shown in Fig. 1a the most stable geometry of BH4ad on Os(1 1 1). The molecule prefers the hcp site and the three B–H bonds span towards the top site in a ‘‘down’’ conﬁguration (thus called hcp-top). The three B–H bonds (B–H1, B–H2, B–H3) are stretched by 19% with respect to the gas-phase. The B–H bond length is 1.49 Å. This bond length is stable as the breaking of B– H bond happens at 2.00 Å with an activation barrier of 0.20 eV. The binding energy is 4.59 eV. Other sites with less binding energies are: fcc-top (4.47 eV), bridge-top (4.35 eV), bridge-hollow (4.33 eV), hcp-bridge (3.54 eV), fcc-bridge (3.38 eV), which are all in ‘‘down’’ conﬁgurations. The ‘‘up’’ conﬁguration is not preferred. These large binding energies are expected since the reference gas phase is not stable. For the most stable conﬁguration (hcp-top), all the three B–H bonds lie almost parallel to the surface. An electron localization  plot shown in Fig. 1b suggests that there is signiﬁcant charge polarity in these bonds. Thus, we verify speculations on the preservation of the charge polarity of the molecule on metal surface . 3.2. Interaction with water molecule on Os(1 1 1) The most stable geometry for the BH4ad + H2Oad adlayer on Os(1 1 1) is shown in Fig. 2a. The H2O monomer is found to be stable since the activation barrier for H2Oad decomposition into Had and OHad on surface is 0.31 eV. Next, the change in the BH4ad structure as a consequence of the BH4ad and H2Oad interaction is: BH4ad ? BH3ad + Had. The H1 is already dissociated from the B atom. The B–H1 distance is 2.43 Å. This process is in fact spontaneous. The activation barrier previously obtained for BH4ad alone disappears in the presence of water monomer. This breaking of B–H1 bond is seen to be due to the electrostatic interaction with H2Oad owing to the high charge polarity of the adsorbed borohydride. This conjecture can be veriﬁed by the density of states projected
Fig. 1. (a) Stable BH4ad conﬁguration on Os(1 1 1). The center of mass of the adsorbate (boron) is at hcp site with the three B–H bonds spanning towards the top sites. (b) Electron localization plot showing the charge polarity of along B–H bonds. ELF value of 1 indicates high electron localizability.
M.C.S. Escaño et al. / Journal of Alloys and Compounds 580 (2013) S6–S9
Fig. 2. (a) The most stable BH4ad + H2Oad adlayer structure on Os(1 1 1). The dissociated H atom distance from B is shown. (b) The density of states projected on the surface atoms (LDOS) before (dashed lines) and after (solid lines) H2O adsorption.
on the surface metal atoms (LDOS) after H2O monomer adsorption (Fig. 2b). We note that the d-states of Os are not modiﬁed upon H2O adsorption. Hence, there is no contribution from the change in the surface electronic structure to the BH4ad structure due to H2O adsorption. The electrostatic nature of the interaction in the BH4ad + H2Oad layer on Os(1 1 1), which involves repulsion of H1 from H of H2Oad sufﬁces to break the B–H1 bond, considering the fact that B–H1 bond has already been signiﬁcantly elongated by Os. A test on adding another water monomer on the surface results to breaking of another B–H bond. Thus, the trend on B–H breaking upon increase of water coverage is predictable based on the knowledge of the effect of one water monomer. The BH4ad + H2Oad layer system therefore sufﬁces in capturing the mechanism of B–H breaking in experiments where high coverage of water molecules can be expected.
3.3. Electric ﬁeld effects The plane-averaged potential in the unit cell at different intensities of the electric ﬁeld is shown in Fig. 3a. We note a potential drop in the vacuum region but none in the bulk region, which indicates that the polarization does not affect charge distribution in the bulk. For instance, a 0.50 V/Å ﬁeld (the negative sign only indicates the direction of the ﬁeld, which is outward from the surface) leads to a potential drop of 2.3 eV on both sides of the slab as shown in the inset ﬁgure. This plot is steep and smooth which indicates a non-spurious charge mixing due to the induced dipole layer. Furthermore, a 1.00 V/Å ﬁeld leads to a potential drop of 4.5 eV which is lower than workfunction of Os(1 1 1). The changes in the structure of BH4ad (focusing on B–H1 bond length) due to applied electric ﬁeld are shown in Fig. 3b. The elongation is also
Fig. 3. (a) Plane-average potential in the unit cell for various intensities of the electric ﬁelds. (b) Change in the B–H1 bond length as a function of electric ﬁeld intensity.
M.C.S. Escaño et al. / Journal of Alloys and Compounds 580 (2013) S6–S9
with water monomer on the surface. This interaction signiﬁcantly changes the BHad molecular structure to BH3ad + Had. However, the water co-adsorption effect is reduced under the inﬂuence of electric ﬁeld due to the stabilization of H2Oad on the surface and the deviation of the O–H bond from the plane, which decreases the electrostatic interaction between BH4ad and H2Oad. In general, the change in the structure due to water and electric ﬁelds suggests important factors to manipulate in order to control hydrolysis of BH4ad relevant for DBFC and hydrogen storage/release applications. Acknowledgments M.C. Escaño would like to thank Japan Society for Promotion of Science (JSPS) and Special Coordination Funds for the Promotion of Science and Technology of the Ministry of Education, Culture and Sports (MEXT) through the Tenure Track Program for Innovative Research, University of Fukui for research funds and R.L. Arevalo would also like to extend gratitude to MEXT for scholarship. Fig. 4. Change in the B–H1 bond length due water co-adsorption in the presence of electric ﬁeld (square data series) and change in the H2O–M (metal) distance with respect to electric ﬁeld intensity (circle data series).
observed in B–H2 and B–H3 bonds, which is the same as in B–H1. We note that the decrease in the B–H1 bond length is 0.05 Å in 1 V/Å increase in electric ﬁeld (i.e. the B–H1 bond changes from 1.49 Å in the ‘‘no-electric ﬁeld’’ case to 1.44 Å in the presence of a strong electric ﬁeld). This is considered minimal. Next, when the electric ﬁeld is applied on BH4ad + H2Oad layer on Os(1 1 1), we note that that the change in the B–H1 bond length with respect to that of BH4ad/Os under the same electric ﬁeld intensity is significant (i.e. decrease of 0.88 Å in 1 V/Å in Fig. 4). For smaller increase in electric ﬁeld such as 0.25 V/Å, the suppression of the effect of the adsorbed water due to the electric ﬁeld can already be noted. Here, the B–H1 bond changes from 2.43 Å in the ‘‘noelectric ﬁeld’’ case to 1.62 Å under this electric ﬁeld intensity. Thus, the breaking of B–H bond can be inhibited upon electric ﬁeld application. The explanation for this is depicted in Fig. 4, which show the H2Oad–metal distance. Here, we note that when the electric ﬁeld is increased, the distance of the water from the surface is decreased, and the O–H bond deviates from the planar conﬁguration due to the negative direction of the electric ﬁeld. As a consequence, the electrostatic interaction between the two adsorbates is reduced, suppressing the B–H1 breaking. 4. Conclusions We conducted periodic density functional theory calculations to investigate the geometric changes in the BHad structure upon H2Oad co-adsorption with and without the presence of homogenous electric ﬁeld. We found a signiﬁcant charge polarity of BH4ad on Os(1 1 1) and such property allows the electrostatic interaction
References                                 
H. Dai, Y. Liang, L. Ma, P. Wang, J. Phys. Chem. C 112 (2008) 15886–15892. B.H. Liu, S. Suda, J. Alloys Comp. 454 (2008) 280–285. E. Gyenge, Electrochim. Acta 49 (2004) 965–978. J. Ma, N.A. Choudhury, Y. Sahai, Renew. Sust. Energy Rev. 14 (2010) 183–199. S.C. Amendola, S.L. Sharp-Goldman, M.S. Janjua, M.T. Kelly, P.J. Petillo, M.J. Binder, J. Power Sources 85 (2000) 186–189. J.S. Zhang, T.S. Fisher, J.P. Gore, D. Hazra, P.V. Ramachandran, Int. J. Hydrogen Energy 31 (2006) 2292–2298. Y. Kojima, K. Suzuki, K. Fukumoto, M. Sasaki, T. Yamamoto, Y. Kawai, H. Hayashi, Int. J. Hydrogen Energy 27 (2002) 1029–1034. U.B. Demirci, F. Garin, J. Alloys Comp. 463 (2008) 107–111. P. Krishnan, T.H. Yang, W.Y. Lee, C.S. Kim, J. Power Sources 143 (2005) 17–23. ~ a-Alonso, A. Sicurelli, E. Callone, G. Carturan, R. Raj, J. Power Sources 165 R. Pen (2007) 315–323. Ö. Sahin, H. Dolas, M. Özdemir, Int. J. Hydrogen Energy 32 (2007) 2330–2336. Ö. Sahin, H. Dolas, M. Kaya, M.S. Izgi, H. Demir, Int. J. Energy Res. 34 (2010) 557–567. M.C. Escaño, E. Gyenge, R. Arevalo, H. Kasai, J. Phys. Chem. C 115 (2011) 19883–19889. M. Simoes, S. Baranton, C. Coutanceau, J. Phys. Chem. C 113 (2009) 13369– 13376. V. Kiran, T. Ravikumar, N.T. Kalyanasundaram, S. Khrishnamurty, A.K. Shukla, S. Sampath, J. Electrochem. Soc. 157 (2010) B1201–B1208. H. Dong, R. Feng, X. Ai, Y. Cao, H. Yang, C. Cha, J. Phys. Chem. B 109 (2005) 10896–10901. V.W.S. Lam, E.L. Gyenge, J. Electrochem. Soc. 155 (2008) B1155. P. Hohenberg, W. Kohn, Phys. Rev. 136 (1964) B864–B871. W. Kohn, L.J. Sham, Phys. Rev. 140 (1965) A1133–A1138. G. Kresse, J. Furthmüller, Comput. Mater. Sci. 6 (1996) 15–50. G. Kresse, J. Furthmüller, Phys. Rev. B 54 (1996) 11169–11186. G. Kresse, J. Hafner, Phys. Rev. B 47 (1993) 558–561. G. Kresse, J. Hafner, Phys. Rev. B 49 (1994) 14251. J.P. Perdew, K. Burke, M. Ernzerhof, Phys. Rev. Lett. 77 (1996) 3865–3868. P.E. Blöchl, Phys. Rev. B 50 (1994) 17953–17979. G. Kresse, J. Joubert, Phys Rev. B 59 (1999) 1758–1775. M. Methfessel, Phys. Rev. B 40 (1989) 3616–3621. M.P. Teter, M.C. Payne, D.C. Allan, Phys. Rev. B 40 (1989) 12255–12263. G.S. Karlber, Phys. Rev. B 74 (2006) 153414–153417. J. Neugebauer, M. Schefﬂer, Phys. Rev. B 46 (1992) 16067–16080. P.J. Feibelman, Phys. Rev. B 64 (2001) 125403–125406. A.D. Becke, K.E. Edgecombe, J. Chem. Phys. 92 (1990) 5397–5404. G. Rostamikia, M.J. Janik, J. Electrochem. Soc. 156 (2009) B86–B92.