Density Functional Theory Study of Hydrogen Electroadsorption on the Pt ( 110 ) surfaces

The hydrogen adsorption on the Pt(110) and Pt(110)-(1x2) electrode surfaces has been investigated. To gain insight into detailed atomistic picture on the equilibrium coverage and structure, we have constructed a lattice gas model by determining the on-site energy and the interaction parameters using the first principles total-energy calculation. Therein atop, fcc, short bridge, long bridge and R, T, F, F’ sites for H/Pt(110) and H/Pt(110)-(1x2) are covered by hydrogen atoms under various coverage conditions 0 ML < θ < 1 ML and the total-energy calculations are done for the (1x1) and (1x2) cells. The surface of (1×2) and (1×1) lateral unit cells. The convergence property with respect to the number of Pt layers and the k-point mesh are found. The comparison between different surface types are done. By comparing the calculated results with two different theoretical simulated data, SIESTA and VASP, we found good agreement between them.

Manuscript Received on July 13 th , 2016.Manuscript Revised December 06 th , 2016.This research is funded by Ho Chi Minh City University of Technology -VNU-HCM, under grand number T-KHUD-2016-71.
Among others, Pt(111) is the simplest surface where calculation can be done most accurately.In doing the theoretical calculation of H/Pt(111), it is worth mentioning that many forgoing calculation [12][13][14][15][16][17][18][19][21][22][23] did not lead to the same conclusion regarding the most stable adsorption site.Some studies showed that the top site is the most stable site [12,15,18], while others found that the fcc is more stable than the top [19,23].This happened despite the fact that those calculations commonly used the density functional theory (DFT) within the generalized gradient approximation (GGA) for the exchange-correlation energy.This is due to insufficient parameters for the DFT-GGA calculation, in particular, insufficient number of kpoints in the Brillouine zone integration and insufficient number of Pt layers for the slab model.Our previous research for H/Pt(111) started from accurate determination of the H adsorption energy within DFT-GGA.The calculated effective H-H interaction, or the g-value, was compared in good agreement with experiment [24].
The comparison nevertheless provides important insight into the H-adsorption, which prompts further theoretical investigation.For the H/Pt(110), the modeling is more complex.For the face-centered cubic FCC(110) surfaces, the unreconstructed (1×1) phase and the reconstructed (1×2) phase with missing-row exist.The (1×1) unit cell contains one substrate atom on the outermost row, the second and third layer atoms are still fairly exposed [24].The (1×2) unit cell contains four more or less exposed Pt atoms [25,26].In practical applications, the Pt catalyst is often finely dispersed in small particles embedded in a matrix and the active sites can be of various types, such as, edges where crystal facets meet.[25][26][27][28][29], reproducing thereby reasonable properties of the most stable adsorption site.Besides, the interaction of hydrogen with the Pt(110)-(1x1) surface has been also studied extensively both experimentally and theoretically [24,25,29,30,31].However, up to now, there is still considerable disagreement as to the chemisorption site of H on Pt(110).The usual assumption of highly coordinated hydrogen [32,33] sitting in the deep troughs of the missing rows was supported by work function measurements [24,31] and vibrational spectroscopy [34], but was challenged by the first direct structure-probing experiment (Helium atom scattering, HAS), which led to the proposal of a highly coordinated subsurface site [35].
Besides, to study the adsorption of H on the missing row Pt(110)-(1×2), Engstrom et al. [24] and Shern [31] carried out low-energy electron diffraction (LEED), temperature-programmed desorption (TPD) and the mirror electron microscope LEED -that can measure the work function change.They supported the usual assumption [17,18,24] of highly coordinated H sitting in the deep troughs of the missing rows.Stenzel et al. [34] also supported the result using the vibrational spectroscopy measurement.However, Kirsten et al. [35] gave another proposal of a highly coordinated subsurface site on the basis of a direct structure-probing experiment (Helium atom scattering, HAS).On the contrary, Zhang et al. [29] performed LEED experiments and DFT calculations to provide an evidence that β 2 -H is chemisorbed at the low coordinated short bridge site on top of the outermost Pt rows.Subsequently, Minca et al. [25] used TPD, quantitative LEED, and DFT to find a chemisorption site, called β 2state, on the outermost close-packed rows under the ideal coverage of 0.5 ML.Adsorption sites on the (111) microfacets, called β 1 -state, are occupied only at higher coverage.Note that the β 1 and β 2 states had been well described in Refs.[24,30,36].Most recently, Gudmundsdóttir et al. [28] used TPD measurements and DFT calculations to confirm that, at low coverage, the strongest binding sites are the low coordination bridge sites at the edge.At higher coverage, on the other hand, H is adsorbed on higher coordination sites either on the micro-facet or in the trough.Those various foregoing researches had motivated us to carefully study the H chemisorption sites.To proceed this study, it will be important to investigate the chemisorption site more thoroughly, including the typical and atypical sites.The first purpose of the present work is to determine the binding sites and obtain the converged DFT data.We then compute the adsorption isotherm for Pt(110)-(1×2) using two different simulation software, the SIESTA and the VASP, and compare their results with those obtained for Pt(110)-(1x1) surface.Our study nevertheless provides important insight into the Hadsorption, which prompts further theoretical investigation.

COMPUTATIONAL METHODS
We used SIESTA (Spanish Initiative for Electronic Simulations with Thousands of Atoms) package simulation.The linear combination of atomic orbitals (LCAO) and pseudopotential scheme implemented in SIESTA [25,26] for the first-principle electronic structure calculations.Then the plane wave and projector augmented wave (PAW) potentials [27,28] scheme implemented in VASP (Vienna Ab initio Software Package) [29,30,31] were used to supplement the SIESTA result.Fig. 1 shows the models and adsorption sites of the DFT calculation used for the calculation.We used the generalized gradient approximation (GGA) to the exchange-correlation functional due to Perdew, Burke, and Ernzerhof (PBE) for the DFT calculation [32].The surface irreducible Brillouin zone was sampled on the k-point mesh generated by the Monkhorst-Pack (MP) scheme [33].We used the repeated slab model to model the surface.The surface slab was separated from its periodic image by 16.6 Å, by which the interaction energy with the image can be reduced to 1 meV.

SIESTA calculation
We have adopted the following computational parameters for the SIESTA calculation.We used the double-zeta polarized (DZP) basic set, the mesh-cutoff of 200Ry.We employed the Fermi Dirac function with the electronic temperature of 300 K in carrying out the Brillouin zone integrations.We used the 200 meV value for the energy shift for the Pt, which determines the cutoff radius per angular momentum channel.For adsorbed H atoms, more extended basis is used in which we used the 60 meV value for the energy shift, and split norm of 0.53 for the second zeta.This ensures for us to obtain correct bond length and energy of H2 molecule in which is important for the long-range interactions [9].These standard computational parameters used in the SIESTA calculation had provided a reasonably accuracy both in the calculation of a bare Pt surface and a H2 molecule [9].The optimized lattice constant of 3.93 Å, which is in good agreement with the experimental bulk value of 3.924 Å [34] were used to construct the slabs.
The calculation of the H adsorbing surfaces was done for the following two sets of configurations.First, one H atom was adsorbed on the missing row Pt(110)-(1×2) surface of (1×2) lateral unit cell and on the Pt(110)-(1x1) surface of (1×1) lateral unit cell (Fig. 1).A vacuum equivalent to a twelve-layer slab separated the Pt slabs where the interlayer spacing was taken as 1.387 Å.The total energy was obtained after relaxing all the H and the Pt atoms of the upper four Pt-layers.This calculation was done mainly for the sake of comparing with previous calculations regarding the stability of the binding sites.Second, the surface of (1×2) and (1×1) lateral unit cells were used to investigate the convergence property with respect to the number of Pt layers and the k-point mesh.We used the spin-polarization calculations for all of the systems.In the Brillouin zone integration, 84 special k-points were used to sample the (12×12×1) MP grids for the (1×2) and (1×1) lateral unit cells.

VASP calculation
The VASP calculation was similarly done for two above sets of H-Pt configurations.Besides, we have used the k-point mesh ranging from (7×7×1) to (24×24×1) MP grids for the (1×2) lateral unit cell of H/Pt(110)-(1×2) system.And the number of Pt layers has changed from 5 to 19 layers when (12×12×1) MP grid was used for H/Pt(110)-(1x1) system.We have used the following computational parameters too.The plane wave cutoff energy was 400 eV, which is large enough to converge the total energy within the order of 1 meV per atom.The Brillouin zone integrations were carried out by employing the Gaussian smearing function with the width 0.02 eV.The optimized lattice constant of the bare missing row Pt(110)-(1×2) and the Pt(110)-(1x1) obtained from the VASP calculation is 3.92 Å.

DFT DESCRIPTION OF H ON THE PT SURFACES
Previous calculations showed that the energy associated with the various binding sites on the surface is strongly dependent on the ΘH.By adding the H-atoms to the surface one at a time, the surface is filled first at the strongest binding sites and finally at the weakest ones [24].In this context, we first tested the order of the adsorption sites where they get filled by calculating the hydrogen adsorption energy where Etot(NH) is the total energy of the Pt surface adsorbed with NH H atoms and EH2 is the total energy of the isolated H2 molecule.

H/Pt(110)-(1x2)
For H/Pt(110)-( 1x2) system, E ads shows that the short bridge site on the ridge (R) is the strongest adsorption site, then the on-top on the micro facet (F), the HCP hollow site (F') and finally the long bridge site in the trough (T) (see Table 1).This result is in agreement with the results of Zhang et al. [18] and Gudmundsdóttir et al. [24].Besides, Gudmundsdóttir et al. has shown that when the ridge has been filled, the preferred sites are the tilted on-top sites on the micro facets (F) followed by adsorption onto the long bridge sites in the trough (T).The filling of the trough sites forces the neighboring H-atoms to move from the on-top sites towards the HCP threefold hollow sites on the (111) micro facet (F').Secondly, we calculated the optimized Pt-H bond lengths for the H on the Pt(110)-(1×2) as shown in Table 2.We have confirmed that the results were affected by less than 1% when changing the number of Pt layers from five to nine.To obtain the converged value, we now investigate in detail the convergence property with respect to the number of Pt layers and k-points.Previous calculation for Pt(111) provided the dependence of the adsorption energy on k-point mesh and number of Pt layers [9].Therefore, in this work, the calculation was done similarly using (1×2) lateral unit cell, on which one H atom was let adsorb either on the R or on the F. Table 3 shows the calculated adsorption energy and Fig. 2 plots the adsorption energy on the F relative to that on the R, Eads.The table shows that the SIESTA calculation provides the adsorption energy systematically larger by 0.15 eV in magnitude, while the figure shows that they provide a similar dependence on the number of Pt layers as it changes from 5 to 19 layers when (12×12×1) MP grid was used.From the Fig. 2 we found that for the low Pt layers (less than 9), the value oscillates with large amplitude, then the oscillation is regular and periodic when taking 9 to 19 layers.It suggests that the converged value has already been determined well around 0.12 eV within the amplitude of the oscillation ( 40 meV) by taking these layers.Fig. 3 plots the dependence on k-points, which shows that the results for various number of Pt layers becomes very close to each other when using (16×16×1) MP grid.From these results we conclude that the converged E ads is located at around 0.12 eV.It means that the R obviously is more stable than F by that amount.This is our conclusion on the theoretical adsorption energy within the UHV surface and DFT-PBE for H/Pt(110)-(1x2) system.

H/Pt(110)
Similar calculation of E ads for H/Pt(110)-(1x2) system shows that the short bridge site is the strongest adsorption site, then the on-top site, the long bridge site and finally the fcc site (see Table 4).From the calculation data, the short bridge site (B) and the on-top site (OT) are the most stable sites on the surface.Therefore, in the next step, we calculated the optimized Pt-H bond lengths only for the B and OT sites of H on the Pt(110)-(1×2) (as shown in Table 5).We have also confirmed that the results were affected by less than 1% when changing the number of Pt layers from five to nine.Besides, base on the successful calculation for the converged value of H/Pt(110)-(1x2) system, we now similarly investigate the convergence property with respect to the number of Pt layers and k-points for H/Pt(110)-(1x1) system.The calculation was done similarly using (1x1) lateral unit cell, on which one H atom was let adsorb either on the B or on the OT.The Fig. 4 plots the adsorption energy on the OT site relative to that on the B site, ∆E ads , when the number of Pt layers was changed from 5 to 18 layers and (12×12×1) MP grid was used.From the Fig. 4 we found that for the low Pt layers (less than 10), the value oscillates with large amplitude, then the oscillation is regular and periodic when taking 10 to 18 layers.It suggests that the converged value has already been determined well around −0.11 eV within the amplitude of the oscillation ( 50 meV) by taking these layers.From these results we conclude that the converged ∆E ads is located at around −0.11 eV.It means that the B obviously is more stable than OT by that amount.

CONCLUSIONS
A converged first-principles DFT-GGA was used to investigate the hydrogen adsorption on the Pt(110) surfaces.It was shown that: for the H/Pt(110)-(1x2) system, the short bridge site on the ridge (R) is the strongest adsorption site, then the on-top on the micro facet (F), the HCP hollow site (F') and finally the long bridge site in the trough (T).The result is in consistent with the LEED experimental and the DFT theoretical results found in the literature.Besides, for the H/Pt(110)-(1x1) system, it was also shown that, the short bridge site is the strongest adsorption site, then the on-top site, the long bridge site and finally the fcc site.These determined sites are playing an important role to study the nature of H adsorbed on Pt surfaces, such as the interaction between hydrogen on the surface, and compare them with experimental data.Therefore, further investigation of the effective H-H interaction on the Pt(110) surfaces is required to compare the theoretical and experimental isotherm.Nghiên cứu lý thuyết phím hàm mật độ về sự hút bám điện tử của hydro trên các dạng bề mặt Pt(110) Trần Thị Thu Hạnh

Figure 3 .
Figure 3. k-point dependence of E ads .