Advertisement

Electrocatalysis

, Volume 8, Issue 6, pp 605–615 | Cite as

Electrodeposition of Ag Overlayers onto Pt(111): Structural, Electrochemical and Electrocatalytic Properties

  • Ludwig A. Kibler
  • Khaled A. Soliman
  • Alan Plumer
  • Christopher S. Wildi
  • Eric Bringley
  • Jonathan E. Mueller
  • Timo Jacob
Original Article

Abstract

Epitaxially grown Ag overlayers have been fabricated by electrochemical deposition onto Pt(111). The electrochemical behaviour of these Ag overlayers has been studied by cyclic voltammetry, and their adsorption properties are significantly influenced by the underlying Pt(111) substrate and markedly different from those of Ag(111). A characteristic voltammetric peak for OH adsorption on pseudomorphic Ag islands has been observed for alkaline solution. A deposition–dissolution hysteresis in the underpotential deposition region for the Ag bilayer suggests exchange processes between subsurface Ag and Pt atoms. Theoretical DFT calculations confirm the stability of a pseudomorphic Ag monolayer. However, it is shown for two and three Ag layers that the formation of sandwich structures is theoretically favoured, i.e. Ag layers tend to be separated by single Pt layers. While Ag displaces hydrogen adsorbed in the underpotential region, the activity of Ag monolayers for the hydrogen evolution reaction (HER) is very close to that of Pt(111). Also, Tafel slopes for HER on the first pseudomorphic Ag monolayer on Pt(111) and for blank Pt(111) are almost identical, whereas thicker overlayers are more Ag-like. It is shown by theoretical calculations for the case of an Ag monolayer on Pt(111) that hydrogen can be adsorbed on the Pt subsurface layer.

Graphical Abstract

Keywords

Pt(111) Ag(111) Single crystals Ag overlayers Cyclic voltammetry Hydrogen evolution reaction 

Introduction

Elementary reaction steps in electrocatalytic reactions are influenced by a huge variety of parameters [1, 2]. For this reason, well-defined and well-characterized model systems are expedient to obtain solid relations between the geometric, electronic and compositional structure of an electrode surface and its electrocatalytic activity for a given reaction. Besides the use of single-crystal electrodes, tangentially strained model surfaces have attracted considerable interest [3, 4]. The latter are easily obtained for bimetallic systems, either by surface enrichment of a component of an alloy single crystal, e.g. PtRu(111) by taking advantage of segregation effects upon heating [5, 6] or by electrochemical deposition of a foreign metal onto a single-crystalline substrate [4, 7].

In both cases, one-component overlayers can be generated, which are often pseudomorphic with the underlying substrate [5, 8]. Thus, metallic surfaces with lattice constants deviating from the bulk values by up to few percent due to compression or expansion are obtained. Differences between the electrochemical properties of these overlayers and those of massive metal surfaces are typically explained in terms of two effects [9]:

The ligand effect, which results from the chemical interaction between substrate and the electrochemically active pseudomorphic monolayer, and the geometric effect, which arises from changes in lateral interatomic spacing, i.e. tangential strain caused by pseudomorphism or epitaxial growth. Both effects alter the electronic structure of the surface component, which causes changes in adsorption properties of reactants, intermediates and chemisorbed spectator species. Tuning the binding energy of adsorbates plays a decisive role in altering and tailoring electrocatalytic activities, because adsorption energies are often related to activation barriers as described by linear free enthalpy relations [10].

Fundamental electrocatalytic studies of segregated surfaces of alloy single crystals are rather scarce [6, 11]. In contrast, examples for the numerous electrochemical and electrocatalytic studies of overlayers with tangential strain electrodeposited onto single-crystal surfaces involve Pd monolayers [4, 12, 13, 14], Pt monolayers [15, 16] and Ag monolayers [17, 18].

In the following, we present a combined theoretical and experimental case study of structural and electrochemical properties of Ag overlayers on Pt(111). We explore the process of electrochemical silver deposition onto Pt(111) surfaces and try to link the electrocatalytic activity for the hydrogen evolution reaction (HER) to the surface structure. While the first pseudomorphic Ag monolayer is stable, it will be seen that the stability of an Ag bilayer is limited and processes of surface alloy formation and structural rearrangement become important.

Experimental

The electrochemical measurements were performed with a conventional three-electrode glass cell. A saturated calomel electrode (SCE) and a platinum wire were used as reference and counter electrodes, respectively. All potentials in this paper are referred to the SCE scale, except for hydrogen evolution reaction (HER) measurements where the RHE scale was used according to E(RHE)/V = E(SCE)/V + 0.242V + 0.059∙pH. In order to protect the solution from chloride contaminations, the SCE was placed in a separate compartment. The solutions were prepared from H2SO4 (Merck, Suprapur), NaOH (Sigma-Aldrich, ≥99.9995%), Ag2SO4 (Fluka, 99.5%), AgNO3 (Sigma-Aldrich, 99.9999%) and ultrapure water (Milli-Q, 18.2 MΩ cm at 25 °C, TOC <1 ppb) and purged with nitrogen prior to each experiment. The diameter of the Ag(111) and Pt(111) crystals used for cyclic voltammetry measurements was 4 mm. The electrodes were annealed before each measurement by inductive heating in the presence of a mixture of nitrogen and hydrogen for less than 1 min and cooled down slowly below 100°C before contacting with 0.1 M H2SO4. After being checked for cleanliness and surface quality, the electrode was transferred to a second cell for Ag deposition. Ag was deposited on Pt(111) in two different solutions: 0.1 M H2SO4 + x M Ag2SO4 for underpotential deposition (upd) experiments and 0.1 M NaOH + 10−6 M AgNO3 for overpotential deposition during potential cycling in the potential range from −0.4 to −0.95 V. The Ag deposition in alkaline solution allowed for the monitoring of the gradual increase in Ag coverage from the change of OH and H adsorption by cyclic voltammetry. In this case, there was no need to transfer the electrode to another cell, e.g. filled with 0.1 M NaOH, to characterize the surface after deposition. The disadvantage of overpotential deposition is the difficulty to form a complete Ag monolayer on Pt(111), because the second layer starts to grow before completion of the first layer. When the desired Ag coverage was achieved, the electrode was thoroughly rinsed with oxygen-free ultrapure water and transferred to another electrochemical cell for further measurements with Ag+-free solutions. Ag adlayers were anodically stripped in 0.1 M H2SO4 to double-check their coverage and to ensure their stability during the measurements.

Computational Methods

All periodic DFT calculations were performed with SeqQuest (Schultz, P.A., unpublished; for a description of the method, see [19]), employing spin polarization and the exchange-correlation functional developed by Perdew, Burke and Ernzerhof (PBE) [20] within the generalized gradient approximation. Surfaces were modelled using infinitely periodic slabs consisting of seven or eight layers of atoms with (111) surface orientations in 1 × 1, 2 × 2 or 3 × 3 surface unit cells. Norm-conserving pseudopotentials replaced the 68 core electrons in Pt and 36 core electrons in Ag [21, 22]. The outermost d and s orbitals were described using a double-ζ plus polarization basis set. Grid spacings of 0.24 Bohr were used for numerical integration in real space and a Brillouin zone sampling equivalent to 12 × 12 in the periodic directions for a 1 × 1 unit cell for reciprocal space. The convergence criterion was set at 0.0005 Ry, and the geometric convergence was set at 0.001 Ry/Bohr.

Deposition energies (∆E dep) were calculated for each system using the following equation:
$$ {\varDelta E}_{\mathrm{dep}}=\frac{1}{N_{\mathrm{Ag}}}\left({E}_{\mathrm{slab}+\mathrm{Ag}}-{E}_{\mathrm{slab}}\right)-{E}_{\mathrm{Ag}} $$
(1)
where N Ag is the number of deposited Ag atoms, E slab + Ag is the energy of the slab with deposited Ag, E slab is the energy of the reference slab and E Ag is the energy of an Ag atom in bulk Ag. A negative value for ∆E dep indicates a thermodynamic driving force for dissolving Ag from an Ag counter electrode and depositing it on the slab being investigated in the absence of a potential difference between the two electrodes. Adsorption energies (E ad) were calculated for each system using the following equation:
$$ {E}_{\mathrm{ad}}=\frac{1}{N_{\mathrm{H}}}\left({E}_{\mathrm{slab}+\mathrm{H}}-{E}_{\mathrm{slab}}\right)-\frac{E_{{\mathrm{H}}_2}}{2} $$
(2)
where N H is the number of adsorbed hydrogen atoms, E slab + H is the energy of the slab with adsorbed H, E slab is the energy of the reference slab and \( {E}_{{\mathrm{H}}_2} \) is the energy of a hydrogen molecule. A positive value for E ad yields a thermodynamically unfavourable structure compared with the slab and H2 reference.

Results and Discussion

To establish structure–activity relations, first, preparation of Ag overlayers on Pt(111) as well as their structure, stability and electrochemical behaviour is described. Knowledge of structural properties and interaction of the Ag/Pt(111) model surfaces with adsorbed hydrogen is then used for a discussion of the hydrogen evolution reaction on Ag(111) overlayers on Pt(111).

Electrochemical Deposition and Characterization of Ag Monolayers on Pt(111)

The underpotential deposition (upd) of Ag onto Pt single crystals in aqueous sulphuric acid is well-known [17, 23, 24, 25, 26, 27, 28, 29, 30, 31]. In the case of Ag upd on Pt(111), two prominent deposition peaks are related with the formation of a first monolayer and an Ag bilayer, as seen in the voltammogram in Fig. 1. In-situ surface X-ray scattering measurements revealed that the first Ag monolayer grows pseudomorphically as expected [29]. However, the second layer was reported to be expanded (!) compared to both the Pt(111) substrate and massive Ag(111) [29].
Fig. 1

Cyclic voltammogram (blue) of Pt(111) in 0.1 M H2SO4 + 1 mM Ag2SO4. Scan rate 1 mV s−1. The red curve represents the integrated charge density starting at 0.5 V in the negative direction. The inset shows an in-situ STM image (25 nm × 25 nm) of Pt(111) in 0.1 M H2SO4 + 1 mM Ag2SO4 at 0.6 V with the structure of (bi)sulphate on the first Ag monolayer

Since the onset of Ag upd overlaps with Pt(111) surface oxidation, a study of the initial stages of the electrocrystallization process on the unoxidized surface is not possible under the present experimental conditions. Therefore, the Pt(111) electrode was typically contacted with the electrolyte at 0.5 or 0.6 V, where the pseudomorphic Ag monolayer is directly formed. The in-situ STM image in Fig. 1 shows the preliminary result for the \( \left(\sqrt{3}\times \sqrt{7}\right) R{19.1}^{{}^{\circ}} \) structure of the (bi)sulphate on the first Ag monolayer. While a few defects are present in the ordered adlayer, the Ag monolayer is found to be complete, with a quite large stability window between 0.45 and 0.75 V. Bi(sulphate) forms the same structure on the Ag monolayer as on Pt(111) [32], which is in variance to sulphate adsorption on Ag(111) [33]. Therefore, the adsorption properties of the Ag monolayer are strongly influenced by the Pt(111) substrate and more Pt-like.

An Ag bilayer is formed after further deposition around 0.4 V, in agreement with a theoretical deposition charge of 240 μC cm−1 for a monolayer (ML). However, after scanning back to 0.5 V, the anodic charge density is around 80 μC cm−1 less as compared to the negative scan. This observation points towards an irreversible process. It is known that alloy formation can take place for Pt(111) covered by an Ag overlayer after thermal annealing [34]. Furthermore, recent surface-physical studies suggest that about 22% of the atoms exchange across the Ag/Pt interface for an Ag bilayer, leading to a complex sandwich-like structure [35]. It might be possible that under electrochemical conditions the second Ag layer is only partially dissolved during scanning back to 0.5 V, while part of the Ag atoms had exchanged places with Pt atoms. Since the exact interfacial structure and possible driving force for the structural rearrangement are not clear, this aspect will be treated below by theory.

It is an experimental challenge to transfer the Pt(111) electrode after Ag deposition to another electrochemical cell for further electrochemical and electrocatalytic studies, without changing the structure and the coverage of the Ag overlayer. The typical electrochemical adsorption processes of Pt(111) in 0.1 M H2SO4 are blocked after the deposition of an Ag monolayer (Fig. 2). This pseudomorphic overlayer AgML/Pt(111) is stable in a broad potential region and anodically stripped around 0.7 V (Fig. 2), before the surface oxidation of Pt(111) proceeds. The dissolution charge for the Ag monolayer is 240 μC cm−1, as expected. In this way, the coverage of Ag overlayers can be double-checked at the end of each measurement. In addition, the Ag coverage was deduced from the suppression of hydrogen adsorption on Pt(111) in 0.1 M H2SO4 [34, 36].
Fig. 2

Cyclic voltammograms of Pt(111) (black line) and AgML/Pt(111) (blue line) in 0.1 M H2SO4. Scan rate 1 mV s−1

While adsorption peaks for Ag overlayers on Pt(111) are absent in acid solution, studies of electrochemical properties for the Ag overlayers show characteristic voltammetric signals in alkaline solution.

The Ag overlayers on Pt(111) show a clearly different behaviour compared to those of massive Ag(111), because of ligand and strain effects. Figure 3a shows voltammograms for an Ag monolayer on Pt(111) in 0.1 M NaOH at 50 mV s−1 with two different methods for Ag deposition. While AgML/Pt(111) is the pseudomorphic monolayer fabricated by upd, AgML */Pt(111) is obtained after partial dissolution of an Ag bilayer at the same potential, where AgML/Pt(111) is formed. It is clearly seen by comparison with the black curve in Fig. 3b that hydrogen adsorption (current plateau below −0.7 V) and OH adsorption (between −0.4 and −0.1 V) on Pt(111) are blocked by the Ag monolayers, indicating a complete blockage of Pt(111) terrace sites. The sharp characteristic current peak at ca. −0.75 V is assumed to show OH adsorption on the first Ag monolayer, since the deposition of Ag onto Pt(111) blocks hydrogen upd completely, as can be seen in Fig. 2 for potentials below 0.05 V. The absence of H upd on the Ag monolayer supports the attribution of the voltammetric peak to OH adsorption. In addition, theoretical calculations show that hydrogen adsorption on Ag overlayers is endothermic (see Fig. 9).
Fig. 3

Cyclic voltammograms for (a) Ag monolayers on Pt(111) prepared by two different methods and (b) 0.2 ML and 4 ML of Ag on Pt(111) in 0.1 M NaOH. Scan rate 50 mV s−1. The curves for Ag(111) and Pt(111) are shown for comparison

It is interesting that the voltammetric behaviour changes after depositing an Ag bilayer and scanning back to the stability region of a single layer. This is indicated by the appearance of an additional small peak at −0.63 V. The adsorption charges for OH on these Ag monolayers are around 90 μC cm−2 in each case. Figure 3b shows voltammograms of 0.2 and 4 ML of Ag on Pt(111) in 0.1 M NaOH at 50 mV s−1. The curve for Pt(111) is shown for comparison. Again, the current peak at ca −0.75 V is related to OH adsorption on pseudomorphic Ag islands. Peaks for hydrogen and OH adsorption on uncovered Pt(111) sites are still present. Four ML of Ag on Pt(111) show a very similar electrochemical behaviour compared to those of massive Ag(111). It is clearly observed that the current density for hydrogen adsorption decreases with increasing Ag coverage. While Ag(111) shows a large overpotential for the hydrogen evolution reaction [33], the Ag monolayers on Pt(111) are significantly more active for the HER showing onset potentials around −1 V (Fig. 3a).

It is not obvious that the structure of Ag overlayers remains stable upon transfer of the electrode from the deposition cell to another electrochemical cell. Therefore, the electrochemical behaviour of Pt(111) in the presence of Ag ions in alkaline solution was studied directly in alkaline solution. In this way, it could be checked if the electrode transfer to another electrochemical cell is possible without structural changes. We found that rinsing the samples with oxygen-free ultrapure water gives very similar results as for direct deposition from alkaline solution, which supports the stability of Ag overlayers on Pt(111) upon electrode transfer to another cell.

Figure 4 shows cyclic voltammograms of Pt(111) in 10−6 M AgNO3 + 0.1 M NaOH solution at 50 mV s−1. Here, the electrochemical behaviour of the Pt(111) surface was monitored between −0.4 and −0.95 V while Ag deposition at overpotentials slowly takes place without significant contribution to the measured current. The current peak at −0.75 V, which is assigned to OH adsorption on the first Ag monolayer on Pt(111), becomes sharper with successive potential cycles, and then decreases along with the growth of a new peak at −0.63 V, indicating that a new surface structure starts to form. It is seen again that the peaks for H adsorption and OH adsorption on Pt(111) in 0.1 M NaOH decrease with increasing Ag coverage and vanish after completion of the first Ag monolayer. In contrast to the Ag upd process in acid solution, in alkaline solution, the second Ag monolayer starts to grow before the first Ag layer is complete. We explain this behaviour by the fact that Ag deposition in alkaline solution proceeds at overpotentials, already. The typical voltammogram of massive Ag(111) appears after 160 min of potential cycling, indicating epitaxial growth.
Fig. 4

Cyclic voltammograms of Pt(111) in 10−6 M AgNO3 + 0.1 M NaOH. Scan rate 50 mV s−1

Energetics of Deposition and Surface Alloying

As described above and is well-known also for gas phase studies [37, 38], Ag can be deposited on Pt(111) to form a pseudomorphic monolayer. To investigate the stability of Ag overlayers on Pt(111) and potential subsequent surface alloying, we performed several series of DFT calculations.

Segregation Behaviour of a Pseudomorphic Ag Monolayer on Pt(111)

The first series of DFT calculations investigates the energetics of segregation and alloying of a pseudomorphic Ag monolayer within the first two layers of a Pt(111) slab. As our reference point, we take a pseudomorphic monolayer of Ag on Pt(111), which is assumed to be the first structure formed upon deposition, is known to be stable [37, 38] and is easily prepared via electrodeposition. The energies associated with swapping Ag atoms with Pt atoms in the first subsurface layer were computed based on energies obtained from DFT calculations.

The difference in energy per supercell, ΔE, between each of the eight structures possible within the constraints of a 2 × 2 supercell and the reference state are shown in Fig. 5 as grey bars and the ΔE per substituted Ag atom as blue bars. The positive energies indicate swapping Ag from the surface into the subsurface is energetically unfavourable and that none of the surface alloys are energetically stable, a conclusion which is corroborated by published experimental observations [39, 40]. Therefore, any surface alloy formation that does occur must be entropically—rather than energetically—driven.
Fig. 5

Energetic cost (ΔE) of 1 ML Ag to form surface alloys is given per 2 × 2 surface cell (grey) and per substituted Ag atom (blue). The average Ag–Ag coordination number is shown at the base of each column

The differing stability of Ag in the surface and subsurface layers is insufficient to explain all of the energy differences between the structures in Fig. 5. Therefore, the situation must be more complicated. Indeed, if this were not the case, then the energies of structures with the same number of substitutions would be identical and those with different numbers of substitutions would be related by a trivial multiplication factor, so that the energy per substituted Ag atom (blue bars) would be constant through the whole range of structures.

A closer examination of the structures suggests that, given an equal degree of surface segregation, the structure with higher average Ag–Ag coordination (see Fig. 5)—consequently also higher average Pt–Pt coordination—is more stable. This behaviour suggests that Ag clusters have a stabilizing effect and that Pt–Ag bonds have a destabilizing effect, perhaps by interrupting the Pt lattice.

Segregation Behaviour of a Pseudomorphic Ag Bilayer on Pt(111)

Having established that there is no energetic incentive for a pseudomorphic Ag monolayer on Pt(111) to diffuse into the subsurface or form an alloy, we proceed to consider a pseudomorphic Ag bilayer. To do this, we computed the relative energies of a second series of nine structures. The new structures were identical to the structures in the first series except for an additional (i.e. second) monolayer of pseudomorphic Ag(111) at the surface that is not substituted (see top views in Fig. 6), making them somewhat analogous to the polycrystalline system studied experimentally by Vaskevich et al. [41].
Fig. 6

ΔE of the second ML Ag to form surface alloys. ΔE for transforming the adsorbed Ag bilayer (top view in the upper left corner) into alloyed structures involving a complete Ag surface layer and Ag/Pt alloys in the second and third layers. ΔE is calculated either per 2 × 2 simulation supercell (grey) or per substituted Ag atom (blue)

While a single deposited Ag layer on Pt(111) is known to be pseudomorphic, experimental evidence in the literature suggests that the newly formed bilayer is not pseudomorphic [29]. Instead, the bilayer surface was reported to undergo a phase transition resulting in an expanded lattice constant, rather than the compressed lattice constant of the Ag monolayer [37, 38]. Since it is prohibitively difficult to model this compression using DFT without comprising the lattice constant of the Pt substrate, we assume that the second layer of Ag deposited is also pseudomorphic.

The difference in energy, ΔE, between each structure and the bilayer reference state is shown in Fig. 6 (grey bars) along with the energy per Ag atom substituted from the second to the third layer (blue bars). The addition of a second Ag layer reduces the energy required to swap an Ag atom from the first Ag monolayer with a Pt atom from the adjacent layer. This reduction in ΔE can be attributed to the second Ag monolayer either destabilizing the reference state or stabilizing the other structures, or more likely to a combination of both. In fact, experimental evidence indicating increased strain in a second Ag adlayer supports the destabilization of the reference state [37, 38]. On the other hand, the unexpected stability achieved by substituting all Ag from the second to the third layer indicates that at least one of the substituted structures is positively stabilized by the second Ag adlayer. Indeed, the fully substituted structure is sufficiently stabilized to make it energetically favourable for two pseudomorphic Ag monolayers at the surface of Pt(111) to separate from one another, forming what we will refer to as a ‘sandwich’ structure, AgPtAg/Pt(111). However, it should be kept in mind that the experimentally observed, expanded bilayer, as it is free of strain, will be a lower-energy system than the pseudomorphic bilayer and, as such, the existence of a stable sandwich structure must be experimentally verified.

Depositing Ag on Pt(111)

To investigate the deposition of Ag on Pt(111), we calculated the energies of deposition of a monolayer and partial monolayer of Ag on Pt(111), Ag/Pt(111) and PtAg/Pt(111) using Eq. (1). Results are shown in Fig. 7.
Fig. 7

(Left) Deposition energy of an Ag adlayer of varying coverage on Pt(111), Ag/Pt(111) and PtAg/Pt(111). (Right) Migration of 4 × 0.25 ML into Pt bulk results in four configurational motifs: (a) uncoordinated Ag without a surface Ag monolayer, (b) uncoordinated Ag with a surface Ag monolayer, (c) coordinated Ag without a surface Ag monolayer and (d) coordinated Ag with a surface Ag monolayer

On all three substrates, the deposition of a complete monolayer is more energetically favourable than that of a partial monolayer as would be intuitively predicted. Therefore, Ag is expected to initially form islands—rather than forming a uniform array of atoms in a partial monolayer—which expand until they form a monolayer covering the entire surface. The formation of islands is in good agreement with the development of the OH adsorption peak for Ag on Pt(111) in alkaline solution (Figs. 3 and 4).

Segregation Behaviour of Three Pseudomorphic Ag(111) Layers in Pt

To further investigate the phenomenon of Ag layers separating within the Pt bulk and to discover whether the phenomenon occurs when more than two Ag layers are present, a series of calculations were performed to calculate the difference in energy between a Pt slab with three surface Ag layers and Pt slabs with the three complete Ag layers dispersed through the upper five layers. This regime gives rise to ten unique structures including the reference state (slab 1). In practice, it is unlikely to be possible to deposit three layers of pseudomorphic Ag on Pt bulk electrochemically as bulk Ag deposition occurs soon after deposition of the second Ag adlayer [29], but it may be possible by a mixture of electrochemical deposition and annealing to create first a sandwich structure upon which a third Ag layer could be deposited.

Figure 8 contains three curves that show how ΔE varies between structures ordered by increasing average depth of Ag. The similarly shaped grey and blue curves represent analogous structures with either no Ag layer separation (the first, fourth and seventh slabs from left to right) or a single Ag layer separation (the second through fifth slabs from left to right). The red curve represents structures with a double Ag layer separation: separation by two Pt layers in the case of the eighth and tenth slabs (from left to right) or, in the case of the ninth slab, two single Ag layer separations, a ‘double-sandwich’ arrangement.
Fig. 8

Relative energy (trilayer structure on the far left set to zero) per surface atom for AgPt(111) slabs with three Ag layers moving between the top five layers

The results show that the least stable structure, in agreement with the conclusions of our previous computational experiments, is the seventh slab from the left, in which Ag is in the third, fourth and fifth layers. In the case of this slab, the Ag layers are not separated and have migrated deep into the structure with respect to the reference (the first slab on the left). As was expected, based on the previous conclusions, the most stable arrangement is a double-sandwich arrangement with Ag in the first, third and fifth layers.

With respect to the reference structure, in which all three Ag layers are at the surface, only those structures which have an Ag surface layer are stable. In general, it appears that Ag moving deeper into the structure, in the absence of any other factor such as a new separation of Ag, is an endothermic process. This can be seen by comparing the similarly shaped grey and blue curves whose points represent analogous structures at different depths in the slab. The structures represented by the blue curve are one layer deeper in the structure than those represented by the grey curve and as such have more positive ΔE values. This observation is in agreement with the conclusion drawn from our previous calculations that migration of Ag into the Pt bulk is unfavourable. Based on these results, we surmise that it is favourable for pseudomorphic Ag layers on Pt(111) to diffuse into the surface in order separate from each other leaving a single Pt layer between each pair of adjacent Ag layers and a single Ag layer at the surface. In the absence of these Ag–Ag interlayer interactions, it is unfavourable for Ag to migrate to the bulk.

Hydrogen Adsorption

There are no characteristic voltammetric adsorption peaks for hydrogen upd on the Ag overlayers deposited onto Pt(111). To get a first look—from a theoretical point of view—into the influence of deposited Ag on the interactions between H and a Pt electrode, we investigated H adsorption. We begin by considering the influence of some of the various surfaces, considered in the previous section, on hydrogen adsorption and then turn things around by considering the influence of adsorbed hydrogen on the stabilities of various surface structures and compositions.

Influence of Surface Composition on H Adsorption

Adsorption energies at face-centred cubic (fcc) sites for three different coverages of H on several of the surfaces considered in previous sections are presented in Fig. 9. The Pt-capped surfaces all exhibit negative energies of adsorption for H (i.e. adsorbed H is energetically preferred over H2), whereas Ag-capped surfaces exhibit positive energies of adsorption for H. Therefore, as one would anticipate, the Pt-covered surfaces are more reactive towards H than the Ag-covered surfaces. In fact, in comparison with the influence of the surface layer, the so-called strain and ligand effects due to the subsurface/substrate are small.
Fig. 9

Adsorption energies for 1/9 ML, 1/4 ML and 1 ML of H adsorbed at fcc sites on seven Ag/Pt(111) surfaces: pure Pt(111), Pt(111) with a pseudomorphic subsurface Ag layer, Pt(111) with a pseudomorphic Ag adlayer, Pt(111) with the pseudomorphic Ag sandwich structure, Pt(111) with two pseudomorphic Ag adlayers, Ag(111) compressed to Pt horizontal lattice constants and Ag(111) with pure Ag lattice constants

Fig. 10

Energies for surface segregation and H adsorption, absorption and desorption involving an adlayer of Ag on Pt(111)

Fig. 11

Current–potential curves of HER at Ag overlayers on Pt(111) in 0.1 M H2SO4. Scan rate 10 mV s−1. Ag(111) is shown for comparison

At low coverages (Θ = 1/9 ML, 1/4 ML), our bottom-of-the-well adsorption energies favour adsorption at fcc sites on all surfaces. However, it is well-known that the inclusion of the zero-point energy shifts the preferred binding site of hydrogen on Pt(111) from an on-top site to an fcc site [42, 43]. At higher coverages (e.g. 1 ML), we find that if the surface is Ag and the subsurface layer is Pt, then H prefers absorbing into the subsurface, where it can adsorb on the more reactive subsurface Pt, rather than on the less reactive surface Ag. Because subsurface H pushes the surface atoms immediately above it up, a complete monolayer of H is required for stability in the subsurface, so that all Ag atoms in the surface layer are pushed up equally to form a flat surface layer.

Influence of Adsorbed H on Surface Structure and Composition

Figure 10 shows a progression of plausible structures, should it be possible to experimentally adsorb a complete monolayer of H on a pseudomorphic Ag adlayer on Pt(111). (Note that the adsorbed H would be thermodynamically unstable with respect to the formation of H2). While H might initially adsorb on the surface (upper central structure in Fig. 10), it ultimately prefers to have half of the H atoms diffuse into the subsurface, with the remaining half sitting directly above the subsurface H on the surface (upper right structure in Fig. 10). However, if the surface can overcome the necessary kinetic barriers to exchange the first and second layers of metal atoms (i.e. Pt and Ag), then more than 0.5 eV per surface atom is to be gained by segregating Pt to the surface (bottom centre structure in Fig. 10). The desorption of H (now energetically unfavourable compared with gas phase H2 if only bottom-of-the-well energies rather than free energies are considered) leaves a thermodynamically unstable surface configuration which can be relaxed by 0.14 eV per surface atom by returning Ag to the surface. Therefore, as has already been demonstrated for NiPt and AuPd alloys [11, 44, 45], adsorbed H has the potential to induce surface segregation in AgPt surface alloys.

Electrocatalytic Activity of Ag(111) Overlayers on Pt(111)

The system under study is of fundamental interest because Ag displaces the so-called Hupd on Pt(111). If it is assumed that both Hupd and the Ag monolayer act as spectators during HER, similar electrocatalytic activities are expected. It is shown above that hydrogen adsorption at a subsurface Pt layer is possible. Figure 11 displays the HER on different Ag overlayers on Pt(111) in 0.1 M H2SO4 at 10 mV s−1. It is seen that the onset potential for HER on an Ag(111) electrode is at ca −0.25 V vs. RHE. This onset potential shifts to −0.05 V vs. RHE for AgML/Pt(111) and AgML */Pt(111), i.e. with different methods for Ag deposition. This indicates that a pseudomorphic Ag monolayer on Pt(111) is far more active than massive Ag(111). Interestingly, the hydrogen oxidation reaction (HOR) still occurs for the Ag monolayer (blue curve in Fig. 11), while HOR on Ag(111) is absent [46]. This might be explained by partially uncovered step sites during Ag deposition, because the OH adsorption peak on Pt(111) at ca −0.25 V disappeared completely. On the other hand, depositing the Ag bilayer and stripping the second ML Ag leads to an overlayer structure AgML*/Pt(111), which is inactive for the HOR (red curve in Fig. 11). Both model systems, AgML/Pt(111) and AgML*/Pt(111), resemble Pt(111) very much with respect to HER activity.

Further increase of the Ag coverage results in a decrease in the activity of the Ag overlayers on Pt(111) towards HER (Fig. 11). It is remarkable that the thick Ag overlayers are still more active than massive Ag(111) (Fig. 11). This might be related to the triangular terrace structure which has been reported for bulk deposition of Ag [25]. Exceeding the potential limits to 0.1 V for 4 ML of Ag on Pt(111) in 0.1 M NaOH induces surface disorder, which leads to an enhancement of HER activity for multilayers (Fig. 11).

Tafel plots (not shown) for the HER at the model systems reveal similar Tafel slopes for the Ag monolayers and Pt(111), around 40 mV dec−1. Tafel slopes for thicker Ag overlayers and for Ag(111) are around 120 mV dec−1. This observation confirms the significantly enhanced electrocatalytic activity of pseudomorphic Ag monolayers on Pt(111).

Conclusions

Various aspects of well-ordered Ag overlayers on Pt(111) electrodes have been investigated in this combined experimental and theoretical study.
  1. 1.

    Preparation and structure: The electrochemical deposition of Ag onto Pt(111) produces surface structures exhibiting several interesting structural properties. (i) A contracted pseudomorphic Ag monolayer is easily formed by underpotential deposition in sulphuric acid solution. The same structure is formed upon deposition from alkaline electrolyte; however, the second Ag monolayer starts to grow before the first monolayer is complete. The deposition–dissolution hysteresis points towards limited stability of an Ag bilayer. Theoretical DFT calculations support the stability of the pseudomorphic Ag monolayer on Pt(111) and predict the existence of an AgPtAg/Pt(111) sandwich structure, which is more stable than an Ag bilayer. A \( \left(\sqrt{3}\times \sqrt{7}\right) R{19.1}^{{}^{\circ}} \) structure of (bi)sulphate on the Ag monolayer on Pt(111) was imaged by in-situ STM.

     
  2. 2.

    Electrochemical behaviour: The presence of Ag islands blocks adsorption processes such as Hupd and OH adsorption on Pt(111). While characteristic voltammetric peaks are absent for Ag overlayers on Pt(111) in sulphuric acid solution, a sharp current peak for OH adsorption on the Ag monolayer in alkaline solution was observed. Theoretical calculations show that hydrogen adsorption is possible on the Pt subsurface layer covered by an Ag monolayer.

     
  3. 3.

    Electrocatalysis: While Hupd on Pt(111) is blocked by Ag, there is still a high activity for the hydrogen evolution reaction for an Ag monolayer. The oxidation of molecular hydrogen is only observed as long as Pt(111) is not completely covered by Ag. The HER activity decreases with increasing Ag overlayer thickness. However, an equivalent of 16 Ag monolayers is still far more active than Ag(111).

     

To better understand the Ag/Pt(111) model system, which serves as a prototype of binary electrocatalysts, future studies such as detailed in-situ surface X-ray scattering analysis of the Pt(111) electrode surface during and after the formation of electrochemically deposited Ag layers might be helpful. The structural information obtained from such measurements can then be used in conjunction with atomistic simulations to elucidate the atomistic structures and mechanisms needed to better explain the electrochemical behaviour.

Notes

Acknowledgements

JEM gratefully acknowledges financial support from the Alexander von Humboldt foundation. CW gratefully acknowledges financial support provided by ERASMUS and the Student Support Trust. EJB gratefully acknowledges financial support under the IREU exchange program coordinated by the American Chemical Society within the REU program of the National Science Foundation under award number IIA# 1261104.

Financial support by the Deutsche Forschungsgemeinschaft (Research Unit For-1376, Ki 787/6-1 and 6-2) and by the Fonds der Chemischen Industrie is gratefully acknowledged. The authors also acknowledge the computer time supported by the state of Baden-Württemberg through the bwHPC project and the DFG through grant number INST40/467-1 FUGG.

References

  1. 1.
    J. Lipkowski, P. N. Ross (eds.), Electrocatalysis (Wiley-VCH, New York, 1998)Google Scholar
  2. 2.
    E. Santos, W. Schmickler (eds.), Catalysis in electrochemistry (Wiley-VCH, New York, 2011)Google Scholar
  3. 3.
    M.T.M. Koper, A. Bandarenka, Chem. Soc. Rev. 42, 5210 (2013)CrossRefGoogle Scholar
  4. 4.
    L.A. Kibler, A.M. El-Aziz, R. Hoyer, D.M. Kolb, Angew. Chem. Int. Ed. 44, 2080 (2005)CrossRefGoogle Scholar
  5. 5.
    H.A. Gasteiger, P.N. Ross, E.J. Cairns, Surf. Sci. 293, 67 (1993)CrossRefGoogle Scholar
  6. 6.
    A.M. El-Aziz, R. Hoyer, L.A. Kibler, Chem. Phys. Chem. 11, 2906 (2010)CrossRefGoogle Scholar
  7. 7.
    L.A. Kibler, M. Kleinert, R. Randler, D.M. Kolb, Surf. Sci. 443, 19 (1999)CrossRefGoogle Scholar
  8. 8.
    M. Takahasi, Y. Hayashi, J. Mizuki, K. Tamura, T. Kondo, H. Naohara, K. Uosaki, Surf. Sci. 461, 213 (2000)CrossRefGoogle Scholar
  9. 9.
    A. Schlapka, M. Lischka, A. Groß, U. Käsberger, P. Jakob, Phys. Rev. Lett. 91, 016101 (2003)CrossRefGoogle Scholar
  10. 10.
    T. Bligaard, J.K. Nørskov, S. Dahl, J. Matthiesen, C.H. Christensen, J. Sehested, J. Catal. 224, 206 (2004)CrossRefGoogle Scholar
  11. 11.
    J. Mueller, P. Krtil, L.A. Kibler, T. Jacob, Phys. Chem. Chem. Phys. 16, 15029 (2014)CrossRefGoogle Scholar
  12. 12.
    N.M. Markovic, C.A. Lucas, V. Climent, V. Stamenkovic, P.N. Ross, Surf. Sci. 465, 103 (2000)CrossRefGoogle Scholar
  13. 13.
    M. Shao, P. Liu, J. Zhang, R. Adzic, J. Phys. Chem. B 111, 6772 (2007)CrossRefGoogle Scholar
  14. 14.
    J. Steidtner, F. Hernandez, H. Baltruschat, J. Phys. Chem. C 111, 12320 (2007)CrossRefGoogle Scholar
  15. 15.
    J. Zhang, M.B. Vukmirovic, Y. Xu, M. Mavrikakis, R.R. Adzic, Angew. Chem. Int. Ed. 44, 2132 (2005)CrossRefGoogle Scholar
  16. 16.
    H.E. Hoster, R.J. Behm, in Fuel cell catalysis: a surface science approach, ed. by M. T. M. Koper. (Wiley-VCH, Chichester, 2009)Google Scholar
  17. 17.
    K.F. Domke, X.-Y. Xiao, H. Baltruschat, Phys. Chem. Chem. Phys. 10, 1555 (2008)CrossRefGoogle Scholar
  18. 18.
    K.A. Soliman, L.A. Kibler, D.M. Kolb, Electrocatalysis 3, 170 (2012)CrossRefGoogle Scholar
  19. 19.
    P.J. Feibelman, Phys. Rev. B 35, 2626 (1987)CrossRefGoogle Scholar
  20. 20.
    J.P. Perdew, K. Burke, M. Ernzerhof, Phys. Rev. Lett. 88, 3865 (1996)CrossRefGoogle Scholar
  21. 21.
    M. Fuchs, M. Scheffler, Comput. Phys. Commun. 119, 67 (1999)CrossRefGoogle Scholar
  22. 22.
    D.R. Hamann, Phys. Rev. B 40, 2980 (1989)CrossRefGoogle Scholar
  23. 23.
    F. El Omar, R. Durand, R. Faure, J. Electroanal. Chem. 160, 385 (1984)CrossRefGoogle Scholar
  24. 24.
    R. Durand, R. Faure, D. Aberdam, S. Traore, Electrochim. Acta 34, 1653 (1989)CrossRefGoogle Scholar
  25. 25.
    N. Kimizuka, K. Itaya, Faraday Discuss. 94, 117 (1992)CrossRefGoogle Scholar
  26. 26.
    J.F. Rodriguez, D.L. Taylor, H.D. Abruna, Electrochim. Acta 38, 235 (1993)CrossRefGoogle Scholar
  27. 27.
    P. Zelenay, M. Gamboa-Aldeco, G. Horanyi, A. Wieckowski, J. Electroanal. Chem. 357, 307 (1993)CrossRefGoogle Scholar
  28. 28.
    D.L. Taylor, H.D. Abruna, J. Electrochem. Soc. 140, 3402 (1993)CrossRefGoogle Scholar
  29. 29.
    N.S. Marinkovic, J.X. Wang, J.S. Marinkovic, R.R. Adzic, J. Phys. Chem. B 103, 139 (1999)CrossRefGoogle Scholar
  30. 30.
    Z. Radovic-Hrapovic, G. Jerkiewicz, in Thin Films: Preparation, Characterization, Applications, Chapter 4, ed. by M.P. Soriaga, J. Stickney, L.A. Bottomley, Y-G. Kim (Springer, US, 2002) p. 53Google Scholar
  31. 31.
    K.A. Soliman, L.A. Kibler, Electrochim. Acta 52, 5654 (2007)CrossRefGoogle Scholar
  32. 32.
    B. Braunschweig, W. Daum, Langmuir 25, 11112 (2009)CrossRefGoogle Scholar
  33. 33.
    M. Schweizer, D.M. Kolb, Surf. Sci. 544, 93 (2003)CrossRefGoogle Scholar
  34. 34.
    J. Clavilier, L.H. Klein, A. Vaskevich, A.A. El-Shafei, J. Chem. Soc. Faraday Trans. 92, 3777 (1996)CrossRefGoogle Scholar
  35. 35.
    K. Ait-Mansour, H. Brune, D. Passerone, M. Schmid, W. Xiao, P. Ruffieux, A. Buchsbaum, P. Varga, R. Fasel, O. Gröning, Phys. Rev. B 86, 085404 (2012)CrossRefGoogle Scholar
  36. 36.
    J.S. Spendelow, Q. Xu, J.D. Goodpaster, P.J. Kenis, A. Wieckowski, J. Electrochem. Soc. 154, F238 (2007)CrossRefGoogle Scholar
  37. 37.
    H. Brune, H. Röder, C. Boragno, K. Kern, Phys. Rev. B 49, 2997 (1994)CrossRefGoogle Scholar
  38. 38.
    J.X. Wang, N.S. Marinković, R.R. Adźić, B.M. Ocko, Surf. Sci. 398, L291 (1998)CrossRefGoogle Scholar
  39. 39.
    P. Durussel, P.J. Feschotte, Alloys Compd. 239, 226 (1996)CrossRefGoogle Scholar
  40. 40.
    H. Röder, R. Schuster, H. Brune, K. Kern, Phys. Rev. Lett. 71, 2086 (1993)CrossRefGoogle Scholar
  41. 41.
    A. Vaskevich, M. Rosenblum, E.J. Gileadi, J. Electroanal. Chem. 383, 167 (1995)CrossRefGoogle Scholar
  42. 42.
    Ş.C. Bădescu, P. Salo, T. Ala-Nissila, S.C. Ying, K. Jacobi, Y. Wang, K. Bedürftig, G. Ertl, Phys. Rev. Lett. 88, 136101 (2002)CrossRefGoogle Scholar
  43. 43.
    Ş.C. Bădescu, K. Jacobi, Y. Wang, K. Bedürftig, G. Ertl, P. Salo, T. Ala-Nissila, S.C. Ying, Phys. Rev. B 68, 205401 (2003)CrossRefGoogle Scholar
  44. 44.
    H. Hoffmannová, M. Okube, V. Petrykin, P. Krtil, J.E. Mueller, T. Jacob, Langmuir 29, 9046 (2013)CrossRefGoogle Scholar
  45. 45.
    M. Okube, V. Petrykin, J.E. Mueller, D. Fantauzzi, P. Krtil, T. Jacob, Chem. Electro. Chem. 1, 207 (2014)Google Scholar
  46. 46.
    R. Tölle, A. Otto, Surf. Sci. 597, 110 (2005)CrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media New York 2017

Authors and Affiliations

  • Ludwig A. Kibler
    • 1
  • Khaled A. Soliman
    • 2
  • Alan Plumer
    • 3
  • Christopher S. Wildi
    • 4
  • Eric Bringley
    • 5
  • Jonathan E. Mueller
    • 6
  • Timo Jacob
    • 1
  1. 1.Institut für ElektrochemieUniversität UlmUlmGermany
  2. 2.Physical Chemistry DepartmentNational Research CentreCairoEgypt
  3. 3.Ecole Nationale Supérieure d’Ingénieurs de CaenCaenFrance
  4. 4.Department of ChemistryUniversity of AberdeenAberdeenUK
  5. 5.University of CambridgeCambridgeUK
  6. 6.Volkswagen AktiengesellschaftWolfsburgGermany

Personalised recommendations