Surface relaxation and surface stress of Au(1 1 1)

Abstract

Changes in surface stress and in the top-layer expansion of Au(1 1 1) electrodes in sulfuric acid have been measured as a function of electrode potential by combining surface stress and X-ray diffraction measurements. Both are linear functions of interfacial charge in the electrode potential range of changing anion coverage. Over this range the surface stress changes by −0.5 N m⁻¹ (compressive direction), while the outward top-layer relaxation decreases from +1.5% to +0.2%. The surface stress changes can be rationalized in terms of a jellium model, while ab initio simulations are needed to explain the top-layer expansion. These simulations yield +1.3% relaxation for the uncharged gold surface, in good agreement with the X-ray diffraction measurements. They also demonstrate that the outward relaxation of the surface is curbed in the presence of an electron withdrawing adsorbate (Cl), which mimics the effects of positive surface charging.

Introduction

The bonding and electronic structure of substrate surfaces can be affected by molecular chemisorption. The bonding in the outmost substrate layers can have significant effects on surface structure and dynamics, leading to phenomena such as surface reconstruction or enhanced substrate mobility [1]. Electrochemists have the unique ability to affect the bond strengths of the substrate surface, by tuning the electrode potential and consequently altering the surface electronic structure. These changes in surface bond strengths are not at present directly measurable, but they result in changes in the surface energy, surface stress, top-layer surface relaxation and in extreme cases surface reconstruction; all of which can now be monitored in situ [2], [3], [4], [5].

In this paper in situ measurements of surface stress and outward surface relaxation are presented and compared. Surface stress has been measured previously at the metal–electrolyte interface using a number of mechanical methods to monitor strain changes in the electrode [4], [6], [7], [8], [9], [10], [11], [12], [13]. Although some of the results on polycrystalline samples were ambiguous, the more recent experiments for single crystals show that the tensile surface stress decreases linearly with increasing electrode potential [14]. Both the natural tensile stress of metals and the increase in surface stress upon positively charging the surface can be explained by a simple qualitative model of the surface bonds, which is described in a review by Ibach [15]. This “simple bonding model” considers that upon cleaving a surface, the bond charge in the missing bonds that are broken by the surface termination is relocated between the surface atoms and their backbonds. The resulting increase in charge density between the surface atoms can be used to explain the tensile stress of metal surfaces [15]. The adsorption of electronegative atoms or positive surface charging removes charge from the surface bonds, effectively weakening them and causing the tensile stress to decrease (compressive stress change). This qualitative bonding model, is described later in this paper in terms of a jellium model for the surface, which is used to quantify the stress changes with surface charging. It is described how this model predicts the correct sign and magnitude of the stress change. The qualitative bonding model described by Ibach also predicts a contraction of the distance between the first and second layers (top-layer relaxation), which should be lessened by positive surface charging [15].

Top layer relaxation is measured by an X-ray diffraction technique to monitor the interlayer spacing at the surface [16], [17]. X-ray diffraction has been employed for some years to monitor surface and adsorbate structures in the electrochemical environment although it has not yet been widely used to monitor top-layer relaxation of single-crystal surfaces. In ultra-high-vacuum (UHV) surface relaxation has been monitored by means of low energy electron diffraction (LEED) I–V analysis and a significant database now exists [16], [17]. Although the outermost atomic layer on most metal surfaces in UHV shows an inward relaxation, tantamount to a reduction of the interlayer spacing between the first two layers, a number of metals exhibit outward relaxation [18]. These include Pd(0 0 1) and Rh(0 0 1) surfaces, with an outward relaxation of (3.0±1.5)% and (0.5±2)%, respectively [19], [20]. Attempts to establish trends in surface relaxation effects by state-of-the-art theoretical calculations have highlighted discrepancies both in the theoretical results and the experimental studies [21].

The Au(1 1 1) surface in aqueous surface acid provides a good model system for studying the effects of electrode potential on the top-layer relaxation and surface stress of Au(1 1 1) electrodes. The surface electrochemistry of Au(1 1 1) in aqueous surface acid is relatively well understood. Sulfate adsorption assists in lifting the $\text{22×}\text{3}$ surface reconstruction and at very positive potentials, close to the onset of gold surface oxidation, the adsorbed sulfate ions form an ordered $\text{(}\text{3}\text{×}\text{7}\text{)}$ structure [22], [23]. In between the potential at which the surface reconstruction is lifted and that at which the ordered anion overlayer is formed, there exists a defined potential region in which the coverage of the adsorbed anion varies with electrode potential on the unreconstructed (1×1) gold surface. As the potential is increased in this region the coverage of the adsorbate anion increases creating a positive image charge in the Au surface that leads to changes in the band structure at the metal surface. These changes leads to the observed changes in surface stress and in the top-layer expansion of Au(1 1 1) electrodes. It is important to note that surface reconstruction can also lead to changes in surface stress and in the top-layer expansion, but the electrochemical environment allows measurements to be made for Au(1 1 1) in the absence of the reconstructed surface.

Experimentally, we were at first puzzled by an unexpected trend in surface stress and layer relaxations, when we analyzed the changes in the atomic structure of a unreconstructed Au(1 1 1) surface in sulfuric acid solution. Using the simple bonding model described in the review by Ibach, a reduction in the tensile stress of the surface (i.e. an increase of compressive surface stress) would be expected to lead to an increased outward relaxation [24]. By contrast, we observed the opposite trend, namely Au(1 1 1), in electrolyte solution, exhibits compressive stress changes (reduced tensile stress) and related inward relaxation of the top layer if it is positively charged. The simple bonding fails to explain this behaviour, since it does not take account of important energy contributions to the electronic structure of the surface. In this respect, it is noted that there is an intricate balance between cohesive energy and band energy contributions of single Au atoms. It is shown in this paper, that this balance in turn shifts the energetic groundstate (energy minimum) of the uncharged surface to the configuration with an extended interlayer distance between the first and second atomic layer. This trade-off between band energy and layer relaxation is also found in other transition metals, e.g. within the 4d series [18]. In this communication we consider how positive charging effects this trade-off of energetic contributions to electronic structure.