Local mesoscopic properties

The utilization of well-defined surfaces facilitates the interpretation of local structural information, but the application of STM is not restricted to such model systems. When irregular surfaces are investigated with STM, the local character of the measurement has to be emphasized and a clear-cut distinction between the well-characterized local morphology and the average surface morphology has to be made.

An example for the variation of the local topography is given in Fig. 5, which shows two images of a flame-annealed polycrystalline gold electrode. Figure 5 a shows a region which exhibits a smooth hill-and-valley structure, while another region, fig. 5 b, shows pronounced crystalline features with monatomic steps, atomically smooth terraces, and islands of monatomic height[ll]. Both kinds of topography were quite often observed on flame annealed gold foils.

Fig. 5. Two different types of morphology of the same flame-annealed polycrystalline gold electrode.

This example is intended to highlight the scaling problem, which is inherent in the correlation of structure and reactivity. The structural characterization is performed on an almost negligible fraction of the sample surface while the reactivity usually refers to the integral sample surface and is an average quantity. In general, mesoscopic structural information can thus not be correlated with the overall electrochemical behavior of an electrode in a straightforward way. STM allows, however, to probe the real local electrochemical behavior of a real sample area with well-characterized mesoscopic topography.

This is visualized in Fig. 6, which shows the bulk deposition of copper on a flame-annealed polycrystalline gold electrode. The experiment was performed on an approximately 100 nm-wide, rather well-ordered microfacet, which is bordered by relatively rough regions of the sample surface. The orientation of the facet is presumably (100), as indicated by the rectangular shapes of islands and supported by

Fig. 6. Copper deposition on polycrystalline gold in 0.01 M HC104 + 0.1 M NaC104 + 0.1 mM Cu2+ (A) £ = 0.1 V vs. Cu/Cu2+ in the range of UPD, (B) E = -0.1 V in the range of bulk deposition, and (C) E = -0.1 V during continuing bulk deposition

more detailed investigations of the substrate surface [11]. The first image of the sequence, Fig. 6 A, was recorded at a potential of +0.1 V versus the Cu/Cu2+ equilibrium potential, where a monolayer of copper is adsorbed on the surface due to the so-called underpotential deposition (UPD). Upon lowering the potential into the region of bulk-copper deposition, E = -0.1 V, after an induction period of approximately 20 s the formation of copper nuclei becomes visible (Fig. 6(B)). The nuclei are about 3 nm in diameter and one to three monolayers in height, and their density is considerably higher in the rough region than on the ordered facet. This gives rise to an inhomogeneous distribution of the deposit on the polycrystalline substrate resulting from the mesoscopic structural properties of the sample. In the course of continuing deposition, Fig. 6(C), the substrate influence is smeared out and the deposit distribution becomes homogeneous.

The dissolution of previously electrodeposited copper clusters from a polycrystalline gold electrode is addressed in Fig. 7. The particles are rather large, ranging from 10 to 50 nm in diameter and 0.5 to 5 nm in height. From a macroscopic point of view a continuous shrinking of the clusters would be expected under the given experimental conditions. Such a continuous shrinking has indeed been observed for cases where the distance between clusters is large [12]. For more densely packed arrangements, however, the size of an individual cluster remains more or less stable for a certain time and then decreases rather fast (on the time scale of the imaging process) until the cluster vanishes. For an agglomerate of clusters this individual behavior averages to an approximately exponential decrease of the number of clusters with time. For the measurement, which is represented in Fig. 7 by six out of a sequence of 25 images,

0 0

Post a comment