Low-energy electron microscopy of CO/Pt(111) surface diffusion by nonequilibrium coverage profile evolution

The adsorption and diffusion of CO molecules on a Pt(111) surface have been studied using low-energy electron microscopy (LEEM). The explicit relationship between LEEM image intensity and CO coverage that is determined during adsorption is used to characterize nonequilibrium CO coverage profiles that are subsequently prepared by laser-induced thermal desorption. Real-time observations of the temporal evolution of these profiles toward equilibrium uniform coverage distributions are analyzed by predictive and inverse solutions of the diffusion equation. These two methods determine consistently the detrimental effect on diffusion of the laser-induced surface damage that is observed directly with LEEM. The inverse method also provides independent information on the coverage dependence of diffusion with high coverage resolution.

Figure 1

(Color online) (a) LEEM image intensity vs incident electron energy, $I\left(V\right)$, curves measured consecutively during continuous exposure of the Pt(111) surface to CO. The accumulated exposure during the measurement of each individual curve was approximately 0.4 L. (b) The ratios of the $I\left(V\right)$ curve for the clean surface to the curves for the CO exposed surface in (a) exhibit a peak that identifies the optimal imaging energy for detecting CO.

Figure 2

(Color online) (a) The continuous variation in the LEEM image intensity at 15.6 eV during CO exposure normalized to the initial intensity for the clean surface is indicated by the solid line. The open symbols (○) are the discrete values of the normalized intensities at 15.6 eV obtained from the $I\left(V\right)$ curves shown in Fig. 1a. (b) The work-function change during CO exposure of the Pt(111) surface is indicated by solid symbols (●). The solid line is a polynomial fit that has no significance other than to highlight the trend. (c) The site-dependent and total coverage vs CO exposure taken from Ref. 24.

Figure 3

The dependence of the normalized LEEM intensity on total CO coverage on the Pt(111) surface is indicated by the solid curve. Attributing the intensity change to atop site coverage only produces the dashed curve. The normalization is with respect to the intensity for the clean surface.

Figure 4

(Color online) (a) LEEM $I\left(V\right)$ curves at the onset of mirror mode were measured consecutively during continuous exposure of the Pt(111) surface to CO. Following the measurement at 0.21 L, the exposure increment is approximately 0.3 L. The energy scale refers to the voltage bias between sample and electron gun cathode. The curves corresponding to the work-function minimum (1.75 L), the initial three (clean, 0.21 L and 0.52 L), and the final (6.1 L) exposures in this set are indicated. (b) First derivatives of selected $I\left(V\right)$ curves in (a) for the clean surface and exposures of 0.21, 0.52, 0.83, 1.75, 2.38, 3.33, and 6.10 L.

Figure 5

(Color) LEEM images show the temporal evolution at 300 K of a nonequilibrium coverage profile of CO on Pt(111) at (a) $t=0$ (immediately after LITD), (b) 6.7, (c) 13.5, (d) 27.5, (e) 55, and (f) 110 s.

Figure 6

(Color online) Radial line scans of CO concentration profiles during profile evolution are shown in comparison with the best-fit solutions of the diffusion equation (solid lines) at times $t=0$, 5.5, 11, 16.5, 22, 33, 66, and 132 s (from bottom to top) for (a) coverage-dependent diffusion coefficient and (b) position and coverage-dependent diffusion coefficient. The optimized spatial component of the diffusion coefficient $f\left(r\right)$ used to fit the data in (b) is shown in (c).

Figure 7

(Color online) Results from the inverse solution method are shown for (a) the position and coverage-dependent diffusion coefficient $D\left(r,\theta \right)=D_{0}f\left(r\right)g\left(\theta \right)$ at 11 s time intervals between $t=0$ and 66 s during profile evolution, (b) the ratio of the spatial component of the diffusion coefficient $f\left(r\right)/f\left(r=1\mu \text{m}\right)$, and (c) the coverage-dependent diffusivity $D_0\left[f\left(r=1\right)/f\left(r\right)\right]g\left(\theta \right)$ at $r=1$ (●), $3(◻)$, 5 (△), 8 (▽), 10 (◇), $15\left(◁\right)$, and $25\mu \text{m}$$\left(▷\right)$.

Figure 8

LEEM images of (a) a typical region on a clean Pt(111) surface following sample preparation, (b) laser beam damage at the core of the desorption region following LITD, and (c) same as (b) with larger field of view.