Second-harmonic microscopy for fluence-dependent investigation of laser-induced surface reactions

Spatially resolved optical second-harmonic generation (SHG), that is, SHG microscopy, is shown to be a versatile tool for the investigation of femtosecond laser-induced surface processes such as adsorbate diffusion and desorption and their dependence on absorbed laser fluence. As an example, time-resolved measurements on the diffusion of O from the steps onto the terraces of a vicinal Pt(111) surface at low substrate temperature were performed. Using the sensitivity of SHG on step coverage and analyzing the signal change across the laser beam profile via SHG microscopy during laser-induced diffusion, the strong nonlinear fluence dependence of the diffusion rate could be determined. Using SHG microscopy in combination with a two-pulse correlation scheme, time-resolved information on the surface reaction was obtained as a function of absorbed laser fluence.

Figure 1

Experimental setup: fs-laser pulses are generated by a Ti:sapphire amplifier; a thin-film polarizer in combination with a $\lambda /2$ plate enables the continuous variation of the peak fluence on the sample. The $800$-nm pulses are slightly focused on the sample; the specular reflected $p$-polarized component of the SH light is imaged by means of a camera lens on the CCD camera. An interferometer containing a delay stage allows for two-pulse correlation measurements. For illustration of the experiment, an intensity distribution across the spot profile and the resulting spatial distribution of laser fluence on the sample is sketched as well. Analyzing regions (1) to (5) of the spatially resolved SH signal will give information on the fluence dependence of the investigated process.

Figure 2

Upper panel: Integrated second-harmonic response of the Pt sample as a function of time during dissociative adsorption of ${\mathrm{O}}_2$ at the steps (oxygen dose) and diffusion of atomic oxygen induced by femtosecond laser pulses. For the laser-induced diffusion, the peak laser fluence was increased from $2.4$ to 4.3 mJ/cm${}^2$. Lower panel: Sequence of false color plots of the generated SH spot, which is recorded by the intensified CCD camera during adsorption (1–3) and diffusion (4–6). Note the different scale for images (1–3) and (4–6). Arrows in the upper panel indicate the time when the respective images in the lower panel have been recorded by the CCD camera.

Figure 3

Upper panel: Spatially resolved SH signal as intensity false color plot, selected areas are hatched in gray. The peak fluence in the center was $F_{\mathrm{abs}}=4.9$ $\mathrm{mJ}/{\mathrm{cm}}^2$. Lower panel: Relative step coverage of oxygen as a function of applied laser shots for the selected areas (inner area: $F_{\mathrm{abs}}=4.9$ $\mathrm{mJ}/{\mathrm{cm}}^2$, high depletion rate; outer areas: $F_{\mathrm{abs}}=2.7$ $\mathrm{mJ}/{\mathrm{cm}}^2$, lower depletion rate). Fits are drawn as solid lines.

Figure 4

Fluence dependence of the extracted hopping probabilities per laser shot of O from fully occupied steps onto the initially empty terraces of the vicinal Pt(111) surface induced by fs-laser pulses. Hopping rates obtained by a geometrical division of the spot are shown as filled dots, data obtained by the binning procedure are shown as open symbols. Data points indicated by the same symbol belong to one single measurement. The inset displays the same data in a double logarithmic scale. The solid lines depict a power law $\propto F_{\mathrm{abs}}^{6.4}$.

Figure 5

Two-pulse correlation traces (data points) for different absorbed fluences; for negative delays, the weaker beam precedes the stronger one. Solid lines show Lorentzian functions fitted to the experimental data to determine the width $\tau$ of the 2PC. For the highest fluence combination, the FWHM $\tau =1.4$ ps is indicated by the arrows.

Figure 6

FWHM $\tau$ of the 2PC shown in Fig. 5 as a function of absorbed laser fluence (filled data diamonds). In comparison, the results from simulations using the empirical friction model of Ref. 25 with a friction coefficient of the form ${\eta }_e={\eta }_0{T}_e^2$ and a diffusion barrier $E_{\mathrm{diff}}=0.8$ eV are drawn as a solid line. Additionally, the FWHM obtained when averaging the SH response over the whole spot is shown (empty diamond, see main text). Inset: 2PC measurement in the center part of the laser spot (filled circles) and modeled 2PC (line) normalized for the same offsets and peak maxima.