DESORPTION-KINETICS OF METHANOL FROM AL2O3(0001) STUDIED USING TEMPERATURE-PROGRAMMED DESORPTION AND ISOTHERMAL DESORPTION

Temperature-programmed desorption (TPD) and isothermal desorption were used to investigate the desorption of methanol from Al2O3(0001) in ul trahigh vacuum. At low coverages, TPD traces for methanol displayed on e broad peak that was attributed to monolayer desorption. A second, mu ltilayer peak was observed at a lower temperature as the coverage was increased. The multilayer peak appeared at coverages well below the sa turation coverage of the monolayer peak, implying that the multilayer was forming before the monolayer was completely full. Isothermal desor ption studies were performed in the multilayer regime as a function of coverage and temperature. The coverage-dependent studies showed that multilayer desorption was of order n = 0.53 +/- 0.12. The temperature- dependent studies showed that the multilayer activation barrier for de sorption was 11.1 +/- 0.5 kcal/mol with an n = 0.53 order preexponenti al of 3.1 x 10(24) s(-1). The approximately half-order desorption from the multilayer is interpreted in terms of a structured multilayer. In the monolayer regime, methanol adsorbs reversibly onto the Al2O3(0001 ) surface. The desorption peak is very broad, and quick temperature ra mps to temperatures within the monolayer TPD peak show that a distribu tion of adsorption sites occur on the surface. The TPD traces were mod eled assuming first-order desorption kinetics and using a Gaussian dis tribution of adsorption sites centered at a desorption activation barr ier of 17.7 kcal/mol with a width (fwhm) of 3.0 kcal/mol. A first-orde r preexponential of 10(13) s(-1) was assumed in the model, and the res ults were in good agreement with the experimental data. The relatively high desorption activation barrier suggests a strong interaction betw een the methanol and the surface, most likely due to oxygen lone pair interactions with aluminum. The distribution of adsorption sites sugge st that aluminum oxide surfaces, even in single crystals, are very inh omogeneous.