J. Nerlov et al., A PHOTOEMISSION-STUDY OF THE COADSORPTION OF CO2 AND NA ON TIO2(110)-(1X1) AND TIO2(110)-(1X2) SURFACES - ADSORPTION GEOMETRY AND REACTIVITY, Surface science, 371(2-3), 1997, pp. 321-336
The coadsorption of CO2 and Na on TiO2(110)-(1 x 1) and -(1 x 2) surfa
ces have been investigated by synchrotron-radiation based core-level a
nd valence band photoemission. We find that the initially adsorbed Na
exhibits a core-level shift of 1.15 eV when the two surfaces are compa
red. From a simple adsorption model this binding energy shift is under
stood in terms of a difference in initial Na adsorption site on these
surfaces. While the (1 x 1) surface seems to favor Na adsorption in a
hollow site ''between'' bridging surface oxygen atoms, it is found tha
t the (1 x 2) surface facilitates a chemically more advantageous Na ad
sorption ''adjacent to'' the bridging oxygen atoms. Valence band measu
rements support this model since Na adsorption on the (1 x 2) surface
leads to emission characteristic of alkali-oxygen-like compounds while
this is not the case for the Na/TiO2(110)-(1 x 1) system. Finally, th
e relatively high resolution of the core-level emission allows in a di
rect way the various features contributing to the Na 2p core-lever emi
ssion to be determined. With respect to adsorption of CO2 we find for
the (1 x 2) surface that CO2 uptake saturates around 0.5 ML Na coverag
e compared to 1 ML for the (1 x 1) surface, indicating that the Na cov
erage required for saturation of CO2 uptake is proportional to the den
sity of protruded oxygen rows present at the surface. The CO2 uptake,
however, increases as the density of the oxygen rows decreases. Valenc
e band photoemission data obtained from both interfaces show that a su
rface carbonate species is formed. At lower coverages/exposures there
are, however, indications of the presence of a CO2- species rather tha
n carbonate, there by suggesting that the carbonate species is formed
through the surface reaction: 2CO(2)(-) --> CO32- + CO.