The CO/CO2 conversion performed over a Ru(0001) surface shows the decisive
role of subsurface oxygen which explains the discrepancy between former UHV
results and high-pressure experiments. At oxygen pressure higher than abou
t 10(-3) mbar the oxygen atoms start to penetrate the bulk. This subsurface
oxygen has to be considered as a species which accompanies the high-pressu
re surface reactions where oxygen is involved. This new oxygen phase exhibi
ts unique properties clearly different from adsorbed oxygen as well as from
regular ruthenium oxides. The presence of subsurface oxygen raises the yie
ld for CO-oxidation reaction by more than two orders of magnitude. Oxygen a
toms forming the subsurface phase seem to be very mobile as can be deduced
from the reaction kinetics which show that the CO2 formation is limited onl
y by the thermal diffusion of subsurface oxygen towards the surface.