T1-relaxation of 129Xe on metal single crystal surfaces-multilayer experiments on iridium and monolayer considerations

J Magn Reson. 2002 Nov;159(1):1-12. doi: 10.1016/s1090-7807(02)00006-x.

Abstract

The surface of a typical laboratory single crystal has about 10(15) surface atoms or adsorption sites, respectively, and is thus far out of reach for conventional NMR experiments using thermal polarization. It should however be in reach for NMR of adsorbed laser polarized (hyperpolarized) 129Xe, which is produced by spin transfer from optically pumped rubidium. With multilayer experiments of xenon adsorbed on an iridium surface we do not only demonstrate that monolayer sensitivity has been obtained, we also show that such surface experiments can be performed under ultra high vacuum conditions with the crystal being mounted in a typical surface analysis chamber on a manipulator with far-reaching sample heating and cooling abilities. With only four spectra summed up we present an NMR signal from at most 4x10(14) atoms of 129Xe, four layers of naturally abundant xenon, respectively. The fact that no monolayer signal has been measured so far is explained by a fast Korringa relaxation due to the Fermi contact interaction of the 129Xe nuclei with the electrons of the metal substrate. T(1)-relaxation times in the order of several ms have been calculated using all electron density functional theory for several metal substrates.