Photoelectron spectroscopy: retrospects and prospects

Abstract

Electron spectroscopy is at present developing towards higher resolution both at low and high energies of the exciting radiation. The limitations are set by quite different reasons in the two regions, i.e. in the v.u.v. and in the X -ray region. In the former region the limit is due to several independent factors: Doppler broadening in the sample gas, small residual electric fields over the sample volume and, of course, possible imperfections of the electron spectrometer. In addition to this the ultimate resolution is set by the inherent line width of the u.v. radiation. This width is dependent on the extremely high temperature of the light-source gas, owing to Doppler effect, and the gas pressure, owing to self-absorption. If a resolution of 5 m eV should be attained all these factors have to be thoroughly studied and separately handled. This leads to arrangements where supersonic molecular jet beams are required, where developments of specially designed light sources are necessary and where extreme care has to be taken over the details of the spectrometer. A new extension of electron spectroscopy is under development. It consists of a simultaneous excitation or fragmentation of the molecules at photoionization. There are several ways to arrange such a spectroscopy, which, for convenience, could be classified as a ‘ dynamic ’ electron spectroscopy in contrast to the usual ‘ static ’ spectroscopy, which deals with the emission of electrons from normal molecules. A particularly attractive scheme for such a dynamic spectroscopy would be a combination between electron and laser spectroscopy. In the X -ray region high resolution cannot be attained unless the selected X -rays are further properly monochromatized towards the limit set by the rocking curve of the diffracting crystal being used. To secure sufficiently high intensity a high-power X -ray source has to be used, either a synchrotron source or a water-cooled, swiftly rotating anode. For the latter, experiments and calculations show that Al Kot and spherically bent quartz crystals (1010, 1st order diffraction) with almost backwards reflection can yield a resolution of 0.15 eV. With nearly identical geometry Ag L a (second order) has a rocking curve of the same crystal, which amounts to 0.08 eV. Sc 3 and Ti K a (third order) have 0.03 eV resolution, and M n K a and C r K p t 3 (fourth order) also 0.03 eV. For most purposes the first choice (Al K a) is the one to be preferred. Some recent investigations related to solids, gases and liquids made by several groups in our laboratory are reviewed.