Absorption Spectrum of Cd I Vapor at 30 Å Using the Background Radiation of the 2.5-GeV Bonn Synchrotron

Abstract

26The results given in this paper for the soluble model are in complete agreement with the asymptotic results obtained previously for the f and g equations of the BBGKY hierarchy.New resonances have been observed in the absorption spectrum of neutral cadmium vapor at wavelengths around 30 k These are believed to be at present the shortest-wavelength metal-vapor absorption features recorded optically. The 2. 5-GeV electron synchrotron at Bonn University was used as a background source of continuous radiation. Hartree-Fock calculations enabled identification of the resonances as caused by excitation from the 3d shell, although somewhat larger differences between computed and observed wavelengths were found than were expected. Following our previous work on the absorption spectra of metal vapors' and the work recently published by Ederer, Lucatorto, and Madden on the Liz absorption spectrum, we report here absorption features caused by excitation of a 3d innershell electron in Cd r. vapor, the structure being more than 400 V above the first ionization potential of Cd z. Our results were obtained in a collaborative experiment between the European Space Research Institute and the Synchrotron Group of the Physikalisches Institut at the University of Bonn.In some earlier work ' using the Baloffet, Rome, nd, and Vodar (BRV) sources to provide a background radiation continuum, spectra of alkalimetal vapors were recorded down to 180 A. At wavelengths below about 100 A, the BRV continuum became too weak to record absorption spectra. The furnace design used in Ref. 2 allowed a metalvapor pressure of several Torr to be achieved in the central region of the furnace while maintaining a pressure of about 10 Torr at the ends, butdid not meet the much more stringent requirements (vacuum better than 10 ' Torr) for operation on a synchrotron. Ederer, Lucatorto, and Madden have used a heat-pipe furnace system together with 1000-Athick aluminum windows to contain Li vapor at pressures of up to 2 Torr. %ith this apparatus, they were able to study the Liz absorption spectrum between 170 and 210 A using the background radiation continuum of the National Bureau of Standards 180-MeV synchrotron.They were not, however, able to observe structures much below 170 A, because of the I, » cutoff in the aluminum filters.Vfe have further developed the furnace described in Ref. 2, essentially by increasing the efficiency of the vapor trapping system. Details of our arrangement will be published separately. Spectra were recorded up to temperatures corresponding to a Cd vapor pressure of 3 Torr at the center of the furnace, the over-all length of the furnace being about 1.5 m. However, owing to the steep pressure gradients in the trapping system, the pressure-path product is probably much smaller than these figures would indicate. Our spectra were recorded using a 2-m grazing incidence spectrograph equipped with a 1200-line/mm Bausch and Lomb platinized replica grating, the entrance slit being set for a 2-deg grazing angle. The slit width was 6. 3 p, ...

Structure in the Zn I absorption spectrum associated with the excitation of the 3s 2 subshell

“…The discrepancy of 1.1 A accords with the percentage error previously found in the case of the 3d shell spectrum of Cd 1 by Connerade et at. (1972) and again, the H artree-Fock wavelength is shorter than the experimental wavelength, in contrast to the usual trend (Connerade 1971(Connerade ,1972 in calculations on highly stripped ions a t similar wavelengths. A better situation exists with respect to the 3s limit: a H artree-Fock calculation places this a t 85.3 A -almost exactly the experimental value.…”

“…Structure in the Zn i absorption spectrum A H artree-Fock calculation performed with the computer code of Froese-Fischer (1969) yielded a wavelength of 87.5 A for the 3s -> 4p transition against an experimental wavelength of about 88.6 A. The discrepancy of 1.1 A accords with the percentage error previously found in the case of the 3d shell spectrum of Cd 1 by Connerade et at. (1972) and again, the H artree-Fock wavelength is shorter than the experimental wavelength, in contrast to the usual trend (Connerade 1971(Connerade ,1972 in calculations on highly stripped ions a t similar wavelengths.…”

“…Increasing the vapour density of the metals results only in an increase of continuous absorption. Again, in Cdi (Connerade et al 1972; Connerade, Mansf ratios of line breadths to intervals between successive members of the 3d inner shell series are much smaller than for the Zn 1 and Hg 1 series mentioned, but of the sam e order as for A ri (Nakamura et al 1968). We can account for these differences by interpreting the breadth as a consequence of autoionization: the 'stability5 of inner shell series against autoionization has been found, from considerable experimental evidence on metal vapour absorption spectra a t longer wavelengths (Garton & Connerade 1969;Connerade 1970Connerade a, 1972Connerade et al, 1971; to depend on the proximity of a long wavelength limit of the same J -value as the 'core5 angular momentum of the inner shell series.…”

The K I absorption spectrum in the vacuum ultraviolet: 2p-subshell excitation

“…A good fit between observations and calculation is immediately obtained by shifting the calculated average energies of each configuration down by about 6500 cm-1 in figure 2. Such shifts in the average energies are often necessary in fitting H artree-Fock calculations to experimental data (see, for example, Mansfield & Connerade 1972). I t is usual for them to be in the same direction and of similar magnitude for each configuration.…”