Spectrum of Platinum homepage

2. Photographic Observations

 
Our observations were made with the 10.7 m normal-incidence vacuum spectrograph at the National Institute of Standards and Technology. Two different gratings were used, the first blazed at 1200  in first order and the second blazed at 3000  in first order. Both gratings were ruled with 1200 lines/mm. All measurements were made in the first order, the plate factor being 0.78 /mm. The slit width was 0.023 mm. With this slit width the resolving limit throughout the region of observation was about 0.020 . Photographic exposures were made on Kodak SWR plates.

Two different light sources were used. The first was a windowless, demountable hollow-cathode lamp having a solid copper cathode containing a helical platinum wire and some chips of silicon and germanium. The general design of the lamp was similar to that of Reader and Davis [14]. In the version used in the present work the O-ring assembly at the front of the lamp was replaced by a large ball joint by which the lamp could be connected directly to the spectrograph. The lamp was operated in series with a 300 Ω ballast resistor at a dc voltage of 250 V and a current of 90 mA. The cathode was cooled with flowing water. The carrier gas consisted of flowing helium with a trace of neon at a total pressure of approximately 266 Pa (2 Torr). With this gas mixture the spectra of both Cu and Pt could be excited simultaneously. This could not be accomplished when only a single gas was used. Exposure times for this lamp were about 15 min.

The second source was a sealed hollow-cathode lamp similar to the one used by GHRS. It has a platinum hollow cathode with neon carrier gas and is sealed with a magnesium fluoride window. The lamp was manufactured by the Westinghouse Corporation (Model WL34045). It was connected to the spectrograph by a quick-disconnect flange. The cathode was located 215 mm from the slit. The lamp was operated with a 5000 Ω ballast resistor at a dc voltage of 310 V and a current of 20 mA. Exposure times ranged from 2 to 150 min.

In the first phase of the wavelength reductions of the photographic data, the spectra of Pt observed with the demountable Pt-Cu lamp were measured with respect to lines of Cu II, Si I, Si II, Ge I, Ge II, Ne I, and Ne II, to determine accurate wavelengths for a select group of Pt lines. Wavelengths for Cu II were Ritz values derived from the level values of Ross [15]. Wavelengths for most Ne I and Ne II lines above 2780  were taken from the Fourier-transform measurements of Palmer and Engleman [16]. Wavelengths for other Ne II lines above 2780  and all Ne II lines below this wavelength were Ritz values given by Persson [17]. Ne I, Si, and Ge wavelengths were taken from the compilation of reference wavelengths by Kaufman and Edlén [18]. The measurements made with the demountable Pt-Cu lamp provided accurate values for about 1500 lines of Pt I and Pt II extending from 1032  to 2885 .

In the second phase of the reductions the spectra of all lines observed with the sealed Pt/Ne lamp were measured with respect to the above group of Pt lines, lines of Ne I and Ne II, and lines of Pt I reported by Engleman [6]. In the region above 2885 , our reference spectra consisted solely of lines of Ne I, Ne II and Pt I with wavelengths taken from the sources cited above.

Next, our values for lines of Pt II with known classifications were combined with values for classified lines of Pt II measured by Engleman by means of Fourier-transform spectroscopy to determine accurate values for 28 even and 72 odd energy levels of Pt II [7]. Using these level values we calculated Ritz-type wavelengths for almost all of the classified lines of Pt II. For some of these levels the energy or J value has been revised as a result of the work of Blaise and Wyart [13]. For those levels that have not been changed, the Ritz values have been substituted for the measured values in the final list of wavelengths.


Previous page Table of Contents Next page