The Raman spectra of diamond and chemical-vapor-deposition (CVD) diamond films in the UV have been excited within the diamond band gap at 228.9 nm for the first time. The lack of fluorescence in the UV-excited Raman spectrum of diamond and CVD diamond films allows Raman spectroscopy to monitor the carbon-hydrogen (C-H) stretching vibrations of the nondiamond components of the CVD film as well as the third-order phonon bands of diamond. The relative intensity of the C-H stretching bands at-2930 cm-' to the diamond first-order phonon band at 1332 cm-' is proportional to the atomic fraction of covalently bound hydrogen in the CVD diamond film. The third-order phonon band intensity and frequency maxima are very sensitive to the size of the diamond crystallite. Its intensity decreases, and the maximum shifts to lower frequency as the size of the diamond crystallite decreases. It is shown here that UV Raman diamond measurements have significantly greater information content than visible Raman measurements.

We report a preliminary investigation of surface analytical methods used to evaluate trace metal contaminants on silicon wafer surfaces. The methods discussed include sputter induced postionization methods, backscattering spectrometry, total reflection x-ray fluorescence, and a sample preparation method known as vapor phase decomposition. In order to initiate the evaluation, a series of silicon wafers was dosed with transition metal contaminants Cu and Fe and then analyzed with total reflection x-ray fluorescence, sputter induced resonant ionization spectroscopy, and heavy ion backscattering spectroscopy. One wafer with low concentrations of Fe was analyzed by time of flight secondary ion mass spectrometry. As a result of this study, we are pursuing improved capability in several areas including standards. Heavy ion backscattering spectrometry is suggested as a means of verifying surface dosed silicon wafer standard.

No abstract

This paper presents surface photovoltage (SPV) applications for the monitoring of chemical cleaning and purity of chemicals through mapping of minority carrier diffusion length, Fe concentration in the bulk, and surface contamination (surface charge and surface recombination). The noncontact, wafer-scale character of the new SPV approach and refined apparatus make this technique uniquely suited for heavy metal monitoring. This method was used to monitor Cu contamination in BHF by measurement of its effect on surface recombination and Fe contamination through it effect on bulk recombination after the rapid thermal annealing step used to drive Fe deposited at the surface during cleaning into the bulk. Fe surface contamination was measured down to the 1 • 109 cm -2 level while the detection limit of this approach is 2 x 108 cm -2. Different Fe contamination levels (1 to 13 ppb) present in different grades of H202 were easily distinguished. This procedure should allow one to monitor Fe contamination in H202 at the 1 ppt level. Cleanliness of incoming chemicals is not always a limiting factor and often is not related to the cleanliness of chemicals at the point of use (in the cleaning station).