Spectroscopic ellipsometry is a non destructive technique, well adapted to the determination of pore volume fraction, pore size distribution and accessible surface area of porous thin films, it is particularly attractive for low-k films. These quantities are deduced from the variation of the refractive index during adsorption and desorption of gas ^1,2. Several difficulties are encountered in the interpretation of the ellipsometric spectra, in particular the knowledge of the density and refractive index of the skeleton, its fraction of micropores, the variation of the film thickness during adsorption/desorption process. We present results obtained on mesoporous oxide films prepared by surfactant templating methods. A combination of spectroscopic ellipsometry measurements in the visible and in the infra red range is shown to solve part of the uncertainties mentioned above, in particular for silica films, where a realistic estimation of the skeleton refractive index can be obtained.

¹M. R. Baklanov, K. P. Mogilnikov, V. G. Polovinkin, F. N. Dultsev, J. Vac. Sci Technol B 18, (2000) 1385.

² C. Wongmanerod, S. Zangooie, H. Harwin, Appl. Surf. Sci.172, (2001) 117.

Nanometer thick carbon films are of high interest for magnetic hard disk coatings and read-write heads. Next generation hard disk drives require carbon overcoats with thickness in the 1-2 nm range, to reach storage densities above 100 Gbit/in². As future head flying heights of about 6 nm promote head-disk contacts, the mechanical characterization of such films is of primary importance. For the measurement of the elastic moduli two techniques, both exploiting surface acoustic waves (SAWs), have been proposed: Surface Brillouin Scattering (SBS) and laser-induced SAWs (LISAW). In this work their performances have been assessed by a round robin test: two sets of carbon films, grown on silicon by two different cathodic arc systems, were independently measured by SBS and LISAW.

Both techniques are non-destructive: they measure the velocity of SAWs, and derive from it the film properties. Their operation is however significantly different: while SBS relies on thermally excited SAWs, LISAW induces them by laser pulses, and allows significantly faster measurements. LISAW analyses pulse propagation in the time domain, while SBS is a spectroscopic technique. LISAW measures propagation in the MHz to hundreds of MHz frequency range, while SBS operates in the tens of GHz range, and is thus potentially more sensitive to perturbations by thin films. However velocity measurements by LISAW are more precise, and these two facts tend to compensate.

The results of the round robin test show a good correlation between the Young's moduli measured by SBS and LISAW. This proves the ability of SAW based techniques to assess the Young's modulus in a thickness range not attainable by conventional techniques, such as nano-indentation. However, some systematic numerical differences are present between the sets of data provided by the two techniques. The causes for such differences are discussed and indications to improve measurements and data analysis are given.