The Influence of Argon Cluster Ion Bombardment on the Characteristics of AlN Films on Glass-Ceramics and Si Substrates

In this paper, the influence of surface modification on the characteristics and properties of AlN thin films on Si and glass-ceramics substrates is investigated. The surface modification was made at various parameters of argon cluster ions. By using XRD and Raman spectroscopy, it was shown that the obtained AlN films have a hexagonal structure with a characteristic direction of texturing along the c axis and slight deviations from it. A comparison of the AlN surface morphology obtained by atomic force microscopy before and after cluster processing was demonstrated. This demonstrated that the cluster ions with low energy per atom (E/N = 10 eV/atom) have a high efficiency of surface smoothing. A decrease in the intensity of the Raman peaks and an increase in their full-width after bombardment with cluster ions were found, which may be caused by a change in the physicochemical state of the surface. The optical properties, the quality of the boundaries, and the distribution map of the thickness of the functional layer of AlN were investigated by the methods of spectral and spatial resolution ellipsometry. By using the cross-sectional SEM, the direction of crystallite texturing was demonstrated. The influence of argon cluster ion bombardment on the stoichiometry of samples was analyzed by EDX spectroscopy. The results obtained demonstrate the efficiency of the cluster ion smoothing of polycrystalline thin films for microelectronics, particularly when creating surface acoustic wave resonators.

Automated summary

This paper investigates the influence of surface modification on the characteristics and properties of AlN thin films on Si and glass-ceramics substrates. It is found that low-energy cluster ions have a high efficiency of surface smoothing. The optical properties and thickness distribution of the functional layer of AlN are also investigated, and the direction of crystallite texturing is observed using cross-sectional SEM.