In this work, a systematic study of the influence of five deposition parameters, i.e., process pressure, substrate temperature, target power, and substrate bias, as well as gas composition on the residual stress in fully textured polycrystalline aluminum nitride thin films deposited on Si(100) wafers using the reactive sputtering method was performed. Post-growth residual stress measurements were obtained indirectly from radius of curvature measurements of the wafer prior to and after deposition. Two different techniques were used to determine the curvature: an optically levered laser beam and an x-ray diffraction method. Stresses in both cases were then evaluated using the Stoney formulation [G.G. Stoney, Proc. R. Soc. (London)A82,172 (1909)]. Both methods give similar results, with slight quantitative differences. The existence of a transition region between tensile and compressive stress previously reported in the literature is also confirmed. The transition is shown to be strongly dependent on the process parameters. Optimal films regarding stress were grown at 2 mtorr, 900 W at the target, a 20/45 Ar/N2gas mixture, and floating potential at the substrate. The substrate temperature did not influence the measured internal stress in the films.
Determination of Stresses in Incrementally Deposited Films From Wafer-Curvature Measurements
