Effect of deposition temperature on surface acoustic wave velocity of aluminum nitride films determined by Brillouin spectroscopy

M. B. Assouar, R. J. Jiménez Riobóo, M. Vila, P. Alnot
2005 Journal of Applied Physics  
Brillouin spectroscopy has been used to study the effect of the deposition temperature on the surface acoustic wave ͑SAW͒ propagation velocity of aluminum nitride ͑AlN͒ films. The results show a dependence of the SAW propagation velocity on the growth temperature of AlN films. The highest value of acoustic velocity was obtained for the film elaborated without heating. Structural characterization of the AlN films synthesized at various deposition temperatures was carried out by x-ray
more » ... y x-ray diffraction. These analyses pointed out that the deposition temperature influences the standard deviation of ͑002͒ AlN film preferred orientation. The growth temperature clearly influences the acoustical and crystalline properties of AlN thin films. Aluminum nitride ͑AlN͒ has been considered as an attractive thin film piezoelectric material for integrated circuit ͑IC͒ compatible surface acoustic wave ͑SAW͒ devices. This compatibility requires a deposition process at relatively low temperatures. 1 Then, the study of acoustical properties of AlN films as a function of deposition temperature becomes crucial to develop AlN films of high crystallinity and acoustical quality. High resolution Brillouin spectroscopy ͑HRBS͒ was used in order to obtain information about the surface acoustic waves propagation velocity in AlN films elaborated at different growth temperatures. Even though there are some HRBS works on AlN elastic properties, 2-7 the study of the SAWs properties is still a rarity 8 and no works about the determination of temperature effects on SAW propagation velocity by Brillouin spectroscopy were carried out. In this work, c-axis oriented aluminum nitride thin films on ͑100͒ silicon substrates were deposited by the reactive rf magnetron sputtering technique at various substrate holder temperatures ͓without heating ͑WH͒ up to 400°C͔ with the same thickness ͑1.3 m͒. The aluminum target ͑purity 99.99%͒ diameter was 107 mm and 6.35 mm thick. The deposition chamber was pumped down to a base pressure of 1 ϫ 10 −7 mbar by a turbomolecular pump prior to the introduction of the argon-nitrogen gas mixture for AlN thin film production. The gas discharge mixture was Ar/ N 2 and the total pressure was kept constant at 5 ϫ 10 −3 mbar. The nitrogen percentage in the Ar/ N 2 gas mixture was 60% and the rf power delivered by the rf generator was 170 W. In order to perform a better comparison between the different samples, the deposition time of AlN films was adjusted to obtain 1.3 m thick films for various deposition temperatures. In order to study the growth thermal conditions on the SAW propagation velocity ͑from now on SAW velocity͒, HRBS was the experimental technique chosen. The experimental set up was already described elsewhere. 9 It can be summarized as follows: The light source was a 2060 Beamlok Spectra Physics Ar + ion laser provided with an intracavity temperature stabilized single-mode and single-frequency z-lok étalon ͑ 0 = 514.5 nm͒. The scattered light was analyzed using a Sandercock-type 3 + 3 tandem Fabry-Pérot interferometer. 10 The incident polarization direction was chosen to be in -plane ͑p polarization͒ while no polarization analysis of the scattered light was made. The typical values for finesse and contrast were 150 and 10 9 , respectively. AlN films are very transparent materials thus making extremely difficult to obtain information of SAWs velocity by means of HRBS. A successful way to enhance the ripple scattering mechanism 11 is to deposit a very thin metallic film on the transparent sample. 8, [12] [13] [14] It has been shown that the thin metallic film will reproduce the main features of the SAWs of the transparent material, as in the case of AlN and synthetic diamond. 8,14 A 40 nm thin Al film was deposited on each of the different AlN film samples via dc magnetron sputtering. The Brillouin spectroscopy was made for all the samples at backscattering with a sagittal angle of 55°. As far as we were only interested in the temperature evolution of the surface acoustic wave velocity, we only needed to fix one acoustic wave vector and follow its changes. In this case kh = 0.8 ͑k is the scattering wave vector and h the thickness of the Al thin film͒. All the recorded spectra are of similar quality and the inset in Fig. 1 shows the typical spectrum for a growth temperature of 250°C, where the Rayleigh mode ͑SRM͒ and a higher order Sezawa mode ͑PSM͒ can be seen. The SAW propagation velocity can be obtained straightforward from the Brillouin frequency shift ͑f͒: JOURNAL OF APPLIED PHYSICS 98, 096102 ͑2005͒
doi:10.1063/1.2121927 fatcat:tlrj4rrldrdgfbdsikbfdrtply