Room temperature grain growth in sputtered Cu films
- PDF / 2,062,908 Bytes
- 6 Pages / 612 x 792 pts (letter) Page_size
- 103 Downloads / 206 Views
W12.6.1
Room temperature grain growth in sputtered Cu films D. Deduytsche°, C. Detavernier°, J. Debaerdemaeker+, R.L. Van Meirhaeghe°, C. Dauwe+, T.S. Kuan* ° Department of solid state physics, Ghent University, Krijgslaan 281/S1, 9000 Gent, Belgium + Department of nuclear physics, Ghent University, Proeftuinstraat 86, 9000 Gent, Belgium * University at Albany, SUNY, Albany, NY ABSTRACT The results are presented of a study on grain growth in magnetron sputter deposited Cu films. Similarly to the well-known case of electroplated Cu, we observed significant changes in the microstructure of sputter deposited films during storage at room temperature, as evidenced by a decrease in sheet resistance (up to 40% decrease in less than 6 hours). Direct evidence for grain growth was observed by FIB and SEM and by a strong decrease in the XRD peak width. Moreover, a decrease in the total defect content of the film was observed as a function of storage time from positron annihilation experiments, probably reflecting the decreasing number of grain boundaries in the film. It is well known that the microstructure of an as-deposited film is strongly dependent on the deposition parameters. This relationship is summarized in the Thornton structure-growth zone diagram [7]. We have studied the kinetics of room temperature grain growth as a function of Ar pressure and substrate temperature for Cu films with a thickness between 50 nm and 1 micrometer. It is found that spontaneous grain growth during storage at room temperature occurs mainly for a zone T microstructure (at low Ar pressure and a substrate temperature below 40°C), while it does not occur for zone I (high Ar pressure, low substrate temperature) or zone II (substrate temperature > 100°C) microstructures. INTRODUCTION It has been shown that the microstructure of electroplated copper films is unstable during storage at room temperature [1-6]. The driving force for this ‘self-annealing’ or spontaneous grain growth originates from the small grainsize of the as-deposited grains (< 0.1 µm), which are pinned by impurities originating from the plating bath. After an incubation time (the length of which depends on the composition of the plating bath, deposition speed and film thickness) some grains break loose from the pinning pressure and consume the neighboring grains. The pinning effect is strongly dependent on the additives that are added to the plating bath [6]. EXPERIMENT The depositions were done on silicon substrates that were covered with a native oxide (formed by RCA cleaning). The samples (about 6 cm2 in size) were mounted onto a home-made substrate holder that could be cooled to 77K by liquid nitrogen and heated to > 800°C by infrared heating lamps. All Cu layers were deposited in a vacuum of 10-6mbar. Prior to Cu deposition, an e-beam source was used to evaporate Ti onto the chamber walls, in order to minimize oxygen contamination in the Cu films. The copper target used for sputtering was 99,95% pure OFHC
W12.6.2
copper (purchased from Goodfellow). For comparison, some
Data Loading...
