Electromigration in Sputter Deposited Copper/Zirconium Alloys
- PDF / 1,540,327 Bytes
- 6 Pages / 417.6 x 639 pts Page_size
- 80 Downloads / 251 Views
Mat. Res. Soc. Symp. Proc. Vol. 563 © 1999 Materials Research Society
zirconium strips were placed. Zirconium was suited to this deposition technique as its high melting point allowed the strips on the surface of the target to withstand the high temperatures generated during sputter deposition. Previous work with lower melting point elements, such as tin, had proven impractical due to this problem. In a series of experiments 3 different copper/zirconium alloy compositions were studied. Alloy composition was varied by changing the number of zirconium strips on the target; 2, 3 and 5 strip arrangements were used. The zirconium strips were evenly distributed, placed as radials across the racetrack. The copper and copper alloy films were deposited directly on oxidised silicon wafers, which had been cleaned immediately prior to processing. A bias of 80 V was applied to the substrate, before and during deposition, to promote adhesion of the copper and copper alloys to silicon dioxide. A number of runs were carried out with each composition ensuring that the results were repeatable and consistent. System base pressure was 3x1 07 mbar, process power was 400 Watts and process pressure was 8xl 0-3 mbar. The presence of the Zr strips did not significantly
influence the deposition rate, which remained the same as that for pure copper, 100 nm/minute. We used Rutherford Back-Scattering (RBS) to obtain an overall zirconium concentration. Since this method does not rely on the availability of standards, it is a good technique to use for novel materials. It is important to know how much zirconium is present in grain boundaries and at surfaces and interfaces. To investigate segregation of zirconium to the top surface and interface, Secondary Neutral Mass Spectrometry (SNMS) was used as it is known to provide better quantification data at interfaces compared to Secondary Ion Mass Spectrometry (SIMS). An attempt was made to employ scanning Auger microscopy in order to observe lateral segregation of zirconium to grain boundaries. An Atomic Force Microscope (AFM) was used to characterise the film growth process by analysing grain-size and grain-size distribution at film thicknesses ranging from 100 to 400 nm. X-Ray Diffraction (XRD) was used to provide texture information by comparing peak intensities. Quantitative stress measurements were obtained using an ADE 'Stressmaster', a noncontact capacitive probe tool, which calculates stress from the change in wafer curvature after film deposition. The technique used to obtain electromigration results has been described previously [6]. The method used was to test twenty lines in parallel structures, using a constant voltage source to ensure that current density in all the lines remains constant throughout the test. Both copper and alloy films were patterned into the test structures by argon ion etching. Lines were 1 mm long, 2 jtrm wide and 100 nm thick. The testing was carried out under nitrogen to ensure minimal oxidation of the samples. Testing temperature was 300TC and the current dens
Data Loading...
