Role of Film Hardness on the Polish Rates of Metal Thin Films
- PDF / 324,478 Bytes
- 6 Pages / 420.48 x 639 pts Page_size
- 8 Downloads / 148 Views
123
Mat. Res. Soc. Symp. Proc. Vol. 566 ©2000 Materials Research Society
EXPERIMENTS 3 gm thick copper films and 1 gm thick tantalum films were sputter-deposited on 125 mm diameter single crystal Si wafers using a TORR International CRC- 150 sputtering system. A 50 nm Ta adhesion/barrier layer was deposited prior to copper deposition. 1 gm thick tungsten coated silicon wafers were purchased from Silica source technology. The hardness of films and metal discs were measured using Nanomechanical test instrument from Hysitron Inc. A Berkovich diamond tip was used in all the measurements. The measurements were made at 10 different peak loads of 300 gN and 600 gN and at intervals of 1000 jiN between 1000 and 8000gN. The loading and unloading were done at a constant rate of 250 gN/s, except at the two lower loads where a rate of 100 gN/s was used. After reaching the peak load, the load was held constant at the peak value for 5 seconds to minimize time dependent plasticity. Figure 1 shows a typical load time sequence. The nanomechanical system provides a
continuous measurement of load and penetration depth. Figure 2 shows the load versus displacement curves for the copper- and tungsten-coated wafers. The hardness (H = Pmax/Ac) 1200 1000
' 800 600 200 0 0
5
10 15 Time (sec)
20
25
Figure 1. Load-Time sequence - peak load 1000 jiN 2500 2000 Z
1500
0 1000 0J
500 0 0
50
100
150
Indentation Depth (nm) Figure 2. Load vs Displacement curves 124
200
is computed from the peak load (Pmax) and contact area A,. The contact area is obtained from the indenter tip shape function, which relates contact area to the contact depth. Contact depth is the depth up to which the indenter is in contact with the sample at peak load. Hardness measurements were carried out at five different locations on the wafer at the same load. The reported values are obtained by averaging the five measurements. The effect of chemicals used during the CMP process on the mechanical properties of copper and W thin films was investigated by dipping the wafers coated with these films in 5 wt % H20 2 and 0.1 M ferric nitrate solutions, respectively. The wafers were dipped for five minutes, washed in de-ionized water and dried. Hardness of these exposed films was measured as above. Polishing experiments were carried out on a bench top Struer's polisher with 3 mm thick metal disks (99.999% pure) of cross sectional area 7.5 cm2. The table speed was 90 rpm and the disk holder was stationary. The applied downward pressure was about 4.3 x l04 N min2 . The solids concentration in the slurries was maintained at 3 % by weight and the slurry feed rate was I ml s-1 for all the experiments. The slurry in the supply tank was stirred continuously with a magnetic stirrer to avoid settling of agglomerated particles. A Suba 500 polish pad was used and was hand conditioned prior to each experiment using a 220 grit sandpaper and a nylon brush. The polish rate was determined from the difference in the weight of the disk before and after polishing for 3 minutes and the r
Data Loading...
