Room Temperature Self-Annealing of Electroplated and Sputtered Copper Films
- PDF / 3,665,876 Bytes
- 8 Pages / 417.6 x 639 pts Page_size
- 101 Downloads / 251 Views
413 Mat. Res. Soc. Symp. Proc. Vol. 564 © 1999 Materials Research Society
EXPERIMENT DC plating with commercially available and in-house developed electrolyte were used to electrochemically deposit the copper films. For the in-house electrolyte, one of the organic additives was varied to study the effect of its concentration on the film self-anneal rate. Blanket Ta and TaN barrier films of 250A in thickness, followed by 2000A copper seed layer, were sputtered on 200mm oxidized silicon wafers using an Electra Ion-Metal Plasma (IMP) chamber [7]. IMP sputtering was used for these experiments since it is the preferred method for deposition of barrier and seed layers for high aspect ratio damascene features, resulting in better step coverage than conventional DC magnetron sputtering. Sheet resistance (Rs) measurements were made on a Tencor RS75. A Phillips X-Pert® diffractometer was used to conduct X-ray diffraction measurements, and Transmission Electron Microscopy (TEM) analysis was performed using a JEOL TEM at Applied Materials. RESULTS AND DISCUSION The sheet resistance change of electroplated copper films with 0.2, 0.5 and 1Lpm thicknesses was compared to examine the film thickness' effect on self-anneal rate. The selfanneal behavior is characterized by the copper grain growth, leading to a change in film sheet resistance. Figure 1 indicates that the Rs for 1ijm of ECP Cu film decreases by approximately 18% over 100 hours at room temperature, while those for 0.2 and 0.5 pm of ECP films drop much more slowly, showing that the rate of self-anneal is higher for thicker films. Sheet Resistance Percent Change over Time SiO2/Ta barier(250A)IMP Cu Seed (2kA)/ECP Cu plated using Electrolyte A 0
upm
•, -5 _ C
c
-a
Cn
-15
i
L.
__ _2 c cu, . 52KAECP Cu ... 5KA ECP Cu
10
_.
.
IOKAECP Cu
-i
-20 1
10
100
1000
"Time(hr) Figure 1. Rs change verses time of ECP Cu films with three thicknesses at room temperature. X-ray diffraction measurements were also made on the above copper films to confirm the recrystallization. The full-width half-maximum (FWHM) of the Cu (111) crystallographic peak of the three different electroplated films as a function of time at room temperature are illustrated in Figure 2. Similar to the results shown for sheet resistance, the FWHM value for 1 lm ECP copper film decreases quickly from 2.80 and stabilized at approximately 2.1' in 100 hours, whereas it was still evolving for the thinner films. It has been reported that residual stress is responsible for self-anneal of ECP films[8]. Stress measurements in this study collected over time showed small difference between films of various thicknesses. These observations indicate that film thickness may play a more significant role in the ECP copper self-anneal rate. Thicker ECP Cu films self-anneal faster than the thinner ones. The film thickness dependence of self414
anneal rate may depend on the grain size verses the film thickness. Self-anneal rate decreases as the grain size gradually approaches the film thickness. In thinner films, gr
Data Loading...
