The Effect of Film Stress on Indentation Modulus/Hardness for Silicon Oxide Films
- PDF / 258,388 Bytes
- 6 Pages / 420.48 x 639 pts Page_size
- 94 Downloads / 217 Views
225 Mat. Res. Soc. Symp. Proc. Vol. 563 © 1999 Materials Research Society
EXPERIMENT AND RESULTS A set of eleven PETEOS films was made. To minimize substrate effects on indentation, the films were deposited to 4 mm in thickness. A 25 nm thin layer of SiN was then deposited in situ to prevent moisture absorption by the PETEOS films. By changing the RF power, the wafers were deposited with different stresses. The residual film stress was obtained by measuring the wafer curvature before and after the deposition. It was also verified that the residual stresses were stable after exposed to the ambient for months indicating that the SiN layer was effective in preventing moisture absorption. Indentation measurements were made using Nano Instrument's Nano-II microindenter with a Berkovich diamond tip and continuous sensing capability so that the modulus and hardness were deduced as a function of depth. The total depth of penetration was about 200 nm, sufficiently deep to eliminate any effects from the 25 nm SiN coating. As shown in Fig. 1, the modulus and hardness were virtually constant beyond about 30 nm of penetration. The modulus and hardness were obtained by averaging the data points between 50 and 150 nm of contact depth, they are plotted in Fig. 2 and 3 with respect to the residual stress, a correlation between the modulus/hardness and the residual stress is evident.
80 70 Modulus
60 CD 50 0.
"I- 40 ou 30
20 Hardness
U
10
-OEOE
*********MMMM UMMMMMMMM EUEEEEEUEEE
UEEEEEEUEE
EuE.U..
0 0
50
100
150
200
Depth (nm) Figure 1. A set of modulus and hardness curves plotted against the penetration depth showing that the modulus/hardness are practically constant when penetration is larger than 30 nm. To determine whether the apparent change of modulus and hardness was a direct consequence of the film stress, one of the wafers with residual stress of 11.1 MPa was cut to pieces of lxi cm and bonded to steel substrates of different thickness at elevated temperatures. Because of the thermal mismatch between silicon and the steel substrates, the silicon pieces were stressed by the steel substrates and therefore provided additional stresses to the SiO2 films on the top surface. FEM calculations were performed to determine the optimal thickness for the steel 226
76 74E
70.25- 0.0326
G0
C. 72
770 o 68 66
•
64
-150
-100
-50
0
50
100
150
Residual Stress GO (MPa) Figure 2. Elastic modulus measured for a set PETEOS films with various residual stresses.
11.5
11
H = 10.3 - 0.0055 Go
I)10.5 I-}
-
10
9.5
-150
-100
-50
0
50
100
150
Residual Stress ao (MPa) Figure 3. Hardness measured for a set PETEOS films with various residual stresses.
227
0.5 0.4 0.3 0.2
S0.1 ~ci0
Lu *-0.1
0.5
1
1.5
2
2.5
-0.2 -0.3
-0.4 -0.5 Steel/silicon Thickness Ratio Figure 4. Normalized stress at the top of silicon piece as a function steel/silicon thickness ratio. substrates. The results plotted in Fig. 4 show the in-plane stress on the top surface of the silicon piece near the center is tensile for relativ
Data Loading...
