Practical Limitations to Indentation Testing of Thin Films
- PDF / 1,323,589 Bytes
- 6 Pages / 414.72 x 648 pts Page_size
- 105 Downloads / 293 Views
91 Mat. Res. Soc. Symp. Proc. Vol. 505 ©1998 Materials Research Society
I6-AI2ON I I()TIgAIN
Figure 1. Bright field TEM image of an AIN film grown on sapphire at 700°C
92
Figure 2. Corresponding SAD pattern showning "alignmentof the AIN with the sapphire substrate
RESULTS AND DISCUSSION Microstructure Evidence of epitaxy is shown in the transmission electron microscopy (TEM) analysis of a 110 nm thick AIN film deposited on 700' C sapphire substrate shown in Figure 1. The corresponding selected area diffraction (SAD) pattern in Figure 2, was taken along the coincident [0001] zones of AIN and A120 3 and shows the alignment of the film and the substrate. The (0001) plane of AIN is parallel to the (0001) plane of A12 0 3 and the [112 0] direction in AMN parallel to the [1010] direction in A120 3. A Moir6 pattern arises in the bright field image in Figure 2 due to the coincidence of the (101 0) AIN reflection with the (112 0) A120 3 reflection. Similar microstructures were noted in films grown on 9000 C substrates. The well aligned grains have a rough, granular surface when examined in the SEM and also using the SFM capability of the Hysitron. Mechanical properties were evaluated for two AIN films grown at substrate temperatures of 7000 and 9000 C. Surface profiles are shown in Figure 3 comparing the two films with an approximate image of the Berkovich tip superimposed over the surfaces. For the 7000 C film, with a peak-valley roughness of 15 nm, the tip is likely to hit either peaks or valleys. For the 9000 C film, with a peak-valley roughness of 30 nm, in addition to the tip hitting either peaks or valleys, it may also occur on a intermediate point.
V
V
.100
.100
50
0
1=
50
4(=
0
1CW
2
mo
4
Figure 3. Profile of surface Roughness of AlN films grown on (a) 7000 C and (b) 9000 C sapphire substrates using Hysitron/NanoScope II with cubic diamond tip. Profiles are shown plotted in nm's. Elastic Moduli and Hardness The method used to calculate elastic modulus and hardness [2] utilizes the load and displacement data taken as the indentation is made (Figure 4). Hardness, H, (eqn. 1) is calculated from the peak load, P. over the contact area, A,. The reduced elastic modulus, E, (eqn. 2) is determined from the slope of the unloading curve (stiffness, S) divided by the square root of the contact area, Ac. The effects of the indenter can be subtracted to give the elastic modulus of the specimen, E. (eqn. 3). All data can be extracted from the unloading curve, except for the contact area. This area is deduced from the contact depth extrapolated from the initial portion of the unloading curve. P max
.
,0.4
LAVA. ..
H
load ng
.__ boo a ing lof
30.1
if 0.2
E
0
r
...
5
..
10
15
Ltact
20
25
(1)
P. A,
2SS47-l [A-
,t
E, = 2gs
(2)
-s
30
-1-=
E
Disp.(nm)
V
,
(3)
+ (-
E
Figure 4. Continuous indentation load (P) vs. displacement data (h).
93
The relationship between the contact depth and the area of contact is obtained by a series of indentations in a known modulus material such as
Data Loading...
