A technique for tensile testing of very small specimens of thin films has been developed and applied to metals, a polymer, and a ceramic-like material. This technique is straightforward and able to reveal clearly the different behavior of the different types of materials. Since the gauge section is 10 by 180 micrometers in plan, the behavior of the aggregate film, including grain interiors and grain boundaries, is sampled. The specimen preparation techniques used to date have relied on photolithographic patterning and removal of a sacrificial substrate. Tensile tests of physical-vapor-deposited (PVD) films of several different specimen materials have been conducted within the scanning electron microscope (SEM).
Our scheme is similar in principle to the apparatus originally reported by Greek et al., with some differences. The key components are all commercially available off the shelf: three piezo-stepper-motor-driven stages and a displacement sensor; for elevated temperature testing a hot stage is also used. Simple hardware is used to assemble the three stepper stages into a three-axis micromanipulator and to support the force probe. The complete procedure for testing is as follows. Specimens are formed by photolithography and freed by etching away the sacrificial substrate beneath the tensile section and the moving end. The fixed end of the specimen remains attached to the substrate, while the moving end is pulled by a tungsten probe tip. By etching to a depth of 50 micrometers beneath the specimen, we are able to avoid interference between the probe tip and the substrate. Force is sensed indirectly via the displacement of flex plates that support the probe holder, measured using a non-contacting eddy current sensor. The displacement of the gauge length is measured using digital image correlation to analyze 100-500 SEM images acquired during the test. Flags patterned into both ends of the specimen provide convenient fiducials for correlation. Temperature is varied by placing a heated stage beneath the substrate, and by a resistive winding on the probe holder. We have reached temperatures up to 150 C, limited by outgassing that degrades the vacuum in the SEM chamber. We have used these techniques in both optical and scanning electron microscopes. For our current specimen sizes, typically 10 by 180 micrometers, the principal benefit of testing in the SEM is its greater depth of focus. We are in the process of attempting to apply this technique to narrower specimens, where the higher magnification of the SEM will be needed to image the specimen. This testing scheme has been applied to pure aluminum films deposited in our laboratory, aluminum films made in a commercial CMOS fab facility, polyimide films, and polysilicon films. The differences among the stress-strain curves for these very different materials were as dramatic as would be expected