Effects of Interlayers on the Scratch Adhesion Performance of Ultra-Thin Films of Copper and Gold on Silicon Substrates
- PDF / 2,059,591 Bytes
- 6 Pages / 414.72 x 648 pts Page_size
- 87 Downloads / 228 Views
be accurately controlled while the stylus moves. Finally, the behavior of the scratched film must
be observed in some way. Traditionally, scratches have been examined by optical microscopy, but more sensitive methods are necessary for work in the nanometer regime. TECHNIQUES FOR NANO-SCALE SCRATCHING
The apparatus used in these experiments was a specially modified nanoindentation system [68]. The system has a load resolution of 0.076 gN and a displacement resolution of 0.04 nm. Its ability to apply controlled, very low loads to the diamond stylus satisfies the first requirement for performing sub-micron scratch tests. To satisfy the second requirement of precision movement of the diamond relative to the film at controlled loads, a precision x-y table with a resolution of 0.1 micron was used to slide the film under the diamond with a minimum of vibration and electrical noise. To measure the lateral forces on the diamond, two separate capacitive displacement gauges were used to sense the lateral displacement of the indenter column in the X and Y directions. Lateral forces were then calculated from the stiffness of the column determined in calibration experiments. Data from these lateral displacement sensors, as well as the normal force and displacement data and table position data, combined to provide an in situ method of determining the behavior of the scratched film. In addition, optical and scanning electron microscopy were used to observe the scratches after testing. The stylus chosen for scratch testing was a standard Berkovich indenter--a pyramidal diamond with a triangular cross-section and a 65.3Y angle between the vertical axis and each of the three faces. Scratches were made with both an edge of the pyramid (a comer of the triangular crosssection) and with a face of the pyramid (a side of the triangular cross-section) facing the direction of scratching. 809 Mat. Res. Soc. Symp. Proc. Vol. 356 ©1995 Materials Research Society
Scratches were made under load controlled conditions. A standard scratch was formed by continuously monitoring and incrementing the load on the diamond such that it was increased linearly in time from some small initial value (approximately 0.02 mN) to its maximum value (0.2 mN to 5.0 mN) over a period of 100 seconds. Simultaneously, the film was translated under the indenter at a rate of 5 microns/second, resulting in a load-ramped scratch 500 microns in length. An additional 100 microns of constant, low load (0.02 mN) scan was usually placed both before and after the scratch to aid in the analysis of the displacement data. A series of standard indentation experiments using the Berkovich diamond was performed on each sample to be scratched. From observation of the load and displacement data, the minimum load required to penetrate the film was estimated for each system. This load then provided a ceiling for the maximum load to be used during a scratch test. In general, tests were conducted at several lower loads prior to testing at the ceiling load. For example, a specimen with a ce
Data Loading...
