Previous studies on very high-cycle fatigue behavior of thin silicon films suggest a strong environmental dependence of the degradation mechanism, the precise nature of which is still subject to debate. In the present study, 2-micron-thick polycrystalline Si notched cantilever beam structures were used to investigate fatigue degradation in a high-temperature (80°C), high-humidity (90%RH) environment. The specimens were subjected to fully reversed sinusoidal loading at resonance (~40kHz) with stress amplitudes ranging from 1.46 to 1.6GPa, resulting in life-spans between 10⁶ and 10⁹ cycles. Comparison to a reference set of S-N data obtained at moderate environmental conditions (30°C and 50%RH) reveals a strong tendency for faster degradation with increasing temperature and humidity. The obtained damage accumulation rates in the 80°C, 90%RH environment exceed the reference by two orders of magnitude. Transmission electron microscopy (TEM) on vertical through-thickness slices reveals oxide thickening after cycling.
The influence of ~20nm Al[subscript2]O₃ deposited on the surface of the fatigue specimens using Atomic Layer Deposition (ALD) technique was also studied. The presence of the alumina coating results in a higher fatigue resistance at 30°C and 50%RH, as well as a drastically different frequency evolution behavior. No oxide thickening was observed in the TEM for coated run-out specimens. A model is proposed to explain the different degradation behavior of the ALD-alumina coated samples.
Thickened oxides after cycling appear consistent with the reaction-layer fatigue mechanism. Finite element modal analysis incorporating surface oxide layers and cracking was employed to relate the damage observed in TEM to the experimentally measured changes in resonant frequency. In conclusion, the reaction-layer mechanism seems capable of describing micron-scale polysilicon fatigue, even though the critical processes such as room-temperature, stress-assisted oxidation remain elusive.
| Identifer | oai:union.ndltd.org:GATECH/oai:smartech.gatech.edu:1853/31766 |
| Date | 19 November 2008 |
| Creators | Budnitzki, Michael |
| Publisher | Georgia Institute of Technology |
| Source Sets | Georgia Tech Electronic Thesis and Dissertation Archive |
| Detected Language | English |
| Type | Thesis |
Page generated in 0.0154 seconds