The Effects of Environment and Fatigue on the Adhesion and Subcritical Debonding of Dielectric Polymers
- PDF / 2,138,912 Bytes
- 6 Pages / 417.6 x 639 pts Page_size
- 44 Downloads / 227 Views
understand the role that mechanical fatigue and environmental conditions play in debond growth velocity. This study will outline testing techniques and present results from an examination of the effects of fatigue and environment on low-K dielectric polymer interfaces. Fatigue and Environment-Assisted Debonding The fatigue life of bimaterial interfaces can be characterized either by debond growth measurement, which measures life after initiation of a debond, or via a more general stress-life approach. This study will use the debond growth technique to examine interfacial debonding. In this analysis, the subcritical growth rate of debonding under cyclic loads, da/dN, (where a is the debond length and N is the number of loading cycles) is related to the applied debond driving force range, AG (AG = Gmax - Groin, where Gmax and Gmin are the maximum and minimum values of the applied strain energy release rate, G. It is well known that many bulk materials possess a sigmoidal curve on a log-log plot of da/dN versus AG, with a linear region over a wide range of growth rates. Recent studies have shown that this same relationship holds for many bimaterial interfaces [1, 2]. The intermediate range of debond growth rates can therefore be described by a simple power law relationship, da dN = C.AGm, (1)
123 Mat. Res. Soc. Symp. Proc. Vol. 565 01999 Materials Research Society
where C and m are empirically derived fitting parameters. For many bulk polymer and metal systems, m has been found to be = 2-6, while for more brittle ceramic systems m > 15. Determination of the parameters C and m and integration of the crack-growth (Eq. 1) are the basis of defect tolerant lifetime predictions [3]. The effect of corrosive environmental species (such as moisture) can be investigated through the use of load relaxation testing. During such static testing, a fixed displacement, corresponding to an initial subcritical debond driving force, G, is applied to a sample and the debond length is monitored as a function of time. This subcritical debonding is most often caused by moisture that interacts with strained bonds near the debond tip. Similar to behavior observed for cyclic fatigue testing, the debond growth rate is often seen to follow the same power-law relationship described above in Eq. 1, but with AG = Gmax and different C and m parameters [4, 5]. EXPERIMENT
BCB/Silicon Interface System Sandwich structures were fabricated with a benzocyclobutene-type polymer (CYCLOTENETM 3022, Dow Chemical Co.) constrained between two silicon substrates. BCB is a low-k dielectric polymer used in interconnect and packaging levels of microelectronic devices. Previous work has characterized the critical adhesion of BCB films to various materials [6], but the subcritical debonding behavior has not been investigated. A 1.0 Ptm film of BCB was spincoated onto a silicon wafer and cured for 1 hr. at 300'C under high purity N 2. The BCB-coated wafer was cleaved into 40 mm squares and the BCB surfaces were bonded together with a thin film of epoxy (= 5 gtm) cur
Data Loading...
