The measurement of ion diffusion in epoxy molding compounds used to encapsulate microelectronic devices

Plastic encapsulation has almost completely replaced hermetic ceramic packages for the protection of semiconductor integrated circuit devices. The plastic encapsulant is not a hermetic seal and is permeable to moisture, gases and ions. When the absorbed moisture is combined with ions, there is an opportunity for electrolytic corrosion of the metal surfaces of the device and package elements. Information regarding the diffusion of ions through commercial encapsulants has not been reported in the literature. In this dissertation three analytical techniques are used to measure ion diffusion in commercial encapsulants. Diffusion cell measurements and time-of-flight secondary ion mass spectroscopy showed that the ions diffuse through the encapsulant as a front. A Fickian diffusion profile was also detected at the leading edge of the diffusing ion front by dynamic secondary ion mass spectroscopy depth profiling.; Diffusion coefficients were calculated for moisture and ions in a commercial encapsulant. The diffusion coefficient for chloride ions in the encapsulant were seven to eight orders of magnitude smaller than the diffusion coefficients for moisture. We concluded that the slow ion diffusion rate in the commercial encapsulants is due to the binding of the ions to ion scavengers and functional groups in the polymer matrix. At the slow ion diffusion rates observed in these studies, ionic contaminants present on the surface of the encapsulant will not diffuse to the surface of the encapsulated semiconductor device in the projected life of the device.; A change in the slope of an Arrhenius plot of the diffusion coefficients for chloride ions as a function of temperature was observed which corresponded to the glass transition temperature of the encapsulant. This increase in ion diffusion above the glass transition temperature has implications for the assignment of acceleration factors for accelerated tests, such as HAST tests, which may be run at temperatures above the glass transition temperature of the encapsulant. The effect of post-mold curing of the encapsulant and the encapsulant formulation on ion diffusion rates in the commercial encapsulants are also addressed. A mathematical model for ion diffusion in the encapsulants is presented.