Increasing demands for more reliability and functionalities in electronic devices have pushed the electronics industry to adopt newly developed materials and reduce interconnect sizes and spacing. These adaptations have led to concerns of reliability failures caused by conductive anodic filament formation (CAF). CAF is a conductive copper-containing salt that forms via an electrochemical process. It is initiated at the anode and grows along the epoxy/glass interface to the cathode, and once CAF reaches the cathode a short circuit will occur. The objective of this research is to evaluate and compare the effects of curing agents (DICY vs. phenolic-cured epoxy) and drilling methods (laser vs. mechanical drilling) on CAF formation using an insulation resistance test at 85 ºC, relative humidity of 85%, and a voltage gradient of 0.4V/µm. Time-to-failure for DICY-cured and phenolic-cured epoxy with laser drilled microvias and mechanically drilled vias were determined using the insulation resistance test. The failed coupons were cross-sectioned and examined using a Scanning Electron Microscope equipped with Energy-dispersive X-ray spectroscopy to verify the existence of CAF. Weibull analysis was used to compare the reliability and identify the failure modes of the failed coupons. Test results show that DICY-cured epoxy is a better CAF resistant material than phenolic-cured epoxy. It is believed that the brittleness of phenolic-cured material might enhance the damage to the epoxy/glass fiber interface during drilling; and hence, facilitate subsequent CAF formation. The study also shows that laser drilled microvias are less prone to CAF formation than mechanically drilled vias, because there is less mechanical damage and lower glass fiber content. Finally, using Weibull analysis, it is determined that laser drilled microvias experienced infant-mortality failure, whereas mechanically drilled vias exhibited a wear-out type failure.

Increasing demands for more reliability and functionalities in electronic devices have pushed the electronics industry to adopt newly developed materials and reduce interconnect sizes and spacing. These adaptations have led to concerns of reliability failures caused by conductive anodic filament formation (CAF). CAF is a conductive copper-containing salt that forms via an electrochemical process. It is initiated at the anode and grows along the epoxy/glass interface to the cathode, and once CAF reaches the cathode a short circuit will occur. The objective of this research is to evaluate and compare the effects of curing agents (DICY vs. phenolic-cured epoxy) and drilling methods (laser vs. mechanical drilling) on CAF formation using an insulation resistance test at 85 ºC, relative humidity of 85%, and a voltage gradient of 0.4V/µm. Time-to-failure for DICY-cured and phenolic-cured epoxy with laser drilled microvias and mechanically drilled vias were determined using the insulation resistance test. The failed coupons were cross-sectioned and examined using a Scanning Electron Microscope equipped with Energy-dispersive X-ray spectroscopy to verify the existence of CAF. Weibull analysis was used to compare the reliability and identify the failure modes of the failed coupons. Test results show that DICY-cured epoxy is a better CAF resistant material than phenolic-cured epoxy. It is believed that the brittleness of phenolic-cured material might enhance the damage to the epoxy/glass fiber interface during drilling; and hence, facilitate subsequent CAF formation. The study also shows that laser drilled microvias are less prone to CAF formation than mechanically drilled vias, because there is less mechanical damage and lower glass fiber content. Finally, using Weibull analysis, it is determined that laser drilled microvias experienced infant-mortality failure, whereas mechanically drilled vias exhibited a wear-out type failure.

Manganese dioxide compounds are preferred curing agents for Polysulphide resins used as sealants in industry. These are required to have consistent setting characteristics and the investigation was initiated to characterise a number of proffered compounds of this type an to establish criteria by which an informed choice could be made of an optimum curing ages for a specific set of conditions. Several different chemical and physical properties were examined and critical parameters were established. A compound - sodium birnessite- was identified as a significant agent in the determination of curing properties. It was synthesised and its curing properties alone and in combination with other manganese dioxide compounds was evaluated. In an effort to find a specific reaction which might be used to characterise manganese dioxide curing agents it was decided to examine the classical reaction between these compounds and potassium chlorate. A literature search revealed major contradictions in the reported conditions under which potassium chlorate undergoes thermal decomposition as result of which it was decided to study the decomposition of potassium chlorate alone and in the presence of manganese dioxide and other catalysts. During this investigation a hitherto unreported high temperature structural change in potassium chlorate at 341° C was identified. The existence of this reversible change was confirmed by Powder Diffraction X-Ray analysis and an orthorhombic (near tetragonal) more open structure was assigned to it. It is suggested that the rapid decomposition of potassium chlorate in the solid state presence of catalysts is related to this change to a more open structure.

The aim of the work reported here was to determine the influence of an amine curing agent, and postcure cycle on the mechanical and thermal properties of diglycidyl ether of bisphenol A (DGEBA) epoxy resin. The results of this initial study were then used as the basis for selecting material to obtain optimum toughness in epoxy/glass fibre systems. These basic materials were further used to make comparisons with the properties of modified resin systems which contained commercial elastomers. Differential Scanning Calorimetry (DSC), Dynamic Mechanical Thermal Analysis (DMTA), Fourier Transform Infrared Spectroscopy (FTIR), flexural and interlaminar shear tests, Instrumented Falling Weight Impact (IFWI), visual observation, Scanning Electron Microscopy (SEM), and Transmission Electron Microscopy (TEM) were all used to investigate various properties and the structures which gave rise to them. The properties of cured products were found to be affected by the amounts of curing agent, curing times and temperatures, and the structure of the elastomers. Not surprisingly the maximum thermal and mechanical properties tended to be found in the stoichiometric (standard) mix systems. However, postcuring at higher than room temperature, which was used as the basic curing temperature, led to more conversion. This effect improved the thermal and mechanical properties of both the unmodified and modified resin systems. The maximum flexural strength of 104 MPa of the unreinforced resins was found in the stoichiometric mix ratio after postcure at 150°C for 4 hr. However, the maximum flexural modulus and glass transition temperature (Tg) were found after postcuring at the same temperature for 48 hr. This was believed to be due to increased crosslinking, but unfortunately the longer curing time led to degradation of the resins. In the systems modified with -20 phr of polyetheramine elastomers, the one modified with the lowest molecular weight (2000) was found to have the highest flexural strength (85.8 MPa) and modulus (2.5 GPa). The impact properties of all the composites with modified resin matrices were found to be higher than the unmodified resin matrix composites. The best impact properties were, however, obtained with the elastomer modifier with a molecular weight of 4000. The impact energy at maximum force increased from 11.9 to 16.4 J, and energy at failure increased from 18.7 to 21.6 J. This increase in impact properties was due to the increase in areas of phase separated elastomer particles over similar systems with lower molecular weight modifier.