Molecular Modeling as a Tool for Adhesive Performance Understanding
- PDF / 930,834 Bytes
- 7 Pages / 417.6 x 639 pts Page_size
- 43 Downloads / 202 Views
Figure 1. Example of bleed after a dynamics calculation between the filler and the substrate surface in a parallel plate configuration. Dynamics calculations at room temperature were then applied and spatial analysis of the results used. Any migration of molecules attracted on the substrate surface outside of the boundaries set by the initial minimized form was identified as "bleed". Figure 1 is an example of the atomistic model after dynamics calculation, showing attraction of the binder molecules to the substrate (bottom) surface and lateral spreading along the substrate. Model sizes averaged around 3000 atoms and varied depending upon the binder and substrate. For adhesion studies, only the oligomeric forms of the cured polymers were modeled on varying surfaces with the limited assumption that interfacial properties control adhesion. In this case, the dynamic calculation used applied a forcing potential in order to shear the oligomer from it's lowest energy state on the substrate. The subsequent total energies of the binder/substrate combination were analyzed for the amount of energy change in the system required to separate the entities. In this way, trends of binder adhesive strength were determined. All components were based upon consituents of JM adhesive materials using mostly cyanate ester type components, but also epoxies and silicones. Because of the proprietary nature of the formulations, reference to specific compounds used must be withheld. RESULTS Surface Spreading ("Bleed") and Underfill Flow Relating the surface spreading or bleed characteristics of the formulation components to the energy contributions, such as the dispersive energies, were initially difficult to correlate until it was realized that large changes in energy were necessary for practical prediction purposes. However, it was found that a bleed trend could be predicted (as shown in Figure 2) using the spatial analysis of the output. In order to pinpoint formulation causal effects to performance, a component analysis using structural elements was done which defined the best structural types that contribute surface spreading. The specific results will be discussed in another paper [9],
24
"D) (D
~0
Normalized Experimental Bleed
Figure 2. Comparison of Experimental Bleed Results to Model Results however, basically it was found that the dynamic calculation in conjunction with structural analysis contributed to the resulting bleed prediction and has helped us choose components and determine the severity of their effects. In general, specific functional group effects were not found to radically affect bleed, rather a combination of shape (as identified by the geometry and mass), chemical structure (as identified by the number and type of bonds) and chemical composition (such as atom types) were found to be important. Surface effects have been one of the most important correlations studied during the modeling. A significant effect on bleed involved the condition of the surface. For instance as shown in Figure 3, when formulations A 700
Data Loading...
