Abstract :
In a fast growing semiconductor market, cost reduction has driven towards package miniaturization with increasing die performance. C90, C65 and C45 wafer technologies with smaller bond pad opening and bond pad pitch have emerged. The requirement of such wafer technologies leaves little or no room for minute die top delamination, as it may result in bond ripped off and electrical failure. To prevent die top delamination, one of the very key elements is to ensure an excellent adhesion between die surface and mold compound. Hence, understanding of die surface passivation behavior after plasma treatment prior to molding is very crucial to ensure high performance package reliability. In this study, TOF-SIMS analysis was carried out on 4 different types of die surface passivation to characterize the surface behavior prior to molding. This was done before and after pre-mold plasma. As widely known, effect of plasma treatment are 2 folds: 1. Removal of surface contaminant from non-organic surface, 2. Removal of surface contaminant and formation of polar groups on organic surface. Formation of polar groups is one of the key desired results as it will help enhance mold compound adhesion through hydrogen bonding. Therefore, the purpose of this study is to characterize the type of plasma effect on each of the 4 different die surface passivations. The 4 types of die surface passivation being studied here were SiON, SiN, Polyimide Type A and Polyimide Type B. These were the 4 most common die surface passivations known to be used in the wafer fabrication. Wafers with the 4 different die surface passivations were sawn and die bonded onto PBGA substrate. All samples went through pre-wire bond plasma, lid attach and lid attach curing. After lid attach curing, samples were split into half, with 1 group without going through pre-mold plasma, and another group was treated with the standard production pre-mold plasma cleaning with Oxygen gas as the main agent. Both groups were then subje- - cted to TOF-SIMS analysis. Heat spreader was removed before analysis to allow analysis to be performed on the center of die top surface. For SiON and SiN passivation, TOF-SIMS result showed that the main effect of plasma treatment was removal of organic contaminants due to chemical reaction with Oxygen radicals and removal as volatile gaseous compound as a result of the chemical reactions. Both passivations showed minimum formation of polar groups (<; 5 groups). For Polyimide Type A, TOF-SIMS result showed that apart from removal of organic contaminants, there were 22 polar groups formation that will help enhance mold compound adhesion. For Polyimide Type B, TOF-SIMS result showed similar plasma effect in removing organic contaminants, but in addition to a formation of 45 polar groups. Number of polar group formation was Polyimide Type B (45 polar groups) >; Polyimide Type A (22 groups) >; SiON & SiN (<;5 groups). Reliability result further confirmed that no die top delamination was found after stressing with Polyimide Type B & Polyimide Type A, even though the polar group numbers was lower for Polyimide Type A die. However, delamination occurred with SiON and SiN die surface passivation. In summary, understanding of die surface passivation behavior is crucial in order to select the desired surface finishing to fulfill the package reliability requirement.
Keywords :
ball grid arrays; bonding processes; microassembling; passivation; semiconductor device reliability; PBGA substrate; TOF-SIMS analysis; bond pad opening; bond pad pitch; die surface passivation; hydrogen bonding; lid attach curing; mold compound; package miniaturization; package reliability; package stressing; pre-wire bond plasma; surface contaminant;