Optimization of laser release layer, glass carrier, and organic build-up layer to enable RDL-first fan-out wafer-level packaging

International Symposium on Microelectronics(2016)

引用 2|浏览8
暂无评分
摘要
Abstract With increasing demand for mobile devices to be lighter and thinner and consume less power while operating at high speed and high bandwidth, many equipment suppliers and assembly participants have invested great efforts to achieve fine-line fan-out wafer-level packaging (FOWLP). However, the inherent warp of reconstituted wafers, which can contribute to poor die placement accuracy and/or delamination at the interface of the build-up layer and carrier, remains a major challenge. In this study, the interactions among laser release layer, glass carrier, and build-up layer were evaluated for optimization of redistribution layer (RDL)–first FOWLP as a foundation to move toward fine-line FOWLP. In this study, a series of experiments incorporating glass carrier, laser release layer, and build-up layers were carried out to determine the optimal setup for RDL-first FOWLP. First, glass carriers (300 mm × 300 mm × 0.7 mm) with coefficients of thermal expansion of 3 and 8 ppm/°C were treated with 150-nm laser release layers. After deposition of 0.1 μm of sacrificial material on the glass carrier, 8-μm build-up layers were coated and patterned by lithography to electroplate Cu interconnections with a density of approximately 10% of the surface area. Subsequent to die attachment, molding compound was applied on top to form a 200-μm protective overcoat. The reconstituted wafer was then separated from the glass carrier through a laser ablation process using a 308-nm laser to complete the design of experiments (DOE). An experiment to study the correlation of glass carrier, laser release layer, build-up layers, and molding compound in RDL-first FOWLP processes is discussed to address full process integration on 300-mm glass substrates. The combination of glass carrier, laser release layer, build-up layer, and molding compound will pave the way for realizing cost-effective RDL-first FOWLP on panel-size substrates.
更多
查看译文
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要