Multi Beam Full Cut Dicing of Thin Si IC Wafers

Abstract

The ongoing trend to thinner wafers which are needed for continuous miniaturization, 3D packaging and IC performance, inevitably means that sole blade dicing evolution is coming to an end. Over the last years several technologies to handle the separation process of thin Si wafer dicing have been evaluated (DBG, Stealth, Plasma, etc). Although they are capable for certain applications to meet the process specifications, they achieve this at expense of flexibility, productivity and process costs. ALSI, the inventor of multi beam dicing for semiconductor materials, has developed a technology using a multi beam laser concept which allows to dice through thin Si IC wafers while achieving a die strength equal or higher than achieved with blade dicing. In this single step process a multi beam laser configuration allows to remove the (ultra) low-K and metal top structures, dice through the Si substrate and recover the die strength (>450MPa for a 70um Si wafer). This technology allows the semiconductor industry to continue with the development of advanced node wafer technology utilizing (ultra) low-K and thick metal structures while having a separation technology that can cope with all these process steps. The presentation will address how the multi beam laser dicing process is an enabling technology and the first process in the world that can meet the die strength criteria without the need of additional process steps which increase the cost and reduce the flexibility and yield of the process. Multi beam laser dicing allows semiconductor manufacturers to execute their technology roadmap in a cost efficient manner. This presentation will address in depth, the challenges and issue's that semiconductor manufacturers are facing with respect to the dicing of thin (ultra) low-K IC wafers. We will present the die strength and quality that has been achieved using the multi beam dicing technology and compare this to other separation technologies. We will disclose how a multi beam process will play a dominant role in achieving an extremely small Heat Affected Zone combined with a significantly higher productivity. It will be demonstrated how a unique combination and optimization of multiple beams, pulse duration, and low pulse energy, can meet the challenging requirements set by the industry. In addition dicing results and achieved productivities will be presented.