This dissertation presents a series of studies related to the study and control of slurry flow, process temperature, and aggressive diamonds in Chemical Mechanical Planarization (CMP). The purpose of these studies is to better understand the fundamentals of CMP and to explore solutions to some of CMP’s greatest challenges. Within-wafer removal rate non-uniformity (WIWRRNU) is a critical parameter to determine film thickness planarity on a wafer-scale level and it grossly impacts yield. Resolving this issue continues to be an area of intense focus in the industry. The first study in this dissertation shows the feasibility of adopting a new method to improve WIWRRNU during copper CMP that is solely based on intentional local temperature manipulation of the pad. A pad surface thermal management system is developed to locally change pad surface temperature. This system consists of one or more thermal transfer modules contacting the pad surface. In this study, the system is employed to adjust the "center-fast" copper removal rate profile to illustrate its effect during the process. Results shows that, when two thermal transfer modules are employed, local removal rates in the wafer center region decrease significantly while maintaining the removal rates near the wafer edge thereby significantly improving WIWRRNU. Another contribution of this dissertation is the investigation of the effect of pad groove design on slurry injection scheme during interlayer dielectric CMP. A novel slurry injector with multiple slurry outlets is designed, which provides optional slurry injection schemes (i.e. one injection point scheme and multi-injection point scheme). These schemes are compared with the standard slurry application method on a concentrically grooved pad and an xy-groove pad, respectively. On the concentrically grooved pad, the one injection point scheme generates significantly higher oxide removal rates (ranging from 22 to 35 percent) compared to the standard slurry application method at different slurry flow rates. On the xy-groove pad, the one injection point scheme still results in higher removal rates (ranging from 3 to 9 percent), however, its removal rate enhancement is not as high as that of the concentrically grooved pad. In order to further improve slurry availability on the xy-groove pad, the multi-injection point scheme is tested. Results show that the multi-injection point scheme results in significantly higher removal rates (ranging from 17 to 20 percent) compared to the standard slurry application method. This work underscores the importance of optimum slurry injection schemes for accommodating particular groove designs. The last contribution of this dissertation involves a study regarding aggressive diamond characterization and wear analysis during CMP. A 3M A3700 diamond disk is used to condition a Cabot Microelectronics Corporation (CMC) D100 pad for 30 hours. The top 20 aggressive diamonds for two perpendicular disk orientations are identified before the polishing, as well as after 15- and 30-hour polishing. The furrow surface area generated by these top 20 aggressive diamonds and their evolution are analyzed and compared. Results show that the original top 20 aggressive diamonds identified before polishing are subjected to wear after the first 15-hour polishing as the furrow surface area that they generate decreases dramatically (by 47%). As these original aggressive diamonds are worn, seven new aggressive diamonds are "born" and join the new top 20 list for both disk orientations. After the second 15-hour wafer polishing, the furrow surface area of these new top 20 aggressive diamonds do not change significantly. The furrow surface area created by all the active diamonds exhibits the same trend as the top 20 aggressive diamonds, confirming that most pad conditioning work is performed by these aggressive diamonds and that the disk loses its aggressiveness in the first 15 hours of polishing and then maintains its aggressiveness during the second 15 hours, albeit to a lesser extent.

In this study, slurry availability and the extent of the slurry mixing (i.e., among fresh slurry, spent slurry, and residual rinse-water) were varied via three different injection schemes. An ultraviolet enhanced fluorescence technique was employed to qualitatively indicate slurry availability and its flow on the pad during polishing. This study investigated standard pad center area slurry application and a slurry injection system (SIS) that covered only the outer half of the wafer track. Results indicated that the radial position of slurry injection and the alteration of fluid mechanics by the SIS played important roles in slurry mixing characteristics and availability atop the pad. Removal rates were found to decrease with slurry availability, while a higher degree of slurry mixing decreased the fraction of fresh slurry and consequently lowered the removal rate. By using a hybrid system (i.e., a combination of slurry injection via SIS and standard pad center slurry application), the polishing process benefited from higher slurry availability and higher fraction of fresh slurry than the conventional pad center slurry application and the shorter SIS, individually. This work underscores the importance of optimum slurry injection geometry and flow for obtaining a more cost-effective and environmentally benign chemical mechanical planarization process.