Fundamentals of Slurry Design for CMP of Metal and Dielectric Materials
- PDF / 518,467 Bytes
- 9 Pages / 612 x 792 pts (letter) Page_size
- 109 Downloads / 307 Views
Fundamentals of
Slurry Design for CMP of Metal and Dielectric Materials
Rajiv K. Singh, Seung-Mahn Lee, Kyu-Se Choi, G. Bahar Basim, Wonseop Choi, Zhan Chen, and Brij M. Moudgil Abstract The formulation of slurries for chemical–mechanical planarization (CMP) is currently considered more of an art than a science, due to the lack of understanding of the wafer, slurry, and pad interactions involved. Several factors, including the large number of input variables for slurries and the synergistic interplay among input variables and output parameters, further complicate our ability to understand CMP phenomena. This article provides a fundamental basis for the choice of chemical additives and particles needed for present-day and next-generation slurry design. The effect of these components on nanoscale and microscale interaction phenomena is investigated. Methodologies are suggested for the development of next-generation slurries required to overcome CMP challenges related to defectivity and the surface topography of soft materials such as low- dielectrics and copper. Keywords: chemical–mechanical planarization, chemical–mechanical polishing, CMP, low- dielectrics, nanoparticles, slurry design.
Introduction Chemical–mechanical planarization (CMP) is widely recognized as the technology of choice for eliminating topographic variations and achieving wafer-level global planarization in ultralarge-scale integration (ULSI) circuits.1–4 The last decade has witnessed significant advancements in CMP technology, both in the development of more sophisticated processing tools and in the formulation of novel slurries to further enhance process performance.5 In spite of these advancements, CMP remains one of the least understood areas in semiconductor processing. Specifically, the proprietary nature of most CMP slurry formulations lends credence to the general belief that the CMP process is more of an art than a science. The information that can be obtained from commercial slurries or gleaned from the patent literature is inadequate for developing an understand-
752
ing of the wafer–pad–slurry interactions that occur during the CMP process. The lack of this understanding is a significant barrier to the development of more sophisticated next-generation slurries. CMP is replete with complexities that further limit our ability to develop a coherent picture of the polishing process. An ideal CMP slurry is expected to simultaneously produce high removal rates (2000–8000 Å/min), excellent global planarity, a good surface finish, low surface defectivity (minimal scratches, corrosion, and other defects), and high selectivity in regard to the underlying layers (particularly in metal polishing and in shallow trench isolation, a process that isolates adjacent devices on a chip by embedding a layer of silicon dioxide between them), all without the introduction of surface topography such as dishing and erosion.2 Other
important slurry issues include ease of handling, waste disposal, long-term stability of the slurry, decreased levels of indu
Data Loading...
