200. EFFECT OF COPPER CMP SLURRY CHEMISTRY ON THE RATE OF AGGLOMERATION OF ALUMINA PARTICLES

Department: NanoEngineering
Faculty Advisor(s): Jan Talbot

Primary Student
Name: Neil A Brahma
Email: nbrahma@ucsd.edu
Phone: 858-534-5663
Grad Year: 2009

Abstract
Chemical mechanical planarization (CMP) is a process used in integrated chip manufacturing to obtain uniformly planarized surfaces by utilizing both mechanical abrasion and chemical etching. During the process, alumina particles, which are used as abrasives, may agglomerate into undesired larger particles that can cause scratches and other defects. By understanding the rate of agglomeration as a function of chemistry, new slurries can be developed that will ultimately reduce scratches and defects. Our current research focus involves the measurement of agglomeration rate of alumina particles in various solution chemistries used in copper CMP. Solutions of de-ionized water with 1 mM KNO3 (for constant ionic strength) with and without copper were tested while the pH was adjusted using KOH and HNO3. Once the desired pH was reached, 2 drops of alumina, from a dispersion containing 40 wt% α-alumina in DI water manufactured by Cabot Corporation, were added to the solution. The alumina primary particles are 20 nm in diameter which form clusters with a median diameter of 150 nm. To study the effects of copper in solution, 0.12 mM of copper nanoparticles was added. Solutions were stirred in a sonicator for 5 minutes. Particle size measurements were made using a ZetaPlus particle size analyzer every minute for the first 25 minutes and then every 30, 45, 60, 120, 240, 1440, and 2880 minutes. The rates of agglomeration for pH of 4 and 10 were zero; there was no apparent change in particle size with time. However at pH of 7.5, the particles agglomerate at an approximate rate of 82 nm/min. The presence of copper increases the rates of agglomeration at pH values of 7.5 and 10. At pH of 2, while there was an increase in agglomerate sizes, there was no clear increase in agglomeration rate. However, for pH of 7.5 and 10, there was a clear increase in agglomeration rate. Work is now being done to model the agglomeration rates as a function of chemistry. Additional agglomeration experiments investigating the effects of additives are also planned.

« Back to Posters or Search Results