A method using high-speed videography is developed for visualizing and quantifying the width of the slurry bow wave formed at the leading edge of a retaining ring during polishing. This new method overcomes many limitations of previous techniques as it employs an inert dye thereby bypassing any photobleaching issues encountered in the past. This allows the use of pads of all types and colors in a normally-lit environment. Image analysis is used to quantify and classify pixel brightness in regions of interest on the CMP pad. By employing suitable baselines, evolution of the slurry bow wave (specifically its width) as a function of polish time was calculated and analyzed. This new method was then applied to retaining ring polishing experiments at two different flow rates and in two different regions where the bow wave was formed. Results consistently show that the bow wave fluctuated with time and its width grew wider at higher flow rates, especially in region closer to the slurry injection point. Fast Fourier Transformation (FFT) was employed to convert the fluctuating component of the bow wave width from time domain to frequency domain. When looking at the spectral signature of the slurry bow wave width, major peaks were observed at 1, 2, 4, 8-9 and 12-14 Hz all of which could be theoretically explained by taking into account ring geometry and polisher kinematics.

Single-run Stribeck+ curves are constructed using real-time, high-frequency, shear force and normal force data from the wafer-slurry-pad interface during copper and cobalt (on ILD wafers) CMP in conjunction with multiple slurries, pads and conditioning discs at various pressures and relative velocities. To avoid having to perform actual polishing experiments to obtain blanket film removal rates, “big data” sets from the same Stribeck+ curves are used to construct new “Kinetic” curves to help infer relative blanket wafer removal rates. The “Kinetic” curves, which are based on the assumption that material removal is Prestonian, are eventually validated with actual removal rate studies involving different wafer types processed at various pressures-velocity combinations with the same pads, conditioning discs and slurries. A strong correlation is seen between the actual and “inferred” removal rates which renders credibility to our new ultra-rapid and ultra-low-budget approach for determining removal rates that does not require any wafer polishing nor any film thickness metrology.

This study investigates the effect of various retaining ring slot designs, conditioning schemes and conditioning disc types on the width of the slurry bow wave formed at the leading edge of the retaining ring during polishing. A method using high-speed videography is employed for visualizing and quantifying slurry bow wave width. In contrast to many limitations associated with previous techniques, this new method allows the use of a concentrically grooved pad with its natural white color commonly used in industrial applications. In general, results show that polishing with a retaining ring having rounded-angle slots, and a full-face conditioner, generates wider slurry bow waves. In contrast, using a retaining ring with sharp-angle slots, in combination with a conditioner having CVD-coated protruding vanes results in narrower bow waves. Compared to ex-situ

We investigated the possibility of employing refractive index (RI) measurements for inline incoming slurry control at the point of use (POU), as an alternative to the widespread densitometry method. As such, it became necessary to determine if RI could detect smaller changes in slurry composition and, therefore, provide a tighter control. Three industrially-relevant silica-based slurries, namely, Fujimi PL-7106, Klebosol 1501-50, and CMC W7801, were characterized using both densitometry and RI measurements. Initial solutions of the three slurries were prepared and increasingly small amounts of ultrapurified water (UPW) were added to study the change in slurry properties. Results showed that both density and RI decreased linearly with the addition of water for all three slurries, with the 1501-50 being the most sensitive to water addition. A linear correlation between the two properties was found, with R2 values that exceeded 0.95 in all cases. Furthermore, the approximate limit of detection of both metrology tools was estimated based on the slope of the fitting line and resolution. When compared to densitometry, RI was found to be the far superior method for detecting smaller changes in water concentration.