Tribological, Kinetic and Thermal Characteristics of Copper Chemical Mechanical Planarization
Keywordscopper chemical mechanical planarization
flash heating model
coefficient of friction
Committee ChairPhilipossian, Ara
MetadataShow full item record
PublisherThe University of Arizona.
RightsCopyright © is held by the author. Digital access to this material is made possible by the University Libraries, University of Arizona. Further transmission, reproduction or presentation (such as public display or performance) of protected items is prohibited except with permission of the author.
AbstractCopper polishing performance depends significantly on the properties of pads, slurries, conditioning, pressure, sliding velocity, slurry flow rate and temperature. A slight variance in each of these parameters will cause significant change in polising results. Various investigations are performed during this study to understand the effect of consumables and other main operating parameters on copper polishing in terms of removal rate, lubrication mechanism, and temperature transients. A modified two-step Langmuir-Hinshelwood removal rate model and a flash heating thermal model are developed to describe the removal mechanism.Results indicate that grain size plays an important role during copper polishing. Smaller grain size may enhance the chemical rate by providing a higher density of favorable reaction sites. However, denser grain boundaries due to smaller grain size may reduce the mechanical rate by increasing the probability of disruption of three body sliding contact. It is found that removal rate increases as slurry flow increases from 60 to 80 cc/min because higher slurry flow rate can provide more reactants to the system. Then removal rate decreases as slurry flow rate is further increased to 140 cc/min, which is due to synergic effects of the wafer temperature, slurry flow and slurry residence time under the wafer. The observed removal rate drop is thought to be due to the change of the wafer temperature at high sliding velocity.Experimental results from eight slurry formulations with various abrasize size and content show that in the case of 13-nm abrasives, the dominant tribological mechanism is that of partial lubrication, while in the case of 35-nm abrasives, the dominant tribological mechanism is that of boundary lubrication. COF values of the slurry with surfactant are generally lower that those of the slurry without surfactant.Logarithmic spiral positive pad, whose spiral groove is at a slight angle to the pad rotation direction, shows the highest average COF. The radial pad results in the smallest average COF. For all types of the grooved pads investigated, CMP is mechanically limited at low pV, and chemically limited at high pV. Non-Prestonian behavior is thought to be due to variations of COF and substrate temperatures.Dual Emission UV Light Enhanced Fluorescence results indicate that during polishing the wafer is tilted towards the center of the pad and that the extent of wafer tilt is a strong function of diamond disc pressure. Increasing the oscillation frequency of the diamond disc or the rotation rate decreases slurry film thickness. Slurry film thickness increases with the slurry flow rate. Also slurry film thickness strongly depends on diamond disc design.
Degree ProgramChemical Engineering