Development of “Soft” Cleaning Chemistries for Enhanced STI Post-CMP Cleaning

Abstract

Due to the emergence of sub-10 nm technologies, limiting device defects has become of utmost importance in maintaining high performing integrated circuits. Shallow trench isolation (STI) chemical mechanical planarization (CMP) uses a synergistic balance of ceria (CeO 2 ) nanoparticles and functional chemistry to modulate critical surface adsorption interactions necessary to remove excess topography. Literature has identified that the surface redox state of the CeO 2 nanoparticle (Ce 3+/4+ ) is a critical factor in controlling CMP performance. More specifically, controlling the oxygen vacancies has been shown to directly impact removal rate (Ce 3+ state), as well as post-CMP cleaning efficiency (Ce 4+ state). Current post-CMP cleaning processes utilize oxidation/reduction reactions, etching, encapsulation, and ultrasonic methods as a means of undercutting the particle from the substrate surface. These methods have proven to be highly effective at particle removal but have shown to potentially induce further defectivity due to the aggressive cleaning conditions. This work aims to develop a post-CMP cleaning chemistry that enhances particle removal via soft encapsulation of undesirable nanoparticle/slurry chemistry non-covalent complexes on the wafer surface. By employing a polyelectrolyte/surfactant (PES) matrix, a surface adsorbing/charge-flipping encapsulation process is proposed as a vehicle to remove slurry contamination at low friction. Investigation into the synergy between slurry chemistry and PES cleaning chemistry, with respect to defect aggregate formation, was correlated to the cleaning efficiency as well as friction. Techniques were developed to study real-time properties such as the defect aggregate formation mechanism via in-line dynamic light scattering (DLS), coefficient of friction (COF), particle removal, and scratch defectivity to draw a correlation between cleaning chemistry structure and post-CMP performance.