General Engineering & Research, L.L.C. (GE&R) has developed slurries containing nano-sized chemical release capsules (nano-CRC) which provide highly controlled processing and efficient planarization. Instead of the traditional mixture of solid nanoparticles and chemicals, nano-CRC based slurries consist of a core-shell nanoparticle where the mechanical and chemical components are combined into a single entity (Figure 1). When the nanoparticle contacts the wafer surface during CMP, the polymer coating is torn away and the chemical payload is released in an area localized to where it is needed. Various abrasive particles are possible (SiO2, Al2O3, etc.) as well as different chemical payloads. In this research effort, nano-CRC particles made from various sized (20nm – 200nm) porous colloidal silica abrasives were loaded with glycine, and then coated with a polymer to encapsulate the glycine. Slurries containing these nano-CRC particles were tested on the copper CMP process. All wafers were polished in house on a Strasbaugh 6EC CMP tool using IC1000 polishing pad and Kinik Diagrid conditioning stone, with wafer downforce of 3psi, head/platen speed of 60/60 rpm and 150ml/min slurry flowrate. All slurries contained 4wt% silica nanoparticles. The nano-CRC slurry surface removal mechanism consists of a simultaneous mechanical abrasion in combination with a localized release of the chemical payload. These slurries do not require any other stabilizers, surfactants, inhibitors, etc., however, because of this, electrophoretic influences can be significant. The nano-CRC interaction with the surface is influenced by solution chemistry, such as pH and oxidizer concentration, which determines the zeta potential of the capsules and the charge on the wafer surface film. The polymer coating on the nano-CRC is thin enough that the capsules typically retain the colloidal properties of the base particle, i.e. the iso-electric-point (IEP) of the silica based nano-CRC slurries is pH~2. Acidic nano-CRC slurries provide good material removal rates (MRR) of copper films (IEP pH~9), however, alkaline solutions provide little to no removal of the copper surface due to the copper surface and nano-CRC particles being similarly charged and repelling each other. Oxidizer concentration in the slurry also affects copper oxide film formation and may inhibit the copper-glycine complex if used in excess. Figure 2 shows the MRR vs. H2O2 concentration for 40nm silica nanoparticles (no loading or polymer coating) and 40nm nano-CRC (glycine loading and polymer coating) slurries, where the MRR of the nano-CRC slurry is very sensitive to H2O2 concentration. The particle size also influences CMP performance. Figure 3 shows the Cu CMP MRR and planarization efficiency (PE) (ratio of step height removed to thickness removed), for slurries containing bare silica particles and the nano-CRC particles. The PE was calculated using a feature with 50um line spacing and 50um line width. Interestingly, of the bare silica slurries, the 80nm particle size provided the highest MRR, however, no improvement in MRR was observed with the 80nm nano-CRC slurry. The 40nm nano-CRC slurry shows both the highest MRR~200nm/min and PE>90%. Both the porosity and the hardness of the particles, which varies with particle size, may influence the CMP performance. Further optimization, particularly increasing the porosity of the nanoparticles, is expected to provide nano-CRC slurries with even higher MRR and high PE. Figure 1
Read full abstract