This Small Business Innovation Research Phase I project is focused on the development of a copper chemical mechanical planarization (CMP) slurry containing chemically impregnated nanoparticles. The CMP process consists of a polishing pad and an aqueous slurry containing chemicals and abrasive nanoparticles. The wafer surface to be planarized is rotated against the polishing pad which is continuously supplied with slurry. The synergistic combination of mechanical abrasion and chemical etching creates high material removal rates (MRR) and rapid planarization. Currently, copper CMP slurry solutions require a complexing agent, such as glycine, to maintain an adequate MRR, which at the same time, also causes static etching and dissolution (i.e. corrosion) of the wafer surface which is difficult to control. If the exposure of the complexing agent to the copper surface could be controlled so that only the abraded areas were affected, then static etching and dissolution could be reduced or eliminated. Therefore, the focus of this proposal is to significantly improve copper CMP slurry technology by combining the complexing agent into the abrasive nanoparticle, so that mechanical and chemical actions of the slurry are localized and simultaneous, thus improving processing efficiency and tolerances.
The broader impact/commercial potential of this project includes advancing the state-of-the-art in CMP slurry formulation and opening the door for functionalized nanoparticles in the CMP slurry market. Specific impacts of the proposed project on copper CMP slurry technology include: increasing planarization efficiency (ratio of step height reduction and removed layer thickness) by decreasing static etching, eliminating a majority of the chemical additives that are currently required in copper CMP slurries, and reducing particle agglomeration by improving the dispersion stability. These advantages will ultimately lead to a reduction in the cost of CMP processing while also improving processing tolerances. Additionally, an enormous opportunity exists to exploit functionalized nanoparticles to improve CMP slurries. Successful implementation of this product will most certainly lead to derivative technologies for other CMP processes.
An important step in semiconductor manufacturing is chemical mechanical polishing (CMP), which uses an aqueous based slurry containing abrasives and chemical components in conjunction with a polishing pad to remove material and to planarize the wafer before patterning additional circuit elements. As minimum device dimensions continue to decrease, next generation CMP slurries must achieve planarization on increasingly higher aspect ratio structures. Current CMP slurries have a limited planarization efficiency (PE) (ratio of step height removed to thickness removed) which introduces restrictions on the design rules that specify the layer thickness and other process parameters during semiconductor design. Without an improvement in PE, CMP is going to become an increasingly expensive and limiting step in the semiconductor manufacturing process. Current commercial CMP slurries consist of homogenous mixtures of abrasives (typically metal oxide nanoparticles) and chemical components. While incremental improvements to CMP performance using this paradigm continue to be obtained, new approaches to CMP slurry formulations will be needed to meet future device requirements. General Engineering & Research, L.L.C. is developing a copper CMP slurry containing nano-sized contact release capsules (nano-CRC). 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 complexing agent (glycine) is released in an area localized to where it is needed. During the Phase I research, nano-CRC particles made from porous colloidal silica abrasives were impregnated with glycine, and then coated with a polymer to encapsulate the glycine. Upon shearing contact with the wafer surface during CMP, the polymer coating is torn away and the glycine is released. The following development goals for a nano-CRC based slurry were met: 1) Chemical payload encapsulated in the nanoparticle pores 2) Stable polymer coating that prevented leakage of payload 3) Polymer coating soft enough to be torn away by the shearing forces during CMP to release the payload 4) Well dispersed particles free from agglomeration 5) Cost effective manufacturing method for competitive CMP slurry pricing To date, no CMP slurry technologies have attempted to combine the mechanical and chemical components of the slurry into a single entity. Initial results have yielded simultaneously high PE and high material removal rate (MRR), which will improve planarity and decrease processing costs. Table 1 summarizes the copper CMP results of the nano-CRC slurry compared to benchmark commercial CMP slurries. Commercial slurries were diluted to achieve approximately the same abrasion loading as the nano-CRC (4wt%). The RL3200 w/10X dilution is an example of a chemical only slurry which is inexpensive but has only moderate performance (substantial etch rate, low MRR and low PE). The CuS1351 has better performance with a negative etch rate and a MRR of 23 nm/min and a PE of 43%. The best performing nano-CRC prototype was able to match the CuS1351 etch rate and provide a 16 nm/min increase in MRR and a 24% increase in PE. PE is a key factor in determining the surface quality after CMP (dishing) and the development of a class of very high PE CMP slurries will strongly impact future CMP utility. Another consequence of high MRR and PE is their effect on the cost, time and waste stream of the CMP process. The volume of slurry (Vol) needed to planarize a 200nm step height (DS) was calculated according to the following equation: Vol = (F ? DS)/ (MRR ? PE) (Table 1). The nano-CRC is predicted to require just 0.3 gal of slurry to achieve planarization compared to 0.8 gal of CuS1351, a 62% reduction in both slurry volume and time to planarize. With further optimization, even greater improvements in the Cu CMP process is anticipated resulting in a product with a very high return on investment for semiconductor manufacturers. The proposed innovation has the potential to advance the state-of-the-art in CMP slurry formulation and enable the continued scale down of next generation microelectronic devices. Specifically, the copper CMP slurry market is currently $340M, and expected to grow to greater than $500M by 2016 due to the increase in the number of Cu CMP steps as devices continue to shrink. With 5% market share, a $4M per year net profit is predicted for the manufacturing company that sells nano-CRC slurry. Further, the nano-CRC technology can be used to impregnate nanoparticles with different chemical payloads, and different base nanoparticles can also be used (i.e. alumina). This indicates the technologies potential to be used as a platform to provide solutions for a variety of industries, both within the realm of CMP where it may be possible to use on other CMP polishing steps, and also for other non-CMP polishing processes, including optical materials, glass polishing, cleaning, etc.
