Simulation and design of planarizing materials and interfacial adhesion studies for step and flash imprint lithography
MetadataShow full item record
Step and flash imprint lithography (SFIL) was developed in 1999 at The University of Texas at Austin as a high resolution, cost-effective alternative to photolithography for nanoscale patterning. Unlike current projection steppers, which are resolution limited by diffraction phenomena, SFIL tools have demonstrated patterning capability down to 20 nm, a resolution currently unattainable using traditional lithographic techniques. The combination of high resolution and low cost of ownership make SFIL a strong candidate for future semiconductor integrated circuit manufacturing. For SFIL to be viable as a high volume process, there are numerous technical issues that need to be resolved. Reverse-tone step and flash imprint lithography (SFIL-R) is a reverse tone variant of SFIL that requires the successful application of a planarizing topcoat over topography through spincoating. Photopolymerizable nonvolatile fluids are ideal topcoat materials because they planarize better than volatile fluids during spincoating and can continue to level after spincoating. Fluid mechanics analyses indicate that complete planarization using capillary force is slow. Therefore, defining the acceptable or critical degree of planarization (DOP[subscript crit]) becomes necessary. Finite difference simulation of the spincoat and post-spin leveling processes was used to determine the planarization time for various topographic and material property combinations. A new material, Si-14, was designed to have ideal planarization characteristics and satisfy SFIL-R processing requirements and was used to validate the models through profilometry and interferometry experiments. During spincoating, minimizing the spin speed generates more planar films, however, this increases the spin time. To rectify this problem, a 2-stage spincoating process -- a first step with high spin speeds to achieve the target thickness quickly and a second step with low spin speeds to improve planarization -- was proposed and experimentally demonstrated. An alternative planarization technique is to generate a reverse-conformal film coating through Marangoni-driven flow. The SFIL process requires the clean separation of a quartz template from a polymer, and the force required to create this separation must be minimized to prevent the generation of defects. Fracture mechanics analyses show that control of the polymer modulus and interfacial fracture energy is the key to minimizing the separation force. Adjusting the crosslinker concentration in the imprint formulation reduces the modulus but has no significant impact on the fracture energy. On the other hand, adding surfactants to the imprint formulation reduces both the modulus and fracture energy. The fracture energy is further decreased by using a nonreactive, liquid surfactant versus a surfactant that reacts with the polymer matrix. Angleresolved X-ray photoelectron spectroscopy (XPS) results indicate that surfactant migration is more effective with a fluorinated surface treatment compared to an untreated quartz or organic surface. However, the fluorinated surface treatment that drives the migration process degrades over multiple imprints. Based on these results, it was concluded that the use of fluorinated surfactants must be accompanied by a surface treatment that is both stable and of a similar energy or polarity to induce migration and to lower the adhesive strength. Mixed-mode fracture affects the separation force, especially if shear stresses are present. Overfilling the templatesubstrate gap causes large amounts of shear stresses during separation; however, this phenomenon can be prevented by controlling the surface energies of the imprint template and substrate.