Understanding nonlinear dissolution rates in photoresists

Abstract

This work focuses on understanding the dissolution phenomenon of surface inhibition, which is observed often in the development of novolac based resists. Many theories have been offered to explain this phenomenon, including a concentration gradient of resist components, oxidation of the surface, formation of a gel layer, and surface roughness effects. This work focuses on theories that propose a concentration gradient in resist components. A technique has been established to separate and analyze individual layers of thin films, and the concentration gradient in many resist components (residual solvent, low molecular weight chains, photoactive compound, density) has been compared to the observed dissolution rate. The results indicate that no significant concentration gradients exist in a 1µm novolac film, and that these hypotheses are inadequate to explain surface inhibition. Several other theories are explored, including oxidation of the surface, surface roughness effects, etc. The critical ionization dissolution model may offer an explanation for why surface inhibition is observed in novolac, but typically not in poly(p-hydroxystyrene).

Characterization of Reactive Dissolution and Swelling of Polymer Films Using a Quartz Crystal Microbalance and Visible and Infrared Reflectance Spectroscopy

“…We presume that its resistance behavior during bulk dissolution stems from factors unrelated to gel formation. While surface roughness has been shown to influence QCM response in liquids, 59 the reported change in roughness of novolac films during development (increasing from ∼0.5 to 2.5 nm) 60 is inconsistent with this being a significant contributor. Possible causes observed resistance shift include changes in surface hydrophobicity, or in the structure of the surface-adjacent water layer, as the polymer undergoes ionization and dissolution.…”

The Multifunctional Role of Base Quenchers in Chemically Amplified Photoresists

“…D. Induction Effects during Resist Dissolution. The dissolution of photoresist is a nonlinear process such that the initial dissolution rate of the resist during an induction period may be much slower than the dissolution rate through the bulk of the film. , The dissolution rate measured by a dissolution rate monitor at half remaining thickness is therefore not necessarily equal to the average dissolution rate of the film as could be calculated by dividing the dissolved thickness by the development time. The duration of the induction period during dissolution was not constant and decreased with increasing exposure dose.…”

Mesoscale Monte Carlo Simulation of Photoresist Processing

“…The goal of a mesoscale film creation simulation is to create a lattice model of a photoresist film that has the desired dimensions, contains appropriate concentrations of the various resist components, and mimics the resist film structurally. While many detailed experiments have been performed to measure concentration gradients generated during film formation, 13,17,18,24 none have detected a concentration gradient of any species within thin resist films prior to exposure. In the simulation, concentration fluctuations exist locally ͑i.e., on small length scales͒ due to stochastic variations arising from the discrete nature of the lattice model.…”

“…It is also necessary to know the copolymer composition. Concentrations of other photoresist components, most notably the PAG loading and presence of residual casting solvent, 17,24 must also be quantified. Knowledge of the chemical composition of the photoresist enables calculation of the appropriate population of each component in the lattice.…”

“…The simulation is based on detailed models of resist function that have been derived from extensive experimental studies, and the simulation inputs are fundamental quantities that can be measured independently of lithographic performance, such as polymer molecular weight, copolymer ratio, photoacid generator ͑PAG͒ loading, activation energy of the polymer deprotection reaction, temperature and duration of bake steps, developer concentration, etc. This simulation framework has now yielded results that are qualitatively correct for all major resist processing steps, namely, spincoat and postapply bake ͑PAB͒, 13,17,18 ultraviolet exposure, 19 postexposure bake ͑PEB͒, 5,20-22 dissolution, 9,15,23 and drying. The reader is referred to the cited literature for a more detailed discussion of the purpose and general function of each of these common processing steps.…”