Simulation of structure profiles in optical lithography of thick DNQ-novolak-based photoresists

Abstract

DNQ-Novolak based photoresists are applied in many fields of microstructure technology. A common feature of such resist structures (>lOjim) described in the literature is a concave profile with the narrowest part at approximately 2/3 of the resist height. By newly including effects of residual solvent concentration, the simulation program accurately describes characteristic resist profiles for AZ4562 between 10 and iøøm resist thickness. The program calculates the vertical distribution of the residual solvent concentration by modeling the evaporation during prebake as a three step process: diffusion of solvent within the resist film to the surface, a transfer to the ambient air described by a phase equilibrium and a transport away from the surface described by convection. Required parameters of the model have been obtained by fitting measured evaporation rates as function of time with the prebake temperature as a parameter to calculated ones. Resist exposure is simulated in vertical and one lateral dimension using the well known Dill's theory with ABC-parameters for DNQ resists. The development process is described by percolation theory where the residual solvent concentration increases the percolation parameter and therefore the local development rate. This paper describes details of the mathematical models used, demonstrates good agreement between experimentally observed profiles and simulated ones and shows that the influence of process parameters such as prebake temperature or time on resist profiles can be accurately simulated.

Moving mask UV lithography for three-dimensional structuring

“…Second, the influence of transport of material by convection on the dissolution rate should be verified. Simulation results may be improved by considering problems specific to thick-photoresist lithography, such as light propagation in thick photoresist [18], the influence of the solvent concentration [20], variation in photoresist properties, such as the refractive index and absorption coefficient, along the thickness direction [21], and reflection from the substrate [22]. Improvement and investigation of these points will be the subject of future work.…”

“…A similar dependence is observed for prebake temperature; for a given pre-bake time, measured depth increases with decreasing pre-bake temperature (figure 4(b)). These photoresist performances are dominated by solvent content, which is varied by changing the pre-bake time and temperature [20].…”

Development and characterization of ultra high aspect ratio microstructures made by ultra deep X-ray lithography

“…To predict the average diluent concentration as a function of the processing parameters, a mathematical model was developed in this work. Chung et al (2000) modelled and Engelke et al (2007) experimentally showed that the residual diluent concentration is very non-uniform over the depth of a thick resist layer. This is because of the diffusive mechanism of molecular transportation in a very viscous prepolymer medium (up to 80,000 cSt, as notified by MicroChem (2007)).…”

A three-dimensional microstructuring technique exploiting the positive photoresist property

“…These reports include the characterization of the DNQ photoresist property [33][34][35][36][37], photolithography simulation models [38][39][40][41][42][43][44][45][46][47][48], and so on. In the past few years, a growing number of research groups reported the DNQ photoresist as a thick photoresist for MEMS applications [49][50][51][52][53][54][55][56][57][58][59]. However, compared to the previously reported studies of the DNQ photoresist with thickness of less than 25 μm, the number of publications on 3D thick-photoresist microstructures using more than 25 μm thickness and its detail experimental study is very limited [20,60].…”