Simulation of UV nanoimprint lithography on rigid and flexible substrates

Date

2016-12

Authors

Jain, Akhilesh

Journal Title

Journal ISSN

Volume Title

Publisher

Abstract

Nanoimprint lithography (NIL) is a low cost, high throughput process used to replicate sub-20 nm feature from a patterned template to a rigid or flexible substrate. Various configurations for NIL are analyzed and classified based on type of template and substrate. The steps involved in pattern transfer using roller template based NIL are identified and models to study these steps are proposed. Important process parameters such as maximum web speed possible, required UV intensity, minimum droplet size and pitch and required force on the roller are calculated. The advantages, disadvantages and optimal process window for the different configurations are identified. Droplet spreading is simulated in NIL with rigid substrates in order to study the effect of droplet size, droplet placement error, gas diffusion and template pattern on throughput and defectivity. Square arrangement is found to be the optimum arrangement for achieving minimum throughput. Large droplet-free regions on the substrate edge and error in droplet placement error have significant impact on the throughput. A fluid flow model with average flow permeability is presented to account for flow in the template patterns. Optimum droplet dispensing for multi-patterned templates is achieved by distributing droplet volume according to local filling requirements. Non-fill defects in NIL are classified into pocket, edge and channel defects. A model to predict the size of non-fill defects based on imprint time and droplet size is presented. Defect characterization is presented for various pattern-types. A model is presented to determine the time required for the encapsulated gas to diffuse into the resist. The coupled fluid-structure interaction in NIL with flexible substrate is studied by simulating the web deformation as the droplet spreads on the substrate. It is found that the flexible substrate can be modeled as a membrane due to the lack of rigidity. RLT variation reduces as the number of droplets or the web tension increases. For the magnitude of RLT variation, thinner residual layers require higher web tension. The position of the template on the substrate is important and template positioned at the corner of the substrate is found to provide the least RLT variation.

Description

LCSH Subject Headings

Citation