Modelling of microelectronic processes and materials

Abstract

Organic electronics promises the creation of electronic components on flexible materials at low temperatures, by fast techniques and more environmentally friendly processes. The research followed two directions. The first part focused on the manufacturing technique nanoimprint lithography (NIL). A comprehensive review was undertaken and process capabilities were compared for trends. It was seen that small feature sizes (< 50 nm) have not been replicated over areas greater than 4 mm2, while aspect ratios greater than 10 have not been achieved. A questionnaire addressing market opportunities suggested NIL is likely to compete for the production of devices that currently use electron beam lithography and laser writing processes that are seeking to change their business strategy from a differentiation base to a cost reduction. NIL must also prove to customers that it is an economical investment. However, improvements in stamp creation, analysis techniques and overlay alignment need to be addressed for a larger share of the microfabrication market. It was apparent that physical limits exist to which imprints can be produced and an analytical model could predict these. A model was created to describe the de-embossing step and to explore how the various material properties and process variables interact. It showed a very strong dependence on the achievable aspect ratio on the pattern area ratio and the interfacial shear stress; that de-embossing using fluorinated coatings and current standard polymers is unlikely to fail for post radii on the order of 100 nm due to adhesion and that large area ratios and aspect ratios are more easily achieved by maintaining the polymer/stamp Young’s moduli ratio (RE) in the range 0.003 to 5.
The second part of the research looked at the formation of crescent singularities in thin sheet materials, which affects the production of polymer electronic based devices produced by the sponsoring company. The author compared an analytical model by Cerda and Mahadevan for the formation of developable cones (d-cones) to a finite element (FE) model and showed that explicit elements could mimic the formation of a d-cone. Different elements were analysed for their suitability and the Belytschko-Lin-Tsay (BT) element was chosen based on its speed, robustness and similarity to the analytical results. An adapted three-point bend test set-up was conceived that would enable specific attributes to be independently varied, to understand their effect on d-cone formation in thin sheets. Digital image correlation (DIC) was used to calculate the displacements and strains. The same set-up was modelled using an FE model with the chosen BT element. The DIC results showed a variation in strain with plunger displacement before the visual appearance of a developable cone and that it occurred between 0.1 and 0.4 % in-plane strain. The FE data showed a similar trend to the DIC results, showing a change in strain once a d-cone began to form. Improvements and suggestions were then made advising how to make the DIC and FE models more accurate.

More information

Identifiers

ORCID for S.M. Spearing: orcid.org/0000-0002-3059-2014

Catalogue record

Export record