(524c) Block Copolymer Directed Self-Assembly Using Chemoepitaxial Guiding Underlayers with Topography | AIChE

(524c) Block Copolymer Directed Self-Assembly Using Chemoepitaxial Guiding Underlayers with Topography

Authors 

Ludovice, P. J. - Presenter, Georgia Institute of Technology
Nation, B., Georgia Institute of Technology
Henderson, C. L., Georgia Institute of Technology
Simulations have been used extensively to analyze the behavior of block copolymers (BCPs) in directed self assembly (DSA) to produce nanoscale features in polymer photoresists. Typically these simulations have targeted patterns of specific chemical makeup (chemoepitaxy), or three dimensional lines (graphoepitaxy). However many applications of these patterns contain both chemoepitaxial and graphoepitaxial features. Recently, hybrid processes for BCP DSA employing both chemoepitaxy and graphoepitaxy have been developed that use underlayer patterns to induce DSA of BCPs. However, little is understood about the correlated effects of both chemoepitaxy and graphoepitaxy in DSA We have carried out coarse-grained molecular dynamics simulations of BCPs using a variety of hybrid underlayer patterns to characterize DSA as a function of these geometries. Simulations of various hybrid DSA approaches currently in use were simulated. These included the Liu-Nealy (LiNE), Surface Modification for Advanced Resolution Technology (SMART), and coordinated line epitaxy (COOL) process flows currently being used in BCP DSA.

It was found that underlayers with vertical sidewalls behave in manners similar to purely graphoepitaxial guiding underlayers, while underlayers with sloped sidewalls behave in a manner similar to chemoepitaxial guiding underlayers. With vertical sidewalls, it is found that larger topographic step heights decrease pattern defectivity, though with diminishing returns. The width of the trench should be an integer multiple of the natural repeat distance of the block copolymer, though the width of the trench can have approximately 10% variation before significantly affecting the defectivity. These results suggest that the width of the guiding regions should be spread out in order to reduce defectivity. It was found that the chemical preference of both the top of the mesa and bottom of the trench have little effect on the defectivity so long as the preference is sufficienty neutral to prevent the formation of horizontal lamellae. A variety of sloped sidewall geometies were explored. Sidewalls that undercut the mesa were found to increase defectivity. It was found that certain trapezoidal shaped topographies can reduce defectivity, especially when the top of the mesa, as well as the width of the sidewalls, were approximately the width of one lamellae. In these cases we observed an optimal height of the topography. The optimal height appears to be the height that makes the effective width of the sidewall the same width as a lamellae. Additionally, it was found that a slightly preferential bottom of the trench and a very preferential top of the mesa helped further lower defectivity. An additional sloped sidewall case, where the topographic feature was triangular in shape was explored. It was found that this triangular topographic feature yields far lower defectivity than a similarly sized rectangular topographic feature.