Paper PS+EM-TuM3
Etching of Sub-10 nm Half-pitch High Chi Block Copolymers for Directed Self-Assembly (DSA) Application

Tuesday, October 22, 2019, 8:40 am, Room B131

As the semiconductor industry approaches the smaller technologic nodes such as the sub-10 nm, conventional lithography technologies have reached their limit. Among the different approaches investigated to continue pattern scaling, Directed Self-Assembly (DSA) of Block Copolymers (BCP) is one of the most promising due to its simplicity, low manufacturing cost and capability to design high density cylindrical or line/space patterns. For the last few years, PS-b-PMMA has been the most used BCP for this application. However, since the minimum feature size for the PS-b-PMMA system is limited to ~13 nm due to its low interaction parameter (also known as “chi”), new systems have been developed to achieve higher resolution by improving its microphase separation strength, thus obtaining the so called “high chi” BCPs.

In this paper, the high chi BCP system investigated is a modified PS-b-PMMA that presents a pitch of 18 nm, which will be referred to as “L18”. One critical step for its integration is the PMMA removal selectively to the PS. The results obtained with both dry and wet etching of the L18 BCP are presented, highlighting the challenges encountered due to its smaller dimensions.

A wet PMMA removal process based on UV exposure followed by IPA rinse will be presented. An Ar/O₂ dry etch step for the brush layer opening was developed and the transfer of the line/space pattern into the SiO₂ and Si underlayers was demonstrated for the L18 BCP. However, this wet development is expected to cause pattern collapse when the BCP will be guided by chemoepitaxy due to capillary forces.

Therefore a complete PMMA removal by dry etching alone was also investigated. Different etching chemistries based on CH₃F/Ar/SO₂, CH₄/N₂ or cyclic CO + CO/H₂ were applied to the modified PS-b-PMMA BCPs with a 30 nm pitch and with an 18nm pitch (L18). For the 30 nm-pitch BCP, all three plasma chemistries allow the complete PMMA and brush layer removal by dry etching and the subsequent pattern transfer into the SiO₂ layer. In contrary, for the L18 BCP, the CH₃F/Ar/SO₂ plasma does not present enough PS budget for pattern transfer due to its low selectivity and the small thickness of the BCP. For the more passivating chemistries such as CH₄/N₂ and cyclic CO + CO/H₂, which have higher selectivity, we observe the formation of bridges that prevent complete pattern transfer. We used ex-situ X-ray Photoelectron Spectroscopy (XPS) to investigate the origin of these bridges and to understand the etching mechanisms present.