There is a limit to the minimum feature size that can be printed using current lithographic techniques. For that reason,
engineers often employ various shrink methods in production to reduce the size of features generated by lithography.
One such technique is the application of a shrink assisted film for enhanced resolution (SAFIER). In such a process, a
SAFIER chemical is coated onto a patterned photoresist and baked. During the bake, the resist expands, and hence, the
patterned spaces in the resist shrink. The shrink process, however, does not necessarily occur uniformly across the
wafer, and some critical dimension (CD) non-uniformity can be introduced during this step. This study investigates the
efficacy of using an intentionally biased SAFIER bake temperature profile to compensate for some of the CD nonuniformities
introduced during the SAFIER process. In the baseline case, patterned wafers underwent a standard
SAFIER process using a thermally uniform bake. The bake temperature of the SAFIER bake was then biased to cancel
out some of the shrink induced CD non-uniformity. Wafers processed through the biased temperature SAFIER bake
showed a 20% improvement in post-SAFIER (critical dimension uniformity) CDU. For comparison, a biased post
exposure bake (PEB) temperature was used to create wafers with a non-uniform starting CD distribution designed to
cancel out some of the CD non-uniformity from the SAFIER process. When these wafers were processed through a
uniform temperature SAFIER process, a 54% improvement in post SAFIER CDU was observed over the baseline case.
This paper investigates the feasibility of using an electrostatic chuck (ESC) on a post exposure bake (PEB) plate in the
track to improve the critical dimension uniformity (CDU) for bowed wafers. Although it is more conventional to
consider vacuum chucking during PEB, electrostatic chucking offers some potential advantages, chief among which is
the fact that electrostatic chucking does not require any type of a seal between the wafer and the PEB plate whereas
vacuum chucking does. Such a seal requires contact and therefore has the potential to generate backside particles on the
wafer. Electrostatic chucking therefore has the potential for a cleaner overall process. Three different PEB plates were
tested in the course of this investigation, a non-chucking PEB plate (SRHP), a PEB plate equipped with a vacuum chuck
(VRHP), and a PEB plate equipped with an ESC (eBHP). It was found that CD uniformities were up to 84 percent lower
for bowed wafers that were chucked during PEB relative to wafers that were not chucked. In every case tested, wafers
processed through chucking PEB plates showed lower CDUs than wafers processed through the non-chucking plate.
CDU results were similar between vacuum chucked wafers and electrostatic chucked wafers. Based on the results
presented in this paper, it can be concluded that electrostatic chucking during PEB is a feasible method for controlling
CD uniformities on bowed wafers.