PhD Dissertation Defense, Min Bai, Monday 3:00pm

Min Bai mbai at
Fri Apr 25 16:59:12 PDT 2003


Speaker:	Min Bai
Advisor:  	Professor R.F.W. Pease

Title:		Insulator Charging in Electron Beam Lithography

Date:		April 28, 2003
Time:		3:00 P.M.
Place:		Center for Integrated Systems Extension (CIS-X) Auditorium

Electron beam lithography (EBL) of masks and wafers results in
charging of the insulating resist films; the charging can deflect the
incoming electron beam, leading to pattern placement errors.  This
error has been believed to be increasingly significant as the design
rules for integrated circuit manufacturing progress to 100 nm and

Models have been developed to understand the mechanism by which
charging occurs during electron beam lithography and to determine
beam placement errors.  First a one-dimensional analytic model was
developed to predict surface potential on electron beam irradiated
resist films.  One sensitive parameter involved in this model was the
electron beam induced conductivity (EBIC).  Secondly an image charge
method was used to predict the beam deflection caused by a specified
voltage distribution on the sample surface.  The results suggested
that under the usual conditions of EBL for mask manufacture the beam
displacement should be negligible.  This is contrary to several
previous reports.

Three independent sets of experiments were employed to quantify the
charging effects.  First a Kelvin probe electrometer was used to
monitor surface potential in resist films immediately following
exposure with electrons.  The results corroborated an earlier
(disputed) observation that the resist could charge either positively
or negatively depending on the film thickness and/or beam energy.
Secondly the pattern displacement induced by charging was directly
measured in an electron beam lithography tool based on a Leica S440
SEM.  The observed pattern positioning error in PMMA resists due to
charging was significantly smaller (< 20 nm) than had been previously
reported.  Thirdly we designed and built a simple secondary electron
collector to monitor the change in specimen surface potential during
electron beam exposure.  These in situ measurements show that
charging doesn't become a significant problem until the resist
thickness is comparable, or larger than the maximum beam penetration

In a fourth set of experiments we measured EBIC in PMMA and SiO2 thin
films.  The SiO2 is almost 100 times more conductive than PMMA under
the same exposure condition.  However, the limited EBIC in PMMA is
yet sufficient to limit the charging of this material.  This is
consistent with the charging experimental results.

>From the results of our modeling and experiment we conclude that
under typical mask making conditions (e.g. 400 nm of resist on
conducting substrate, exposed with > 10 keV electrons), resist
charging is much less serious in EBL than had been previously
reported.  We believe that those earlier beam displacement results
may have been due to other factors.  However, in other scanning
electron beam applications such as inspection, the charging problem
could still be significant because the low energy secondary electrons
used as the signal are much more sensitive to spurious fields and the
samples may include insulating films thicker than the primary
electron range.

More information about the labmembers mailing list