Browse Prior Art Database

Extremely Accurate Two Dimensional Length Measuring Standard

IP.com Disclosure Number: IPCOM000110646D
Original Publication Date: 1992-Dec-01
Included in the Prior Art Database: 2005-Mar-25
Document File: 2 page(s) / 80K

Publishing Venue

IBM

Related People

Wagner, D: AUTHOR

Abstract

This article describes an extremely accurate two-dimensional length measuring standard which is superior to prior-art measuring means.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 52% of the total text.

Extremely Accurate Two Dimensional Length Measuring Standard

       This article describes an extremely accurate
two-dimensional length measuring standard which is superior to
prior-art measuring means.

      As the integration density of semiconductor circuits in the
submicron range increases, there are stricter overlay requirements.
The measuring devices developed for checking the overlay of masks and
wafers necessitate a two-dimensional length standard for calibration
and stability control.

      Currently, it is assumed that semiconductor lithography for
chips of up to 256 MBit will use light-optical means to measure the
critical layers.  The problems connected with the 4 - 5 x masks to be
employed for such chips are relatively few.  In contrast to this,
X-ray lithography of chips of 1 GBit will require 1 x masks for which
a coordinate measuring accuracy of 3 s = 8 nm is indispensible.  This
accuracy also applies to systematic error detection.

      For two-dimensional length standards, the optimum measuring
tolerance presently is 2 sM = 40 nm for a measuring length of 150 mm.
X-ray lithography to be used for future applications requires much
higher standards.

      An analysis of the measuring tolerance of 2 sM = 40 nm has
shown that some 30 nm of this value are attributable to topographical
variations of the interferometer mirror.  The next item of 8 nm is
attributable to mask bending tolerances.  These tolerances are
noticeably reduced by the new standard.

      Measuring the mirror planarity is difficult in so far as a
single measuring step provides only lump sum values with regard to
planarity and mask defects.  This problem cannot be resolved by
rotating the mask by 180o about an axis perpendicular to the mask
plane, as in this case only another mirror area would be measured.
An alternative would be a second measuring step...