Browse Prior Art Database

Fast, Accurate Proximity Correction Algorithm for 0.25 Mm Electron- Beam Lithography

IP.com Disclosure Number: IPCOM000120089D
Original Publication Date: 1991-Mar-01
Included in the Prior Art Database: 2005-Apr-02
Document File: 2 page(s) / 75K

Publishing Venue

IBM

Related People

Ashton, CJ: AUTHOR [+3]

Abstract

A fast, highly accurate proximity correction algorithm is needed in order to successfully expose and develop 0.25 mm and smaller structures using electron-beam lithography. A fast algorithm for long range back-scattered electron proximity correction coupled with an approximate algorithm for forward scattered correction has previously been disclosed. However, this algorithm, hereafter referred to as DOSEB, is limited in accuracy for 0.25 mm lithography.

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

Fast, Accurate Proximity Correction Algorithm for 0.25 Mm Electron-
Beam Lithography

      A fast, highly accurate proximity correction algorithm is
needed in order to successfully expose and develop 0.25 mm and
smaller structures using electron-beam lithography.  A fast algorithm
for long range back-scattered electron proximity correction coupled
with an approximate algorithm for forward scattered correction has
previously been disclosed.  However, this algorithm, hereafter
referred to as DOSEB, is limited in accuracy for 0.25 mm lithography.
A modified self-consistent proximity correction algorithm (1), which
uses the Gauss-Seidel method for solving the simultaneous equations,
hereafter referred to as DOSEF, gives accurate results, but it is too
slow for large data sets, especially at the higher electron-beam
voltages (where the range of the back-scattered electron exposure is
higher) which will be used for X-ray mask making and by advanced
electron-beam lithography systems under development. Finally, it has
recently been shown that multiple Gaussians are required to describe
the radial absorbed energy distribution in the resist for accurate
proximity correction at 0.25 mm and below (2,3).  In addition,
multiple Gaussians are required to describe the energy distributions
for structures with high atomic number substrates and at higher beam
voltages on X-ray mask membranes (2,3).

      The proposed proximity correction method models the exposure as
a sum of N Gaussians, where N may range from 2 to 10 or more.
One of the Gaussians has a long range and represents the exposure
from back- scattered electrons.  The remaining Gaussians have much
shorter ranges, and represent the exposure from forward scattered and
high- angle-back-scattered electrons.

      The first step i...