In electron beam projection systems there are competing requirements for co
lumn length. Shorter lengths lead to reduced space charge interactions whil
e Longer columns allow larger field sizes. Other parameters include the con
vergence angle and beam voltage. We have used a generic 4:1 reduction telec
entric system as a test vehicle to study these tradeoffs. To determine the
lens aberrations we employed a thick lens configuration with constant magne
tic fields; space charge blur was estimated with a Monte Carlo simulation u
sing a thin lens approximation. For a given beam voltage V, held size F, co
nvergence angle alpha, and current I, there is a maximum value of column le
ngth L that satisfies a given resolution requirement. For example, if V=100
kV and 1=20 mu A then L must be less than 30 cm for 100 nm resolution. The
use of a retarding field just upstream of the target can lead to greatly i
mproved performance in terms of resolution, field size, and throughput; e.g
., the minimum resolved feature size is approximately proportional to the r
atio of landing voltage over beam voltage. (C) 1999 American Vacuum Society
. [S0734-211X(99)09406-8].