
Proceedings Paper
Effect of mask erosion on process latitudes in bilayer lithographyFormat | Member Price | Non-Member Price |
---|---|---|
$17.00 | $21.00 |
Paper Abstract
Mask erosion is not significant in tn-layer lithography where the etching mask is Si02; however,
mask erosion may significantly affect etching profiles and process latitudes during the pattern
transfer step in bi-layer lithography. We discuss the effect of selectivity, wall angle, layer thickness,
and etching anisotropy on line-width loss in bilayer lithography. Predictions from the kinetic
theory of bombardment induced interface evolution are compared to experimental pattern transfer
results for electron beam exposed poly(trimethylsilylmethyl methacrylate-co-m,pchloromethylstyrene).
The interface evolution model describes bombardment induced etching by
an axisymmetric angular distribution of energetic particles where the yield per incident particle is a
function of energy and angle relative to the surface normal. Kinetic theory is used to relate the
energy and angular distributions incident on the substrate to the pressure, the voltage drop across
the plasma sheath, the sheath thickness, and the cross sections describing the collision processes.
The rate controlling step for the organosilicon polymer is assumed to be Si02 sputtering as
predicted by the steady-state model of organosilicon polymer etching. The 5i02 sputtering yield is
assumed to be a separable function of.energy and angle with a sputtering threshold of 50 eV. The
etching yield of the organic planarizing layer is assumed to be proportional to bombardment energy
and independent of angle. The angle dependent yields in the upper layer result in a facet that propagates
down from the upper edge of the resist profile. This facet has no effect on process latitudes
if less then F30% of the upper layer is consumed during the pattern transfer step; however, faceting
results in rapid line-width loss if more than 30% of the upper layer is consumed. Under
nearly optimized exposure and etching conditions, the experimental line-width loss during pattern
transfer was less than 20% of the nominal width for features with nominal dimensions down to 0.25
zm.
Paper Details
Date Published: 1 June 1990
PDF: 16 pages
Proc. SPIE 1262, Advances in Resist Technology and Processing VII, (1 June 1990); doi: 10.1117/12.20093
Published in SPIE Proceedings Vol. 1262:
Advances in Resist Technology and Processing VII
Michael P. C. Watts, Editor(s)
PDF: 16 pages
Proc. SPIE 1262, Advances in Resist Technology and Processing VII, (1 June 1990); doi: 10.1117/12.20093
Show Author Affiliations
Charles W. Jurgensen, AT&T Bell Labs. (United States)
Anthony E. Novembre, AT&T Bell Labs. (United States)
Anthony E. Novembre, AT&T Bell Labs. (United States)
Eric S. G. Shaqfeh, AT&T Bell Labs. (United States)
Published in SPIE Proceedings Vol. 1262:
Advances in Resist Technology and Processing VII
Michael P. C. Watts, Editor(s)
© SPIE. Terms of Use
