The use of evolutionary resist processing techniques designed to improve profiles and extend process latitude, is of general
interest to lithographers. The more commonly known methods (HARD, LENOS) utilise the phenomenon of alkali catalysed
crosslinking, to create additional surface inhibition during development. Such techniques whilst undoubtedly useful, are by
their very nature limited to improving only the very top of the resist profile. Additionally under certain process conditions the
formation of a top hat profile can be observed, this is particularly noticeable on smaller features. The use of an interrupted
development cycle, or of multiple interrupted development cycles (PRISM) has also been reported to exert a beneficial effect
on profiles. However our own tests, utilising a single interrupted development cycle, are unable to support such claims, and
have in fact demonstrated degraded profiles, in comparison with a standard double puddle process.
We report here on an alternative technique, the so called intermediate development bake or 1DB, whereby a short low
temperature baking step is introduced midway during the development cycle. The distributed nature of this technique
discourages the formation of top hat profiles, whilst improving the upper part of the developed profile.
Previously reported work concentrated on the use of infra red baking. The present paper investigates this further, whilst
additionally covering the use of warm air and warm water treatments. The use of warm water in particular is potentially very
attractive, and we demonstrate its effectiveness on a range of i-line specific, g-line and multiwavelength resists, exposed
using i-line. A particularly close study has been made of Olin's HiPR 6512 resist using both metal ion and metal ion free
Based on a number of observations including the use of FTIR and XPS, we advance a mechanism to explain our results.