IMMERSION DEVELOP RINSE WATER INJECTION PROCESS TO ELIMINATE POSITIVE PHOTORESIST SCUMMING
Original Publication Date: 1988-Oct-01
Included in the Prior Art Database: 2001-Oct-29
Publishing Venue
Motorola
Related People
Abstract
A simple modification to a positive resist immersion develop system's dump and fill rinse tank is described which alleviates resist scumming problems. The resist scumming is attributed to the precipitation of dissolved resist and its subsequent adhesion to the wafer surface. The resist precipitation is the result of the rapid change in pH experienced by the exposed positive resist during its transfer from the highly alkaline develop tank to the deionized water dump rinse tank.
MOlOROLA Technical Developments Volume 8 October 1988
IMMERSION DEVELOP RINSE WATER INJECTION PROCESS TO ELIMINATE POSITIVE PHOTORESIST SCUMMING
by Paul M. Winebarger
ABSTRACT
A simple modification to a positive resist immersion develop system's dump and fill rinse tank is described which alleviates resist scumming problems. The resist scumming is attributed to the precipitation of dissolved resist and its subsequent adhesion to the wafer surface. The resist precipitation is the result of the rapid change in pH experienced by the exposed positive resist during its transfer from the highly alkaline develop tank to the deionized water dump rinse tank.
DISCUSSION
The introduction of immersion develop systems in MOS 2 brought on sporadic instances of resist scum- ming. The scumming areas would range from microscopic to 50% of the wafer surface. Under microscope in- spection the scumming appeared as thin layers of resist deposited both in developed out areas as well as on top of patterned lines.
Different surface layers showed varying degrees of suceptibility to resist scumming. CuSiAl was most susceptible followed by polysilicon, nitride, and oxide layers. Fully loaded wafer carriers were found to be more prone to scumming than partially loaded carriers with the wafers spread out.
The severity of the problem was increased by the fact that the scumming is difficult to see visually, even with the aid of a bright light. Because the scumming tended to lie in isolated splotches, finding the problem by microscope inspection was very unreliable; however, after pattern etching, the problem would be very evi- dent as unetched areas on the wafer.
The scumming was not due to incomplete develop since decreasing exposure and/or develop time did not induce the problem. This fact led to investigation of the rinse process.
THEORY
The resist scumming was found to be easily induced by acidifying the rinse water. Addition of bubbling CO* or any acid solution to the rinse tank would greatly magnify the scumming problem.
Basic photoresist chemistry indicates that this result should not be surprising. Exposure of positive photoresist results in the photolysis of hydrophobic diazoquinone, which results in the formation of an organic carboxylic acid.'
Diazooxide lndene Carboxylic Acid
0 Motorola, Inc. 1988 9
m MOlYlROLA Technical Developments Volume a October 1998
Positive photoresist developer solutions are merely dilute ionic or organic alkaline solutions which preferen- tially dissolve the resulting carboxylic acid. The acidification of the rinse water used after d...