As discussed in Chapter 1, the mask pattern is transferred to the wafer by means of optical projection onto photosensitive materialsâphotoresists. These materials clearly play a critical role in lithography. The chemistries of the most common classes of photoresists are outlined in this chapter. However, the emphasis is on operational performance, i.e., the ways in which resists behave, as observable by the practicing lithographer (in contrast to the resist chemist).
3.1 Positive and negative resists
Resists are broadly classified as positive or negative. Unexposed positive resists normally have very low solubility in developer and become soluble by exposure to light. Negative resists behave in the opposite manner; unexposed negative resists are soluble in developer, and lose their solubility upon exposure to light. Negative resists were used predominately prior to the advent of wafer steppers. Resists most widely used in the early years of the semiconductor industry, such as Kodak's KTFR resist, were based upon the photo-induced cross-linking of low-molecular-weight cyclized poly-isoprene (Fig. 3.1). Azide additives were used to improve their sensitivity to light and to facilitate cross-linking (Fig. 3.2). Cross-linking reduced the material's solubility in organic solvents such as xylene. After exposure, an organic solvent was used to develop the resist, by removing the low-molecular-weight unexposed resist, and leaving behind cross-linked, high-molecular weight material.
These early negative resists had a number of deficiencies. During development, the cross-linked material would absorb some of the organic solvent used as the developer, causing swelling. Because of the swelling, closely spaced geometries would come into contact during develop, and would sometimes remain stuck together, resulting in a patterning defect. This swelling limited the thickness and resolution potential of these resists.
Online access to SPIE eBooks is limited to subscribing institutions.