Processing and modeling optimization for grayscale lithography

Thomas Dillon; Mathew Zablocki; Janusz Murakowski; Dennis Prather

doi:10.1117/12.773213

Translator Disclaimer

4 April 2008 Processing and modeling optimization for grayscale lithography

Thomas Dillon, Mathew Zablocki, Janusz Murakowski, Dennis Prather

Author Affiliations +

Proceedings Volume 6923, Advances in Resist Materials and Processing Technology XXV; 69233B (2008) https://doi.org/10.1117/12.773213
Event: SPIE Advanced Lithography, 2008, San Jose, California, United States

Abstract

Grayscale lithography is an extension of the conventional binary lithographic process for realization of arbitrary three-dimensional features in photoresist materials, with applications especially in micro-optics fabrication. The grayscale photomask possesses a spatially varying transmission that modulates the exposure dose received in the photoresist. By using a low contrast photoresist, such as those based on diazonaphthoquinone (DNQ), the material is only partially removed during development in proportion to the local exposure dose received. In this way, an arbitrary surface topography can be sculpted in the photoresist material. It is common practice in grayscale lithography to encode the transmission levels of the photomask by using the photoresist contrast curve to determine the exposure dose required for a given photoresist thickness at each lateral point in the pattern. This technique is adequate when the surface topography is slowly varying and the photoresist film is thin. However, it is inaccurate when these conditions are not met, because the technique essentially represents a one-dimensional approximation to the lithographic process where the isotropy of the development and the diffractive imaging of the photomask are neglected. Currently we are applying grayscale lithography to the fabrication of a fiber-to-waveguide coupler based on the parabolic reflector, where the efficiency of the device is quite sensitive to fabrication errors in the coupler geometry. In this case the thin photoresist and slowly varying topography conditions are not met, and we turn to more comprehensive process models to determine the appropriate transmission levels to encode in the photomask. We demonstrate that the photomask can be optimized, based on simulation of the lithography process, to produce the required three-dimensional photoresist pattern.

Citation Download Citation

Thomas Dillon, Mathew Zablocki, Janusz Murakowski, and Dennis Prather "Processing and modeling optimization for grayscale lithography", Proc. SPIE 6923, Advances in Resist Materials and Processing Technology XXV, 69233B (4 April 2008); https://doi.org/10.1117/12.773213

ACCESS THE FULL ARTICLE