| Summary: | The Microelectronics industry has continuously pushed the limit of critical dimensions to sub-20 nm. One of the challenges is pattern collapse, caused by unbalanced capillary forces during the final rinse and drying process. The use of surfactants offers a convenient method to reduce capillary forces but causes another deformation issue. This thesis work focuses on alternative approaches that are compatible with lithographic processes to mitigate pattern collapse. First, an e-beam lithography pattern with a series of varying line and space widths has been specifically designed in order to quantitatively study pattern collapse behavior. This pattern generates increasing stress in the pairs of resist lines as one moves across the pattern array and eventually a sufficiently small space value (critical space, S1c) is reached in each array such that the stress applied to the resist exceeds the critical stress (σc) required for pattern bending and subsequently feature deformation and collapse occurrs. The patterns we designed allow us to qualitatively and quantitatively study pattern collapse and obtain consistent, reproducible results.
In the first part of the thesis work, a quick surface crosslink (called a reactive rinse) that involves the strengthening of the resist using crosslinking via carbodiimide chemistry while the resist structures are still in their wet state, has been developed and demonstrated. This technique provides efficient and significant improvement on the pattern collapse issue. In the second part of the thesis work, a triethoxysilane compound, vinyl ether silane (VE), has been successfully synthesized. It can be used to modify the silicon or silicon nitride substrates and form a covalent bond with the resist film instead of manipulating the surface energies using common HMDS. Compared to traditional Hexamethyldisilazane (HMDS) vapor primed surfaces, the implementation of the VE adhesion promoter resulted in a significant improvement in the adhesion and resistance to adhesion based pattern collapse failure in small sub-60 nm resist features. In the third part of the thesis work, the effect of drying rates and drying methods has been systematically studied. SEM analysis and critical stress results showed that fast drying appear to reduce the resist collapse. The line pair orientations in each pattern array with respect to the wafer radius reveal an apparent effect of fluid flow and centrifugal forces on collapse. Finally, a comprehensive pattern collapse model that incorporates adhesion based pattern failure and elastoplastic deformation-based failure, and dimensionally dependent resist modulus properties has been developed. This model provides such an excellent prediction of the experimental data and supports the idea that this level of combined adhesion-failure and elastoplastic-failure based pattern collapse modeling, where one explicitly considers the dimensionally dependent mechanical properties of the resist can be quantitatively predictive and useful for understanding the pattern collapse behavior of polymeric nanostructures.
|
|---|
