(207g) Modifying Properties of Polymer Nanostructure through Liquid Phase Reactive Treatments

As pattern sizes shrink below 50 nm in semiconductor devices, the resist film thicknesses and resist feature sizes used in the lithography steps for patterning these devices are also rapidly decreasing in size. This reduction in film thickness was originally moivated by limitations in the depth of focus of the optical projection exposue tools used to pattern such resist films. However, it has also become clear that such film thickness reductions are important in mitigating pattern collapse problems that occur due to capillary force induced collapse of such fine polymeric resist nanostructures during wet processing and drying of the resist pattern. This reduction in film thickness is a concern as film thicknesses approach and shrink below 100 nm. The sub-100 nm film thickness regime is where ultra-thin film effects have already been shown to become a major issue due to the confinement of the material at the interfaces and the large surface to volume ratio in such thin films; i.e. the interfacial layer between the substrate and the resist and the interfacial layer at the free surface of the resist film become significant relative to the total film thickness. These ultra-thin film effects have shown to result in dramatic changes in the properties of the materials such as raising or lowering the glass transition temperature, changing the mobility of the material, and orders of magnitude reduction in the diffusivity of small molecules such as acid and water. Among the most important changes for resist and lithography applications may be the potential change in the mechanical properties of these films as the film thickness shrinks and/or the patterned feature size shrinks. We have carefully studied the effect of polymer film thickness and patterned polymer feature size using carefully designed pattern collapse test structures that are used to determine the effective resist polymer modulus and critical stress at the point of pattern collapse for varying film thicknesses, pattern sizes, and types of photoresists. High resolution e-beam lithography was used to generate 2-D line pattern collapse tet structures that provide the ability to calculate the capillary forces generated during drying the patterned features from their as-developed wet state. These experiments have clearly shown that effective polymer modulus decreases with both decreasing film thickness and polymer nanostructure width. What is needed is a way to mitigate this reduction in mechanical properties that tend to exacerbate pattern collapse problems in a resist material as both feature size and film thickness are reduced.

To that end, we have developed a new liquid phase reaction chemistry based on carbodiimide chemistry that covalently cross-links the solid phase resist nanostructures on its exposed surface while still in the wet processing state. The goal in this process is to use the high surface to volume ratios in such thin films and nanostructures to allow for significant modification and imporvement of the effective modulus of the nanostructure through only this surface cross-linking. Chemistry has been developed that will cross-link either pendant alcohols or carboxylic acids, meaning that this reactive rinse can be applied to a majority of all commercial photoresists. The liquid phase reaction is carried out by simply rinsing a patterned photoresist with a water solution with the reagents dissolved into it. Due to the very high surface area of the patterned surface, the reaction is carried out very quickly (i.e. less than one minute) at room temperature. Using the pattern collapse test structures mentioned earlier, we have shown that the use of this reactive rinse provides a significant improvement in the pattern collapse behavior in photoresists and a large measurable increase in the effective modulus of the polymer nanostructure. Since collapse occurs due to capillary forces on the resist features during drying, the reactive rinse can be applied at the end of the normal wet resist processing sequence to prevent pattern collapse but yet not affect any of the previous processing of the resist and also not adding any significant complexity to the resist process. The influence of this process on resist performance will also be discussed in detail including: (1) spectroscopic studies have been carried out to examine the extent of cross-linking during the reactive rinse process and to determine its location in the polymer nanostructure, (2) the effect of the rinse on polymer swelling and distortion, and (3) the effect of the rinse of nanostructure roughness.