(4bu) Rational Design of Advanced Organic Materials

In the late 1800's, Thomas Edison tried thousands of different materials to find a suitable filament material for his incandescent light bulb. Over a hundred years later, this ?Edisonian? approach is still one of the dominant methods used in the development of new materials. While this method may eventually yield improved materials, it is far too costly and slow for many advanced applications such as fuel cell membranes, organic electronics, organic photovoltaics, photoresists, biomaterials, and many other important purposes. Modern materials require modern, interdisciplinary approaches to their design and development. It is no longer sufficient to make thousands of different materials to identify a few superior materials. Instead, it would be much preferred to utilize predictive models to guide material design and selection and utilize a more limited experimental data set to calibrate and refine these model predictions. This is a daunting task even at the level of simply predicting the bulk properties of a material, but unfortunately in many emerging applications (e.g. organic electronics) the manner in which the material is processed and used can strongly affect its physiochemical properties. Thus, coupling in this information about processing effects will also be important in many advanced material applications. Applying these fundamentals to the synthesis of new materials combined with intelligent synthetic schemes quickly yield new and better materials.

Our work has piloted the use of such a rational materials design scheme in the area patterning materials for microelectronics processing. More specifically, we have used a combination of atomistic and mesoscale modeling to guide the design of novel molecular glass photoresist materials that can be used for high resolution patterning in semiconductor fabrication. Conventional state-of-the-art photoresist materials are composed multi-component blends of polymers with small molecule photoacid generators, dissolution inhibitors, and base quenchers. While this design approach has worked for the patterning of multiple generations of smaller and smaller transistors, it can be used for future generations because this design is unable to print features small enough with sufficient smoothness (line edge roughness) and sensitivity. This is especially true since the patterned feature sizes have truly approached the scale of individual polymers. Custom written kinetic Monte Carlo mesoscale simulations of the lithographic process have allowed us to investigate multiple different materials design strategies and processing conditions that lead to greatly improved material performance and helped guide the design of better materials. These simulations not only accurately model current performance; they have successfully predicted the improved performance of new materials. We have also developed thermodynamic models of molecular glass materials that successfully predict the glass transition temperature and accurately predict the full range of dissolution properties of these materials. These models allow the reduction of thousands of candidate materials and conditions to tens of materials. Even if excellent materials are designed, they are only useful if they can be made and made economically. To that end, we have created synthetic strategies that have allowed us to synthesize several generations of these materials, each with improved performance. After modeling and synthesizing new materials, we have improved their performance even further thorough the use of fundamental reaction engineering. By modifying the kinetics of cross-linking of one negative tone resist, the resolution was improved down to 25 nm. This rational design approach has yielded truly superior performing materials.