(376d) Multi-Objective Optimization for Selective Atomic Layer Deposition: A Case Study with Zirconia Deposition on Silicon Copper Composite.

Zirconia is a ceramic material with a high dielectric constant that has various applications in electronics and biomedical industries. Atomic layer deposition (ALD) is surfacing as a promising ‘bottom-up’ technique in fabrication of the electronic materials. Zirconia can replace silicon dioxide as an insulating material and can be used in designing complementary metal oxide semiconductors with a very small feature size. However, zirconia needs to be selectively deposited alongside copper interconnects for this application. In our previous work, a novel ALD process was developed that selectively deposits ZrO₂ on silicon on a silicon copper composite. In this work, a theoretical model was developed that uses modified UNIFAC group contribution method to estimate selective deposition of zirconia. The interaction parameters between the functional groups of the precursor and that of the substrate were estimated using the experimental data and the model was able to show a selective deposition rate of ZrO₂ on silicon with little to no deposition on copper. The resultant interaction parameters of the functional groups thus found are used in a multi-objective optimization method to determine a temperature window that will maximize deposition on silicon while minimizing that on copper. Furthermore, the model is also enhanced to implement a combinatorial multi-objective optimization method for CAMD of novel precursor materials that will maximize deposition on silicon while minimizing that on copper. Development of such a model is highly desirable as the experiments involve expensive materials. Moreover, this model predicts ALD growth based on chemisorption of the precursor only and is free from design constraints of the reactors and offers flexibility to include various experimental parameters such as temperature to assess the impact on the deposition process.