Using molecular modeling to uncover the origins of subtle solvation-based film defects

Summary form only given. A series of lithographic metallization steps are used to build up the interconnect network on the chip during wafer fabrication and one often overlooked film technology that is an integral part of the process is the bottom anti-reflective coating (BARC), used to enable the production of uniform fine features in the resist. In order to create these features, near-perfect thin films must be produced, and any surface defect found during the spin-coating process is unacceptable, especially as interconnect densities become finer and finer. Recently, a defect was found that seemed to be sporadic in manifestation and proved to be difficult to explain from normal inspection of the BARC manufacturing process, which included inspection of every step of the process from incoming raw materials through final filtration and packaging. At that point, molecular modeling was called in to simulate a mechanism that could cause the defect. This molecular modeling encompassed a variety of methods, including thermodynamic calculations of the reaction steps using quantum mechanics and several molecular dynamics techniques to calculate solvation compatibilities (a cohesive energy density calculation), cohesive modulus on suspect oligomers, and simulation of temperature history-dependent pathways of the cohesive energy density on a high suspect oligomer. This final simulation uncovered a temperature sensitivity which explained the origin of the sporadic defects. This discovery initiated corrective actions that led to a final resolution. This paper will describe the modeling that was done and show the often forensic nature of uncovering the origin of this defect using molecular modeling. This example demonstrates how molecular modeling's role can be expanded by being used in a purely investigative manner rather than a predictive one.