Engineering self-assembly pathways in polymer blends for nanopattern optimization

Extreme ultraviolet lithography (EUV) has been pivotal to the continued pattern density scaling for microelectronics but is challenged by stochastic photon shot noise which results in pattern roughness and non-uniformity. Block copolymer (BCP) directed self-assembly (DSA) has emerged as a promising strategy for reducing pattern roughness associated with EUV lithography. The successful application of BCP DSA requires both near zero defectivity and minimal self-assembled pattern roughness. Minimal defectivity requires fast self-assembly kinetics, which is achieved in weakly segregated BCP systems, whereas low pattern roughness entails sharp BCP domain interfaces that are present only in strongly segregated BCP systems. This talk will present a potential solution using a two-step annealing process for engineering self- assembly pathways for symmetric BCPs and their blends with homopolymers. The first step involves annealing in a solvent with near neutral selectivity to both BCP blocks, imposing weaker segregation for rapid self-assembly of perpendicular lamellae line patterns. In the second step the samples are thermally annealed to enable short-range polymer chain rearrangement that sharpens domain interfaces under strong segregation conditions. The added homopolymers act as plasticizers for accelerating self-assembly kinetics in the first step and redistributes rapidly within the domain during the second step facilitating chain rearrangements thus allowing flexibility to self-assembly process pathway design. Key pattern characteristics including domain period (pitch), degree of pattern order, domain interface width, and line edge/width roughness is determined by scanning electron microscopy (SEM) and grazing-incidence small-angle x-ray scattering (GISAXS). The process window across a range of blend compositions and annealing conditions that maximize assembly kinetics and minimize pattern roughness in both directed and undirected self-assembly is identified.