Solvent Vapor Annealing, Defect Analysis, and Optimization of Self-Assembly of Block Copolymers Using Machine...

Area of Science:

• Materials Science
• Nanotechnology
• Chemical Engineering

Background:

• Block copolymer (BCP) self-assembly offers high-resolution, cost-effective nanopatterning for microelectronics.
• High defectivity in BCP nanopatterns limits their application, often due to sensitivity to processing variations.
• Optimizing solvent vapor annealing (SVA) is critical for achieving low-defect BCP patterns.

Purpose of the Study:

• To develop a precise, flow-controlled SVA system integrated with design of experiments (DOE) and machine learning (ML).
• To optimize the self-assembly of poly(styrene-b-dimethylsiloxane) (PS-b-PDMS) for hexagonal dot arrays with minimal defects.
• To establish a generalized methodology for identifying optimal processing conditions for BCP nanopatterning.

Main Methods:

• Utilized a flow-controlled SVA system for precise control over annealing parameters.
• Employed DOE to systematically investigate the impact of processing variables on BCP self-assembly.
• Applied ML to analyze defectivity, quantify pattern quality, and map the parameter space for optimal conditions.

Main Results:

• Identified critical dependencies of defectivity on film thickness and swelling degree during SVA.
• Quantified defect formation and pattern quality, including grain coarsening and macroscale phase formation.
• Developed a figure of merit (FOM) to assess pattern quality and mapped optimal SVA conditions using ML.

Conclusions:

• The integrated DOE and ML approach significantly accelerates the identification of optimal SVA conditions for low-defect BCP nanopatterning.
• This methodology enables a rational, less resource-intensive route to producing high-quality BCP dot arrays.
• The approach is scalable and applicable to various BCP systems for microelectronic fabrication.