Accelerated HKUST-1 Thin-Film Property Optimization Using Active Learning.

Area of Science:

• Materials Science
• Chemical Engineering
• Nanotechnology

Background:

• Solution shearing is a promising method for fabricating large-area, void-free metal-organic framework (MOF) thin films.
• Full coverage is critical for MOF thin films in applications such as separations, membranes, and sensors.
• Optimizing the numerous parameters (e.g., concentrations, temperature, speed) for solution shearing is complex and time-consuming.

Purpose of the Study:

• To incorporate active learning into the solution-shearing process for fabricating HKUST-1 thin films.
• To gain control over film coverage and thickness using active learning.
• To improve the understanding of high-quality MOF thin-film formation via solution shearing.

Main Methods:

• An active learning approach was integrated with the solution-shearing technique.
• Processing parameters for HKUST-1 thin films were systematically explored and correlated with film coverage.
• A redox-active molecule, 7,7,8,8-tetracyanoquinodimethane (TCNQ), was used to confirm film coverage.

Main Results:

• Active learning successfully guided the optimization of solution-shearing parameters for HKUST-1 thin films.
• A large-area, fully covered HKUST-1 thin film with a minimized thickness of 2.2 μm was fabricated.
• The TCNQ@HKUST-1 thin film exhibited comparable electrical conductivity to previous thicker films, representing a 60% reduction in thickness.

Conclusions:

• Active learning effectively navigates complex processing parameter spaces in multicomponent systems.
• This approach accelerates the optimization of thin-film fabrication, especially when experiments are costly.
• The developed method enables efficient production of high-quality, thin MOF films for advanced applications.