Autonomous High-Throughput Characterization of Liquid-Liquid Phase Behavior.

Area of Science:

• Materials Science
• Chemical Engineering
• Physical Chemistry

Background:

• Traditional methods for characterizing liquid-liquid miscibility are slow, labor-intensive, and provide limited chemical insight.
• Accurate phase behavior data is crucial for formulation stability and efficacy but lags behind autonomous discovery methods.
• Existing techniques lack the throughput and chemical generality needed for modern formulation science.

Purpose of the Study:

• To develop an automated, high-throughput platform for comprehensive liquid-liquid phase behavior characterization.
• To enable rapid screening of miscibility, phase separation kinetics, and thermodynamic mapping.
• To support autonomous formulation discovery by providing scalable phase behavior data.

Main Methods:

• Integration of an asymmetric capacitance sensor for phase boundary detection and multi-angle turbidimetry for emulsion stability.
• Development of a flow-through chamber enabling continuous screening of diverse fluid chemistries.
• Application of Gaussian-process-based active learning and nonlinear programming for autonomous ternary phase diagram mapping.

Main Results:

• Successful classification of chemically diverse binary mixtures and real-time resolution of phase separation kinetics.
• Accurate identification of partial miscibility across varying compositions and temperatures.
• Autonomous mapping of ternary phase diagrams in approximately 2 hours and extraction of tie lines in approximately 5 minutes per line.

Conclusions:

• The automated platform unifies miscibility classification, kinetic characterization, and thermodynamic mapping in a single workflow.
• This integrated approach provides comprehensive phase behavior screening at unprecedented scale and throughput.
• The developed system is essential for accelerating autonomous formulation discovery and development.