Accurate and Robust Static Hydrophobic Contact Angle Measurements Using Machine Learning.

Area of Science:

• Materials Science
• Surface Science
• Computational Science

Background:

• Static contact angle measurement is crucial for characterizing surface properties.
• Current methods like Young-Laplace fitting are accurate but time-consuming.
• Existing techniques can be prone to human error and limitations in handling complex surface conditions.

Purpose of the Study:

• To develop a rapid and accurate machine learning model for static contact angle measurement.
• To overcome limitations of traditional goniometry methods.
• To provide an open-source tool for high-throughput surface analysis.

Main Methods:

• Trained a machine learning model on over 7.2 million half-drop contours derived from Young-Laplace equation solutions.
• Incorporated factors like surface roughness, gravity, drop size, and reflections into the training data.
• Developed an automated image and contour processing pipeline.

Main Results:

• The ML model, Conan-ML, achieves an estimated error of 1° for contact angles >110°.
• Conan-ML is two orders of magnitude faster than Young-Laplace fitting.
• Demonstrated superior accuracy compared to existing methods on an experimental dataset.

Conclusions:

• The Conan-ML approach offers a robust and reproducible method for high-throughput contact angle analysis.
• This open-source tool facilitates advancements in goniometry and surface characterization.
• The speed and accuracy of this ML model pave the way for new research and industrial applications.