Machine Learning Inverse Design Reveals a Double Narrow-Band Absorption Approach for Effective Colored Radiative...

Area of Science:

• Materials Science
• Nanotechnology
• Optical Engineering

Background:

• Colored radiative cooling paints are crucial for aesthetic thermal management.
• Previous designs were limited by cost, intuition, or idealized optical spectra.

Purpose of the Study:

• To develop a machine-learning-enabled inverse design framework for optimal radiative cooling paints.
• To bridge the gap between theoretical spectra and practical paint formulations.
• To achieve desired color and high cooling performance simultaneously.

Main Methods:

• Integration of photon Monte Carlo simulations with surrogate modeling.
• Machine-learning-enabled inverse design framework.
• Systematic exploration of the HSL color space.

Main Results:

• Uncovered a "double narrow-band absorption" strategy for reduced solar heating.
• Achieved reductions in solar heating power of up to 193 W/m² compared to conventional methods.
• Determined that 24% of colors can achieve subambient cooling under realistic material constraints.

Conclusions:

• The automated pipeline provides practical guidelines for designing high-performance, aesthetically tailored radiative cooling coatings.
• Demonstrated the feasibility of achieving significant cooling with colored paints.
• Highlights the potential of machine learning in materials design for thermal management.