Optimization and Multimachine Learning Algorithms to Predict Nanometal Surface Area Transfer Parameters for Gold and...

Area of Science:

• Nanotechnology
• Materials Science
• Physical Chemistry

Background:

• The nanometal surface energy transfer (NSET) mechanism explains interactions between metallic nanoparticles and molecular dyes.
• Expanding the NSET model to diverse metal compositions is crucial for developing novel nanosensors.

Purpose of the Study:

• To modify and test the NSET formula for gold, silver, copper, and platinum nanoparticles of varying sizes.
• To investigate the predictive power of NSET compared to Förster resonance energy transfer (FRET) for different metals.

Main Methods:

• Modified the NSET formula, adjusting size-dependent dampening and skin depth terms.
• Tested the model with gold, silver, copper, and platinum nanoparticles.
• Employed an exhaustive grid search and artificial intelligence/machine learning algorithms (multilayer perception, least absolute shrinkage and selection operator regression) for model optimization.

Main Results:

• Modified NSET model showed closer agreement with experimental data for metal nanoparticles than FRET.
• Scattering effects were observed for nanoparticles with diameters around 20 nm.
• AI/ML algorithms achieved a correlation coefficient (R²) greater than 0.97, indicating robust model performance.

Conclusions:

• The adapted NSET model provides a viable framework for understanding energy transfer with diverse metal nanoparticles.
• Further research into physics-informed machine learning is warranted to refine NSET equations.
• This work supports the development of advanced nanosensors for various applications.