Systematic Development of a Plasmon Ruler Using Dithiolate-Grafted Gold Nanoparticle for High-Throughput Screening of Anti-Icing Activity

Abstract

Here, we performed a systematic development of an assay based on a gold nanoparticle (AuNP) plasmon ruler to detect the activity of anti-icing reagents with superior stability and consistency in different phases (e.g., solid and liquid) applicable for standardized high-throughput screening (HTS). We found that dithiolate-grafted AuNPs indeed act as plasmon molecular rulers to measure the AuNPs interparticle distance during freezing. The plasmon ruler reflects the changes in ice crystallization in the presence of anti-icing reagents and can be used for standardized HTS of anti-icing reagents. We found that dithiolate ligands on the AuNP surface play a critical role in the assay performance. Molecular dynamics simulations showed that the anti-icing effects of the AuNPs are mostly sensitive to the ligand size and charges during cooling. Our optimized AuNP probe demonstrated higher colloidal stability and superior buffer resilience than previous versions, making it suitable for various types of target molecules, including chemicals, proteins, and organometals, with a high dynamic range of sensing. We investigated how the localized surface plasmon resonance peak or absorption spectrum correlates to ice formation and how this correlation can be used to evaluate the reagents’ anti-icing efficacy. We deconvoluted the measured spectra with theoretically derived unit spectra of a quantized AuNP cluster using COMSOL Multiphysics to calculate an apparent interparticle distance of AuNPs during ice formation and its changes in the presence of anti-icing reagents. We also proposed quantifying a AuNP-based colorimetric assay using dose–response Hill’s equation and an IC₅₀ calculation, which is essential for a standardized HTS for anti-icing efficacy measurement. Then, we compared AuNP results with established assays, ice recrystallization inhibition, and dynamic ice, of which the dynamic changes of size and shape over time were analyzed using machine learning.