Thiolate-protected metal nanoclusters (TPNCs) have attracted tremendous interest owing to their unique physicochemical properties, which enable applications from imaging cancer cells and bacteria to many forms of catalysis. TPNCs make up a unique class of atomically-precise nanomaterials that possess high stability at specific “magic size†compositions (i.e. M
n(SR)
m). Since TPNCs can be modulated by varying ligand and metal type, they are ideal systems to formulate structure-property relationships (SPRs) through systematic studies on the growing catalogue of structures. Herein, we employ Density Functional Theory and machine learning to develop SPRs that can accelerate TPNC design for targeted applications. First, we developed a simple, generalized framework that captures TPNC solubility behavior
1. The work leverages the molecular nature of TPNCs to reveal that ligand type and structural symmetry dictate solubility. Of note, we then leveraged these molecular properties to rationalize crystallization behavior in a series of trimetallic TPNCs
2. Next, we introduce structure-electronic-property relationships that capture electron affinity and ionization potential across the TPNC space
3. Capturing size and ligand effects, the models rely only on simple structural characteristics, and thus avoid the need for ab initio calculations. Importantly, we apply these SPRs to an atom-by-atom evolution series of TPNCs, rationalizing the dopant-based stability that is observed under experiment
4. Finally, we investigate composition and chemical ordering effects on the stability of heterometal-doped TPNCs. The work extends our previously developed genetic algorithm that predicts bimetallic nanoparticle stability
5 and sets the ground work for complete structural prediction of TPNCs
6. Our new model captures explicit metal-ligand interactions to determine stable dopant concentrations and positions in TPNCs. Overall, our computational studies rationalize a large series of experimental observations.
1. Cowan, M. J.; Higaki, T.; Jin, R.; Mpourmpakis, G., Understanding the Solubility Behavior of Atomically Precise Gold Nanoclusters. J. Phys. Chem. C 2019, 123 (32), 20006-20012.
2. Li, Y.; Cowan, M. J.; Zhou, M.; Taylor, M. G.; Wang, H.; Song, Y.; Mpourmpakis, G.; Jin, R., Heterometal-Doped M23 (M = Au/Ag/Cd) Nanoclusters with Large Dipole Moments. ACS Nano 2020, 14 (6), 6599-6606.
3. Cowan, M. J.; Mpourmpakis, G., Structure–Property Relationships on Thiolate-Protected Gold Nanoclusters. Nanoscale Adv. 2019, 1 (1), 184-188.
4. Li, Y.; Cowan, M. J.; Zhou, M.; Luo, T.-Y.; Song, Y.; Wang, H.; Rosi, N. L.; Mpourmpakis, G.; Jin, R., Atom-by-Atom Evolution of the Same Ligand-Protected Au21, Au22, Au22Cd1, and Au24 Nanocluster Series. J. Am. Chem. Soc. 2020, 142 (48), 20426–20433.
5. Dean, J.; Cowan, M. J.; Estes, J.; Ramadan, M.; Mpourmpakis, G., Rapid Prediction of Bimetallic Mixing Behavior at the Nanoscale. ACS Nano 2020, 14 (7), 8171–8180.
6. Cowan, M. J.; Mpourmpakis, G., Towards elucidating structure of ligand-protected nanoclusters. Dalton Trans. 2020, 49 (27), 9191-9202.
