Surface energy is a fundamental property of metallic nanoparticles (MeNPs), which plays a crucial role in nucleation and growth and has strong implications for the application and environmental impact of MeNPs. Surface energy (J m−2) can be size dependent, but experimental data on surface energy trends for MeNPs are inconclusive. Computational chemistry may resolve the issue, but the location and area of the surface used for scaling, which dramatically influences the outcome and interpretation, has not been properly investigated. The size dependency of the surface energy can only be determined by scaling to the thermodynamic surface of tension. To identify this, we have derived a generalized Tolman approach for non-spherical particles, which is used to analyze the thermodynamic consistency of various surface definitions. Only the physical surface, defined here, is consistent with the surface of tension. Scaling of recent computational data for faceted MeNPs to this surface yields a low size dependency of surface energy, in good agreement with the Tolman lengths corresponding to its interfacial position. We find Tolman lengths of −0.03 nm for icosahedra and −0.04 nm for cuboctahedra of gold or silver. With this result, our approach can be used to quantify the twinning energy for icosahedral nanoparticles, being ∼0.06 J per m2 twin area. To understand the unorthodox negative Tolman lengths, we have analyzed the surface energetics of the solid–gas interface of metals in relation to the liquid–vapor interface of water. Surface entropy is found to be imperative in determining the size dependence of surface free energy. At room temperature, the influence of surface entropy on surface enthalpy is much smaller for metals than for water. It explains why these two interfaces have opposite size dependencies of the surface Gibbs free energy and opposite signs of the Tolman length. For water, forming nanodroplets or nanobubbles, the Tolman length is negative (∼−0.014 nm) for the surface enthalpy, but positive (∼+0.06 ± 0.02 nm) for the surface Gibbs free energy. For MeNPs at room temperature, both entities are negative, but at high temperature, the increased surface entropy term may cause the size dependency of surface Gibbs free energy to become reversed.
Oxidative dissolution has large implications for the environmental fate and toxicity of silver nanoparticles (AgNPs). In this study, we quantify the kinetics, pH, and size dependency of silver ion (Ag+) release from AgNPs and explain our results in a consistent manner with a mechanistic view. Pristine AgNPs are covered by partially oxidized silver present in a single layer of subvalent Ag3OH groups that will be released by oxidative dissolution via two different pathways. Undersaturation of a solution, created by acidification, will initiate a fast oxidative dissolution process in which a pristine surface can be opened at particular points that grow laterally until a full layer of Ag is stripped off. At the newly exposed surface, Ag3OH is reformed. The opening of new spots stops due to increasing Ag+ concentrations. Via another pathway, the initial Ag3OH can be released by oxidative dissolution while simultaneously a new stable surface state is built with subvalent silver in two layers. This process is initiated by dilution and is visible around neutral pH values and may release a maximum of 30 ± 1 μmol Ag+ m−2. Its equilibration can be well described with a formulated thermodynamic model. The equilibrium constant (log K) is linearly related to the specific surface area of the AgNPs used, but can be shifted by the type of capping agent. The particle size dependency of the log K can be attributed to a surface Gibbs free energy contribution of 0.7 ± 0.1 J m−2. Ag+ release by stripping is relatively fast (∼1 day) in contrast to the process that leads to equilibration of two types of surface species that differ in the amount of subvalent silver. For this process, a kinetic Langmuir model has been developed in which the rate of Ag+ release is governed by adsorbed molecular oxygen that can be activated via a proton, while adsorption of molecular oxygen by itself can become rate limiting in the initial stage of dissolution with high rates of release. In our study with data of different kinds, the overall release and equilibration by AgNPs has been interpreted successfully with a coherent, overarching mechanistic view.