Key points are not available for this paper at this time.
Small particles at the nano- and micron-scale have found many applications made possible by their specific size, size distribution, and generally large surface-to-volume ratio. Yet, approaches to systematically designing these particles and their size distribution based on mechanistic principles─that is, being able to choose reaction conditions based on an established mechanism of formation─have so far largely eluded the field. Herein, we present a methodology that allows for optimizing reaction conditions in silico, using a well-characterized, prototype system of iridium nanoparticles as an example. We show that given a model of nanoparticle formation that we have previously vetted against experimental data, statistical estimates for the parameters in this model also previously obtained, and a suitable optimization algorithm, we can predict experimental conditions (such as initial concentrations and reaction end times) for which the resulting particle size distribution both closely matches a desired mean value, and is very narrow─that is, it allows for outcomes that enable many currently unattainable applications. Moreover, our methodology accomplishes this optimization task while also accounting for parameter uncertainty. The combination of model, parameter estimates, and optimization algorithm is generic and is applicable to many other nanoparticle systems as well, as long as a reliable model of the formation of these particles is available. Thus, we contribute a full analysis workflow where an understanding of kinetics equations is the only prerequisite mathematical knowledge.
Long et al. (Mon,) studied this question.