(438a) Control of Crystal Nucleation By Manipulating the Concentration and Properties of the Nucleation Precursors

Control of crystal nucleation is among the grandest fundamental challenges facing modern materials science. Classical nucleation theory (CNT) identifies three parameters that govern the rate of crystal nucleation: the solution supersaturation, the surface free energy of the interface between the nucleus and the solution, and the solute concentration. Modest supersaturation variations often completely suppress nucleation or invoke dramatic shifts to unwanted crystal morphologies and polymorphs. The free surface energy can only be lowered by adsorbents (increases of are banned by the Gibbs adsorption law). An alternative means to enhance crystal nucleation is to introduce heterogeneous substrates that support the emerging crystal nuclei. CNT suggests that nucleation assisted by heterogeneous substrates proceeds faster by orders of magnitude.

Numerous industrial tasks require suppression of nucleation. Elaborate structures can grow by precise nucleation control that consists of suppression of nucleation throughout and enhanced nucleation at a perfectly selected location. Robust suppression of nucleation cannot be understood or designed within the realm of classical theory. Strategies to suppress nucleation, however, naturally arise for two-step nucleation. Nucleation can be inhibited by disrupting the assembly of disordered precursors that host crystal nuclei or by engineering the cluster properties to constrain crystal nucleation.

We show that the nucleation of hematin crystals can be suppressed and fine-tuned via the properties of a population of nucleation precursors extant in the solution. Hematin crystals form in the digestive vacuoles of malaria parasite, where they sequester toxic hematin released as the parasite catabolizes hemoglobin from the patient’s erythrocytes. Suppression of hematin crystallizations is the most productive pathway to kill the parasites and fight malaria. The addition of four antimalarial drugs, heme-artesunate adduct (H-ARS), mefloquine (MQ), chloroquine (CQ), and pyronaridine (PY), known to control hematin crystallization invokes one of three outcomes: accelerated nucleation, suppressed nucleation, or no effect.

The nucleation enhancement and the passivity of two modifiers can be understood within CNT. The nucleation suppression falls outside of the realm of CNT and suggests that hematin crystal nucleation may follow a nonclassical pathway. Transmission electron microscopy (TEM) characterization of supersaturated hematin solutions supports this hypothesis. The micrographs reveal the presence of amorphous particles with diameters ranging from 70 to 250 nm that assist the nucleation of elongated hematin crystals, analogously to previous observations on crystal nucleation. The electron diffraction patterns of several particles represent superpositions of amorphous and crystal signatures and imply that the amorphous particles host crystal nucleation.

We characterized the amorphous particles that host crystal nucleation by oblique illumination microscopy. OIM revealed hematin aggregates that exhibit a relatively narrow size distribution with an average diameter d ca. 70 nm. Such aggregates may hold ca. 100,000 packed hematin molecules. The observed ds are consistent with the diameters of the amorphous aggregates observed by TEM. Both d and the number of aggregates per unit volume, N, are steady for at least one hour, behaviors that stand in contrast to expectations for domains of classical liquid or solid condensates due to a first-order phase transition.

These behaviors of the hematin aggregates cohere with previous observations of mesoscopic solute-rich clusters of other organics, such as olanzapine and numerous proteins. We conclude that the hematin aggregates are mesoscopic solute-rich clusters. According to recent models of the mesoscopic clusters, they exist owing to the accumulation of transient dimers. The cluster size is determined by the balance between the lifetime of the transient dimers and their rate of outward diffusion from the cluster core and is, hence, independent of the solute concentration and steady in time. By contrast, the amount of solute captured in the clusters, and the related number of clusters and cluster population volume, increases exponentially with the solute concentration as a consequence of the thermodynamic equilibrium between the clusters and the bulk solution. The mesoscopic clusters of hematin appear to remarkably well comply with the predictions of this model.

In further compliance with the model, the addition of the modifiers CQ, MQ, PY, and H-ARS does not affect the clusters size. These additives, however, invoke disparate responses of the cluster concentration N. Remarkably, the responses of the cluster population to the four modifiers run parallel to the responses of crystal nucleation, coherently with the role of clusters as crystal nucleation sites. Collectively, the TEM observation of cluster-assisted nucleation and the parallel trends of additive activity on the cluster population and crystal nucleation support a mechanism of nucleation control employing additives that modulate the nucleation precursors.