Owing to the potential applications of metal and metal oxide nanoparticles, this class of materials has attracted considerable attention in recent years. One of the major challenges is to be able to control their morphology (shape). Alpha-Al2O3 nanoparticles are probably the most studied system. Comparing with the conventional calcination method to produce alpha-Al2O3 nanoparticles (at least 950 C), both recently developed hydrothermal and glycothermal methods significantly improve the synthesis environment (300 C and 4 MPa). However, it is still difficult, under such conditions, to carry out fundamental studies on factors that could influence the morphology of alpha-Al2O3 nanoparticles.

Growth habits of crystals, especially the relative growth rates of various crystal faces bounding the crystal, play an important role in determining the final morphology of nanoparticles. The growth habit is determined both by the internal structure factors of a given crystal and by the external conditions such as temperature, super-saturation, solution composition, etc. Early theoretical models that were developed to predict crystal growth habits emphasize entirely on the internal structure factors and disregard the effects of external crystallization conditions. Hence the growth habits of crystals predicted by these theories are generally not in agreement with the experimental results. Based upon the observation that varying hydrothermal conditions could yield different morphology for a variety of metal oxide crystals, Li et al. presented a set of rules for controlling the growth of nanoparticles. In particular, they attributed the growth rates of various crystal faces to the types and number of elements that are present in the coordination polyhedrons at the interface. According to their rules, they found that the growth habit of alpha-Al2O3 particles is platy (001), and the growth rates in different directions are: V<001> < V<113> < V<110>.

The objective of this research is to study the effect of solvents on the relative growth rates of various crystal faces of alpha-Al2O3. Molecular dynamics simulation was used. Here, we present the results on the interactions between 1, 4-butanediol and four commonly emerged alpha-Al2O3 crystal faces. Simulation results show that the corresponding adsorption energy could be compared in the following sequence: Eads(001) > Eads(102) > Eads(113) > Eads(110). This result shows that 1, 4-butanediol molecules preferentially adsorb onto the crystal faces that have higher number of octahedral faces per unit area. Consequently, they prevent the growth units of alpha-Al2O3 from binding onto the crystal faces. In addition to the adsorption energy, our results also show that the growth habits are also influenced by the mobility of 1, 4-butanediol near the crystal growth faces. In fact, the computed diffusion coefficient of 1, 4-butanediol near the (102) face is the lowest comparing to those near other crystal faces. This explains why (102) face does not emerge unless moderate stirring is applied in practice. As clearly demonstrated here, atomistic level simulation can be used to gain useful insights into the growth habits of the alumina nanoparticles in a solvent environment that is extremely difficult to obtain experimentally. Hopefully, the simulation approach would lead us to develop criteria for preparing metal and metal oxide nanopartilces into the desired morphology.