Simulated crystallisation

Seeded crystallisation entails introducing a pre rystallised seed (perhaps generated using symmetry) into an amorphous sea of ions. The seed then nucleates the crystallisation of the amorphous regions. 

A strategy, which is perhaps more aligned with experiment, is simulated templated crystallisation. Here, one uses a (crystalline) substrate material to template the crystallisation of an overlying amorphous thin film. In particular. Fig. 5.9 shows the structure of a thin film of MgO supported on BaO. Such simulations, similar to atom deposition, attempt to simulate processes that occur during, for example, chemical vapour deposition or molecular beam epitaxy, where at some point the substrate will help nucleate and template the crystallisation of the amorphous thin film deposited thereon. 

Another approach to help facilitate crystallisation is to bias the tr ectories of all the atoms moving in the (molten/amorphous) configuration to favour crystallisation. Specifically, order parameters (which can be bond distances, coordination numbers, nearest neighbour densities and radial distribution functions) may be used as a gauge of crystallinity and used to help drive the simulation (trajectory) to favour maximising the order parameter and, ultimately, induce crystallinity. 

One of the problems associated with these types of simulations, is that MD tr ectories, which do not lead to crystalline structures, may be explored repeatedly. To prevent such occurrence and reduce computational effort, a tr ectory history can be recorded and compared with the current tr ectory. If a match is found, then the tr ectory can be modified to ensure that the simulation always explores a new phase space and structure. Clearly this approach can be overwhelmed by the sheer number of configurations that are possible. However, the technique has been used successfully by Quigley 

There are a wealth of other innovative techniques to circumvent such challenges including accelerated dynamics, metadynamics and hyperdynamics. A review of the challenges and state of the art of 

---

In this second section of this chapter we explore simulation methods, which can loosely be described as simulating synthesis including simulated annealing, atom deposition and simulated crystallisation. 

A third approach is to simulate crystallisation. In particular, all (crystalline) material synthesis involves some kind of crystallisation process. This may be crystallisation in solution or crystallisation... 

Another approach is the simulated moving-bed system, which has large-volume appHcations in normal-paraffin separation andpara- s.yXen.e separation. Since its introduction in 1970, the simulated moving-bed system has largely displaced crystallisation ia xylene separations. The unique feature of the system is that, although the bed is fixed, the feed point shifts to simulate a moving bed (see Adsorption,liquid separation). 

The current state of theory is quite crude state-of-the-art density functional theory calculations (e.g. PRISM ) have yielded scant clues about the renormalisation of isomeric state potentials due to density, and phenomenological theories have introduced an admittedly ad hoc coupling. Hence there is a clear need for theory which captures the minimal necessary molecular detail of orientation, conformation and packing. A possible route is simulation computational speed can almost handle the necessary melt simulations, while simulations of solution crystallisation have demonstrated how orientational correlations grow as crystallisation in theta solvent conditions. ... 

Figure 7.62 Scanning transmission elec- trailing edge of the transformation front tron microscope (STEM) image of transfer- well-crystallised HAp is formed whereas at mation of ACP to crystalline phases on con- the leading edge nanocrystalline HAp pretact with revised simulated body fluid (r- vails (see text) (Heimann, 2007).

Note that simple flashes can simulate a number of simple equilibrium devices, as evaporators, decanters or crystallisers. The flash units can also be used to check thermodynamic options before more sophisticated separations, or to prepare tables and diagrams of properties, as temperature-enthalpy in heat integration. 

Figure 4 Simulation of typical concentration changes during the cooling of a NaCl solution showing undercooling of water, ice nucleation and growth, transient supersaturation and final NaCl crystallisation, with the re-establishment of equilibrium conditions...

Ryan studied the rejection process during crystallisation (9). A wave of rejected additive builds up ahead of the growing spherulite as shovm in Figures 2 and 3. The shape of this wave depends on the diffusion coefficient of the additive and the growth rate of the spherulite and has been compared with computer simulations of the rejection. 

---