Semi-batch fractional evaporative crystallization involves the removal of multiple solutes from a feed solution by the progressive achievement of supersaturation (through evaporation) and concomitant nucleation and growth of each species. The feed solution to the evaporator is expected to contain several sodium salts, among them sodium nitrate, sodium carbonate, sodium sulfate, sodium oxalate, sodium fluoride and sodium chloride. It is anticipated that evaporation from the feed solution produces burkeite (Na2CO3∙2Na2SO4), sodium carbonate monohydrate (Na2CO3∙H2O) and sodium nitrate (NaNO3), as well as other sodium salts, which do not interfere with the crystallization kinetics of the three main species.
The present work explores the application of evaporative fractional crystallization on a multi-species solution leading to the nucleation of several sodium salts, each with distinct nucleation points, crystallization kinetics, and resultant crystal morphologies. Moreover, interactions among species are likely to influence the kinetics and crystalline habit of each.
Experiments were performed to investigate the influence of several operating variables on the characteristics of the final crystal product. Sampling was performed on a regular basis, and analysis of PLM images of the samples allowed determination of the nucleation point of the different species and comparing them to thermodynamic simulations. Results showed that the interactions between species modified the nucleation points of each of the main species (sodium nitrate, sodium carbonate and burkeite).
In addition, growth and nucleation rates were estimated for the overall crystal population along with those of specific crystalline species. An especially noteworthy observation was that the sodium nitrate-burkeite interactions had an effect on the crystal growth and potentially on the morphology of sodium nitrate crystals.
Finally, the evolution of the crystal population through the experiment was determined and presented as a series of histograms. Useful information was obtained concerning the correlation between population and mass distributions for multiple-species crystallization and the decomposition of overall distributions into single-species distributions. Results obtained on this very complex system represent the first step leading toward the modeling of complex multiple species crystallization with species interactions.