Abstract
Distillation is the most common method employed to facilitate the separation of chemical species. However, distillation is considered impractical when separating mixtures of alkane and aromatic compounds, due to the formation of azeotropes. Therefore, separation processes like solvent extraction must be employed. The success of any solvent extraction process is heavily dependent on the selection of the most appropriate solvent. An appropriate solvent is characterised by its selectivity, capacity, and its effect on health, safety, and the environment, among other factors. Due to the difficulty of identifying a single solvent with all ideal properties, mixtures of solvents are often considered.
This research aimed to evaluate a solvent mixture of N-methylpyrrolidone (NMP) and glycerol as a potential solvent for the separation of the model alkanes and aromatics, specifically n-heptane and toluene, respectively. To evaluate the solvent mixtures’ effectiveness of separation, liquid–liquid equilibrium (LLE) data were measured at 298.15 K and 313.15 K under 1 atm for the novel systems: n-heptane + (NMP + 10% glycerol) + toluene, n-heptane + (NMP + 30% glycerol) + toluene, and n-heptane + (NMP + 50% glycerol) + toluene. Furthermore, solvent selectivity and capacity were computed. The analytical technique was used for the LLE measurements, and the composition analysis was performed by gas chromatography. The plait points were determined using the graphical Coolidge method.
It was found that maximum selectivities, capacities and areas of separation, for the solvent mixtures studied, in decreasing order of magnitude were: NMP + 50% glycerol > NMP + 30% glycerol > NMP + 10% glycerol. The high selectivity associated with glycerol was observed with increasing glycerol content in the solvent mixtures. However, the poor capacities were not observed. Temperature had an insignificant effect on the separation at the temperatures 298.15 and 313.15 K, as there was no significant change observed in the extent of the two-phase region.
In the final section of the research, the binodal curve data were correlated with the Hlavatý, β function and the log γ equations. The tie-line data were correlated with the NRTL and UNIQUAC activity coefficient models. The data was found to be better correlated with the β function equation and the NRTL activity coefficient model. The best NTRL correlation was reported for αij = 0.2. Root mean square deviation (rmsd) values of 0.0058, 0.0066, 0.0031,
iv
0.0033, 0.0044, and 0.0030 were recorded for the n-heptane + (NMP + glycerol) + toluene system at 10%, 30%, and 50% glycerol concentrations, each evaluated at 298.15 and 313.15 K under 1 atm, respectively.
Among the solvent mixtures investigated, NMP + 50% glycerol demonstrated the best selectivity with increasing aromatics content in the feed as well as a good capacity. However, the high glycerol content of this solvent mixture can be accompanied by high viscosities, which translates to high operational costs. Therefore, the most suitable solvent would be a trade-off between high selectivity, good capacity and high operational costs.