| Osmotically driven membrane processes show promise for water and wastewater treatment, desalination, and power generation. Forward osmosis (FO), in particular, has drawn attention recently as a potential energy-efficient desalination technology. However, successful implementation of FO is hindered by the lack of a suitable membrane and a draw solution. While significant advances have been made in the development of FO membranes, additional work is needed to develop suitable draw solutes that overcome the difficulties of draw solution regeneration and reverse draw solute permeation. Reverse draw solute permeation, which is inevitable in FO, is highly undesirable. This work aims at improving our understanding of the performance-limiting phenomenon of reverse draw solute transport in FO by systematically investigating the transport of a range of draw solute types.;In the first part, the reverse transport of a strong electrolyte is quantified and described. The resulting expressions successfully predict the experimentally observed reverse permeation and the reverse flux selectivity of the strong electrolyte. Reverse flux selectivity was found to be independent of draw solution concentration, but solely dependent on the pure water permeability of the membrane, the selectivity of the active layer, and the ability of the draw solute to generate an osmotic pressure. The model developed demonstrated a simple methodology for selecting a membrane and strong electrolyte draw solute system that could potentially maximize the water flux while minimizing the loss of draw solute.;In the second part, the reverse transport of neutral solutes is investigated and modeled. The transport model takes into account the formation of an additional external boundary layer on the feed side, since neutral solutes are not as effectively rejected by the membrane. The reflection coefficient of the membrane and the solute is incorporated to accurately predict water flux and reverse flux selectivity. The necessity to account for this term suggests the presence of solvent-solute coupling effects, possibly due to the high rate of reverse solute permeation into the feed.;In the third part of this work, the reverse transport of another type of draw solute, a weak electrolyte, was investigated and characterized. The solute reverse permeation and its dependence on the draw solution pH and the dissociation equilibrium is quantified and modeled. Under solution conditions where the pH of the draw favored the uncharged protonated species of the weak electrolyte, the reverse draw permeation increased significantly due to the rapid transport of the uncharged species through the membrane. However, in these cases, only moderate change in the water flux was observed, and is attributed to the inability of dominant uncharged weak electrolyte species to generate an appreciable additional osmotic pressure. This effect underscores the importance and relevance of operating at an optimal pH to minimize reverse draw solute flux whilst achieving a desired water flux, when utilizing a weak electrolyte draw solution.;In the final portion of this work, a new high-performance thin film composite FO (TFC-FO) membrane was used with two very different high osmotic pressure generating draw solutions: sodium chloride and ammonia-carbon dioxide. The pure water permeability of the TFC-FO membrane was approximately ten times higher than that of the previously used cellulose triacetate FO membrane. Both solutions were evaluated for reverse permeation and reverse flux selectivity. Using previously developed models, the reverse permeation rates at varying temperatures were predicted fairly accurately. The impact of operating temperature of feed and draw solutions on FO performance was also investigated for the ammonia-carbon dioxide system. Results showed higher water fluxes at higher temperatures. However, reverse permeation also increased significantly and reverse flux selectivity decreased, underscoring the importance of operating temperature on FO performance.;Reverse permeation of draw solutes is detrimental to the efficiency and performance of FO processes. This work purposefully advances the understanding of this performance-limiting phenomenon for a range of draw solute types. The results and models developed in this work can serve as a basis for the selection, development, and refinement of the myriad of potential draw solutions available for FO. These models can also be used in parallel with other design heuristics to help optimize a forward osmosis process. |
