Removal of non-ionic organic pollutants from water via liquid-liquid extraction.

Juan C López-Montilla, Samir Pandey, Dinesh O Shah, Oscar D Crisalle

Index: Water Res. 39(9) , 1907-1913, (2005)

Full Text: HTML

Abstract

The removal of model pollutants bromocresol green (BG) and phenol from water is demonstrated via two liquid-liquid extraction methods. Both methods exploit selective interactions established by the pollutant molecule with a surfactant, oil, or alcohol, and are variants of the more general Winsor systems where the phases are in contact along an extremely large interfacial area. In the first method the surfactant and the co-surfactant move from a predominantly oil-in-water microemulsion (Winsor I), to a middle phase microemulsion (Winsor III), and finally to a water-in-oil microemulsion (Winsor II), as the physicochemical conditions of salinity, temperature or hydrophilic-lipophilic balance of the surfactant system are varied. This method achieves better than 99% removal of the pollutant BG from water. It is argued that the removal is produced upon increasing the salinity of the system because the interaction of BG with a medium chain-length alcohol drives it to move along with the alcohol to another phase. The second method, which is scalable to industrial levels, uses a spontaneously produced water-in-oil microemulsion with large interfacial area that appears after bringing in contact water and a pre-formed Winsor II or Winsor III microemulsion system containing different surfactants and oils. The method is applied to the removal of phenol from water, and it is found that systems with polar oils such as ethyl butyrate or with cationic surfactants such as stearyl trimethylammonium chloride are more efficient in removing phenol than systems with normal alkanes or anionic surfactants. It is also shown that a microemulsion formed using a polar oil performs better than using only the polar oil as the extraction solvent. Finally, the efficiency of the second liquid-liquid extraction method can be increased from 69% in a single-stage process to 83% in a two-stage process, using the same total amount of extraction solvent.