(607g) Stripping of Metal Ions for Ionic Liquid Recycling in a Self-Optimizing, Continuous Process

Ionic liquids (ILs), as green alternatives to some traditional volatile organic compound solvents, are widely employed in a variety of organic reactions and inorganic nanoparticle syntheses; however, the high cost of ILs often hinders their extensive applications. It is essentially favorable to recycle the expensive IL solvent from post-reaction solution for the purpose of minimizing the loss of materials. Also, there are yet very limited studies on IL extraction and separation, especially in the flow. We have designed and implemented a continuous, microfluidic process where extraction to wash the loaded IL and successive phase separation happen.

As mentioned, ILs as solvents can be used in colloidal metal nanoparticle syntheses, which makes metal ions one of the main impurities in the solution. We examined Fe(III) as an impurity in BMIM-Tf2N, a common IL in nanoparticle syntheses, using water as the washing agent to strip Fe ions from the IL phase. In the process, two infusing streams, Fe-loaded IL and water, formed slug flow in a PTFE tubing from a T-junction to allow mass transfer between the two phases. The biphasic slug flow then entered an SLA-printed membrane separator to conduct phase separation, where the aqueous phase (wetting phase) was able to permeate through the hydrophilic nylon membrane while the IL phase (non-wetting phase) was retained. Realtime Fe concentration was monitored by inline spectrophotometry which detected the transmittance at 305 nm, the wavelength of Fe(III) absorbance. The transmittance represented the extraction performance, while separation performance was evaluated by a pair of IR emitters and detectors that could distinguish different slugs in the flow.

To screen the experimental space, we utilized Design of Experiment (DoE) statistics approach. We also defined a parameter “score” to mirror the overall performance of the process as the output of a response function. The resulting fitting response function revealed where high scores resided in the experimental space. We also used a realtime, self-optimizing simplex algorithm to locate the optimal flow rate conditions for high scores with good extraction and perfect separation.

Another kind of IL, Cyphos 104, was applied to the as-established extraction-separation process as an application. Neodymium, one of the rare-earth metals, was dissolved in water first, then was transferred to the Cyphos IL by batch extraction and phase separation. Nd(III)-loaded IL was then washed with 1M nitric acid in flow using a similar configuration as the BMIM-Tf2N system. Nd was stripped off from the IL phase under the optimal flow rate combinations discovered by simplex.