Faculty Advisor: William Phillip, Department of Chemical and Biomolecular Engineering

Modeling and Experimentally Verifying a Multistep Nonhomeogenous Diafiltration Cascade for Application in Lithium Ion Battery Recycling (Summer 2024)

Membranes have been used commercially due to their energy efficiency (as compared to traditional commercial separation processes), modular design, and operational simplicity. Membrane nanofiltration
(NF) technology is an application of commercial membranes that may allow for more specific control over solute rejection properties by allowing for the fractionation of similarly sized ions, which is not possible in reverse osmosis or ultrafiltration setups. The commercial application of NF membranes for separations is rooted in their ability to purify solutions as a mode of filtration, by modulating the flux of the solute in solution. The principal way by which flux is modulated is through the rejection of a specific solute to concentrate the retentate solution while purifying the permeate.

While enhancing the fractionation of solutes is a problem typically addressed by developing ‘better’ membrane materials, it can also be accomplished by staging multiple membrane modules into a diafitration apparatus, which takes a mixed feed solution and separates it using multiple membrane module “stages” which will split the stream into a permeate and retentate solution. The modules are connected in a countercurrent fashion such that the effluent streams of module one become the influent streams of the subsequent/previous module, depending on whether the stream is retentate or permeate (Figure 1). A paper Staged Diafiltration Cascades Provide Opportunities to Execute Highly Selective Separations (Ind. Eng. Chem. Res. 2021, 60, 15706-15719) investigated the efficacy of utilizing diafiltration cascades with multiple modules (stages) to separate multivalent cations from monovalent ones. This efficacy was shown to be incredibly high, though notable shortcomings can be noted and attributed to the study assuming a large difference in rejection between the two solutes being separated, which is not necessarily true or accurate for most commercial membranes. Nevertheless, the study led to these cascades being extrapolated to lithium-ion battery recycling in a 2022 publication from the Phillip and Dowling Labs titled Optimal Diafiltration Membrane Cascades Enable Green Recycling of Spent Lithium-ion Batteries (ACS Sustainable Chem. Eng. 2022, 10, 12207-12225) which modelled an approach to determine the efficacy of a diafiltration based recycling method for Lithium-ion batteries. This efficacy was theorized through the model to be extremely high, as a three stage cascade was capable of recovering up to 95% of the lithium in the feed stream with a purity exceeding 93%. In both studies the efficacy was defined as the purity of the effluent streams with respect to the intended solute for each stream.

The proposed study will advance our understanding of diafiltration cascades by first testing the rejection properties of previously synthesized charge-functionalized membranes for various concentrations and mixture ratios of lithium and cobalt salt. A model will then be developed, which will iterate through combinations of membranes in order to derive the optimal arrangement for solute recovery and develop heuristics that will guide the development of future devices. The model will take into account real membrane rejection and throughput data that will work to move away from the shortcomings of previous studies of diafiltration cascades, who assumed large differences in respective rejection between solutes. The model’s validity will finally be experimentally verified using a physical 4-stage cascade utilizing the model’s proposed arrangements. In this project, a mixed solution of lithium and cobalt will be fed to an apparatus that will effectively separate the two solutes into two streams of purified lithium permeate and concentrated cobalt retentate.