Capacitive deionization (CDI) is a class of electrochemical desalination technologies which desalinate via ion storage in electric double-layers. CDI has received renewed attention in recent years due to the ability to couple energy storage with salt separation. During galvanostatic operation, a current is applied between porous carbon electrodes until a limiting voltage is reach. The cell is then discharged by applying a reverse current, generating brine and recovering stored charge. However, CDI desalination and energy efficiency can be limited by parasitic side reactions, and selective adsorption of counter-ions (anions at the positive electrode, and cations at the negative electrode).

Several material additions and electrode configurations have been proposed to overcome these limitations, with the most prominent being the addition of ion exchange membranes (IEMs) promote counter ion flux out of the desalination chamber and incorporation of carbon slurry electrodes to increase system adsorption capacity. Likewise, functionalization of carbon electrodes has been studied to improve counter-ion adsorption within EDL micropores. While the incorporation of functionalized carbon, IEMs in membrane capacitive deionization (MCDI), and the use of slurry electrodes in flow capacitive deionization (FCDI) have successfully reduced energy consumption or increased ion adsorption capacity in CDI systems, these modifications are often evaluated under limited conditions on the basis of specific performance enhancements. Additional clarity is necessary to evaluate the associated performance and cost tradeoffs across the design space.

In this study, an equivalent circuit (M)CDI model, with porous electrode sub-models, was used to measure the sensitivity of CDI performance to material selection, design, and operating choices. In order to investigate the performance of FCDI, pulse-flowed electrodes of high capacity and electronic resistances were incorporated into the existing model. Similarly, fixed charge in the anodic and cathodic micropores was studied to investigate functionalized carbon. Constrained system parameters were randomly selected via latin hypercube sampling (LHS) across multiple electrode geometries and influent salt concentrations. The resultant model outputs were then correlated with input values to quantify parameter sensitivity.

Our sensitivity analysis shows that operating current density, electrode specific capacitance, contact resistance, and fixed charge density were the parameters which most significantly dictated (M)CDI performance. These parameters where then used to construct an operational space for CDI, MCDI, and FCDI. The results of the sensitivity analysis and technology comparison were used to evaluate the current effectiveness of CDI technologies as well as near-term projections for improvements in material performance. Through this analysis we are able to set targets for performance and material properties necessary for each CDI configuration to favorably operate.