Abstract and subjects
Capacitive deionization (CDI) is a promising technology that has gained interest for the desalination of brackish water. Hierarchically porous carbons are commonly used as electrodes for CDI due to their high surface areas and controlled pore size distributions that maximize ion adsorption capacity and rate. Electrospinning is an effective way of generating carbon nanofibers with high inter-fiber macroporosity that can be further modified to improve surface area, total pore volume, and pore size distribution. This work describes the use of sacrificial mesopore formers in tandem with a micropore etching technique to induce hierarchical porosity in electrospun fibers. Mesopores are formed via the dissolution of silica nanoparticles that are introduced into the fibers during the electrospinning step. After mesopore formation, micropores are etched into the resulting surface through KOH impregnation and thermal activation. This sequential technique creates a hierarchical network of pores from the inherent macroporosity of the fiber network, to the mesopores, and finally micropores to simultaneously maximize surface area and accessibility. Micropore formation is optimized to maximize specific surface area while maintaining physical integrity of the fibers. The combination of mesopores and micropores enables fast ion adsorption rates and capacity. Carbon fiber electrodes fabricated in this method achieve specific surface areas exceeding 1400 m2 g−1, with pore volumes exceeding 1.0 cc g−1. The pore size distributions are highly controlled, with 80 % of total pore volume coming from pores <20 nm in radius. In 500 ppm constant voltage CDI tests, these fiber electrodes obtain a salt adsorption capacity of over 14 mg g−1 at a salt adsorption rate of ∼4 mg g−1 min−1, showcasing the high capacity matched with high rate of these easily fabricated, inexpensive materials.
•Hierarchically porous carbon fibers were prepared by electrospinning.•Fibers were tested as electrodes for capacitive deionization.•Pore scale modeling provided insights into effect of pore size distribution.•Optimization of mesopore and micropore content provided maximum performance