Selection and characterization of manganese oxide electrodes for harnessing salinity gradient energy
Open Access
- Author:
- Fortunato, Jenelle E
- Graduate Program:
- Environmental Engineering
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- October 01, 2020
- Committee Members:
- Christopher A Gorski, Dissertation Advisor/Co-Advisor
Christopher A Gorski, Committee Chair/Co-Chair
Bruce Logan, Committee Chair/Co-Chair
William D Burgos, Committee Member
Peter J Heaney, Outside Member
Patrick Joseph Fox, Program Head/Chair
Bruce Logan, Dissertation Advisor/Co-Advisor - Keywords:
- salinity gradient energy
manganese oxide
energy - Abstract:
- The potential energy contained in the controlled mixing of waters with different salt concentrations (i.e., salinity gradient energy) can theoretically provide a substantial fraction of the global electrical demand. One method for generating electricity from salinity gradients is to use electrode-based reactions in electrochemical cells. The performance of such salinity-gradient cells is enhanced when moving beyond high-surface-area carbons, that rely on double-layer capacitance, to faradaic electrode materials, such as manganese oxides, that store charge via pseudocapacitance. Manganese oxides are a compelling electrode material to use in electrochemical systems due to the low cost of manganese and the ability to tune synthesis protocols to control crystal structure, particle size, and morphology. Manganese oxides have shown to associate with Na+ ions, where Na+ uptake/release accompanies Mn3+/4+ redox reactions for charge compensation in the oxide. These redox reactions can occur at the manganese oxide surface via pseudocapacitive mechanisms, or they can occur within the manganese oxide structure via intercalation processes that involve structural tunnels or interlayers. In chapter 2, I characterized the electrochemical, structural, and morphological properties for twelve synthetic manganese oxides (β-, γ-, α-, δ-MnO2, birnessite, Mn2O3, and sodium manganese oxide), and subsequently tested each manganese oxide electrode in a salinity gradient flow cell to measure the power densities they produced with 0.02 M and 0.5 M NaCl solutions. The aim here was to make progress towards developing a rational framework for selecting electrode materials used to harvest salinity gradient energy. Power correlated with materials’ specific capacities, suggesting that cyclic voltammetry may be a simple method to screen possible materials. The highest power densities were achieved with manganese oxides capable of intercalating sodium ions when their potentials were pre-poised prior to power production. In chapter 3, I reported on a systematic investigation into the effects of manganese oxide mass loading on power production in the salinity-gradient flow cell using 0.02 M and 0.5 M synthetic NaCl solutions. Anodic electrodeposition was used to coat carbon cloth substrate with nanostructured Akhtenskite-type manganese oxide at controlled incremental mass loadings. Average power density measurements with the resulting manganese oxide electrodes show a positive correlation with manganese oxide mass loading, reaching a peak average power density of 2.21 ± 0.01 W/cm2 at 1.90 mg/cm2 loading. Flow cell data and results from electrochemical analysis suggest that performance of the manganese oxide electrode became limited by resistive losses in the poorly conductive manganese oxide coating at the highest mass loading/thickness. In chapter 4, I employed in-situ electrochemical Raman spectroscopy using a novel electrochemical capillary cell to probe the reversible phase transformations in sodium birnessite and calcium birnessite as a function of cyclic voltammetry cycling duration, electrochemical scan rate, and electrolyte composition, with the goal of understand the long term cycling stability of sodium birnessite. In-situ Raman spectroscopy revealed that sodium birnessite was stable for 15 cycles of sodium intercalation/deintercalation in 3 M NaCl at a slow scan rate (v = 2 mV/s), but the kinetically slow intercalation reaction was inhibited in CVs performed at a scan rate of 10 mV/s. Furthermore, when cycled in 3 M CaCl2, the Raman spectra of sodium birnessite became distorted and did not behave like sodium birnessite or calcium birnessite. This is in contrast to calcium birnessite, which adopted the Raman spectra of triclinic sodium birnessite and hexagonal birnessite when cycled in 3 M NaCl.