Frontiers in Chemical Engineering (Sep 2021)
En Route to a Unified Model for Photoelectrochemical Reactor Optimization. II–Geometric Optimization of Perforated Photoelectrodes
Abstract
Results have been reported previously of a model describing the performance of photoelectrochemical reactors, which utilize semiconductor | liquid junctions. This model was developed and verified using SnIV-doped α-Fe2O3 as photoanodes. Hematite films were fully characterized to obtain parameter inputs to a model predicting photocurrent densities. Thus, measured photocurrents were described and validated by the model in terms of measurable quantities. The complete reactor model, developed in COMSOL Multiphysics, accounted for gas evolution and desorption in the system. Hydrogen fluxes, charge yields and gas collection efficiencies in a photoelectrochemical reactor were estimated, revealing a critical need for geometric optimization to minimize H2-O2 product recombination as well as undesirable spatial distributions of current densities and “overpotentials” across the electrodes. Herein, the model was implemented in a 3D geometry and validated using solid and perforated 0.1 × 0.1 m2 planar photoanodes in an up-scaled photoelectrochemical reactor of 2 dm3. The same model was then applied to a set of simulated electrode geometries and electrode configurations to identify the electrode design that would maximize current densities and H2 fluxes. The electrode geometry was modified by introducing circular perforations of different sizes, relative separations and arrangements into an otherwise solid planar sheet for the purpose of providing ionic shortcuts. We report the simulated effects of electrode thickness and the presence or absence of a membrane to separate oxygen and hydrogen gases. In a reactor incorporating a membrane and a photoanode at 1.51 V vs RHE and pH 13.6, an optimized hydrogen flux was predicted for a perforation geometry with a separation-to-diameter ratio of 4.5 ± 0.5; the optimal perforation diameter was 50 µm. For reactors without a membrane, this ratio was 6.5 and 8.5 for a photoanode in a “wired” (monopolar) and “wireless” (photo-bipolar) design, respectively. The results and methodologies presented here will serve as a framework to optimize composite photoelectrodes (semiconductor | membrane | electrolyte), and photoelectrochemical reactors in general, for the production of hydrogen (and oxygen) from water using solar energy.
Keywords
