Ph.D Thesis

Ph.D StudentPiekner Ifat
SubjectLight-Matter Interaction in Photoelectrodes for
Photoelectrochemical Solar Water Splitting
DepartmentDepartment of Energy
Supervisor PROF. Avner Rothschild


Conversion of intermittent solar power to fuel is a promising path for solar energy storage. One way to do this is by solar-powered photoelectrolysis: splitting water molecules to hydrogen and oxygen using solar power. Hematite (α-Fe2O3) is considered as a leading photoanode candidate, primarily since it is stable in alkaline electrolyte and absorbs visible light. Its photocurrent is estimated to potentially reach 12.6 mA cm-2, assuming every photon with above band gap energy is absorbed and produces charge carriers that contribute to the water splitting reaction. However, hematite is known for its fast charge carrier recombination, which is commonly perceived as the main cause preventing reaching the theoretical photocurrent limit. In order to overcome this problem, nanostructuring and light trapping in ultrathin films were widely investigated, yielding partial success. Less than half of the limit has been reached by the champion hematite photoanodes reported to date.

Previous work in our group showed that the photoconversion efficiency can be increased by implementing strong interference in ultrathin film photoanodes on metallic back reflectors. However, metallic back reflectors cannot be used in stacked tandem cells because they don’t transmit photons to the bottom cell. By replacing them with semitransparent distributed Bragg reflectors (DBRs), we introduced a way to utilize strong interference in ultrathin film top absorbers in a tandem cell configuration with a hematite photoanode in front of a silicon PV cell. Constructing this PEC−PV tandem cell enabled demonstrating unassisted solar water splitting. Additionally, we showed that by optimizing the optical stack and employing a V-shaped configuration, the light harvesting efficiency can be increased substantially.

However, despite overcoming the charge carrier collection problem by using ultrathin films, the photocurrent enhancement lagged behind the optical absorption enhancement, suggesting there is an elusive factor that limits hematite’s photoconversion efficiency, besides recombination. We hypothesize that some optical transitions do not produce mobile charge carriers that can contribute to photocurrent generation. In this thesis we introduced two methods to extract the photogeneration yield spectrum, defined as the wavelength-dependent ratio between the absorption of photons that generate electron-hole pairs that can ultimately contribute to the photocurrent and the overall absorption, using optical and photoelectrochemical external quantum efficiency measurements. This provided a more realistic estimation of the maximum photoconversion efficiency of hematite photoanodes.

In the first method, no assumptions were made to extract the photogeneration yield shape. However, this method doesn’t resolve its absolute value since it is multiplied by the average spatial collection efficiency, and is applicable only to ultrathin films. The second method is not limited to ultrathin films and extracts both the photogeneration yield spectrum and the spatial collection efficiency using complementary measurements and minimal a priory assumptions regarding the charge carrier collection efficiency profile. These methods show that more than half of the absorbed photons in hematite are lost for non-contributing absorption. This observation, and the methods developed in this thesis, may direct future efforts to explore new ways to achieve high photoconversion efficiency in hematite and other correlated electron materials.