Apr 9, 2025
9:00am - 9:15am
Summit, Level 3, Room 321
John Hylak1,Harry Atwater1
California Institute of Technology1
We report the first growth of single phase epitaxial FeWO
4 thin films, using oxygen plasma assisted molecular beam epitaxy. The FeWO
4 films are grown on c-plane sapphire (0001) at 650
○C substrate temperature by co-evaporating elemental Fe and molecular WO
3 from effusion cells with atomic O sourced from an rf plasma cell. The FeWO
4 films are oriented in (100) growth direction and exhibit 3 rotational twin variants where FeWO
4 [001] is aligned in the growth plane to sapphire [100] equivalent lattice directions, respectively. Epitaxial growth of FeWO
4 is driven by lattice match for a supercell consisting of 3 FeWO
4 unit cells stacked along the [010] direction. Additionally, matching hexagonal oxygen sublattices is a factor guiding the epitaxial growth.
Iron tungstate (FeWO
4) is a promising complex oxide semiconductor material for solar fuels photoanode devices due to its favorable band gap below 2 eV and chemical stability in bulk aqueous electrolyte. However, the ability to achieve phase pure FeWO
4 in thin film geometries is currently limited due to inherent complexities of the material, which is comprised of two low vapor pressure cations in an intermediate oxidation state. Typically, polycrystalline FeWO
4 thin films containing various Fe/W oxide phase impurities are achieved, which is a barrier to the further study of oxidation dependent optical absorption and electrical conductivity properties relevant to devices.
We also report on the structural, optical, and electronic transport properties of epitaxial FeWO
4 thin films as a function of oxidation state. The film oxidation state is controlled by manipulating the rf power to the atomic O plasma cell in the range of 60-120 W, with the cation flux conditions held constant. From X-ray diffraction (XRD), we observe 80-100 W films are structurally optimized with minimal strain and impurity phases/orientations. Comparatively, 60 W films exhibit some degree of epitaxial breakdown to polycrystalline FeWO
4, while 120 W films are compressed along the growth axis and potentially contain an Fe
3+ oxide phase impurity. X-ray photoelectron spectroscopy (XPS) indicates the structural trends are caused by a shift in the Fe
3+:Fe
2+ cation ratio from 2:3 in 60 W films to 3:2 in 120 W films. Optical absorption measurements indicate the structurally optimized 80-100 W films have a 1.8 ± 0.1 eV band gap with an additional interband transition at 3.1 ± 0.1 eV, with similar transitions in 60 W films. However, in 120 W films the higher lying 3.1 eV transition is shifted to 2.7 ± 0.1 eV due to the excess Fe
3+. Electrical transport measurements show the resistivity decreases over 2 orders of magnitude from 10
4-10
5 Ω cm to 120 ± 10 Ω cm as rf power is increased. The most conductive variants at 120 W are n-type with 2 ± 1 cm
2V
-1s
-1 mobility and 3 ± 1 × 10
16 cm
-3 charge carriers.
Our initial study of oxygen stoichiometry in FeWO
4 suggests electron polaron hopping is a dominant transport mechanism in over-oxidized FeWO
4 variants containing Fe
2+/Fe
3+ cation mixtures. It also elucidates the potential for an under-oxidized growth regime where Fe
0/Fe
2+ mixtures are present, where a different conductivity rule could apply based on oxygen vacancies. The synthesis of such under-oxidized variants, as well as synthesis on electrically conductive substrate materials, remain open opportunities for future FeWO
4 synthesis work.