Hematite is predicted to have a high photoelectrochemical water splitting efficiency but currently is limited by a low polaron carrier conductivity. Consequently, doping hematite is necessary to increase the carrier concentration and thereby the conductivity. However, at modest dopant concentrations of 1.0% and above, a significant improvement is not observed. Recent theoretical calculations have predicted that there is a critical concentration, ${\mathrm{C}}_{\mathrm{crit}}$, of dopants that can be incorporated in the hematite structure for a given annealing temperature and oxygen partial pressure, above which precipitates may begin to form, thereby limiting the substitution of dopant into the hematite lattice. This critical concentration varies with the type of dopant, as well as the annealing temperature. In this paper we test the validity of these predictions by examining four dopants; Sn, Ge, Zr, and Nb, for samples synthesized in air at 800 ${}^{\circ }\mathrm{C}$. We find in good agreement with theory, that dopant-Fe-oxide or dopant-oxide precipitates do form at moderate dopant concentrations, reasonably close to the calculated critical concentrations (0.22%, 0.44%, 0.03%, and 0.07%, respectively) and explore the evolution of these oxide formations as the concentration increases above the critical point. Using EXAFS data analysis, we first show that at low total dopant concentrations the dopant substitutes uniformly throughout the hematite structure on the Fe site. As the concentration increases and oxide precipitates form we modeled the EXAFS data as a sum of a fraction, f1, of doped hematite and a second fraction, f2, of oxide. Surprisingly, the dopant-Fe-oxides/dopant-oxides that form are quite different for different dopants—and for Zr and Nb two different nanocrystal precipitates develop; two polymorphs of ${\mathrm{ZrO}}_2$ (orthorhombic and tetragonal) for Zr, and ${\mathrm{NbFeO}}_4$ and ${\mathrm{Nb}}_{12}{\mathrm{O}}_{29}$ for Nb. At moderate concentrations (0.3%) one oxide is present (orthorhombic ${\mathrm{ZrO}}_2$ and ${\mathrm{NbFeO}}_4$). As the dopant concentration increases further, these fractions saturate while the second oxide begins to grow. Finally, the fraction of dopant going into the oxide phase first increases logarithmically with increasing total dopant concentration and then begins to saturate. This means that as the total concentration increases above the critical concentration, a significant fraction of the dopant still substitutes into the hematite lattice until the total concentration is 3–5 times the critical concentration. This is important for applications that need to maximize parameters such as the catalytic efficiency.

Close

PrevNext

View figure in article

• Chemical kinetics, dynamics & catalysis
• X-ray absorption spectroscopy