Novel experimental technique opens the way to a better understanding of polarons on the hematite Fe₂O₃ surface

In a new publication, TACO researchers present a novel experimental technique to study polarons on the surface of the common mineral hematite (Fe₂O₃). Polarons are quasi-particles that form in the atomic lattice of a material when extra charges, such as an electron or a hole, i.e., the lack of an electron, interact with the surrounding ions. Polarons can have significant effects on the material properties such as electrical transport, optical properties, and surface reactivity. Studies of polarons generally rely on indirect experimental evidence as their existence and properties are typically deduced from such macroscopic material properties. For a better understanding of their behaviour, it is, however, crucial to study polarons directly at the single-particle limit. The new TACO publication presents novel experimental methods in exactly that regard and some surprising results.

Traditionally, polarons are introduced in a material by doping, i.e., by adding a small amount of an extra chemical element to a material, or by shining light on the material. However, the first method also affects the material’s crystal structure and creates trap sites for the polarons, and the second method creates both electron and hole polarons at the same time, making it impossible to study them separately. In the new method, the researchers added single polarons of a chosen type to the material’s surface by injecting them via the extremely narrow tip of a noncontact atomic force microscope (ncAFM). The electrons tunnel the tiny gap between the ncAFM tip and the surface one by one, and it is possible to count how many have tunnelled. After the injection, the researchers were able to image the polarons with the ncAFM tip that created them and study how electric fields, heating, and light affected the polarons.

The researchers grew thin films (50-100 nm) of hematite on natural Fe₂O₃ single crystals and purposely doped them with Ti (n-doping) or Ni (p-doping). The material is insulating at cryogenic temperatures (4.7 K). Thus, any charges that are injected at the surface are “frozen” out in the lattice at the spot of injection. As the material is heated up after injection, one can “watch” with the ncAFM how the cloud of polarons diffuses. 

As the experimental results show, doping significantly affects the stability of the electron polaron cloud. When doped with 3% Ti, the cloud scatters completely already at temperatures as low as 10 K, whereas undoped samples require 40 K for the same effect. This is somewhat surprising because one would expect the presence of dopants to create more lattice defects that usually impede the free movement of electrons in materials. This suggests that dopants in hematite do not serve as traps for electrons. Hole polarons, on the other hand, turned out to be much more stable with temperature than the electron polarons and get trapped at Ni dopant sites.

The article’s authors also applied density functional theory (DFT) and kinetic Monte Carlo (KMC) simulations to interpret the experimental results of electron transport in the doped hematite. Polaron diffusion is typically dominated by hopping, i.e., the electron polaron stays at a certain lattice atom for a longer while before hopping to one of the nearest neighbour atoms. The simulations suggest that the electrons acquire a transient metallic character during the hopping process. Like in metals, the electron delocalizes, albeit only for a short moment before it localizes again in an available Fe site nearby. This kind of diffusion via an intermediate metallic state is also known as “random flight”. 

The methodology introduced by this work may open new paths to understanding fundamental charge transport mechanisms that are essential in catalysis. Hematite is of special interest as it has several technical applications. In particular, it is one of the most promising semiconductor materials for photocatalytic water splitting, which is in the focus of attention for a “green”, CO₂-free energy source. The new study could help to learn how doping could help improve the photocatalytic properties of hematite for its use in green energy conversion.

This work was only possible by the strong collaboration between subprojects in TACO. Subproject P02, headed by Ulrike Diebold, grew the films; Gareth Parkinson’s P04 contributed to the understanding of the Fe₂O₃ surface, and P07, led by Cesare Franchini, did the DFT calculations. In addition, the groups of first author Jesus Redondo and last author Martin Setvin at Charles University Prague performed many of the AFM measurements and KMC simulations.