We have been studying the reactivity of cerium dioxide (CeO2) using DFT+U (density functional theory corrected for on-site Coulomb interactions). Using U = 5 eV, we showed that formation of oxygen vacancies leads to formation of localised Ce3+ ions. The formation of oxygen vacancies is key to applications of ceria. For example, in oxidation reactions, such as CO oxidation to CO2 oxygen from the lattice is used. Therefore, the ease with which oxygen can be removed is important.
More recently, we have applied hybrid DFT, in the shape of the screened exchange HSE06 functional to oxygen vacancies in the (110) and (100) surfaces of ceria.
Recent work has investigated doping of the (110) ceria surface with a number of metals, in particular Ti and Zr, La, Au (and here), and trivalent dopants.
With Ti and Zr as dopants, the energy cost to form an oxygen vacancy is substantially reduced and the surface is more reactive to CO adsorption, forming a mono-dentate adsorbate that easily breaks up to form CO2.
La doping of the (111) and (110) surfaces of CeO2, we identified an unexpected defect. With La doping, a 3+ ion substitutes for the 4+ Ce ion, leaving behind an oxygen "hole", i.e. an O- ion. The O- ion is localised on an oxygen near the dopant, as in the image. In metal oxides with lower valent dopants, it is generally found that the stable defect is on in which two dopants are accompanied by an oxygen vacancy. The oxygen vacancy acts to compensate the two oxygen holes, by donating two electrons to the oxide. This is well known for the case of Li-doped MgO.
For La-CeO2, we investigated the anion vacancy compensation mechanism by doping with 2 La ions and removing an oxygen vacancy. We found (i) the oxygen vacancy has an energy cost of +0.64 eV, so that it will not necessarily form spontaneously and (ii) upon oxygen vacancy formation we form a Ce3+ ion and an oxygen hole. This is in contrast to the expected situation in which the two oxygen holes are compensated. Experimental efforts are underway to establish the presence of this defect centre.
For other 3+ dopants, namely, Al, Sc, In and Y, we have applied hybrid DFT. The polaron description of a single dopant is found with both DFT+U and hybrid DFT. For dopant compensation, we find that Al behaves classically, while the other dopants are similar to La. There is thus a dopant ionic radius effect in trivalent doping.
In 2011, we studied divalent dopants: Ni2+ and Pd2+ in the (111) and (110) surfaces of ceria. These dopants are useful in improving the reactivity of ceria to CO oxdiation and methane dehydrogenation. As divalent dopants, their substitution into CeO2 can result in either formation of hole states (as a result of there being fewer electrons on the dopant than on the host cation, see here and here), or formation of a charge compensating oxygen vacancy, whereby formation of the oxygen vacancy compensates the lower valencey of the dopant. Here both DFT+U and hybrid DFT (HSE06) were used, with the reliability of the latter for doped oxides being well established, we were able to investigate the ability of DFT+U in this situation.
We found (i) that Ni and Pd substitute in square planar-like configurations at the Ce site in both surfaces, (ii) charge compensating oxygen vacancies formed in both cases (i.e. the formation energy of the oxygen vacancy is negative) with both DFT approaches and (iii) the formation of the next (i.e. active) oxygen vacancy was smaller than in the corresponding undoped surface.
Contact enquiry (at) tyndall (dot) ie for all Business Development enquiries