Dominic Ryan, McGill University

On the face of it, the cubic bronze EuPd3S4 seemed like an interesting material for a quick look.  Relatively easy to make, but with an odd feature — the europium was present as a random ∼50:50 mix of 2+ and 3+, even though the compound exists for all of the trivalent rare earths and none of the divalent alkaline earths.

Then covid hit, the world shut down, and I had a lot of time on my hands… 

Lots of samples were made (doping with Ca, Sr, Y and La) to see how robust the valence distribution was. Single-crystal x-ray diffraction showed that the assigned crystal structure was wrong and neutron diffraction confirmed that the europium is valence ordered, rather than random — at least up to 340K. 

Hydrostatic pressure was expected to drive all of the europium trivalent (smaller ion), but it also enhances the ordering temperature and the system goes tetragonal as the magnetic order is lost and all of the europium converts to the trivalent state.  

All was going well until we tried to confirm that the structural change at 340K was associated with an averaging of the europium valence; indications of which had been seen in the lanthanum-doped samples. Understanding how (and why) the valence can randomize spatially at 340K without affecting the apparent valence identity of the europium until about 650K provides a final, and unexpected challenge.