It would be hard to envisage anything more ‘academic’ than spending a week in one of Oxford’s oldest colleges. While at one time, the lack of television may have resulted in plenty of time for self reflection, high-speed ethernet ensures distractions are there when required.
The event was a workshop on "Density functional theory for experimental spectroscopy", which could easily be subtitled "CASTEP 101", with CASTEP being the UK's premier plane wave density functional theory code. They definitely have some nice spectroscopic functionalities that go beyond what VASP can do natively; some healthy competition is always a good thing. One extra advantage of CASTEP is that, although licensed to Accelrys for use in Materials Studio, it remains free to UK-based academics.
After working through practicals on solid-state NMR, core level excitations, and vibrational spectra (using density functional perturbation theory), I got a nice feel for the basic usage of CASTEP. The biggest current drawback is that different pseudopotentials are required for difference functionalities, i.e. a mix of norm-conserving, ultra-soft and PAW may be needed, although their new on-the-fly pseudopotential generation is trying to overcome this. On the positive side, their implementation of hybrid functionals seems quite robust, making good use of symmetry and enabling non-self consistent band structure calculations (unlike VASP).
A shiny new PRL on the origins of p-type behaviour in cuprous oxide is here. Now it's time to prepare for two September conferences in London and Dublin.