After an intensive week of talks, I made it back to London alive. Nuggets of wisdom during the last few days included “Theorists needs to provide error bars in their work; where are the error bars?” in relation to structure prediction (CCP5, London) and “The defect is localised... it’s not rocket science” (Defects in Oxides, Dublin).
Structure prediction is a new area to me, and the first day of CCP5 gave a nice introduction, with a fresh stack of papers on energy landscapes and global optimization on my desk waiting to be read. Defects in oxides, I am more familiar with, and there was a nice array of talks covering everything from catalysis to surface chemistry. Messages of the day: (i) an explanation for d0 magnetism in oxides is still elusive (with d0 being representative of a nominally closed shell system with no transition metal ions present); (ii) the band gap is a ground state property, so it is as valid to fit hybrid (exact-exchange) DFT functionals to this as anything other descriptor; (iii) GW theory remains lost in the realm of having no total energy or forces; (iv) TiO2 is still receiving more than its fair share of attention, both theoretically and experimentally.
At a well organized conference, informal conversations tend to surpass the presented material, and this was again the case at Trinity College. The conference dinner, at Fire, definitely finished things on a good note: 'Wood Fired Goats Cheese', 'Panfried Lemon Sole' and 'Traditional Apple Charlotte'.