In the chemical zoo this week: yttriumferricyanide YFe(CN)6. Otherwise a truly unremarkable compound but for one property: its compressibility. Duyker et al. report about it in Nature Chemistry (DOI). Measuring it is one thing: 25% at 1.5 Gpa - by neutron powder diffraction - but figuring out how exactly the crystalline material compresses is another. Regular solid caesium has a similar compressibility but entirely on account of reduced bond lengths. In Yttrium ferricyanide it is expected that actual bond distortion is involved but which ones? The material is a coordination framework akin a MOF with octahedral FeC6 units in a hexagonal superstructure with trigonal prismatic YN6 units. It turns out the FeC6 units are very rigid and that the YN6 units do all the work: they go from trigonal to octahedral and the linear Y-N-C-Fe bonds get seriously bent. In fact the bending motion costs energy but the twisting of the YN6 unit is energetically favourable, the unit is compared to a pre-wound mainspring. Watch the movie!