The synthetic and structural chemistry of polynuclear manganese, cobalt, nickel and copper carboxylate complexes, stabilized by various forms of di-2-pyridyl ketone, is discussed. The structural diversity displayed by the described complexes stems from the ability of the doubly and singly deprotonated forms of the gem-diol form of di-2-pyridyl ketone, or the monoanion of the hemiacetal form of this ligand, to adopt a variety of coordination modes. The nuclearities of the clusters vary from four to fourteen. Perhaps the most aesthetically pleasing families are the "flywheel Cu 12 clusters, and the Co 9 and Ni 9 complexes in which the nine metal ions adopt a topology of two square pyramids sharing a common apex. A means of increasing the ground-state total spin value of a polynuclear 3d-metal cluster is also proposed. The approach is based on the replacement of hydroxo bridges, that most often propagate antiferromagnetic exchange interactions in clusters, by the end-on azido ligand, which is a well known ferromagnetic coupler. This approach involves "true" reactivity chemistry on pre-isolated clusters and the products are not undergone significant structural changes, except for the azido-for-hydroxo substitution, compared to the starting materials/clusters.

In this review we approach the emerging field of sensors based on luminescent metal–organic frameworks from the perspective of the most commonly encountered pitfalls and we suggest best practices so that they can be avoided.