Although the development of single-molecule magnets (SMMs) is rapid, there are only two families of high energy barrier (Ueff ) dysprosium(III) SMMs known so far: the cyclopentadienyl (Cp) family with a sandwich structure and the pentagonal-bipyramidal (PB) family with D5h symmetry. These high-barrier SMMs, which usually possess Ueff >500 cm-1 allow the separate study of the four magnetic relaxation paths, namely, direct, quantum tunnelling, Raman and Orbach processes, in detail. Whereas the first family is chemically more challenging to modify the Cp rings, it is shown herein that the latter family, with the common formulae [DyX1 X2 (Leq )5 ]+ , such as X1 /X2 =- OCMe3 , - OSiMe3 , - OPh, Cl- or Br- ; Leq =THF/pyridine/4-methylpyridine, can be readily fine-tuned with a range of axial and equatorial ligands by simple substitution reactions. This allows unambiguous confirmation that the Ueff mainly depends on the identity of X1 and X2 , rather than on Leq . More importantly, the fitted parameters are barrier dependent. If X1 is an O donor and X2 is a halide, 500<Ueff <600 cm-1 , log τ0avg (s)=-10.66, log Cavg (s-1 K-n )= -5.05, navg =4.1 and TH =9 K (in which τ0 is the pre-exponential factor for the Orbach relaxation process, C and n are parameters used to describe Raman relaxation, and TH is the highest temperature at which magnetic hysteresis is observed). For cases in which both X1 and X2 are O donors, 900<Ueff <1300 cm-1 , log τ0avg (s)=-11.63, log Cavg (s-1 K-n )= -6.03, navg =4.1 and 18<TH <25 K. Based on these results, it can be further concluded that Ueff not only has a linear correlation to the axial Dy-X bond lengths, but also to TH for these PB SMMs. This represents the first systematic study of a family of lanthanide SMMs and derives the first magneto-structural correlation in Dy SMMs.
Read full abstract