An isostructural series of dinuclear chromium-(III)−lanthanide(III) clusters is formed by fluoride abstraction of cis-[CrF 2 (phen) 2 ] + by Ln 3+ resulting in LnF 3 and methoxide-bridged Cr−Ln clusters (Ln = Nd (1), Tb (2), Dy (3 )) of formula [Cr I I I (phen) 2 ( μ -MeO) 2 Ln-(NO 3 ) 4 ]·xMeOH (x = 2−2.73). In contrast to fluoride, methoxide bridges in a nonlinear fashion, which facilitates chelation. For 3, X-ray magnetic circular dichroism (XMCD) provides element-specific magnetization curves that are compared to cluster magnetization and susceptibility data acquired by SQUID magnetometry. The combination of XMCD and SQUID is able to resolve very small magnetic coupling values and reveals a weak Cr III −Dy III coupling of j = −0.04(3) cm −1 . The Dy III ion has a ground-state Kramers doublet of m J = ±13/2, and the first excited doublet is found to be m J = ±11/2 at an energy of δ = 57(21) cm −1 . The Cr III ion exhibits a uniaxial anisotropy of D Cr = −1.7(1.0) cm −1 . Further, we observe that a weak anisotropic coupling of dipolar origin is sufficient to model the data, suggesting that methoxide bridges do not play a significant role in the magnetic coupling for the present systems.
■ INTRODUCTIONMolecular nanomagnets 1 have attracted a lot of interest because of their fascinating properties and their prospects in future applications in quantum information processing and molecular spintronics. 2 In particular, there is increasing activity regarding 3d−4f single-molecule magnets, and numerous species have been reported so far. 3,4 The complete understanding of their magnetic behavior is, however, lagging behind, and for instance, the key to the necessary ingredients that mediate properties such as slow relaxation of magnetization in 3d−4f clusters is still missing. 5 Difficulties in finding this key are aggravated by the complexity of the magnetism of 4f ions with orbital contribution to their ground state. 6 In particular, it is difficult to obtain precise values of the magnetic exchange coupling between 3d and 4f ions: In 3d clusters, this information can be readily extracted from dc magnetic susceptibility measurements; however, the large anisotropy splittings resulting from the 4f ligand-field states can strongly influence the temperature dependence of the magnetic susceptibility. Hence, exchange coupling deduced solely from susceptibility data may be subject to large uncertainties. The problem can be circumvented by using additional methods such as diamagnetic substitution, 7,8 electron paramagnetic resonance, 8 or inelastic neutron scattering studies. 9 X-ray magnetic circular dichroism 10 (XMCD) is able to obtain absolute values of element-specific magnetization, which can in turn be exploited to obtain a sensitive measurement of very small magnetic coupling values, and thus it is among the techniques of choice to quantify such coupling in 3d−4f clusters. 11 Precise knowledge about the magnetic exchange coupling is the prerequisite for obtaining its dependence along the homologous series ...