Review pubs.acs.org/CR Lanthanide Single-Molecule Magnets Daniel N. Woodruff,† Richard E. P. Winpenny,†,‡ and Richard A. Layfield*,† † ‡ School of Chemistry, and The Photon Science Institute, The University of Manchester, Oxford Road, Manchester M13 9PL, United Kingdom liquid-helium temperatures, retain magnetization for long periods of time in the absence of an external magnetic field.1 The famous dodecametallic manganese-acetate cage [Mn12O12(OAc)16(H2O)4] (Mn12Ac) became the progenitor of a large family of magnetic materials known as single- molecule magnets (SMMs).2 Notwithstanding the huge intrinsic interest in SMMs, it was also realized that they could in principle be developed for new technological applications. SMMs can be considered as molecular analogues of classical bulk ferromagnets; hence, it might be possible to develop them for applications involving the storage and CONTENTS processing of digital information. However, in contrast to bulk magnets currently used for this purpose, such as 1. Introduction A neodymium−iron boride magnets,3 the molecular nature of 2. Designing and Characterizing Lanthanide SMMs B SMMs offers unique attributes that may allow information to be 2.1. Characterization of Ln-SMMs C stored with much higher densities, and to be processed at 3. Survey of Lanthanide SMMs C unprecedented speeds.4 Completely new applications of SMMs 3.1. Monometallic Ln-SMMs E have also been envisaged, including in the development of 3.1.1. Monometallic Ln-SMMs with Phthalo- molecular spintronics.5 Ultimately, however, SMM-based cyanine Ligands E technology can only be realized when two major problems 3.1.2. Non-phthalocyanine Monometallic Ln- have been solved. First, the unique properties of SMMs are SMMs H currently only accessible using liquid helium cooling; therefore, 3.2. Polymetallic Ln-SMMs P either the operating temperatures need to rise significantly, or 3.2.1. Bimetallic SMMs P applications so novel and important need to be discovered that 3.3. Trimetallic SMMs W temperature ceases to be an issue. Second, depositing and 3.3.1. Dysprosium Triangles and Vortex Spin addressing individual molecules of SMMs on surfaces have only Chirality W been explored with very few examples. One of the grand 3.3.2. Dy3 Chains as SMMs Y challenges in this field is still, therefore, to design and to 3.4. Tetrametallic SMMs Z synthesize efficient SMMs that function at temperatures likely 3.4.1. Tetrametallic Chain Ln-SMMs Z to be of practical use, or which show physics that goes beyond 3.4.2. Tetrametallic Square Ln-SMMs AA what can be achieved with classical magnets. fl 3.4.3. Butter y- or Diamond-Shaped Ln4 The success (or not) of an SMM can be measured in more SMMs AA than one way. First, the magnetic blocking temperature, TB,is 3.4.4. Cube-like Ln4 SMMs AC the highest temperature at which an SMM displays hysteresis in 3.5. Pentametallic and Hexametallic SMMs AC plots of magnetization (M) versus magnetic field (H). It is 3.5.1. Pentametallic Ln-SMMs AD important to note that the value of TB strongly depends on the 3.5.2. Hexametallic Ln-SMMs AD sweep rate of the magnetic field; hence, comparing the blocking 3.6. Heptametallic and Higher-Nuclearity SMMs AF temperatures of different SMMs should be done cautiously. 4. Conclusions AH Gatteschi, Villain, and Sessoli have proposed TB as being the Author Information AI temperature at which the time (τ) taken for the magnetization Corresponding Author AI to relax is 100 s:2 it would be useful if this definition was Notes AI fi universally adopted. Second, the coercive magnetic eld, Hc,is Biographies AI the strength of the magnetic field needed to drive the Acknowledgments AJ magnetization of an SMM back to zero following saturation. References AJ Third, the effective energy barrier to reversal of the magnet- Note Added in Proof AM ization (also called the anisotropy barrier), Ueff, is the energy required to convert an SMM back into a simple paramagnet. By far the most popular parameter is Ueff, which is used in the vast 1. INTRODUCTION In the early 1990s, great excitement followed the discovery that Received: January 15, 2013 a molecular transition metal coordination compound could, at © XXXX American Chemical Society A dx.doi.org/10.1021/cr400018q | Chem. Rev. XXXX, XXX, XXX−XXX Chemical Reviews Review majority of SMM studies, and to observe SMM behavior at transition metal SMMs, and this article is recommended fi 13 higher temperatures Ueff should be large. Although M(H) reading for those new to the eld. hysteresis and measurements of coercive fields have been used There has been a growing realization that single-ion to characterize some SMMs, both TB and Hc are used less anisotropy is the crucial property to consider when designing 14 frequently than Ueff, due largely to the phenomenon of SMMs with large anisotropy barriers; perhaps the clue was quantum tunneling of the magnetization that is particularly always in the name. Since 2003, and especially in the last five prominent in Ln-SMMs (see section 2.1). The blocking years, considerable attention has therefore focused on the temperature is also dependent on the technique used for the elements whose single-ion anisotropies are unrivalled through- measurement, and so is not an ideal parameter for judging the out the periodic table: the lanthanides and actinides. SMMs quality of an SMM. based on coordination compounds of the f-elements, To date, the largest anisotropy barrier claimed in a transition particularly those of the lanthanides, have accounted for some metal SMM is 67 cm−1, which was obtained from studies of the of the most eye-catching recent advances in molecular magnetism.15 Possibly of even greater significance is that cobalt(II) complex [Co(hfpip)2{D2py2(TBA)}]2, where hfpip is hexafluoro-4-(4-tert-butylphenylimino)-2-pentanoate and lanthanide SMMs (Ln-SMMs) have already shown consid- 6 erable potential to be developed for surface deposition and D2py2(TBA) is a diazo-dipyridyl ligand. Very recently, larger − fi (104−181 cm 1) barriers have been determined in various device applications. Herein, we review the rst decade of fi progress in studies of SMMs based solely on complexes of the applied elds for a series of linear two-coordinate complexes of − iron(II).7 However, for the first 15 years, the SMM field was lanthanides. Hybrid d f compounds constitute an important dominated by the Jahn−Teller ion high-spin manganese(III).8 class of SMM, but coverage of this area is beyond the scope of our Review.16 Actinide SMMs represent a small-but-growing In 3d-SMMs, the reversal of the magnetization is blocked by a 17 combination of two properties, the Ising-type magnetic class of SMM, but we also do not cover them in this Review. anisotropy, which can be expressed as the axial zero-field splitting parameter, D, and the total spin on the molecule, S. 2. DESIGNING AND CHARACTERIZING LANTHANIDE | | 2 | | 2 − SMMS The simple equations Ueff = D S and Ueff = D (S 0.25) then allow Ueff to be determined for SMMs with integer or A recent review by Rinehart and Long provides a lucid account noninteger total spin, respectively. From the outset, almost all of how f-element electronic structure can in principle be ff 15a e orts to generate SMMs with large Ueff values focused on manipulated to develop new SMMs. In this section, we synthesizing exchange-coupled cages with the largest possible summarize the important features of lanthanide electronic spin. However, as the field matured, it became apparent that structure that should be appreciated to interpret the properties this strategy might not necessarily produce the desired of Ln-SMMs. Irrespective of the type of metal, the two strict outcome, as three important examples illustrate. prerequisites for a molecule to be an SMM are that the −1 The original Mn12Ac was determined to have Ueff =51cm , electronic ground state must be bistable, and that magnetic − which arises from the product of S = 10 and D = −0.51 cm 1.1 anisotropy must be present. For lanthanide ions with ground 1 8 One of the largest anisotropy barriers in a 3d-SMM (measured electronic terms other than S0 and S7/2, the orbital in zero applied magnetic field) occurs in the hexametallic contribution to the magnetic moment is large and unquenched, fi ff manganese(III) cage [Mn6O2(sao)6(O2CPh)2(EtOH)4] and ligand eld e ects in lanthanide complexes can be regarded fi 18 ({Mn6}) (saoH2 = 2-hydroxybenzaldehye oxime), where a as a small-but-signi cant perturbation. In contrast, for 3d − −1 − combination of S = 12 and D = 0.43 cm results in Ueff =62 transition metals, spin orbit coupling is subordinate to ligand −1 9 fi ff ff cm . However, a key result was the Mn19 cage eld e ects, and Ln-SMMs therefore di er fundamentally from 2+ [Mn19O8(N3)8(HL)12(MeCN)6] , or {Mn19}(H3L = 2,6- transition metal SMMs in the nature of their bistable ground bis(hydroxymethyl)-4-methylphenol), which has a large total state. For transition metal SMMs, the total spin S and the 2 spin of S = 83/2. If the strategy of maximizing the total spin of a ensuing [2S +1]mS substates lead to ground-state bistability. cage to increase U is generally correct, then {Mn } could In contrast, for Ln-SMMs, ground-state bistability arises from eff 19 − reasonably be expected to be an SMM with a large anisotropy the [2J +1]mJ microstates within the spin orbit-coupled −1 10 “ ” 2S+1 barrier: instead, {Mn19} has Ueff =4cm . The problem is ground term, LJ.