- Email: [email protected]
[TmIII(hfac)3(NITPhOPh)]∞: A new member of a lanthanide-based Single Chain Magnets family
Inorganica Chimica Acta 360 (2007) 3807–3812 www.elsevier.com/locate/ica
[TmIII(hfac)3(NITPhOPh)]1: A new member of a lanthanide-based Single Chain Magnets family K. Bernot a
a,b
, L. Bogani a, R. Sessoli
a,*
, D. Gatteschi
a
Department of Chemistry and INSTM (Udr Firenze), Universita` degli Studi di Firenze, Via della Lastruccia 3, 50019 Sesto Fiorentino, Italy b Sciences Chimiques de Rennes, UMR 6226 CNRS - INSA Rennes, Equipe ‘‘Mate´riaux Inorganiques : Chimie Douce et Re´activite´’’, INSA-Rennes, 20 Avenue des buttes de Coe¨smes, CS 14315, 35043 Rennes Cedex, France Received 21 October 2006; accepted 3 December 2006 Available online 12 December 2006
Abstract The polymeric coordination compound of formula [Tm(hfac)3(NITPhOPh)]1 (where NITPhOPh is a nitronyl-nitroxide radical) has been synthesized and found to belong to the only reported family of isostructural Single Chain Magnets. Both static and dynamic magnetic measurements have been performed, and a dependence of the out-of-phase signal on the frequency is observed below 3 K. Scaling procedures indicate Ising magnetic anisotropy. Comparison of the extracted parameters with those of the previously reported isostructural compounds confirms a trend along the lanthanide series. 2006 Elsevier B.V. All rights reserved. Keywords: Nitronyl-nitroxides; Single chain magnets; Lanthanides; Magnetic anisotropy
1. Introduction For years the only known class of superparamagnetic molecular materials was that of the zero-dimensional derivatives called single molecule magnets, or SMMs [1]. The observation, a few years ago, of slow relaxation of the magnetization in 1D-materials has opened a new field of great interest for both physicists [2] and chemists [3]. In analogy to SMMs these new compounds have been called Single Chain Magnets (SCMs) [4]. Slow dynamics of the magnetization in anisotropic 1D-systems was originally predicted by Glauber in the 60s [5] and his model has since then been applied to a variety of different systems. The two key requirements of the model, that need to be satisfied in the realization of SCMs, are a strong Ising anisotropy of the magnetic centers and a very low ratio between inter-chain and intra-chain interactions. It is thus interesting to investigate how the variation of both these requirements in a *
Corresponding author. Tel.: +39 0 554573268; fax: +39 0 554573372. E-mail address: roberta.sessoli@unifi.it (R. Sessoli).
0020-1693/$ - see front matter 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.ica.2006.12.002
family of related systems affects the magnetic behavior. In this context the use of the chemical tuning offered by nitronyl-nitroxide radicals has provided major advances [5] and, in particular, linear chains based on Cobalt(II) [6,7] and Dysprosium(III) ions [8] have been obtained. In the last case the use of a bulky R substituent on the radical has reduced inter-chain dipolar interactions preventing the onset of three-dimensional magnetic order. In a previous work we showed how lanthanide-based SCMs can be used to investigate the role of anisotropy on the dynamics, by designing a family in which the substitution of the magnetic ion leads to a modulation of the magnetic anisotropy without structural modifications [9]. Our previous study was performed on chains containing the rare earths (REs), Eu, Gd, Tb, Dy, Ho, Er, Yb, for which a non-monotonous trend in the magnetic anisotropy is observed. The study was performed on chains of formula [M(hfac)3(NITPhOPh)]1, where M= Eu, Gd, Tb, Dy, Ho, Er, Yb, and NITPhOPh is the nitronyl-nitroxide radical 2,4 0 -benzoxo-4,4,5,5-tetramethyl-imidazoline-1-oxyl-3oxide, depicted in. In this family, despite the internal dis-
3808
K. Bernot et al. / Inorganica Chimica Acta 360 (2007) 3807–3812
position of f-electrons, which leads to a weak intra-chain coupling, a very remarkable behavior can be observed. The contribution to the energy barrier of the strong Ising anisotropy significantly compensates the weak exchange interaction that characterizes 4f-ions. A very good isolation of the chain, and the huge anisotropy provided by LnIII ions give rise to a SCM behavior for the Tb, Dy and Ho derivatives. Both static and dynamic properties of the system are affected by the change of the magnetic ion. From the static magnetic properties, effects are seen on the intra-chain correlation length at low temperature and on the presence of steps in the magnetization curves. The most interesting feature is displayed by the dynamic behavior of the family. Peaks in v00 versus T curves appear and shift in temperature as the anisotropy of the ion used changes. As very little is known about the magnetic anisotropy of TmIII ion in low symmetry environment, in this work we study the Tm-based chain 1, [Tm(hfac)3(NITPhOPh)]. We thus complete the family of heavy 4f-SCMs and give a last evidence of the crucial influence of the anisotropy of Ising ions in SCMs. 2. Synthesis
Fig. 1. Powder X-ray spectrum of the chain [Tm(hfac)3NITPhOPh]1 compound (upper graph) and simulated spectrum using data from the reported system [8]. In the inset we show the view of the crystal structure of [Dy(hfac)3NITPhOPh]1. Fluorine and hydrogen atoms are omitted for clarity.
needle-like crystals. Anal. Calc. for C34H24O9N2F18Tm: C, 36.61; H, 2.17; O, 12.91. Found: C, 36.58; H, 2.13; O, 12.93%.
2.1. General procedures and materials
2.3. Crystallography
All chemical and solvents were used in reagent grades. The radical 2,4 0 -benzoxo-4,4,5,5-tetramethyl-imidazoline1-oxyl-3-oxide, named NITPhOPh in the following, has been synthesized according to the literature methods [10].
A single crystal of [Tm(hfac)3NITPhOPh1 was characterized with a cell measurement, revealing that the main crystallographic parameters are: a = 14.400(4), b = 16.681(2), c = 17.120(8), a = 90.03(7), b = 89.96(5), c = 89.95(9), at room temperature. As the isostructural Dybased compound has already been reported by some of us [8], we did not proceed to a full structure determination. Anyway, as magnetic measurements were performed on microcrystalline powders, a powder X-ray diffraction pattern was recorded at room temperature and superimposed to the simulated powder pattern of the [Dy(hfac)3NITPhOPh]1. The two spectra, which match very well, are depicted in Fig. 1, together with the view of the structure of the polymeric compound.
2.2. Synthesis of [Tm(hfac)3(NITPhOPh)]1 The reacting salt [Ln(hfac)3 Æ 2H2O] was synthesized using the following procedure: 1.78 ml (24 mmol) of a 25% ammonia solution was added drop-wise to 3.39 ml (24 mmol) of hexafluoroacetylacetone (hfac) in 100 ml of ether at 2 C. Then 8 mmol of [Tm(Cl)3 Æ 5H2O] in 10 ml of water were added to the solution. After half an hour of strong stirring, 20 ml of water were added. Then organic and aqueous phases were separated and washed with water and ether, respectively. The organic phase was dried on magnesium sulfate, filtered, and then concentrated to form yellowish oil. Twenty milliliter of hexane was then added and the solution was then heated up to 55 C for some minutes. After filtration, in order to separate the formed solid, the compound was kept in the freezer at 15 C, where it crystallized. The mean overall yield was about 50%. The chain [Tm(hfac)3NITPhOPh]1 was synthesized by the following procedure: 1 mmol of [Tm(hfac)3 Æ 2H2O] was dissolved in 30 ml of dry boiling n-heptane. The solution was left to boil for 20 min and then cooled down to 75C, when 1 mmol of the crystalline solid radical NITPhOPh was added under stirring together with 3 ml of CH2Cl2. The final solution was then cooled down to room temperature and was left still for about 24 h, to give dark
3. Physical measurements Temperature-dependent dc magnetic susceptibility measurements were performed on solid polycrystalline samples with a Cryogenic Ltd. S600 SQUID magnetometer and were corrected for the diamagnetic contribution, as calculated with Pascal’s constants, and for the diamagnetism of the sample holder, as independently determined. Magnetization curves were measured with an Oxford VSM system after inclusion in Apiezon grease to prevent orientation of the crystallites in the applied magnetic field. Data were corrected for the magnetism of the grease, which was independently determined at the same temperature and fields. A 12 kOe/min field sweep rate was used, while acquiring continuously. The ac magnetic susceptibility was measured
K. Bernot et al. / Inorganica Chimica Acta 360 (2007) 3807–3812
3809
using a homemade probe [11]. Powder X-ray diffraction patterns were recorded on a Powder diffractometer Bruker D8 Advance. Powder diffraction pattern of Dy compound was simulated from CIF file using Powder Cell software. 4. Magnetic characterization 4.1. Static magnetic measurements The vT product versus T was recorded at temperatures between 2 and 300 K and is displayed in Fig. 2. The compound has a room temperature susceptibility of 7.60 emu K/mol, slightly lower than that expected for one TmIII ion, which has a 3H6 ground state, plus an uncoupled S = 1/2 radical center. On cooling down, the value is constant until 130 K where a slight decrease starts, then becomes abrupt around 30 K. The curve reaches its lower value, 4.67 emu K/mol, at 3.4 K. It then markedly increases to reach 7.02 emu K/mol at 2 K. Fig. 3 shows the magnetization versus field curve at 1.55 K. No hysteresis was observed down to 1.55 K. Nevertheless the compound displays a non-trivial behavior at low field, where a slight step is visible. In order to better appreciate this feature, the derivative of the magnetization is depicted in the inset. The step is found around 6.10 kOe. One can notice that this step becomes sharper on cooling and is independent from the magnetic field sweep rate. The magnetization value is 23920 emu/mol (4.28 lB) at 120 kOe, lower than the expected 8 lB value. This is, however, in agreement with previous results on similar compounds [8,9,12]. 4.2. Dynamic magnetic measurements As first investigation of the dynamical behavior of 1, zero-field-cooled and field-cooled (ZFC–FC) curves were recorded in low temperature region and are found to be perfectly superimposable, thus indicating that down to
Fig. 3. Magnetization versus field measurement for the [Tm(hfac)3(NITPhOPh)]1 chain as a polycrystalline sample at 1.55 K and with a field sweep rate of 12 kOe/min. In the inset we show an enlargement of the derivative in the low field region.
Fig. 4. Temperature dependence of the imaginary v00 and real v 0 components of the ac susceptibility measured in zero applied field for 15 logarithmically spaced frequencies in the range 110 Hz (light gray) to 19 010 Hz (dark gray). In the inset the imaginary component v00 vs. T is shown, depicting only the eight highest frequency for clarity. The lines are a guide to the eye.
the lowest investigated temperature the relaxation time is shorter than a few seconds. Measurements of ac susceptibility were performed with 15 logarithmically spaced frequencies in the 110– 19 010 Hz range between 1.55 and 7 K, in the zero external field. The temperature dependence of both in-phase (v 0 ) and out-of-phase signals (v00 ) is depicted in Fig. 4. v 0 ranges from 1.23 emu mol1 at 5 K to 9.96 emu mol1 at 1.55 K for the highest frequency. The v 0 /v00 ratio is rather high but, in spite of that, some frequency dependence of the out-of phase susceptibility (inset of Fig. 4) can be found under 4 K. For the sake of clarity only the eight highest frequencies are depicted for v 0 . 5. Results and discussion
Fig. 2. Temperature dependence of the vT product extracted from SQUID magnetic measurements. Symbols are experimental data (empty circles are recorded in an external field of 0.1 kOe to avoid saturation effects, squares in 1 kOe and half-filled triangles in 10 kOe). In the inset is reported the scaling of the data using an Ising temperature dependence of the susceptibility (see text) and plotting ln (vT mol K1 emu1) vs. 1/T. The solid line represents the linear fit in the 2.0–2.5 K temperature range.
5.1. Static properties The association of lanthanides and nitronyl-nitroxide radicals has led to a huge variety of compounds [5b] and has been widely investigated from the magnetic point of view [12a]. In those systems the simulation of magnetic
3810
K. Bernot et al. / Inorganica Chimica Acta 360 (2007) 3807–3812
Table 1 Main magnetic data are extracted from the static properties of the chain compounds Rare earth
vT@300 K (emu K/mol)
M@120 kOe (emu/mol)
Step field (kOe)
Steepness @ Step (emu mol1 Oe1)
Slope (Sl) (K)
Tb**
11.10 12.2 14.4 11.37 7.60
26 700 30 900 34 300 32 000 23 920
13.77 16.23 13.08 6.26 6.10
1310 1680 1080 790 1065
9.1 ± 0.5 18.2 ± 0.3 3.4 ± 0.3 4.6 ± 0.2 3.7 ± 0.3
Dy* Ho** Er** Tm
The field value of the steps and their steepness are extracted from the first derivatives of the magnetization curves. The slope is extracted from the logarithmic scaling of the vT curves. Parameters are extracted from data reported in Refs. [8]* and [10]**.
data is hampered by the peculiarity of the ground states of the rare-earth ions: their orbitally degenerate states show strong spin–orbit coupling. The ground multiplet, characterized by J = L S for fn configurations with n < 7 and J = L + S when n > 7, is split by the crystal field. That is why in the last decade the attention mainly focused on GdIII and EuII derivatives, which have a f7 configuration and an orbitally non-degenerate ground state. Nevertheless the orbital component, while complicating the analysis of the magnetic properties, is rather important, as it also affords considerable anisotropy to those ions. They are therefore good candidates for designing one-dimensional magnetic materials where the anisotropy can be tuned at will, thus offering the possibility of exploring its key role in the dynamics of Ising systems. The case of the Tmbased compound is discussed hereafter and comparison with already reported values for the family is made in Table 1. The dc magnetic measurements show the expected behavior for this type of compounds. The first effect is due to the crystal field, i.e on cooling down the system, the vT product decreases, as the Stark sub-levels of the TmIII are progressively depopulated. The following abrupt increase of the vT product at low temperature is associated with the intra-chain coupling involving both TmIII ions and the S = 1/2 radicals, and is connected to the growth of the magnetic correlation length n inside the chain. Previous investigations of the static properties of chains designed with nitronyl-nitroxide radicals and rare-earth ions reveal the simultaneous presence of both NearestNeighbor (NN) and Next-Nearest-Neighbor (NNN) exchange interactions between the magnetic centers [12a], with antiferromagnetic NNN interactions. Data on the magnetic anisotropy of TmIII ions are extremely scarce, given that it is a non-Kramers ion and often EPR-silent, except for a few reports suggesting an Ising-like character of TmIII ion [13]. The anisotropy varies widely depending on the structural conformation of the ligands around the magnetic ion, and is hardly modellizable in low-symmetry environments. Moreover in nitronyl-nitroxide based compounds it is not possible to separate crystal field effects from the ones due to the exchange interaction with the spin of the coordinated radicals. Thus, to obtain some information about the kind of anisotropy involved in our compound, we choose to perform a scaling procedure of the low temperature data, in the region where n starts to grow
[14]. This is due to the fact that the susceptibility of a ferromagnetic Ising chain follows the growth of the correlation and, for an Ising system an exponential behavior is expected [15]: vT / n ¼ e2JS i S iþ1 =kB T where J is the intra-chain exchange energy, and with Si we indicated the spin coordinates. Thus, if the system has a truly Ising nature, plotting the logarithm of vT versus T1 should afford a straight line, with the slope directly providing the exchange energy. In our case a straight line is found for very low temperatures (between 2 and 2.5 K) as reported in the inset of Fig. 2. The linear fit gives a slope of 3.7 ± 0.3 K with an agreement factor R = 0.9994. This is comparable with reported values observed on Ho and Er derivatives (3.4 ± 0.3 K and 4.6 ± 0.2 K, respectively) and suggests a weak exchange interaction along the chain. The magnetization curve recorded as a function of the applied field is displayed in Fig. 3. Like other compounds based on anisotropic ions in the family, the Tm chain presents steps, clearly evidenced by the plot of the derivative of the magnetization, as reported in the inset. The Gd and Yb chains on the contrary do not show such an anomaly. The step for Tm is found at 6.10 kOe, close to the one of Er compound (6.25 kOe). It is however sharper, with a steepness of 1065 emu mol1 Oe1, close to the value found for the Ho derivative. As shown in Table 1, the trend in the steepness previously observed in the reported family is in this way confirmed. A comparison of the derivatives of the magnetization between the reported compounds and the Tm-based one is available as Supplementary information. 6. Dynamic magnetic properties The theory explaining the dynamic behavior of 1D systems involving Ising-type magnetic centers has been developed by Glauber in the 60s [15]. In particular he demonstrated that an ideal chain material should show slow relaxation of the magnetization at low temperature, with the presence of a hysteresis cycle below a blocking temperature Tb. The phenomena observed in SCMs are presently explained by this theory. Glauber’s theory predicts that the Ising chain should show exponential divergence of the relaxation time of the magnetization on decreasing the temperature. This exponential divergence
K. Bernot et al. / Inorganica Chimica Acta 360 (2007) 3807–3812
3811
7. Conclusions
Fig. 5. Comparison of the normalized temperature dependence of the imaginary v00 components of the ac susceptibility between the reported compounds and the Tm derivative. The measurement is performed in a zero applied field for 15 logarithmically spaced frequencies in the range 110 Hz (light gray) to 19 010 Hz (dark gray). Curves are ranging from the Tb derivative (up) to the Er one (bottom).
can be written as an Arrhenius law: s ¼ s0 eD=kB T , where D is four times the exchange energy between two adjacent spins. In the Glauber notation, the Ising spins can assume only ±1 values and D = 4J, with J the intra-chain exchange coupling constant. The value of D can be extracted from standard ac magnetometry but, unfortunately, 1 displays only the onset of the frequency dependence of the out-of-phase signal. As the maxima in v00 occur at lower temperatures or higher frequencies than those investigated here, no reliable estimation of D can be extracted from the measurement. The behavior of 1 is in agreement with the general feature of the family. In fact in this family of compounds only very anisotropic ions lead to a chain behaving as an SCM above 2 K. The Tm chain fits quite well in this analysis and implements our first conclusions. Therefore, the temperature range where the frequency dependence of v00 versus T signal is found, follows a general trend. The Tb derivative, which is well known to show a strong Ising anisotropy has the highest temperature range and the Er, which shows more similar gi and g^ values, has the lowest one. In order to illustrate it, the normalized out-of-phase signals versus T of the reported family plus the Tm derivative is depicted in Fig. 5. The frequency dependence of the v00 versus T starts at slightly higher temperature for Tm than for Er. Thus the trend does not follow the lanthanide series (. . .Tb, Dy, Ho, Er, Tm,. . .). Interestingly Tm and Er derivatives show also a similar field value for the step in the magnetization curve. The value of the magnetization at saturation and of the susceptibility is however significantly smaller for Tm. These two features do not appear to be correlated either with the field of the step and with the blocking temperature, suggesting that dipolar interactions are not playing any key role in this family of compounds.
In this work a new Tm-based compound is added to a family of isostructural SCMs and is magnetically studied. Very little data are available on the magnetic anisotropy of TmIII which is a non-Kramer ion and often EPR-silent [16]. Magnetic measurements confirm a sizeable magnetic anisotropy. In particular dc measurements confirm the trend observed in the position and in the width of unusual steps in magnetization curves, which is characteristic of the anisotropic RE ions. They also give access, thanks to a scaling procedure, to a value of the intra-chain exchange interaction. Dynamic measurements agree with a trend on the Tb of the family, with the Tm compound fitting between Ho and Er derivatives. It confirms the importance of the single-ion anisotropy in building one-dimensional lanthanide-based magnets. It also validates that fine tuning of the behavior of those isostructural magnets can be achieved by choosing the appropriate LnIII ion. Acknowledgements The authors thank Dr. C. Sangregorio and S. Ciattini for help with the structural characterization. We acknowledge financial support from Italian MURST (FIRB and PRIN Grants), from the EC through the Human Potential Program RTN-QUEMOLNA (MRTN-CT-2003-504880), from the NE-MAGMANET (NMP3-CT-2005-515767) and from German DFG (SPP1137). Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.ica.2006. 12.002. References [1] (a) See, for reviews: D. Gatteschi, R. Sessoli, Angew. Chem. 115 (2003) 278; (b) D. Gatteschi, R. Sessoli, Angew. Chem. Int. Ed. 42 (2003) 268; (c) D. Gatteschi, R. Sessoli, J. Villain, Molecular Nanomagnets, Oxford University Press, Oxford, 2006. [2] (a) L. Thomas, F. Lionti, R. Ballou, D. Gatteschi, R. Sessoli, B. Barbara, Nature (London) 383 (1996) 145; (b) J.R. Friedman, M.P. Sarachik, J. Tejada, R. Ziolo, Phys. Rev. Lett. 76 (1996) 3830; (c) W. Wernsdorfer, R. Sessoli, Science 284 (1999) 133; (d) W. Wernsdorfer, N. Aliaga-Alcade, D.N. Hendrickson, G. Christou, Nature 416 (2002) 406; (e) L. Bogani, R. Sessoli, M.G. Pini, A. Rettori, M.A. Novak, P. Rosa, M. Massi, M.E. Fedi, L. Guintini, A. Caneschi, D. Gatteschi, Phys. Rev. B 72 (2005) 064406; (f) L. Bogani, A. Caneschi, M. Fedi, D. Gatteschi, M. Massi, M.A. Novak, M.G. Pini, A. Rettori, R. Sessoli, A. Vindigni, Phys. Rev. Lett. 92 (2004) 207204. [3] (a) R. Lescouezec, J. Vaissermann, C. Ruiz-Perez, F. Lloret, R. Carrasco, M. Julve, M. Verdaguer, Y. Dromzee, D. Gatteschi, W. Wernsdorfer, Angew. Chem. 115 (2003) 1521;
3812
K. Bernot et al. / Inorganica Chimica Acta 360 (2007) 3807–3812
(b) R. Lescouezec, J. Vaissermann, C. Ruiz-Perez, F. Lloret, R. Carrasco, M. Julve, M. Verdaguer, Y. Dromzee, D. Gatteschi, W. Wernsdorfer, Angew. Chem. Int. Ed. 42 (2003) 1483; (c) T. Liu, D. Fu, S. Gao, Y. Zhang, H.L. Sun, G. Su, Y. Liu, J. Am. Chem. Soc. 125 (2003) 13976; (d) L.M. Toma, R. Lescouezec, F. Lloret, M. Julve, J. Vaissermann, M. Verdaguer, Chem. Commun. (2003) 1850; (e) E. Pardo, R. Ruiz-Garcia, F. Lloret, J. Faus, M. Julve, Y. Journaux, F. Delgado, C. Ruiz-Perez, Adv. Mater. 16 (2004) 1597; (f) A. Maignan, V. Hardy, S. Hebert, M. Drillon, M.R. Lees, O. Petrenko, D.M. Paul, D. Khomskii, J. Mater. Chem. 14 (8) (2004) 1231; (g) M. Ferbinteanu, H. Miyasaka, W. Wernsdorfer, K. Nakata, K. Sugiura, M. Yamashita, C. Coulon, R. Clerac, J. Am. Chem. Soc. 127 (9) (2005) 3090; (h) J.P. Costes, J.M. Clemente-Juan, F. Dahan, J. Milon, Inorg. Chem. 43 (2004) 8200; (i) L.M. Toma, R. Lescoue¨zec, J. Pasa`n, C. Ruiz-Pe´rez, J. Vaissermann, J. Cano, R. Carrasco, W. Wersndorfer, F. Lloret, M. Julve, J. Am. Chem. Soc. 128 (14) (2006) 4842. [4] R. Clerac, H. Miyasaka, M. Yamashita, C. Coulon, J. Am. Chem. Soc. 124 (2002) 12837. [5] (a) A. Caneschi, D. Gatteschi, P. Rey, Prog. Inorg. Chem. 39 (1991) 331; (b) D. Luneau, P. Rey, Coord. Chem. Rev. 249 (2005) 2591, and references therein. [6] A. Caneschi, D. Gatteschi, N. Lalioti, C. Sangregorio, R. Sessoli, G. Venturi, A. Vindigni, A. Rettori, M.G. Pini, M.A. Novak, Angew. Chem. 113 (2001) 1810.
[7] A. Caneschi, D. Gatteschi, N. Lalioti, C. Sangregorio, R. Sessoli, G. Venturi, A. Vindigni, A. Rettori, M.G. Pini, M.A. Novak, Angew. Chem. Int. Ed. 40 (2001) 1760. [8] L. Bogani, C. Sangregorio, R. Sessoli, D. Gatteschi, Angew. Chem. Int. Ed. 44 (2005) 5817. [9] K. Bernot, L. Bogani, A. Caneschi, D. Gatteschi, R. Sessoli, J. Am. Chem. Soc. 128 (2006) 7947. [10] E.F. Ullman, J.H. Osiecki, D.G.B. Boocock, R. Darcy, J. Am. Chem. Soc. 94 (1972) 7049. [11] S. Midollini, A. Orlandini, P. Rosa, L. Sorace, Inorg. Chem. 44 (2005) 2060. [12] (a) C. Benelli, D. Gatteschi, Chem. Rev. 102 (2002) 2369; (b) C. Benelli, D. Gatteschi, A. Caneschi, R. Sessoli, Inorg. Chem. 32 (1993) 4797; (c) C. Benelli, A. Caneschi, D. Gatteschi, R. Sessoli, J. Appl. Phys. 73 (1993) 5333; (d) C. Benelli, A. Caneschi, D. Gatteschi, D.L. Pardi, P. Rey, Inorg. Chem. 29 (1990) 4223. [13] R.L. Carlin, Magnetochemistry, Springer-Verlag, New York, 1986. [14] (a) C. Coulon, R. Clerac, L. Lecren, W. Wernsdorfer, H. Miyasaka, Phys. Rev. B 69 (2004) 132408; (b) M. Ferbinteanu, H. Miyasaka, W. Wernsdorfer, K. Nakata, K. Sugiura, M. Yamashita, C. Coulon, R. Clerac, J. Am. Chem. Soc. 127 (2005) 3090. [15] R.J. Glauber, J. Math. Phys. 4 (1963) 294. [16] A. Abragam, B. Bleaney, Electron Paramagnetic Resonance of Transition Ions, Dover Publication, New York, 1986.
[TmIII(hfac)3(NITPhOPh)]1: A new member of a lanthanide-based Single Chain Magnets family K. Bernot a
a,b
, L. Bogani a, R. Sessoli
a,*
, D. Gatteschi
a
Department of Chemistry and INSTM (Udr Firenze), Universita` degli Studi di Firenze, Via della Lastruccia 3, 50019 Sesto Fiorentino, Italy b Sciences Chimiques de Rennes, UMR 6226 CNRS - INSA Rennes, Equipe ‘‘Mate´riaux Inorganiques : Chimie Douce et Re´activite´’’, INSA-Rennes, 20 Avenue des buttes de Coe¨smes, CS 14315, 35043 Rennes Cedex, France Received 21 October 2006; accepted 3 December 2006 Available online 12 December 2006
Abstract The polymeric coordination compound of formula [Tm(hfac)3(NITPhOPh)]1 (where NITPhOPh is a nitronyl-nitroxide radical) has been synthesized and found to belong to the only reported family of isostructural Single Chain Magnets. Both static and dynamic magnetic measurements have been performed, and a dependence of the out-of-phase signal on the frequency is observed below 3 K. Scaling procedures indicate Ising magnetic anisotropy. Comparison of the extracted parameters with those of the previously reported isostructural compounds confirms a trend along the lanthanide series. 2006 Elsevier B.V. All rights reserved. Keywords: Nitronyl-nitroxides; Single chain magnets; Lanthanides; Magnetic anisotropy
1. Introduction For years the only known class of superparamagnetic molecular materials was that of the zero-dimensional derivatives called single molecule magnets, or SMMs [1]. The observation, a few years ago, of slow relaxation of the magnetization in 1D-materials has opened a new field of great interest for both physicists [2] and chemists [3]. In analogy to SMMs these new compounds have been called Single Chain Magnets (SCMs) [4]. Slow dynamics of the magnetization in anisotropic 1D-systems was originally predicted by Glauber in the 60s [5] and his model has since then been applied to a variety of different systems. The two key requirements of the model, that need to be satisfied in the realization of SCMs, are a strong Ising anisotropy of the magnetic centers and a very low ratio between inter-chain and intra-chain interactions. It is thus interesting to investigate how the variation of both these requirements in a *
Corresponding author. Tel.: +39 0 554573268; fax: +39 0 554573372. E-mail address: roberta.sessoli@unifi.it (R. Sessoli).
0020-1693/$ - see front matter 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.ica.2006.12.002
family of related systems affects the magnetic behavior. In this context the use of the chemical tuning offered by nitronyl-nitroxide radicals has provided major advances [5] and, in particular, linear chains based on Cobalt(II) [6,7] and Dysprosium(III) ions [8] have been obtained. In the last case the use of a bulky R substituent on the radical has reduced inter-chain dipolar interactions preventing the onset of three-dimensional magnetic order. In a previous work we showed how lanthanide-based SCMs can be used to investigate the role of anisotropy on the dynamics, by designing a family in which the substitution of the magnetic ion leads to a modulation of the magnetic anisotropy without structural modifications [9]. Our previous study was performed on chains containing the rare earths (REs), Eu, Gd, Tb, Dy, Ho, Er, Yb, for which a non-monotonous trend in the magnetic anisotropy is observed. The study was performed on chains of formula [M(hfac)3(NITPhOPh)]1, where M= Eu, Gd, Tb, Dy, Ho, Er, Yb, and NITPhOPh is the nitronyl-nitroxide radical 2,4 0 -benzoxo-4,4,5,5-tetramethyl-imidazoline-1-oxyl-3oxide, depicted in. In this family, despite the internal dis-
3808
K. Bernot et al. / Inorganica Chimica Acta 360 (2007) 3807–3812
position of f-electrons, which leads to a weak intra-chain coupling, a very remarkable behavior can be observed. The contribution to the energy barrier of the strong Ising anisotropy significantly compensates the weak exchange interaction that characterizes 4f-ions. A very good isolation of the chain, and the huge anisotropy provided by LnIII ions give rise to a SCM behavior for the Tb, Dy and Ho derivatives. Both static and dynamic properties of the system are affected by the change of the magnetic ion. From the static magnetic properties, effects are seen on the intra-chain correlation length at low temperature and on the presence of steps in the magnetization curves. The most interesting feature is displayed by the dynamic behavior of the family. Peaks in v00 versus T curves appear and shift in temperature as the anisotropy of the ion used changes. As very little is known about the magnetic anisotropy of TmIII ion in low symmetry environment, in this work we study the Tm-based chain 1, [Tm(hfac)3(NITPhOPh)]. We thus complete the family of heavy 4f-SCMs and give a last evidence of the crucial influence of the anisotropy of Ising ions in SCMs. 2. Synthesis
Fig. 1. Powder X-ray spectrum of the chain [Tm(hfac)3NITPhOPh]1 compound (upper graph) and simulated spectrum using data from the reported system [8]. In the inset we show the view of the crystal structure of [Dy(hfac)3NITPhOPh]1. Fluorine and hydrogen atoms are omitted for clarity.
needle-like crystals. Anal. Calc. for C34H24O9N2F18Tm: C, 36.61; H, 2.17; O, 12.91. Found: C, 36.58; H, 2.13; O, 12.93%.
2.1. General procedures and materials
2.3. Crystallography
All chemical and solvents were used in reagent grades. The radical 2,4 0 -benzoxo-4,4,5,5-tetramethyl-imidazoline1-oxyl-3-oxide, named NITPhOPh in the following, has been synthesized according to the literature methods [10].
A single crystal of [Tm(hfac)3NITPhOPh1 was characterized with a cell measurement, revealing that the main crystallographic parameters are: a = 14.400(4), b = 16.681(2), c = 17.120(8), a = 90.03(7), b = 89.96(5), c = 89.95(9), at room temperature. As the isostructural Dybased compound has already been reported by some of us [8], we did not proceed to a full structure determination. Anyway, as magnetic measurements were performed on microcrystalline powders, a powder X-ray diffraction pattern was recorded at room temperature and superimposed to the simulated powder pattern of the [Dy(hfac)3NITPhOPh]1. The two spectra, which match very well, are depicted in Fig. 1, together with the view of the structure of the polymeric compound.
2.2. Synthesis of [Tm(hfac)3(NITPhOPh)]1 The reacting salt [Ln(hfac)3 Æ 2H2O] was synthesized using the following procedure: 1.78 ml (24 mmol) of a 25% ammonia solution was added drop-wise to 3.39 ml (24 mmol) of hexafluoroacetylacetone (hfac) in 100 ml of ether at 2 C. Then 8 mmol of [Tm(Cl)3 Æ 5H2O] in 10 ml of water were added to the solution. After half an hour of strong stirring, 20 ml of water were added. Then organic and aqueous phases were separated and washed with water and ether, respectively. The organic phase was dried on magnesium sulfate, filtered, and then concentrated to form yellowish oil. Twenty milliliter of hexane was then added and the solution was then heated up to 55 C for some minutes. After filtration, in order to separate the formed solid, the compound was kept in the freezer at 15 C, where it crystallized. The mean overall yield was about 50%. The chain [Tm(hfac)3NITPhOPh]1 was synthesized by the following procedure: 1 mmol of [Tm(hfac)3 Æ 2H2O] was dissolved in 30 ml of dry boiling n-heptane. The solution was left to boil for 20 min and then cooled down to 75C, when 1 mmol of the crystalline solid radical NITPhOPh was added under stirring together with 3 ml of CH2Cl2. The final solution was then cooled down to room temperature and was left still for about 24 h, to give dark
3. Physical measurements Temperature-dependent dc magnetic susceptibility measurements were performed on solid polycrystalline samples with a Cryogenic Ltd. S600 SQUID magnetometer and were corrected for the diamagnetic contribution, as calculated with Pascal’s constants, and for the diamagnetism of the sample holder, as independently determined. Magnetization curves were measured with an Oxford VSM system after inclusion in Apiezon grease to prevent orientation of the crystallites in the applied magnetic field. Data were corrected for the magnetism of the grease, which was independently determined at the same temperature and fields. A 12 kOe/min field sweep rate was used, while acquiring continuously. The ac magnetic susceptibility was measured
K. Bernot et al. / Inorganica Chimica Acta 360 (2007) 3807–3812
3809
using a homemade probe [11]. Powder X-ray diffraction patterns were recorded on a Powder diffractometer Bruker D8 Advance. Powder diffraction pattern of Dy compound was simulated from CIF file using Powder Cell software. 4. Magnetic characterization 4.1. Static magnetic measurements The vT product versus T was recorded at temperatures between 2 and 300 K and is displayed in Fig. 2. The compound has a room temperature susceptibility of 7.60 emu K/mol, slightly lower than that expected for one TmIII ion, which has a 3H6 ground state, plus an uncoupled S = 1/2 radical center. On cooling down, the value is constant until 130 K where a slight decrease starts, then becomes abrupt around 30 K. The curve reaches its lower value, 4.67 emu K/mol, at 3.4 K. It then markedly increases to reach 7.02 emu K/mol at 2 K. Fig. 3 shows the magnetization versus field curve at 1.55 K. No hysteresis was observed down to 1.55 K. Nevertheless the compound displays a non-trivial behavior at low field, where a slight step is visible. In order to better appreciate this feature, the derivative of the magnetization is depicted in the inset. The step is found around 6.10 kOe. One can notice that this step becomes sharper on cooling and is independent from the magnetic field sweep rate. The magnetization value is 23920 emu/mol (4.28 lB) at 120 kOe, lower than the expected 8 lB value. This is, however, in agreement with previous results on similar compounds [8,9,12]. 4.2. Dynamic magnetic measurements As first investigation of the dynamical behavior of 1, zero-field-cooled and field-cooled (ZFC–FC) curves were recorded in low temperature region and are found to be perfectly superimposable, thus indicating that down to
Fig. 3. Magnetization versus field measurement for the [Tm(hfac)3(NITPhOPh)]1 chain as a polycrystalline sample at 1.55 K and with a field sweep rate of 12 kOe/min. In the inset we show an enlargement of the derivative in the low field region.
Fig. 4. Temperature dependence of the imaginary v00 and real v 0 components of the ac susceptibility measured in zero applied field for 15 logarithmically spaced frequencies in the range 110 Hz (light gray) to 19 010 Hz (dark gray). In the inset the imaginary component v00 vs. T is shown, depicting only the eight highest frequency for clarity. The lines are a guide to the eye.
the lowest investigated temperature the relaxation time is shorter than a few seconds. Measurements of ac susceptibility were performed with 15 logarithmically spaced frequencies in the 110– 19 010 Hz range between 1.55 and 7 K, in the zero external field. The temperature dependence of both in-phase (v 0 ) and out-of-phase signals (v00 ) is depicted in Fig. 4. v 0 ranges from 1.23 emu mol1 at 5 K to 9.96 emu mol1 at 1.55 K for the highest frequency. The v 0 /v00 ratio is rather high but, in spite of that, some frequency dependence of the out-of phase susceptibility (inset of Fig. 4) can be found under 4 K. For the sake of clarity only the eight highest frequencies are depicted for v 0 . 5. Results and discussion
Fig. 2. Temperature dependence of the vT product extracted from SQUID magnetic measurements. Symbols are experimental data (empty circles are recorded in an external field of 0.1 kOe to avoid saturation effects, squares in 1 kOe and half-filled triangles in 10 kOe). In the inset is reported the scaling of the data using an Ising temperature dependence of the susceptibility (see text) and plotting ln (vT mol K1 emu1) vs. 1/T. The solid line represents the linear fit in the 2.0–2.5 K temperature range.
5.1. Static properties The association of lanthanides and nitronyl-nitroxide radicals has led to a huge variety of compounds [5b] and has been widely investigated from the magnetic point of view [12a]. In those systems the simulation of magnetic
3810
K. Bernot et al. / Inorganica Chimica Acta 360 (2007) 3807–3812
Table 1 Main magnetic data are extracted from the static properties of the chain compounds Rare earth
vT@300 K (emu K/mol)
M@120 kOe (emu/mol)
Step field (kOe)
Steepness @ Step (emu mol1 Oe1)
Slope (Sl) (K)
Tb**
11.10 12.2 14.4 11.37 7.60
26 700 30 900 34 300 32 000 23 920
13.77 16.23 13.08 6.26 6.10
1310 1680 1080 790 1065
9.1 ± 0.5 18.2 ± 0.3 3.4 ± 0.3 4.6 ± 0.2 3.7 ± 0.3
Dy* Ho** Er** Tm
The field value of the steps and their steepness are extracted from the first derivatives of the magnetization curves. The slope is extracted from the logarithmic scaling of the vT curves. Parameters are extracted from data reported in Refs. [8]* and [10]**.
data is hampered by the peculiarity of the ground states of the rare-earth ions: their orbitally degenerate states show strong spin–orbit coupling. The ground multiplet, characterized by J = L S for fn configurations with n < 7 and J = L + S when n > 7, is split by the crystal field. That is why in the last decade the attention mainly focused on GdIII and EuII derivatives, which have a f7 configuration and an orbitally non-degenerate ground state. Nevertheless the orbital component, while complicating the analysis of the magnetic properties, is rather important, as it also affords considerable anisotropy to those ions. They are therefore good candidates for designing one-dimensional magnetic materials where the anisotropy can be tuned at will, thus offering the possibility of exploring its key role in the dynamics of Ising systems. The case of the Tmbased compound is discussed hereafter and comparison with already reported values for the family is made in Table 1. The dc magnetic measurements show the expected behavior for this type of compounds. The first effect is due to the crystal field, i.e on cooling down the system, the vT product decreases, as the Stark sub-levels of the TmIII are progressively depopulated. The following abrupt increase of the vT product at low temperature is associated with the intra-chain coupling involving both TmIII ions and the S = 1/2 radicals, and is connected to the growth of the magnetic correlation length n inside the chain. Previous investigations of the static properties of chains designed with nitronyl-nitroxide radicals and rare-earth ions reveal the simultaneous presence of both NearestNeighbor (NN) and Next-Nearest-Neighbor (NNN) exchange interactions between the magnetic centers [12a], with antiferromagnetic NNN interactions. Data on the magnetic anisotropy of TmIII ions are extremely scarce, given that it is a non-Kramers ion and often EPR-silent, except for a few reports suggesting an Ising-like character of TmIII ion [13]. The anisotropy varies widely depending on the structural conformation of the ligands around the magnetic ion, and is hardly modellizable in low-symmetry environments. Moreover in nitronyl-nitroxide based compounds it is not possible to separate crystal field effects from the ones due to the exchange interaction with the spin of the coordinated radicals. Thus, to obtain some information about the kind of anisotropy involved in our compound, we choose to perform a scaling procedure of the low temperature data, in the region where n starts to grow
[14]. This is due to the fact that the susceptibility of a ferromagnetic Ising chain follows the growth of the correlation and, for an Ising system an exponential behavior is expected [15]: vT / n ¼ e2JS i S iþ1 =kB T where J is the intra-chain exchange energy, and with Si we indicated the spin coordinates. Thus, if the system has a truly Ising nature, plotting the logarithm of vT versus T1 should afford a straight line, with the slope directly providing the exchange energy. In our case a straight line is found for very low temperatures (between 2 and 2.5 K) as reported in the inset of Fig. 2. The linear fit gives a slope of 3.7 ± 0.3 K with an agreement factor R = 0.9994. This is comparable with reported values observed on Ho and Er derivatives (3.4 ± 0.3 K and 4.6 ± 0.2 K, respectively) and suggests a weak exchange interaction along the chain. The magnetization curve recorded as a function of the applied field is displayed in Fig. 3. Like other compounds based on anisotropic ions in the family, the Tm chain presents steps, clearly evidenced by the plot of the derivative of the magnetization, as reported in the inset. The Gd and Yb chains on the contrary do not show such an anomaly. The step for Tm is found at 6.10 kOe, close to the one of Er compound (6.25 kOe). It is however sharper, with a steepness of 1065 emu mol1 Oe1, close to the value found for the Ho derivative. As shown in Table 1, the trend in the steepness previously observed in the reported family is in this way confirmed. A comparison of the derivatives of the magnetization between the reported compounds and the Tm-based one is available as Supplementary information. 6. Dynamic magnetic properties The theory explaining the dynamic behavior of 1D systems involving Ising-type magnetic centers has been developed by Glauber in the 60s [15]. In particular he demonstrated that an ideal chain material should show slow relaxation of the magnetization at low temperature, with the presence of a hysteresis cycle below a blocking temperature Tb. The phenomena observed in SCMs are presently explained by this theory. Glauber’s theory predicts that the Ising chain should show exponential divergence of the relaxation time of the magnetization on decreasing the temperature. This exponential divergence
K. Bernot et al. / Inorganica Chimica Acta 360 (2007) 3807–3812
3811
7. Conclusions
Fig. 5. Comparison of the normalized temperature dependence of the imaginary v00 components of the ac susceptibility between the reported compounds and the Tm derivative. The measurement is performed in a zero applied field for 15 logarithmically spaced frequencies in the range 110 Hz (light gray) to 19 010 Hz (dark gray). Curves are ranging from the Tb derivative (up) to the Er one (bottom).
can be written as an Arrhenius law: s ¼ s0 eD=kB T , where D is four times the exchange energy between two adjacent spins. In the Glauber notation, the Ising spins can assume only ±1 values and D = 4J, with J the intra-chain exchange coupling constant. The value of D can be extracted from standard ac magnetometry but, unfortunately, 1 displays only the onset of the frequency dependence of the out-of-phase signal. As the maxima in v00 occur at lower temperatures or higher frequencies than those investigated here, no reliable estimation of D can be extracted from the measurement. The behavior of 1 is in agreement with the general feature of the family. In fact in this family of compounds only very anisotropic ions lead to a chain behaving as an SCM above 2 K. The Tm chain fits quite well in this analysis and implements our first conclusions. Therefore, the temperature range where the frequency dependence of v00 versus T signal is found, follows a general trend. The Tb derivative, which is well known to show a strong Ising anisotropy has the highest temperature range and the Er, which shows more similar gi and g^ values, has the lowest one. In order to illustrate it, the normalized out-of-phase signals versus T of the reported family plus the Tm derivative is depicted in Fig. 5. The frequency dependence of the v00 versus T starts at slightly higher temperature for Tm than for Er. Thus the trend does not follow the lanthanide series (. . .Tb, Dy, Ho, Er, Tm,. . .). Interestingly Tm and Er derivatives show also a similar field value for the step in the magnetization curve. The value of the magnetization at saturation and of the susceptibility is however significantly smaller for Tm. These two features do not appear to be correlated either with the field of the step and with the blocking temperature, suggesting that dipolar interactions are not playing any key role in this family of compounds.
In this work a new Tm-based compound is added to a family of isostructural SCMs and is magnetically studied. Very little data are available on the magnetic anisotropy of TmIII which is a non-Kramer ion and often EPR-silent [16]. Magnetic measurements confirm a sizeable magnetic anisotropy. In particular dc measurements confirm the trend observed in the position and in the width of unusual steps in magnetization curves, which is characteristic of the anisotropic RE ions. They also give access, thanks to a scaling procedure, to a value of the intra-chain exchange interaction. Dynamic measurements agree with a trend on the Tb of the family, with the Tm compound fitting between Ho and Er derivatives. It confirms the importance of the single-ion anisotropy in building one-dimensional lanthanide-based magnets. It also validates that fine tuning of the behavior of those isostructural magnets can be achieved by choosing the appropriate LnIII ion. Acknowledgements The authors thank Dr. C. Sangregorio and S. Ciattini for help with the structural characterization. We acknowledge financial support from Italian MURST (FIRB and PRIN Grants), from the EC through the Human Potential Program RTN-QUEMOLNA (MRTN-CT-2003-504880), from the NE-MAGMANET (NMP3-CT-2005-515767) and from German DFG (SPP1137). Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.ica.2006. 12.002. References [1] (a) See, for reviews: D. Gatteschi, R. Sessoli, Angew. Chem. 115 (2003) 278; (b) D. Gatteschi, R. Sessoli, Angew. Chem. Int. Ed. 42 (2003) 268; (c) D. Gatteschi, R. Sessoli, J. Villain, Molecular Nanomagnets, Oxford University Press, Oxford, 2006. [2] (a) L. Thomas, F. Lionti, R. Ballou, D. Gatteschi, R. Sessoli, B. Barbara, Nature (London) 383 (1996) 145; (b) J.R. Friedman, M.P. Sarachik, J. Tejada, R. Ziolo, Phys. Rev. Lett. 76 (1996) 3830; (c) W. Wernsdorfer, R. Sessoli, Science 284 (1999) 133; (d) W. Wernsdorfer, N. Aliaga-Alcade, D.N. Hendrickson, G. Christou, Nature 416 (2002) 406; (e) L. Bogani, R. Sessoli, M.G. Pini, A. Rettori, M.A. Novak, P. Rosa, M. Massi, M.E. Fedi, L. Guintini, A. Caneschi, D. Gatteschi, Phys. Rev. B 72 (2005) 064406; (f) L. Bogani, A. Caneschi, M. Fedi, D. Gatteschi, M. Massi, M.A. Novak, M.G. Pini, A. Rettori, R. Sessoli, A. Vindigni, Phys. Rev. Lett. 92 (2004) 207204. [3] (a) R. Lescouezec, J. Vaissermann, C. Ruiz-Perez, F. Lloret, R. Carrasco, M. Julve, M. Verdaguer, Y. Dromzee, D. Gatteschi, W. Wernsdorfer, Angew. Chem. 115 (2003) 1521;
3812
K. Bernot et al. / Inorganica Chimica Acta 360 (2007) 3807–3812
(b) R. Lescouezec, J. Vaissermann, C. Ruiz-Perez, F. Lloret, R. Carrasco, M. Julve, M. Verdaguer, Y. Dromzee, D. Gatteschi, W. Wernsdorfer, Angew. Chem. Int. Ed. 42 (2003) 1483; (c) T. Liu, D. Fu, S. Gao, Y. Zhang, H.L. Sun, G. Su, Y. Liu, J. Am. Chem. Soc. 125 (2003) 13976; (d) L.M. Toma, R. Lescouezec, F. Lloret, M. Julve, J. Vaissermann, M. Verdaguer, Chem. Commun. (2003) 1850; (e) E. Pardo, R. Ruiz-Garcia, F. Lloret, J. Faus, M. Julve, Y. Journaux, F. Delgado, C. Ruiz-Perez, Adv. Mater. 16 (2004) 1597; (f) A. Maignan, V. Hardy, S. Hebert, M. Drillon, M.R. Lees, O. Petrenko, D.M. Paul, D. Khomskii, J. Mater. Chem. 14 (8) (2004) 1231; (g) M. Ferbinteanu, H. Miyasaka, W. Wernsdorfer, K. Nakata, K. Sugiura, M. Yamashita, C. Coulon, R. Clerac, J. Am. Chem. Soc. 127 (9) (2005) 3090; (h) J.P. Costes, J.M. Clemente-Juan, F. Dahan, J. Milon, Inorg. Chem. 43 (2004) 8200; (i) L.M. Toma, R. Lescoue¨zec, J. Pasa`n, C. Ruiz-Pe´rez, J. Vaissermann, J. Cano, R. Carrasco, W. Wersndorfer, F. Lloret, M. Julve, J. Am. Chem. Soc. 128 (14) (2006) 4842. [4] R. Clerac, H. Miyasaka, M. Yamashita, C. Coulon, J. Am. Chem. Soc. 124 (2002) 12837. [5] (a) A. Caneschi, D. Gatteschi, P. Rey, Prog. Inorg. Chem. 39 (1991) 331; (b) D. Luneau, P. Rey, Coord. Chem. Rev. 249 (2005) 2591, and references therein. [6] A. Caneschi, D. Gatteschi, N. Lalioti, C. Sangregorio, R. Sessoli, G. Venturi, A. Vindigni, A. Rettori, M.G. Pini, M.A. Novak, Angew. Chem. 113 (2001) 1810.
[7] A. Caneschi, D. Gatteschi, N. Lalioti, C. Sangregorio, R. Sessoli, G. Venturi, A. Vindigni, A. Rettori, M.G. Pini, M.A. Novak, Angew. Chem. Int. Ed. 40 (2001) 1760. [8] L. Bogani, C. Sangregorio, R. Sessoli, D. Gatteschi, Angew. Chem. Int. Ed. 44 (2005) 5817. [9] K. Bernot, L. Bogani, A. Caneschi, D. Gatteschi, R. Sessoli, J. Am. Chem. Soc. 128 (2006) 7947. [10] E.F. Ullman, J.H. Osiecki, D.G.B. Boocock, R. Darcy, J. Am. Chem. Soc. 94 (1972) 7049. [11] S. Midollini, A. Orlandini, P. Rosa, L. Sorace, Inorg. Chem. 44 (2005) 2060. [12] (a) C. Benelli, D. Gatteschi, Chem. Rev. 102 (2002) 2369; (b) C. Benelli, D. Gatteschi, A. Caneschi, R. Sessoli, Inorg. Chem. 32 (1993) 4797; (c) C. Benelli, A. Caneschi, D. Gatteschi, R. Sessoli, J. Appl. Phys. 73 (1993) 5333; (d) C. Benelli, A. Caneschi, D. Gatteschi, D.L. Pardi, P. Rey, Inorg. Chem. 29 (1990) 4223. [13] R.L. Carlin, Magnetochemistry, Springer-Verlag, New York, 1986. [14] (a) C. Coulon, R. Clerac, L. Lecren, W. Wernsdorfer, H. Miyasaka, Phys. Rev. B 69 (2004) 132408; (b) M. Ferbinteanu, H. Miyasaka, W. Wernsdorfer, K. Nakata, K. Sugiura, M. Yamashita, C. Coulon, R. Clerac, J. Am. Chem. Soc. 127 (2005) 3090. [15] R.J. Glauber, J. Math. Phys. 4 (1963) 294. [16] A. Abragam, B. Bleaney, Electron Paramagnetic Resonance of Transition Ions, Dover Publication, New York, 1986.