Slow Magnetic Relaxation in Uranium(III)

Communication
pubs.acs.org/Organometallics
Slow Magnetic Relaxation in Uranium(III) and Neodymium(III)
Cyclooctatetraenyl Complexes
Jennifer J. Le Roy, Serge I. Gorelsky, Ilia Korobkov, and Muralee Murugesu*
Chemistry Department and the Centre for Catalysis Research and Innovation, University of Ottawa, 10 Marie Curie, Ottawa, ON
K1N 6N5, Canada
S Supporting Information
*
ABSTRACT: The synthesis, structure, and magnetic properties of
a uranium(III) sandwich complex, [Li(DME)3][UIII(COT″)2]
(COT″ = bis(trimethylsilyl)cyclooctatetraenyl dianion), and its
coordinatively analogous tetravalent equivalent, [UIV(COT″)2],
were investigated. Additionally, a full structural and magnetic
comparison to the isostructural and isoelectronic lanthanide
complex, [Li(DME)3][NdIII(COT″)2], is provided. DFT calculations reveal that the UIII complex leads to weaker ligand-to-metal
donation as compared with the tetravalent equivalent complex.
Alternating current magnetic susceptibility results reveal slow
magnetic relaxation in both UIII and NdIII complexes. The enhanced magnetic performance of the UIII congener further
encourages the use of actinides in the design of single-molecule magnets.
important to explore how new ligand fields effect the magnetic
properties of actinides.
With this in mind we have turned our attention toward
uranocene-type sandwich complexes. The influence of a
sandwich-type ligand field is unknown on the SMM properties
of uranium. Moreover, strong ligand donation has been well
established in uranocene [UVI(COT)2], which is a desirable
property in a magnetic building block.7 However, magnetically
UIII is preferable to UIV, where UIII has been recognized to have
the required components such as anisotropy for attaining SMM
behavior.5g Here, we report the use of 1,4-bis(trimethylsilyl)cyclooctatetraenyl dianion (COT″) as the ligand to isolate
uranocene-type molecules with unsymmetrical coordination
geometries and multiple oxidation states. We describe the
synthesis, structure, and magnetic properties of [Li(DME)3][UIII(COT″)2], 1, and its coordinatively analogous tetravalent
equivalent, [UIV(COT″)2], 2. The magnetic properties of several
lanthanide-COT″ complexes have been established, with the ErIII
analogue exhibiting a high blocking temperature of 8 K.2a,6a,8
Isostructural and isoelectronic complexes in the lanthanide and
actinide series provide a unique opportunity to elucidate subtle
differences such as how atomic mass (z) influences magnetic
performance. Therefore, in addition to the computational study
on the metal−ligand covalency of complexes 1 and 2, the
magnetic properties of 1 are compared to structural analogue
[NdIII(COT″)2][Li(DME)3], 3.
Uranocene is unreactive to lithiation,9 and all reported
trivalent uranocene-type complexes have been prepared via
reduction of [UIV(COT-R)2] (R = CH3,10 1,4-(tBuMe2Si)211) by
T
remendous effort has been put forth toward improving
lanthanide single-molecule magnets (SMMs) for their use
in magnetic materials.1 Many of the best candidates thus far have
been single-ion lanthanide complexes, which display remarkable
magnetic properties due to unquenched orbital angular
momentum.2 However, the operating temperature of even the
best lanthanide SMMs is well below what is required for practical
applications. One of the main challenges toward higher
temperature lanthanide SMMs is synthesizing magnetic
complexes with strong superexchange-type metal−metal interactions, where metals are strongly coupled through diamagnetic
bridging ligands. This requires strong metal−ligand covalency,
which is challenging due to the poor radial extension of 4f
orbitals.
Uranium complexes have tremendous potential as SMMs, as
they possess the key physical properties of large intrinsic total
ground state spin (S) and more importantly uniaxial magnetic
anisotropy (D) required for magnet-like behavior of slow
relaxation of the magnetization.3 Akin to lanthanide ions, the
heavy-element nature of uranium can additionally result in
significant spin−orbit coupling constants.4 However, unlike
lanthanides, actinides have enhanced 5f radial extension, which
leads to substantial metal−ligand covalency, allowing for strong
exchange coupling (J) between metals in multimetallic
complexes.
Despite the many obvious advantages, only a handful of
actinide SMMs have been reported, none of which have shown
enhance magnetic properties over the best performing
lanthanide SMMs.5 It is well established that even subtle
variations in the ligand field can drastically change the magnetic
properties of lanthanide SMMs.6 Therefore, in order to isolate
new uranium SMMs with large anisotropic barriers, it is
© 2015 American Chemical Society
Received: November 29, 2014
Published: April 7, 2015
1415
DOI: 10.1021/om501214c
Organometallics 2015, 34, 1415−1418
Communication
Organometallics
(nearest C−H distance 2.76 Å (1); 2.74 Å (3)) undoubtedly
contributes to the difference in tilt angle (Figure S5).
Also of interest, 2 is only the second example of a uranocene
with a bent structure where the (ring centroid)−U−(ring
centroid) angle deviates from perfect linearity (180°) by 6.5°.14
Lorenz et al. recently reported a uranocene with bulky
[C8H6(SiPh3)2]2− ligands leading to a large 11.3° (ring
centroid)−U−(ring centroid) linear deviation.14 The less
sterically bulky TMS groups account for the reduced bend in 2
in comparison to SiPh3 groups. Interestingly complex 1 has a
(ring centroid)−U−(ring centroid) angle of 172.8°; we can
therefore conclude that the oxidation has little effect of the bend
angle.
To fully understand the consequence of oxidation state on
uranium−COT″ bonding, the electronic structure of 1 and 2 was
probed using DFT calculations conducted at the spinunrestricted B3LYP/TZVP level of theory (the SDD basis set
and effective core potential for the U atom). Single-point
calculations were conducted using Gaussian 09 with the crystal
structure geometries in which COT″ and methyl C−H bond
distances were adjusted from the X-ray model values to 1.07 and
1.08 Å, respectively. The optimized wave functions for the
ground states (with S = 1 for UIV and S = 3/2 for UIII) were used
to evaluate bonding contributions. Figure S6 illustrates the spin
density distribution of the electronic ground states.
In complex 1, the UIII atom carries a spin density of 3.026 au
due to three singly occupied 5f orbitals of the UIII ion (α-spin
HOMO−2, HOMO−1, and the HOMO of the complex, Figure
S7), while the COT″ ligands demonstrate weak spin polarization
(a spin density of −0.013 au per ligand). Overall, each dianionic
COT″ ligand donates ∼0.92 e− to UIII, resulting in the +1.16 au
charge for the U atom. The Mayer bond order between UIII and
each COT″ ligand is 1.73; α- and β-spin occupied orbitals
contribute 0.98 and 0.75, respectively, to the total metal−ligand
bond order. The analysis of the wave function in terms of
contributions from fragment orbitals (the metal cation being one
fragment and the two anionic ligands being the other two
fragments) indicate that only charge donation from the COT″
ligands to the UIII contribute to the covalent bonding in complex
1. Five occupied π orbitals of the COT″2− ligands (Figure S8),
namely, the highest occupied fragment orbital (HOFO),
HOFO−1, HOFO−2, HOFO−3, and HOFO−7, participate
significantly (change in orbital population is greater than 3%) in
covalent bonding with the UIII ion. The changes in the
populations of these fragment molecular orbitals are 11.3%,
11.5%, 3.3%, 4.1%, and 6.8% for α-spin manifold and 8.0%, 8.2%,
3.0%, 3.7%, and 6.2% for β-spin manifold, respectively. The
analysis of populations of UIII atomic orbitals (Table S2)
indicates that metal s, p, d, and f unoccupied orbitals participate
in ligand-to-metal donation, with the 6d orbitals receiving almost
half of the total electron density (1.84 e−) transferred from the
two ligands.
In 2, the UIV atom carries a spin density of 2.306 au due to two
singly occupied 5f orbitals of the UIV (α-spin HOMO−3 and the
HOMO of the complex, Figure S7), while the COT″ ligands
demonstrate a more significant spin polarization (a spin density
of only −0.153 au per ligand). This is due to a difference in charge
donation from the dianionic COT″ ligands to UIV through α- and
β-spin orbitals (see below). Overall, each dianionic COT″ ligand
donates ∼1.5 e− to UIV, resulting in the +1.01 au charge for the
UIV atom. The Mayer bond order between UIV and each COT″
ligand is 2.42, showing a significant increase in the metal−ligand
covalency as compared to the corresponding UIII complex. The
potassium metal. Our synthetic strategy is a direct method where
reaction of [UIIII3(1,4-dioxane)1.5]12 with
[Li4(COT″)2(THF)4]8a in DME (dimethoxyethane) yields 1,
the first [UIII(COT”)2]− lithium salt. Single crystals of 1 suitable
for X-ray diffraction were grown from a 50:50 mixture of DME/
hexanes. Complex 2 was produced in a two-step synthesis where
[Li4(COT″)2(THF)4]8a and [UIIII3(1,4-dioxane)1.5]12 were first
combined in THF to produce [UIII(COT″)2][Li(THF)4],
following which the metal was oxidized to UIV using FeCl2.
Upon concentrating the solution, large green block crystals of 2
were isolated, suitable for single-crystal X-ray diffraction.
[NdIII(COT″)2][Li(THF)4], 3, was synthesized with minor
modifications to a previously published method.13 Single crystals
of 3 were grown in a concentrated solution of a 50:50 mixture of
DME/hexanes.
Complex 1 crystallizes in the triclinic P1̅ space group where the
UIII ion is bound to two COT″ ligands in a η8-fashion to form a
distorted sandwich complex. The asymmetric η 8-COT″
coordination is reflected by UIII−CCOT″ bond distances ranging
from 2.726(0) to 2.755(1) Å (Figure 1). The lithium counterion
Figure 1. Molecular structures of 1−3. Thermal ellipsoids are drawn at
50%. Hydrogen atoms are omitted for clarity. Color code: Si = green,
UIII = pink, UIV = purple, NdIII = blue, C = gray, O = red, Li = light blue.
in the crystal lattice adopts an octahedral coordination
environment filled by three DME molecules. Complex 2
crystallizes in a monoclinic P2/c space group, and the UIV ion
is also asymmetrically coordinated by two η8-COT″ ligands with
UIV−CCOT″ bond distances ranging from 2.646(3) to 2.669(3) Å.
It is noteworthy, upon oxidation of the UIII ion to UIV, that
metal−COT″ distances become shorter by an average of 0.07 Å
as a consequence of the smaller atomic radius and possibly an
enhanced covalent interaction between the COT″ and the higher
oxidation state UIV ion. Complex 3 is isostructural to 1,
crystallizing in a triclinic P1̅ space group. As expected, complexes
1 and 3 have very similar average metal−carbon distances of 2.74
and 2.73 Å, respectively. X-ray structures and a table containing
the structural details of 1−3 are provided in the Supporting
Information (Figures S1−3, Table S1).
When viewing all three complexes from above, the CCOT″
atoms in each layer are staggered with CCOT″ atoms in other
layers, most likely in order to sterically accommodate the bulky
TMS (trimethylsilyl) groups (Figure S4).14 Complexes 1, 2, and
3 also have distinctly different tilt angles of 4.4°, 6.9°, and 3.8°,
respectively. This difference in tilt angle can be due to several
factors such as ionic radii, oxidation state15 of the metal ion, and
intermolecular interactions. For complexes 1 and 3, the close
contact between the sandwich molecule and the DME attached
to the Li counterion through an asymmetrical steric interaction
1416
DOI: 10.1021/om501214c
Organometallics 2015, 34, 1415−1418
Communication
Organometallics
α- and β-spin occupied orbitals contribute 1.34 and 1.08,
respectively, to the total metal−ligand bond order. Similar to the
UIII complex, only charge donation from the COT″ ligands to
UIV contribute to the covalent bonding in the UIV complex, and
five occupied π orbitals of the COT″2− ligands (Figure S8)
participate in covalent bonding with the UIV ion. The changes in
the populations of the HOFO, HOFO−1, HOFO−2, HOFO−3,
and HOFO−7 of the COT2− ligand are 22.7%, 22.5%, 6.0%,
5.3%, and 7.2% for α-spin manifold and 14.8%, 14.8%, 5.1%,
4.4%, and 7.6% for β-spin manifold, respectively. When going
from UIII to UIV, the energies of the metal orbitals were lowered
due to a higher ionic charge and a new 5f orbital became empty
and thus opened up for donation. Thus, ligand-to-metal donation
became stronger (Table S2), with the 5f and 6d orbitals receiving
most of the total electron density (2.99 e−) transferred from the
two ligands.
As can be seen from the electronic structure descriptors such
as bond orders, atomic charges, and changes in populations of
fragment orbitals, the U−C bonds in the UIV complex are more
covalent than the U−C bonds in the UIII complex. The
observation of shorter UIV−C bonds (average value of 2.66 Å)
in the X-ray structure of 2 relative to the UIII−C bonds (average
value of 2.74 Å) in the X-ray structure of 1 is consistent with a
stronger, more covalent metal−ligand interaction in 2.
Although UIII−COT″ covalency in 1 is less than that of 2, the
Mayer bond order of 1 is still far greater than the 1.2 calculated
for structurally analogous Dy III complex [Li(DME) 3 ][DyIII(COT″)2].8a This confirms the radial extension of actinide
ions is greater than that of the lanthanide ions. This increase
signifies 1 has potential as a desirable SMM building block unit
for creating larger UIII-SMMs with U ions coupled via a
superexchange mechanism. One additional advantage of utilizing
actinides as SMMs is the theoretical increase in spin−orbit
coupling constant with atomic mass.4 However, to understand
the effect a higher z has on the magnetic properties of a SMMs,
isoelectronic complexes such as 1 and 3 need to be carefully
examined. Only one direct comparison between isostructural and
isoelectronic 4f/5f molecular magnets has been presented, where
UIII and NdIII trispyrazolylborate (Tp) complexes exhibited slow
magnetic relaxation with Ueff = 2.84(7) and 3.81(8) cm−1,
respectively, under a 100 Oe applied dc field.5h Reported UTp3
exhibited only slightly enhanced magnetic properties over the
NdTp3, where temperature dependence of the magnetization in
both complexes was very small due to the distorted triangular
dodecahedron geometry. A major challenge in the rational design
of UIII SMMs is the limited predictability of how different ligand
field environments would affect SMM properties. Baldovi et al.
recently suggested complexes with ligand electron density along
the symmetry axis as well as a trigonal prismatic geometry may
provide ideal conditions to harness UIII SMM behavior.16
The sandwich architecture of complexes 1−3 provides the
strictly axial coordination environment that is magnetically
favorable for elements with a prolate f-electron distribution like
erbium.2a,17 The oblate nature of the f-electron distribution of
NdIII and UIII likely means that the magnetic properties of 1 and 3
will behave similarly to the recently published DyIII analogue,
which exhibited a spin reveal barrier of Ueff = 25 K under zero
applied direct current (dc) field.6a
Direct current magnetic susceptibility measurements of 1−3
were performed in the temperature range 1.8−300 K under a
1000 Oe applied dc field (Figure S9). The room-temperature χT
values of 1.13 cm3·K·mol−1 (1) and 1.50 cm3·K·mol−1 (2) are
within the range typically reported for UIII and UIV monomers,
respectively.5,18 The room-temperature χT value of 1.63 cm3·K·
mol−1 for 3 is in good agreement with the theoretical value of
1.64 cm3·K·mol−1 for a mononuclear NdIII complex.
The temperature dependence of the χT product of 1 shows a
slight decrease in magnetic susceptibility with decreasing
temperature to reach a minimum value of 0.33 cm3·K·mol−1 at
1.8 K. The χT product of 2 decreases gradually with decreasing
temperature; below 50 K there is a sharp decrease to a minimum
value of 0.14 cm3·K·mol−1 at 1.8 K (Figure S9). The sharp
decrease in the χT product is most likely due to the thermal
depopulation of higher excited states upon decrease of
temperature, as often seen for a UIV ion (ground term 3H4).
The quenching of the magnetic moment could also result form
the low symmetry and loss of Kramer degeneracy for UIV. The
temperature dependence of the χT product of 3 decreases
gradually with decreasing temperature, with a sharp decrease
below 10 K to reach a minimum value of 0.33 cm3·K·mol−1 at 1.8
K. The final decrease at low temperature is most likely due to
single-ion effects such as significant anisotropy as often
encountered in LnIII compounds rather than intermolecular
interactions (closest NdIII−NdIII distance is 10.53 Å).
The magnetization as a function of field in 1 and 3 (Figure
S10−S12) shows no saturation even at low temperature and high
fields. The reduced magnetization plot shows nonsuperimposition of iso-temperature lines, indicating significant magnetic
anisotropy in both complexes. In the case of 2, the vanishingly
small values are most likely arising from the low-lying excited
states, which can be easily populated even at 1.8 K (Figure S11).
Although no hysteresis was observed for 1−3, to probe slow
magnetization relaxation dynamics, alternating current (ac)
susceptibility measurements were performed. Under zero dc field
no ac signal was observed for 1−3. To compare the slow
relaxation of complexes 1 and 3, all ac susceptibility measurements were performed under an identical dc field (1000 Oe).
Frequency-dependent studies reveal a strong frequency-dependent in-phase (χ′) and out-of-phase magnetic susceptibility (χ″) in
both 1 and 3 (Figure 2) indicative of field-induced SMM
Figure 2. Out-of-phase magnetic susceptibility of 1 (left) and 3 (right)
under a 1000 Oe applied dc field between indicated temperatures.
behavior. The anisotropic energy barrier, Ueff, can be obtained
from the high-temperature regime of the slow magnetic
relaxation data where it is thermally induced (Arrhenius law τ
= τ0 exp(Ueff/kBT)). Effective energy barriers of 27 and 21 K with
pre-exponential factors (τ0) of 4.6 × 10−6 and 5.5 × 10−5 s for 1
and 3, respectively, were extracted from the ac data (Figure S13).
This obtained value for 1 is consistent with other recently
reported UIII single-ion magnets (SIMs) (Ueff = 4−29 K).5 The
obtained Ueff of 21 K for 3 is significantly higher than the 3.9 K
(2.8 cm−1) previously reported for a NdIII SMM.5h In the case of
2 no SIM behavior was observed, once more confirming the
nonmagnetic ground state of UIV ions.18
1417
DOI: 10.1021/om501214c
Organometallics 2015, 34, 1415−1418
Communication
Organometallics
(3) Gatteschi, D.; Sessoli, R.; Villain, J. Molecular Nanomagnets; Oxford
University Press: Oxford, U.K., 2006.
(4) (a) Khudyakov, I. V.; Serebrennikov, Y. A.; Turro, N. J. Chem. Rev.
1993, 93, 537. (b) Meihaus, K. R.; Long, J. R. Dalton Trans. 2015, 44,
2517.
(5) (a) Rinehart, J. D.; Long, J. R. J. Am. Chem. Soc. 2009, 131, 12558.
(b) Mougel, V.; Chatelain, L.; Pecaut, J.; Caciuffo, R.; Colineau, E.;
Griveau, J.-C.; Mazzanti, M. Nat. Chem. 2012, 4, 1012. (c) Mills, D. P.;
Moro, F.; McMaster, J.; van Slageren, J.; Lewis, W.; Blakel, A. J.; Liddle,
S. T. Nat. Chem. 2011, 3, 454. (d) Coutinho, J. T.; Antunes, M. A.;
Pereira, L. C. J.; Bolvin, H.; Marçalo, J.; Mazzantic, M.; Almeida, M.
Dalton Trans. 2012, 41, 13568. (e) Antunes, M. A.; Pereira, L. C. J.;
Santos, I. C.; Mazzanti, M.; Marçalo, J.; Almeida, M. Inorg. Chem. 2011,
50, 9915. (f) King, D. M.; Tuna, F.; McMaster, J.; Lewis, W.; Blake, A. J.;
McInnes, E. J. L.; Liddle, S. T. Angew. Chem., Int. Ed. 2013, 52, 4921.
(g) Moro, F.; Mills, D. P.; Liddle, S. T.; Slageren, J. V. Angew. Chem., Int.
Ed. 2013, 125, 1. (h) Rinehart, J. D.; Long, J. R. Dalton Trans. 2012, 41,
13572. (i) Rinehart, J. D.; Harris, T. D.; Kozimor, S. A.; Bartlett, B. M.;
Long, J. R. Inorg. Chem. 2009, 48, 3382. (j) Rinehart, J. D.; Meihaus, K.
R.; Long, J. R. J. Am. Chem. Soc. 2010, 132, 7572. (k) Magnani, N.;
Apostolidis, C.; Morgenstern, A.; Colineau, E.; Griveau, J.-C.; Bolvin,
H.; Walter, O.; Caciuffo, R. Angew. Chem., Int. Ed. 2011, 50, 1696.
(l) Coutinho, J. T.; Antunes, M. A.; Pereira, L. C. J.; Marçalo, J.; Almeida,
M. Chem. Commun. 2014, 50, 10262. (m) Pereira, L. C. J.; Camp, C.;
Coutinho, J. T.; Chatelain, L.; Maldivi, P.; Almeida, M.; Mazzanti, M.
Inorg. Chem. 2014, 53 (22), 11809. (n) Meihaus, K. R.; Minasian, S. G.;
Lukens, W. W., Jr.; Kozimor, S. A.; Shuh, D. K.; Tyliszczak, T.; Long, J.
R. J. Am. Chem. Soc. 2014, 136 (16), 6056. (o) Antunes, M. A.; Santos, I.
C.; Bolvin, H.; Pereira, L. C. J.; Mazzanti, M.; Marcalo, J.; Almeida, M.
Dalton Trans. 2013, 42, 8861.
(6) (a) Le Roy, J. J.; Jeletic, M.; Gorelsky, S. I.; Korobkov, I.; Ungur, L.;
Chibotaru, L. F.; Murugesu, M. J. Am. Chem. Soc. 2013, 135, 3502.
(b) Petit, S.; Pilet, G.; Luneau, D.; Chibotaru, L. F.; Ungur, L. Dalton
Trans. 2007, 40, 4582. (c) Cucinotta, G.; Perfetti, M.; Luzon, J.; Etienne,
M.; Car, P.-E.; Caneschi, A.; Calvez, G.; Bernot, K.; Sessoli, R. Angew.
Chem., Int. Ed. 2012, 51, 1606.
(7) (a) Dallinger, R. F.; Stein, P.; Spiro, T. G. J. Am. Chem. Soc. 1978,
100, 7865. (b) Chang, A. H. H.; Pitzer, R. M. J. Am. Chem. Soc. 1989,
111, 2500.
(8) (a) Jeletic, M.; Lin, P.-H.; Le Roy, J. J.; Korobkov, I.; Gorelsky, S. I.;
Murugesu, M. J. Am. Chem. Soc. 2011, 133, 19286. (b) Le Roy, J.;
Korobkov, I.; Murugesu, M. Dalton Trans. 2014, 43, 2737.
(9) Seyferth, D. Organometallics 2004, 23, 3562.
(10) Boussie, T. R.; Eisenberg, D. C.; Bigsbee, J.; Streitwieser, A., Jr.;
Zalkin, A. Organometallics 1991, 10, 1922.
(11) Parry, S. J.; Cloke, F. G. N.; Coles, S. J.; Hursthouse, M. B. J. Am.
Chem. Soc. 1999, 121, 6867.
(12) Monreal, M. J.; Thomson, R. K.; Cantat, T.; Travia, N. E.; Scott, B.
L.; Kiplinger, J. L. Organometallics 2011, 30, 2031.
(13) Edelmann, A.; Lorenz, V.; Hrib, C. G.; Hilfert, L.; Blaurock, S.;
Edelmann, F. T. Organometallics 2013, 32, 1435.
(14) (a) Lorenz, V.; Schmiege, B. M.; Hrib, C. G.; Ziller, J. W.;
Edelmann, A.; Blaurock, S.; Evans, W. J.; Edelmann, F. T. J. Am. Chem.
Soc. 2011, 133, 1257. (b) Tsoureas, N.; Summerscales, O. T.; Cloke, F.
G. N.; Roe, S. M. Organometallics 2013, 32, 1353.
(15) Boussie, T. R.; Eisenberg, D. C.; Bigsbee, J.; Streitwieser, A., Jr.;
Zalkin, A. Organometallics 1991, 10, 1922.
(16) Baldovi, J. J.; Cardona-Serra, S.; Clemente-Juan, J. M.; Coronado,
E.; Gaita-Arino, A. Chem. Sci. 2013, 4, 938.
(17) Rinehart, J. D.; Long, J. R. Chem. Sci. 2011, 2, 2078.
(18) (a) Siladke, N. A.; Meihaus, K. R.; Ziller, J. W.; Fang, M.; Furche,
F.; Long, J. R.; Evans, W. J. J. Am. Chem. Soc. 2012, 134, 1243.
(b) Castro-Rodrí guez, I.; Meyer, K. Chem. Commun. 2006, 1353.
(c) Seaman, L. A.; Pedrick, E. A.; Tsuchiya, T.; Wu, G.; Jakubikova, E.;
Hayton, T. W. Angew. Chem., Int. Ed. 2013, 52, 10589. (d) Lewis, A. J.;
Williams, U. J.; Carroll, P. J.; Schelter, E. J. Inorg. Chem. 2013, 52, 7326.
The observed relaxation barriers for both 1 and 3 are smaller
than the one observed in our analogous DyIII compound;
however it is important to emphasize that the inherent spin value
for the DyIII ion is much higher than for the UIII and NdIII
compounds.2a Thus, from the direct comparison of isostructural
and isoelectronic complexes of 1 and 3, it is reasonable to
conclude that in the same ligand field environment the inherent
magnetic anisotropy of the UIII ion is larger.
In conclusion, we have successively synthesized two
structurally similar U(COT″)2 sandwich complexes. Through
the employment of TMS-substituted COT″ ligands, structural
characterization of trivalent and tetravalent U complexes was
obtained. Contrary to previous understanding, the comparative
structural study in our case reveals the oxidation state of the
metal ion has little effect on the (ring centroid)−U−(ring
centroid) bend angle. A detailed DFT study reveals ligand-tometal donation is stronger for the UIV ion compared to UIII due to
the additional vacant metal orbital. The observed difference in
low-temperature magnetic behavior is due to the presence of
significant spin−orbit coupling in the UIII ion. The latter effect
not only enhances the anisotropy of the system but also
contributes toward the magnet-like behavior seen in 1. Although
this first UIII sandwich SIM has a smaller Ueff barrier than the
analogous DyIII complex, the covalent bonding of the UIII ion
with the COT″ ligand suggests these mononuclear units can be
ideal for the preparation of multinuclear SMMs with large energy
barriers.
■
ASSOCIATED CONTENT
S Supporting Information
*
NMR spectra and energies of the computed species. This
material is available free of charge via the Internet at http://pubs.
acs.org.
■
AUTHOR INFORMATION
Corresponding Author
*E-mail: [email protected]
Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS
We thank the University of Ottawa, NSERC (Discovery and RTI
grants); ERA, CFI, and ORF.
■
REFERENCES
(1) (a) Layfield, R. A. Organometallics 2014, 33, 1084. (b) Woodruff,
D. N.; Winpenny, R. E. P.; Layfield, R. A. Chem. Rev. 2013, 113, 5110
and references therein. (c) Fahrendorf, S.; Atodiresei, N.; Besson, C.;
Caciuc, V.; Matthes, F.; Blügel, S.; Kögerler, P.; Bürgler, D. E.;
Schneider, C. M. Nat. Commun. 2013, 4, 2425. (d) Mannini, M.;
Pineider, F.; Danieli, C.; Totti, F.; Sorace, L.; Sainctavit, P.; Arrio, M. A.;
Otero, E.; Joly, L.; Cezar, J. C.; Cornia, A.; Sessoli, R. Nature 2010, 468,
417.
(2) (a) Le Roy, J. J.; Korobkov, I.; Murugesu, M. Chem. Commun. 2014,
50, 1602. (b) Ungur, L.; Le Roy, J. J.; Korobkov, I.; Chibotaru, L. F.;
Murugesu, M. Angew. Chem., Int. Ed. 2014, 53, 4413. (c) Jiang, S.-D.;
Wang, B.-W.; Sun, H.-L.; Wang, Z.-M.; Gao, S. J. Am. Chem. Soc. 2011,
133, 4730. (d) Ishikawa, N.; Sugita, M.; Ishikawa, T.; Koshihara, S.;
Kaizu, Y. J. Am. Chem. Soc. 2003, 125, 8694. (e) Ganivet, C. R.;
Ballesteros, B.; de la Torre, G.; Clemente-Juan, J. M.; Coronado, E.;
Torres, T. Chem.Eur. J. 2013, 19, 1457. (f) Zhang, P.; Zhang, Li.;
Wang, C.; Xue, S.; Lin, S.-Y.; Tang, J. J. Am. Chem. Soc. 2014, 136, 4484−
4487.
1418
DOI: 10.1021/om501214c
Organometallics 2015, 34, 1415−1418
`