Copper Lanthanide Phosphonate Cages: Highly Symmetric {Cu3Ln9P6} and {Cu6Ln6P6} Clusters with C 3v and D3h Symmetry

Article abstract of DOI:10.1021/acs.inorgchem.5b00649

Two families of copper lanthanide phosphonate clusters have been obtained through reaction of [Cu₂(O₂C^tBu)₄(HO₂C^tBu)₂] and either Ln(NO₃)₃·nH₂O or [Ln₂(O₂C^tBu)₆(HO₂C^tBu)₆] and tert-butylphosphonic acid or an amino-functionalized phosphonic acid. The clusters, with general formula [Cu(MeCN)₄][Cu₃Ln₉(μ₃-OH)₇(O₃P^tBu)₆(O₂C^tBu)₁₅] and [Cu₆Ln₆(μ₃-OH)₆(O₃PC(NH₂)Me₂)₆(O₂C^tBu)₁₂], were structurally characterized through single crystal X-ray diffraction and possess highly symmetric metal cores with approximately C_3v and D_3h point symmetry, respectively. We have investigated the possible application of the isotropic analogues in magnetic cooling, where we were able to observe that up to around 70% of the theoretical magnetic entropy change is obtained. Simulation of the magnetic data shows antiferromagnetic coupling between the spin centers, which explains the magnetic entropy value observed. (Chemical Equation Presented).

Full text of DOI:10.1021/acs.inorgchem.5b00649

Article
pubs.acs.org/IC
Copper Lanthanide Phosphonate Cages: Highly Symmetric
{Cu3Ln9P6} and {Cu6Ln6P6} Clusters with C3v and D3h Symmetry
Eufemio Moreno Pineda,†,‡ Christian Heesing,§ Floriana Tuna,† Yan-Zhen Zheng,†,∥ Eric J. L. McInnes,*,†
Jurgen Schnack,§ and Richard E. P. Winpenny*,†
̈
†School of Chemistry and Photon Science Institute, The University of Manchester, Oxford Road, Manchester M13 9PL, U.K.
§Faculty of Physics, University of Bielefeld, Universitatsstrasse 25, D-33615 Bielefeld, Germany
̈
S
* Supporting Information
ABSTRACT: Two families of copper lanthanide phosphonate
clusters have been obtained through reaction of
[Cu2(O2CtBu)4(HO2CtBu)2] and either Ln(NO3)3·nH2O or
[Ln2(O2CtBu)6(HO2CtBu)6] and tert-butylphosphonic acid or an
amino-functionalized phosphonic acid. The clusters, with general
formula [Cu(MeCN)4][Cu3Ln9(μ3-OH)7(O3PtBu)6(O2CtBu)15] and
[Cu6Ln6(μ3-OH)6(O3PC(NH2)Me2)6(O2CtBu)12], were structurally
characterized through single crystal X-ray diffraction and possess
highly symmetric metal cores with approximately C3v and D3h point
symmetry, respectively. We have investigated the possible application
of the isotropic analogues in magnetic cooling, where we were able to
observe that up to around 70% of the theoretical magnetic entropy
change is obtained. Simulation of the magnetic data shows
antiferromagnetic coupling between the spin centers, which explains
the magnetic entropy value observed.
anisotropy make Gd(III) (8S7/2) an ideal candidate for magnetic
INTRODUCTION
■
cooling applications;13 highly anisotropic lanthanides such as
The serendipitous assembly of polymetallic compounds can lead
to interesting physical properties, even where the synthetic
chemist does not possess full chemical control over the product
o f t h e r e a c t i o n s . A n e x a m p l e i s t h e
Tb(III) (7F6), Dy(III) (6H15/2), Ho(III) (5I8), and Er(III)
(4I15/2), which impose slow relaxation characteristics, are the
preferred lanthanides for single molecule magnets;14 the high
anisotropy makes them unsuitable as components of magnetic
coolants.
[Mn12O12(O2CMe)16(H2O)4] cage, that was first proposed in
the early 1920s by Weinland and Fischer,1 but which was not
To achieve high nuclearity cages, numerous ligands have been
fully structurally characterized until 1980,2 and where the
employed.10,11 Phosphonates15,16 have become increasingly
magnetism was not understood until the 1990s.3 This happy
accident, best known as {Mn12}, was the first molecule to show
slow relaxation of magnetization, and hence the first single
molecule magnet (SMM). Since then, many aesthetically
pleasant cages, obtained from serendipitous self-assembly, have
been reported, along with several proposals of futuristic
popular due to the convenience of tuning the R groups and the
diversity of binding modes observed. Introduction of preformed
metal cages and coligands in the synthesis have proven successful
in producing discrete species.16 We have recently reported the
synthesis of several 3d−4f clusters with considerable high
magnetic entropy (−ΔSm) such as {Ni6Gd6P6},15a
{Fe6 Gd6 P6 },1 5 b {Cr6 Gd2 P2 },1 5 c {Mn9 Gd9 P1 2 },1 5 d
{Mn4Gd6P6},15d and {CoxGdyPz},15e,f where phosphonates
prove successful in coordinating several metal centers into a
single molecule. Furthermore, the introduction of further
functionalities to the R group of the phosphonates led to two
molecules, namely, {Co4Ln10P10} and {Co6Ln4P6}, showing that
functionalization of phosphonate ligands is a very promising
route to obtain further 3d−4f cages.15g Lately this method has
also afforded discrete species composed of homometallic
lanthanide phosphonates.16
applications, such as in data storage devices,4 spintronics,5 and
quantum computing.6 Recently, substantial work has been
devoted to the implementation of molecular cages as magnetic
coolants to reach the sub-kelvin regime T < 1 K, for which the
rare and expensive 3He is currently used.7,8 Such implementation
is based upon the magnetocaloric effect, which is dependent on
the spin degeneracy of the system through ∑R ln(2S + 1) (where
R is the gas constant, and S is the spin of the ground state).7 Ideal
molecular magnetic coolants are characterized by high spin
degeneracy through weak ferromagnetic exchange interactions,
and low magnetic anisotropy.9
3d−4f mixed metal systems have been extensively studied due
to the ferromagnetic interactions exhibited between them.12
Weak interactions, high spin degeneracy, and negligible
Received: March 24, 2015
© XXXX American Chemical Society
A
DOI: 10.1021/acs.inorgchem.5b00649
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
solution was refluxed for 3 h. It was then allowed to cool to room
temperature and filtered. Blue block-shaped X-ray quality crystals of
[Cu6Ln6(μ3-OH)6(O3PC(NH2)Me2)6(O2CtBu)12] (Ln = Gd(III), 6;
Tb(III), 7; Dy(III), 8; and Ho(III), 9) were obtained after 2 weeks of
slow evaporation.
Here we report further 3d−4f phosphonate cages. In one
family the copper precursor was [Cu2(O2CtBu)4(HO2CtBu)2]
reacted with Ln(NO3)3·nH2O and tert-butylphosphonic acid
producing a highly symmetrical family of {Cu3Ln9P6} cages with
general formula [Cu3Ln9(μ3-OH)7(O3PtBu)6(O2CtBu)15][Cu-
(MeCN)4] where Ln = Gd(III), 1; Tb(III), 2; Dy(III), 3;
Ho(III), 4; and Er(III), 5. Functionalization of the R group with
an amino functionality yielded a second family of cages with
general formula [Cu6 Ln6 (μ3 -OH)6 (O3 PC(NH2 )-
Me2)6(O2CtBu)12] (where Ln = Gd(III), 6; Tb(III), 7; Dy(III),
8; and Ho(III), 9) possessing D3h symmetry. Here we report the
magnetic behavior of these systems and explore the possibility of
the application of such compounds in magnetic cooling. We have
modeled the magnetic behavior of the Gd(III)-containing
systems despite the gigantic Hilbert space involved. Quasi-
exact methods allow the determination of the exchange
interaction for {Cu3Gd9P6}, while for the {Cu6Gd6P6} analogue,
we have taken advantage of the symmetry to model the magnetic
data.
X-ray Data Collection and Structure Solution. The data of 1, 4,
5, 7, and 8 were collected on an Agilent SUPERNOVA diffractometer
with Mo Kα radiation (λ = 0.71073 Å). Single crystal X-ray diffraction
measurement for 3 and 6 were carried out on a Bruker SMART CCD
diffractometer with Mo Kα radiation (λ = 0.71073 Å) at 100 K. Data for
3 was collected on a Rigaku Saturn724+ diffractometer (synchrotron, λ
= 0.68890 Å) at beamline I19 at DIAMOND Light Source, U.K. Data
reduction and unit cell refinement for all systems were performed with
Crysallis software. The structures were solved by direct methods using
SHELXS20a and were refined by full-matrix least-squares methods using
Olex2.20b In all cases the crystals were mounted on a tip using
crystallographic oil and placed in a cryststream. Data were collected
using ϕ and ω scans chosen to give a complete asymmetric unit. All non-
hydrogen atoms were refined anisotropically, while hydrogen atoms
were calculated geometrically riding on their respective atoms.
Large solvent accessible voids were found for complexes 6 to 9.
However, the location of discrete solvent molecules could not be
determined by simple refinement, therefore the contents of the large
voids on the crystal structure were determined using the solvent-
masking procedure SQUEEZE using Platon.20c
Crystal data and refinement parameters are given in Table S2 in the
Supporting Information and have been deposited with CCDC codes
1053725−1053732. These data can be obtained free of charge via www.
ccdc.cam.ac.uk/conts/retrieving.html (or from the Cambridge Crys-
tallographic Data Centre, 12 Union Road, Cambridge CB21EZ, U.K.;
fax (+44)1223-336-033; or deposit@ccdc.cam.ac.uk).
Magnetic Measurements. The magnetic behavior of complexes 1
to 9 was studied with a Quantum Design MPMS-XL7 SQUID
magnetometer employing ground samples placed in a gel capsule and
fixed with a small amount of eicosane to avoid movement during the
measurement. The diamagnetic contribution from the gel capsule and
the eicosane were corrected using experimental data obtained for such
diamagnetic matrices and the diamagnetic contribution from the
complexes calculated from Pascal constants.21
EXPERIMENTAL SECTION
■
Starting Materials. [Cu2(O2CtBu)4(HO2CtBu)2]17 and
[Ln2(O2CtBu)6(HO2CtBu)6]18 (Ln = Gd(III), Tb(III), Dy(III),
Ho(III), and Er(III)) were prepared by literature methods. Solvents
and starting materials were used as purchased. Elemental analyses were
performed by the microanalytical service of The University of
Manchester (Table S1 in the Supporting Information).
Synthesis of {Cu3Ln9P6}. Compounds 1 to 5 were obtained by
mixing [Cu2(O2CtBu)4(HO2CtBu)2] (18 mg, 0.024 mmol) and
Ln(NO3)3·nH2O (0.1 mmol) (Ln = Gd(III), Tb(III), Dy(III), Ho(III),
or Er(III)) in MeCN (16 mL) for 15 min. After this period a solution of
H2O3PtBu (14 mg, 0.1 mmol) and Et3N (0.1 mL, 1 mmol) in MeCN (4
mL) was added. The resulting solution was refluxed for 3 h, and then
filtered. Blue hexagonal-shaped X-ray quality crystals of [Cu3Ln9(μ3-
OH)7(O3PtBu)6(O2CtBu)15][Cu(MeCN)4] (Ln = Gd(III), 1; Tb(III),
2; Dy(III), 3; Ho(III), 4; or Er(III), 5) were collected after 2 weeks of
slow evaporation crystallization.
Synthesis of (1-Amino-1-methylethyl)phosphonic Acid.19
A
solution of acetone (6.3 mL, 0.085 mol), diethyl phosphite (11 mL,
0.085 mol), and NH3/EtOH (50 mL, 0.1 mol) was heated in a sealed
ampule to 100 °C for 3 h, followed by removal of the excess NH3 and
EtOH under reduced pressure. The resulting oily solution was a mixture
of diethyl (1-amino-1-methylethyl)phosphonate and diethyl phosphite.
The oily product (13 g) was then dissolved in acetone (38 mL), and a
solution of oxalic acid (5.4 g, 0.06 mol) in acetone (13.7 mL) was added.
The mixture was refrigerated overnight, and the precipitated product
was separated by filtration and washed with cold acetone (3 × 25 mL).
The dry product was treated with a saturated solution of NaHCO3 and
then extracted with dichloromethane (DCM) until the extracted DCM
became colorless (see Scheme 1). Then the solution was concentrated in
vacuo to give pure diethyl (1-amino-1-methylethyl)phosphonate (8 g,
48.2%).
RESULTS AND DISCUSSION
■
Synthetic Description. Reflux reactions of
[Cu2(O2CtBu)4(HO2CtBu)2], Ln(NO3)3·nH2O, and tert-butyl-
phosphonic acid led to {Cu3Ln9P6} in reasonable yield (17−
31%), obtained through slow evaporation crystallization. Single
crystal X-ray diffraction identified the products as [Cu-
(MeCN)4][Cu3Ln9(μ3-OH)7(O3PtBu)6(O2CtBu)15]. The pres-
ence of [Cu(MeCN)4]+ could be explained as the reduction
product of the [Cu2(O2CtBu)4(HO2CtBu)2] in the presence of
free carboxylic acid, as previously seen.22 Similarly a family of
highly symmetrical {Cu6Ln6P6} were obtained after refluxing
[Cu2(O2CtBu)4(HO2CtBu)2], [Ln2(O2CtBu)6(HO2CtBu)6],
and an amino-functionalized phosphonate source, (1-amino-1-
methylethyl)phosphonic acid. Single crystal X-ray analysis
revealed the products to be [Cu6Ln6(μ3-OH)6(O3PC(NH2)-
Me2)6(O2CtBu)12] in (20−36%) yield.
S y n t h e s i s o f { C u 6 L n 6 P 6 } .
A
s o l u t i o n o f
[Cu2(O2CtBu)4(HO2CtBu)2] (0.165 g, 0.22 mmol) and
[Ln2(O2CtBu)6(HO2CtBu)6] (Ln = Gd(III), Tb(III), Dy(III), and
Ho(III)) (0.3 mmol) in MeCN (18 mL) was prepared and stirred for 15
min. (1-Amino-1-methylethyl)phosphonic acid (0.082 g, 0.6 mmol) and
iPrNH2 (0.2 mL, 2.4 mmol) in H2O (2 mL) was added, and the resulting
In our approach we have used two phosphonate sources, which
led to high nuclearity discrete compounds. First, reaction of tert-
butylphosphonic acid along with a preformed copper carboxylate
dimer and a simple lanthanide salt gives oligonuclear {Cu3Ln9P6}
species. In our second approach we employed a functionalized
phosphonate, (1-amino-1-methylethyl)phosphonic acid, that
Scheme 1. Schematic Description of Synthesis of 1-Amino-1-
methyl-ethylphosphonic Acid
t
due to its higher polarity presented compared to BuPO3H2
required the addition of a small amount of water to the reaction.
We first react (1-amino-1-methylethyl)phosphonic acid with
[Cu2(O2CtBu)4(HO2CtBu)2], Ln(NO3)3·nH2O, and Et3N.
B
DOI: 10.1021/acs.inorgchem.5b00649
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Under these conditions a gel-like material was formed, which we
were unable to characterize, probably due to the formation of a
polymeric material due to the extra binding mode of the amino
group present in the phosphonate source; we therefore used a
preformed lanthanide carboxylate dimer as a substrate to prevent
the formation of the insoluble polymeric material and replace the
O). Finally, each Dy(III) atom in the inner triangle is bridge by
one μ3-OH at a Dy···O distance of 2.387(1) Å.
{Cu6Ln6P6}. Structural X-ray analysis of compounds 6 to 9
reveals a family of cages of formula [Cu6Ln6(μ3-OH)6(O3PC-
(NH2)Me2)6(O2CtBu)12] (Ln = Gd(III), 6; Tb(III), 7; Dy(III),
8; and Ho(III), 9) (Figure 2 and Figure S2 in the Supporting
Information). All clusters crystallize in the space group Fddd with
Z = 16 and an overall D3h symmetry with one half of the molecule
in the asymmetric unit. The structure of 1 is described as
representative.
i
EtN3 by PrNH2 to optimize the reaction. This leads to
{Cu6Ln6P6} species, where the metal sites are again arranged
in a highly symmetrical array. It is important to stress that while
the presence of a mild base such Et3N in the reaction of
{Cu3Ln9P6} is to deprotonate the phosphonate source, in the
The {Gd6P6} core (Figure 2) resembles the double-six-ring
(D6R or 6−6) building block unit found in zeolites.24 The
interaction between the oxygen atoms of the phosphonate and
the hydrogens of the iPrNH2 template the metal core, featuring a
cylindrical drum with a hexagonal prism form, whose top and
bottom faces are made up of 12-membered {Gd3O6P3} rings
(including the oxygen atoms). The top and bottom faces are
linked by six oxygens interconnecting one Gd(III) and one
phosphonate (Gd−O−P), to give an alternated {Gd6O18P6}
hexagonal prism. All aminophosphonates exhibit a 4.2111
coordination mode. In each case the amino group is terminally
bound to Cu(II), while one O-donor bridges between a Cu(II)
and a Gd(III). Two μ3-OH connect two Cu(II) ions and a
Gd(III) ion with the top and bottom face of the cluster. If just the
Cu(II), Gd(III), μ3-OH, and P are considered, this fragment
resembles a distorted cubane. Three other 2.11 pivalates
interconnect one Gd(III) ion with a Cu(II). 6 is formed by
three distorted cubane fragments, with two crystallographically
different {Cu2Gd2P2} moieties (see Table S3 in the Supporting
Information). For example, in 6 two Cu···Cu distances (2.958(1)
and 2.986(1) Å) and three Cu···O···Cu angles (96.2(2), 97.3(2),
and 98.3(2)°) are observed. All Gd(III) ions exhibit O-bound
capped octahedral geometry with one 1.11 pivalate capping the
octahedron, while all Cu(II) ions exhibit a square pyramidal
geometry.
i
case of the synthesis of {Cu6Ln6P6}, the PrNH2 may also
template the molecule through hydrogen bonding with the
oxygens of the phosphonate groups.
Structural Description. {Cu3Ln9P6}. Single crystal X-ray
diffraction of 1 to 5 shows these to be [Cu3Ln9(μ3-
OH)7(O3PtBu)6(O2CtBu)15][Cu(MeCN)4] (Ln = Gd(III), 1;
Tb(III), 2; Dy(III), 3; Ho(III), 4; and Er(III), 5) (Figure 1). The
F i g u r e 1 . ( a ) C r y s t a l s t r u c t u r e o f [ C u 3 D y 9 ( μ 3 -
OH)7(O3PtBu)6(O2CtBu)15]− viewed along the C3 axis; (b) coordina-
tion modes of phosphonates. Color code: Dy, purple; Cu, blue; P, green;
Magnetic Description. Magnetic susceptibility studies on
1−9 were performed on polycrystalline samples in the
temperature range 2−300 K under an applied dc magnetic
field (H) of 0.1 kOe. Magnetization as a function of applied field
was investigated in the field and temperature ranges 0−7 T and
2−10 K, respectively.
t
O, red; C, gray. Bu groups on carboxylates and phosphonates and
hydrogens were removed for clarity.
For compounds 1−9 room temperatures χMT values (where
χM is molar susceptibility) are close to those expected for non-
interacting ions (results are summarized in Table 1 along with
relevant parameters (S, L, J, gJ)). For all four complexes there is a
minimum in χMT at a few Kelvins; this may be a signature of weak
intramolecular magnetic interactions.
A gradual decrease of the χMT with decreasing temperature is
observed for all nonisotropic lanthanide containing cages, due to
the depopulation of the Stark levels and/or possible presence of
antiferromagnetic interactions (Figures S5 and S6 in the
Supporting Information). Field dependent magnetization
measurements were performed at between 1.8 and 5 K, and
the values obtained for complexes 1−9 at 2 K and 7 T are
summarized in Table 1.
Understanding the magnetic behavior of polymetallic
compounds containing Tb(III), Dy(III), Ho(III), and Er(III)
is difficult due to their anisotropic character as depopulation of
the Stark levels has a similar effect on the variable temperature
magnetic properties as antiferromagnetic exchange interactions.
Modeling the magnetic behavior of complexes containing the
isotropic Gd(III) is far easier, although still challenging due to the
very large size of the problem. For compound 1, the room
clusters crystallize as anions with [Cu(MeCN)4]+ cations (Figure
S1 in the Supporting Information), in the space group R3 with Z
= 3, and have an overall C3v symmetry. The crystallographic
description of the anion of 3 is given as representative. Three
Dy(III) atoms with Ln···Ln distances of 3.884(1) Å form a
triangle inside a {Cu3Dy6} ring (Figure 1). In the nonagon, a
Cu(II) is found between each two Dy(III) ions with Cu···Dy
distances ranging between 3.455(3) and 3.767(3) Å, and Dy···Dy
of 4.276(2) Å. The nine Dy(III) sites have dodecahedral
coordination geometries, and three Cu(II) sites have five-
coordinate square pyramidal geometries (Figure 1).
The metal ions in the anionic cage are bridged by a total of 15
pivalates, six fully deprotonated phosphonates, and seven
hydroxides. The metal ions are sandwiched between two
phosphonate triangles. One of the triangles is composed of
phosphonates with a bridging mode 4.222 while the other has a
4.221 mode (Harris notation).23 Pivalate groups adopt 1.11,
2.11, or 2.21 coordination modes, the 2.21 pivalate chelating to a
Dy(III) site and bridging to a copper. Each copper is also bridged
by two hydroxides to two adjacent Dy(III) and one Dy(III) of
the inner triangle with distances of 1.945(9) and 1.99(1) Å (Cu···
C
DOI: 10.1021/acs.inorgchem.5b00649
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 2. (a) Crystal structure of [Cu6Gd6(μ3-OH)6(O3PC(NH2)Me2)6(O2CtBu)12] viewed along the C3 axis; (b, d) 4f-phosphonate metal core with
D6R-like core; (c) phosphonate coordination mode in the cage. Color code: Gd, purple; Cu, blue; P, green; O, red; N, light blue; C, gray. tBu groups on
carboxylates and hydrogens were removed for clarity.
Table 1. Magnetic Data for 1−5
d
d
d
1
2
3
4
5
6
7
8
9
S (Cu2+)
S (Ln3+)
L (Ln3+)
J (Ln3+)
gJ (Ln3+)
g (Cu2+)
χMT (calcd)
1/2
7/2
0
1/2
3
1/2
5/2
5
1/2
2
1/2
3/2
6
1/2
7/2
0
1/2
3
1/2
5/2
5
1/2
2
3
6
3
6
7/2
2
6
15/2
4/3
2
8
15/2
6/5
2
7/2
2
6
15/2
4/3
2
8
3/2
2
5/4
2
3/2
2
5/4
2
2
2
b
72.0
72.1
34.6
64.0
107.4
106.2
32.2
45.2
128.6
128.3
64.7
49.1
127.7
124.4
40.1
48.7
104.4
104.9
54.1
52.2
49.5
48.7
41.2
41.3
73.1
71.7
33.5
28.5
87.2
86.4
57.5
32.8
86.6
85.1
42.2
35.9
b
χMT (obs)
b
χMT (at 2 K)
c
M (obs)
a
S (total spin angular momentum), L (total orbital angular momentum), and J (total angular momentum), of the ground multiplet. gJ is the Lande
́
b
c
d
factor. Values of χMT are given in emu mol−1 K. Values of M are given in μB. Magnetization measured at 1.8 K.
temperature value is close to that expected for the non-
interacting ions (Table 1). Upon cooling the χMT stays
practically constant up to about 40 K, where it sharply decreases.
The M(H) was also investigated at different applied fields and
temperatures revealing no saturation at the highest field and
lowest temperature, i.e., 7 T and 2 K. This observation is
consistent with the behavior observed in the χMT at low
temperatures. The magnetic data χMT(T) and M(H) could be
fitted despite the challenging dimension of the Hilbert space
((2Si + 1)n × (2Sj + 1)m = 1073741824, where n = 9 and m = 3).
Modern quasi-exact approximations, such as finite-temperature
Lanczos method, which is a Krylov-space method, has proven to
be very accurate to solve such problems.25
Many exchange paths are in principle possible. The C3v
symmetry suggests at least four exchange interactions J1, J2, J3,
J4. However, it turned out that a parameter set where nearest
neighbor interaction pathways are equivalent (J1 = J3, J2 = J4, see
inset in Figure 3a) is sufficient to model the data. Then the
Hamiltonian reads
̂
̂
̂
̂
̂
H = −2J ∑ SGd ·SGd − 2J2 ∑ SCu ·SGd
1
i
j
i
j
i,j=n.n.
i,j=n.n.
̂
̂
+ μBBz(gGd ∑ Sz + gCu ∑ Sz
)
Cu
j
Gd
i
i
j
(1)
D
DOI: 10.1021/acs.inorgchem.5b00649
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 3. (a) Molar magnetic susceptibility χMT vs T plot and simulation for 1 under 1 kG dc field. (b) Molar magnetization M(H) and simulation for 1.
Red traces are simulation with parameter in the text.
Figure 4. (a) Molar magnetic susceptibility χMT vs T plot and simulation for 6 under 1 kG dc field, (b) Molar magnetization M(H) and simulation for 6.
Red traces are simulation with parameter in the text.
Figure 5. M/μB vs H at different temperatures (a) for compound 1 and (b) for compound 6. Magnetic entropy change (c) for compound 1 and (d) for
compound 6.
Here the first term accounts for the exchange interaction
between the nearest Gd(III) neighbors, the second term takes
into account the interactions between the Cu(II) ions and their
nearest Gd(III) neighbors, and the third term is the Zeeman
interaction. Agreement between experiment and simulations is
achieved with two exchange interactions; J1 = J3 = −0.05 cm−1
and J2 = J4 = −0.21 cm−1 (Figure 3) assuming an isotropic g-
factor for Cu(II) and Gd(III) of g = 2.0.
The magnetic data of 6 have been modeled using a different
approach. Due to the symmetry of the system, D3h (crystallo-
E
DOI: 10.1021/acs.inorgchem.5b00649
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
graphically half of the molecule is present in the asymmetric unit,
with two very similar {Cu2Gd2} fragments), and the long
distances observed between Gd(1)···Gd(1)′ and Gd(2)···
Gd(2)′ we have simulated the data of a third of the fragment
assuming three independent {Cu2Gd2} moieties. Simultaneous
fitting26 of χMT(T) and M(H) allowed determination of the
parameters of the following Hamiltonian (eq 2) with SCu = 1/2,
paramagnetic ions in a single entity. Similarly the functionaliza-
tion of a phosphonate with an extra coordinating group leads to a
family of cages, both families possessing highly symmetric motifs,
the first family being described as a nonagon, with C3v symmetry,
while the second possesses a D6R-like metal core with D3h point
group symmetry. We have studied the magnetic properties of
both families and the application of the isotropic cages into
applications such as magnetocaloric. Simulation of the isotropic
systems reveals antiferromagnetic interactions operating within
each system, leading to a diminished magnetic entropy change
compared to that of the non-interacting ions. The antiferro-
magnetic exchange interactions mean that these compounds are
not of interest for magnetic cooling.
g
Cu = 2.0, SGd = 7/2, and gGd = 2.0:
̂
̂
̂
̂
̂
̂
̂
H = −2J (SCu1·SCu2) − 2J2(SCu1·SGd2 + SCu2·SGd1
)
1
4
̂
̂
̂
̂
− 2J3(SCu1·SGd1 + SCu2·SGd2) + gμBB ∑ Si
(2)
i=1
In the Hamiltonian (eq 2) the first term is the isotropic
ASSOCIATED CONTENT
■
exchange interaction between Cu(1)···Cu(2) bridged by two μ3-
OH with an average distance of 2.977(1) Å, while the second
term represents the interaction between Cu(1)···Gd(2) and
Cu(2)···Gd(1) bridged by one pivalate, a μ3-OH, and the amino
group of the phosphonate at an average distance of 3.402(1) Å,
and the third term provides the interaction between Cu(1)···
Gd(1) and Cu(2)···Gd(2) bridged by μ3-OH and the amino
group of the phosphonate at an average distance of 3.974(1) Å
(see inset of Figure 4a). The fourth term represents the Zeeman
interaction with the applied magnetic field. A good agreement
was found between the experimental data and the simulation
using the following parameters: J1 = −5.37 cm−1, J2 = −0.85
cm−1, and J3 = −0.40 cm−1 (Figure 4). A relatively weak J1 was
found, a value that can be rationalized by a combination of the
long Cu···O distances and small Cu···O···Cu angles, and similar
values have been reported.27 Hatfield and Hodgson have related
the Cu···O···Cu angles to the nature of the exchange interaction,
i.e., ferromagnetic (for angle <97.5°) or antiferromagnetic (for
angle >97.5°);28 in 6 we see angles very close to this value (see
Table S3 in the Supporting Information), and hence a very small
exchange interaction would be expected.
S
* Supporting Information
Elemental analysis tables, crystallographic tables, figures of the
packing of clusters, and figure of magnetic data and simulations
for all compounds. The Supporting Information is available free
of charge on the ACS Publications website at DOI: 10.1021/
acs.inorgchem.5b00649.
AUTHOR INFORMATION
■
Corresponding Authors
*E-mail: eric.mcinnes@manchester.ac.uk.
*E-mail: Richard.Winpenny@manchester.ac.uk.
Present Addresses
‡E.M.P.: Institute of Nanotechnology, Karlsruhe Institute of
Technology, D-76344, Eggenstein-Leopoldshafen, Germany.
∥Y.-Z.Z: Center for Applied Chemical Research, Frontier
Institute of Science and Technology, Xi’an Jiaotong University,
Xi’an 710054, China.
Notes
The authors declare no competing financial interest.
Ac susceptibility tests were performed for all anisotropic
lanthanide-containing compounds. However, no out-of-field
component, a typical feature of SMMs, was observed for the
clusters above 2 K.
The high nuclearity, the isotropic nature of Cu(II) and
Gd(III), and the relatively small exchange interaction led us to
examine compounds 1 and 6 for possible magnetocaloric
applications. The magnetic entropy changes of 1 and 6 were
obtained from isothermal magnetization measurements from 0
to 7 T in the temperature range 2−10 K (Figure 5).
ACKNOWLEDGMENTS
■
E.M.P. thanks the Panamanian agency SENACYT-IFARHU. J.S.
thanks the Deutsche Forschungsgemeinschaft (SCHN/615-15)
for continuous support. Supercomputing time at the LRZ
Garching is gratefully acknowledged. R.E.P.W. thanks the Royal
Society for a Wolfson Merit Award. We also thank Dr. David
Allan and his team for help in use of X-ray at the synchrotron at
Diamond Light Source, and we thank DLS for beam-time.
The maximum values observed for the complexes 1 and 6 at 3
K and 7 T are −ΔSm = 26.9 and 27.4 J kg−1 K−1 respectively. The
magnetic entropy for non-interacting centers is given by ∑nR
ln(2s + 1) (R being the gas constant), leading to values of 20.8R
(40.4 J kg−1 K−1, for nine Gd(III) and three Cu(II)) for 1 and
16.6R (38.7 J kg−1 K−1, for six Gd(III) and six Cu(II)) for 6.
Comparison of the experimental −ΔSm with the theoretical value
clearly shows that 66 and 71% (at 3 K and 7 T) of the total
magnetic entropy is accessed. This fact is most likely attributable
to the antiferromagnetic exchange operating within the
molecules.
REFERENCES
■
(1) Weinland, R. F.; Fischer, G. Z. Anorg. Allg. Chem. 1921, 120, 161−
180.
(2) Lis, T. Acta Crystallogr., B 1980, 36, 2042−2046.
(3) (a) Caneschi, D.; Gatteschi, D.; Sessoli, R. J. Am. Chem. Soc. 1991,
113, 5874−5876. (b) Sessoli, R.; Tsai, H. L.; Schake, A. R.; Wang, S.;
Vincent, J. B.; Folting, K.; Gatteschi, D.; Christou, G.; Hendrickson, D.
N. J. Am. Chem. Soc. 1993, 115, 1804−1816. (c) Sessoli, R.; Gatteschi,
D.; Caneschi, A.; Novak, M. A. Nature 1993, 365, 141−142.
(4) (a) Gatteschi, D.; Sessoli, R.; Villain, J. Molecular Nanomagnets;
Oxford University Press: Oxford, U.K., 2006; pp 306−317. (b) Mannini,
M.; Pineider, F.; Sainctavit, P.; Danieli, C.; Otero, E.; Sciancalepore, C.;
Talarico, A. M.; Arrio, M.-A.; Cornia, A.; Gatteschi, D.; Sessoli, R. Nat.
Mater. 2009, 8, 194−197.
(5) (a) Sanvito, S. Chem. Soc. Rev. 2011, 40, 3336−3355. (b) Bogani,
L.; Wolfgang, W. Nat. Mater. 2008, 7, 179−186.
(6) (a) Leuenberger, M. N.; Loss, D. Nature 2001, 410, 789−793.
(b) Troiani, F.; Affronte, M. Chem. Soc. Rev. 2011, 40, 3119−3129.
CONCLUSIONS
■
The synthesis and characterization of mixed-metal 3d−4f cages
have been achieved through the synthesis of preformed
carboxylate cages with phosphonates under reflux conditions,
phosphonates proving successful ligands to bring several
F
DOI: 10.1021/acs.inorgchem.5b00649
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(7) (a) Giauque, W. F.; MacDougall, D. P. Phys. Rev. 1933, 43, 768−
783. (b) Sharples, J. W.; Collison, D.; McInnes, E. J. L.; Schnack, J.;
Palacios, E.; Evangelisti, M. Nat. Commun. 2014, 5, 5321−5326.
(8) Evangelisti, M.; Candini, A.; Ghirri, A.; Affronte, M.; Brechin, E. K.;
McInnes, E. J. L. Appl. Phys. Lett. 2005, 87, 072504.
(9) (a) Evangelisti, M.; Brechin, E. K. Dalton Trans. 2010, 39, 4672−
4676. (b) Garlatti, E.; Carretta, S.; Schnack, J.; Amoretti, G.; Santini, P.
Appl. Phys. Lett. 2013, 103, 202140. (c) Evangelisti, M.; Lorusso, G.;
Palacios, E. Appl. Phys. Lett. 2014, 105, 046101/1. (d) Garlatti, E.;
Carretta, S.; Schnack, J.; Amoretti, G.; Santini, P. Appl. Phys. Lett. 2014,
105, 046101.
different donor atoms. See Supporting Information and also: Coxall, R.
A.; Harris, S. G. D.; Henderson, K.; Parsons, S.; Tasker, P. A.; Winpenny,
R. E. P. J. Chem. Soc., Dalton Trans. 2000, 2349−2356.
(24) (a) Yang, Y.; Walawalkar, M. G.; Pinkas, J.; Roesky, H. W.;
Schmidt, H.-G. Angew. Chem., Int. Ed. 1998, 37, 96−98. (b) Walawalkar,
M. G.; Roesky, H. W.; Murugavel, R. Acc. Chem. Res. 1999, 32, 117−126.
̌ ̌
(25) (a) Jaklic, J.; Prelovsek, P. Phys. Rev. B 1994, 49, 5065−5068.
(b) Schnack, J.; Wendland, O. Eur. Phys. J. B 2010, 78, 535−541.
(c) Schnack, J.; Heesing, C. Eur. Phys. J. B 2013, 86, 46.
(26) Chilton, N. F.; Anderson, R. P.; Turner, L. D.; Soncini, A.;
Murray, K. S. J. Comput. Chem. 2013, 34, 1164−1175.
(27) (a) Real, J. A.; Munno, G. D.; Chiappetta, R.; Julve, M.; Lloret, F.;
Journaux, Y.; Colin, J.-C.; Blondin, G. Angew. Chem., Int. Ed. Engl. 1994,
33, 1184−1186. (b) Sletten, J.; Sørensen, A.; Julve, M.; Journaux, Y.
Inorg. Chem. 1990, 29, 5054−5058. (c) Charlot, M. F.; Khan, O.;
Jeannin, S.; Jeannin, Y. Inorg. Chem. 1980, 19, 1410−1411.
(28) Hodgson, D. J. Prog. Inorg. Chem. 1975, 19, 173. (b) Hatfield. W.
E. In Magneto-Structural Correlations in Exchange coupled Systems;
Willett, R. D., Gatteschi, D., Kahn, O., Eds.; Reidel: Dordrecht, The
Netherlands, 1984; pp 555.
(10) Accorsi, S.; Barra, A.; Caneschi, A.; Chastanet, G.; Cornia, A.;
Fabretti, A. C.; Gatteschi, D.; Mortalo, C.; Olivieri, E.; Parenti, F.; Rosa,
P.; Sessoli, R.; Sorace, L.; Wernsdorfer, W.; Zobbi, L. J. Am. Chem. Soc.
2006, 126, 4742−4755.
(11) (a) Brechin, E. K. Chem. Commun. 2005, 41, 5141−5153.
(b) Glaser, T. Chem. Commun. 2011, 47, 116−130. (c) Cowan, M. G.;
Brooker, S. Coord. Chem. Rev. 2012, 256, 2944−2971. (d) Escuer, A.;
Aromí, G. Eur. J. Inorg. Chem. 2006, 23, 4721−4736.
(12) See: Winpenny, R. E. P. Chem. Soc. Rev. 1998, 27, 447−452 and
references therein.
(13) Hooper, T. N.; Schnack, J.; Piligkos, S.; Evangelisti, M.; Brechin,
E. K. Angew. Chem., Int. Ed. 2012, 51, 4633−4636.
(14) (a) Sorace, L.; Benelli, C.; Gatteschi, D. Chem. Soc. Rev. 2011, 40,
3092−3104. (b) Woodruff, D. N.; Winpenny, R. E. P.; Layfield, R. A.
Chem. Rev. 2013, 113, 5110−5148.
(15) (a) Moreno-Pineda, E.; Tuna, F.; Zheng, Y.-Z.; Winpenny, R. E.
P.; McInnes, E. J. L. Inorg. Chem. 2013, 52, 13702−13707. (b) Moreno-
Pineda, E.; Tuna, F.; Zheng, Y.-Z.; Teat, S. J.; Winpenny, R. E. P.;
Schnack, J.; McInnes, E. J. L. Inorg. Chem. 2014, 53, 3032−3038.
(c) Zangana, K. H.; Moreno-Pineda, E.; Vitorica-Yrezabal, I. J.;
McInnes, E. J. L.; Winpenny, R. E. P. Dalton Trans. 2014, 43, 13242−
13249. (d) Zheng, Y.-Z.; Moreno-Pineda, E.; Helliwell, M.; Winpenny,
R. E. P. Chem.Eur. J. 2012, 18, 4161−4165. (e) Zheng, Y.-Z.;
Evangelisti, M.; Winpenny, R. E. P. Chem. Sci. 2011, 2, 99−102.
(f) Zheng, Y.-Z.; Evangelisti, M.; Tuna, F.; Winpenny, R. E. P. J. Am.
Chem. Soc. 2012, 134, 1057−1065. (g) Moreno-Pineda, E.; Tuna, F.;
Pritchard, R. G.; Zheng, Y.-Z.; Regan, A. C.; Winpenny, R. E. P.;
McInnes, E. J. L. Chem. Commun. 2013, 49, 3522−3524. (h) Zheng, Y.-
Z.; Breeze, B. A.; Timco, G. A.; Tuna, F.; Winpenny, R. E. P. Dalton
Trans. 2010, 39, 6175−6177. (i) Zheng, Y.-Z.; Evangelisti, M.;
Winpenny, R. E. P. Angew. Chem., Int. Ed. 2011, 50, 3692−3695.
(j) Zheng, Y.-Z.; Winpenny, R. E. P. Sci. China: Chem. 2012, 55 (6),
910−913.
(16) (a) Zangana, K. H.; Moreno-Pineda, E.; Schnack, J.; Winpenny, R.
E. P. Dalton Trans. 2013, 42, 14045−14048. (b) Zangana, K. H.;
Moreno-Pineda, E.; McInnes, E. J. L.; Schnack, J.; Winpenny, R. E. P.
Chem. Commun. 2014, 50, 1438−1440. (c) Zangana, K. H.; Moreno-
Pineda, E.; Winpenny, R. E. P. Dalton Trans. 2014, 43, 17101−17107.
(17) lljina, E.; Korjeva, A.; Kuzmina, N.; Troyanov, S.; Dunaeva, K.;
Martynenko, L. Mater. Sci. Eng. 1993, 234−236.
(18) (a) Fomina, I. G.; Kiskin, M. A.; Martynov, A. G.; Aleksandrov, G.
G.; Dobrokhotova, Z. V.; Gorbunova, Y. G.; Shvedenkov, Y. G.;
Tsivadze, A. Y.; Novotortsev, V. M.; Eremenko, I. L. Zh. Neorg. Khim.
2004, 49, 1463−1474. (b) Zoan, T. A.; Kuzmina, N. P.; Frolovskaya, S.
N.; Rykov, A. N.; Mitrofanova, N. D.; Troyanov, S. I.; Pisarevsky, A. P.;
Martynenko, L. I.; Korenev, Y. M. J. Alloys Compd. 1995, 225, 396−399.
(19) Medved, T. Y.; Kabachnik, M. I. Russ. J. Chem. B 1954, 3, 255−
258.
(20) (a) Sheldrick, G. M. Acta Crystallogr. 2008, A64, 112−122.
(b) Dolomanov, O. V.; Bourthis, L. J.; Gildea, R. L.; Howard, J. A. K.;
Puschmann, H. J. Appl. Crystallogr. 2009, 42, 339−341. (c) Sluis, P. v. d.;
Spek, A. L. Acta Crystallogr. 1990, A46, 194−201.
(21) Bain, G. A.; Berry, J. F. J. Chem. Educ. 2008, 85, 532−536.
(22) Lopez-Dellamary, T.; Sampedro, J. A.; Carlos, G.; Fontes, M.;
Ogura, T. Transition Met. Chem. 1978, 3, 342−344.
(23) Harris notation describes the binding mode as [X·Y1Y2···Yn],
where X is the overall number of metals bound by the whole ligand, and
each value of Y refers to the number of metal atoms attached to the
G
DOI: 10.1021/acs.inorgchem.5b00649
Inorg. Chem. XXXX, XXX, XXX−XXX