Ligand design for heterobimetallic single-chain magnets: Synthesis, crystal structures, and magnetic properties of MIICuII (M = Mn, Co) chains with sterically hindered methyl-substituted phenyloxamate bridging ligands

Article abstract of DOI:10.1002/chem.200600992

Two new series of neutral oxamato-bridged heterobimetallic chains of general formula [MCu(L^x)₂]· mDMSO (m = 0-4) (L ¹ = N-2-methyl-phenyloxamate, M = Mn (1a) and Co (1b) ; L² = N-2,6-dimethylphenyloxamate, M = M

Full text of DOI:10.1002/chem.200600992

DOI: 10.1002/chem.200600992
Ligand Design for Heterobimetallic Single-Chain Magnets: Synthesis, Crystal
Structures, and Magnetic Properties of MIICuII (M=Mn, Co) Chains with
Sterically Hindered Methyl-Substituted Phenyloxamate Bridging Ligands
Emilio Pardo,[a] Rafael Ruiz-García,[b, f] Francesc Lloret,*[a] Juan Faus,[a] Miguel Julve,[a]
Yves Journaux,[c] Miguel A. Novak,[d] Fernando S. Delgado,[e] and Catalina Ruiz-PØrez[e]
Abstract: Two new series of neutral ox-
oxamate bridge is 5.296(1) Š in 2b’
and 5.301(2) Š in 3b’, while the short-
est interchain Co···Co distance is
5.995(5) Š in 2b’ and 8.702(3) Š in 3b’,
that is, the chains are well isolated in
the crystal lattice due to the presence
of the bulky methyl-substituted phenyl
groups. Although both MnIICuII and
CoIICuII chains exhibit ferrimagnetic
behaviour with moderately strong in-
trachain antiferromagnetic coupling
tures (TB <3.5 K), which is characteris-
tic of single-chain magnets (SCMs) be-
cause of the high magnetic anisotropy
of the CoII ion. The blocking tempera-
tures TB along this series of chains vary
according to the steric hindrance of the
aromatic substituent of the oxamate
ligand in the series L1 <L2 <L3. Analy-
sis of the SCM behaviour for 3b and
3b’ on the basis of Glauberꢀs theory
amato-bridged heterobimetallic chains
of
general
formula
[MCu(Lx)2]·
mDMSO (m=0–4) (L1 =N-2-methyl-
phenyloxamate, M=Mn (1a) and Co
(1b);
L2 =N-2,6-dimethylphenyloxa-
mate, M=Mn (2a) and Co (2b); L3 =
N-2,4,6-trimethylphenyloxamate, M=
Mn (3a) and Co (3b)) have been pre-
pared by reaction between the corre-
sponding anionic oxamatocopper(II)
complexes [Cu(Lx)2]2ꢀ with Mn2+ or
Co2+ cations in DMSO. The crystal
for
a one-dimensional Ising system
(ꢀJMn,Cu =24.7–27.9 cmꢀ1 and ꢀJCo,Cu
=
showed a thermally activated mecha-
nism for the magnetic relaxation (Ar-
rhenius law dependence). The activa-
tion energies Ea to reverse the magnet-
isation direction are 38.0 (3b) and
16.3 cmꢀ1 (3b’), while the preexponen-
tial factors t0 are 2.3”10ꢀ11 (3b) and
4.0”10ꢀ9 s (3b’).
P
35.0–45.8 cmꢀ1; H=
ꢀJM,CuSM,iSCu,i),
only the CoIICuII chiains show slow
magnetic relaxation at low tempera-
structures of [CoCu(L2)2
and [CoCu(L3)2
(H2O)2]·4H2O (3b’)
A
ACHTREUNG
have been solved by single-crystal
X-ray diffraction methods. Compounds
2b’ and 3b’ adopt zigzag and linear
chain structures, respectively. The in-
trachain Cu···Co distance through the
Keywords: cobalt · copper · mag-
netic properties · manganese · N,O
ligands · supramolecular chemistry
Introduction
[a] Dr. E. Pardo, Prof. Dr. F. Lloret, Prof. Dr. J. Faus, Prof. Dr. M. Julve
Departament de Química Inorgànica
Instituto de Ciencia Molecular, Facultat de Química
Universitat de Val›ncia
Magnetic chains have been actively investigated for the
design and synthesis of molecular magnets, that is, molecule-
based compounds exhibiting spontaneous magnetisation
below a critical temperature TC.[1,2] This novel molecular ap-
proach to magnetic materials consists of assembling ferri-
magnetic heterospin chains in a ferromagnetic fashion with
Paterna, Val›ncia, 46980 (Spain)
Fax : (+34)96-354-4322
E-mail: francisco.lloret@uv.es
[b] Dr. R. Ruiz-García
Departament de Química Orgànica
Instituto de Ciencia Molecular, Facultat de Química
Universitat de Val›ncia
[e] Dr. F. S. Delgado, Prof. Dr. C. Ruiz-PØrez
Laboratorio de Rayox X y Materiales Moleculares
Departamento de Física Fundamental II
Facultad de Física, Universidad de La Laguna
Tenerife, 38204 (Spain)
Paterna, Val›ncia, 46980 (Spain)
[c] Dr. Y. Journaux
Laboratoire de Chimie Inorganique et MatØriaux Moleculaires
UniversitØ Pierre et Marie Curie–Paris6
UMR 7071, Paris, 75005 (France)
[f] Dr. R. Ruiz-García
Fundació General de la Universitat de Val›ncia
Universitat de Val›ncia
Val›ncia, 46002 (Spain)
[d] Prof. M. A. Novak
Instituto de Física
Universidade Federal do Rio de Janeiro
Rio de Janeiro-RJ, 21945-970 (Brazil)
Supporting information for this article is available on the WWW
under http://www.chemeurj.org/ or from the author.
2054
ꢁ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 2054 – 2066

FULL PAPER
the possibility of achieving long-range three-dimensional
(3D) magnetic order through interchain interactions.[3,4]
Such an approach was first applied successfully by Kahn
et al. in the oxamato-bridged manganese(II)–copper(II)
ferromagnetic interactions through oximato and phenolato
bridges, respectively;[9] 2) [{FeL(CN)4}2Co
(H2O)2]·nH2O
A
(L=2,2’-bipyridine (bpy), 1,10-phenanthroline (phen) with
n=4, and L=2,2’-bipyrimidine (bpym) with n=6), which is
made up of cyanide-bridged 4,2-ribbonlike chains with ferro-
magnetic coupling between the low-spin FeIII and high-spin
chain [MnCu
G
G
propylenebis
ACHTREUNG
other heterobimetallic chains of the same family with vary-
ing results. For instance, the cobalt(II)–copper(II) chain ana-
CoII ions;[10a,c] 3) [{Fe
N
N
which is a related cyanide-bridged double 4,2-ribbonlike
chain with coexistence of intrachain ferro- and interchain
antiferromagnetic interactions;[10b] 4) [Co(bt)(N3)2] (bt=2,2’-
bithiazoline) which is composed of ferromagnetic azide-
bridged helical chains of high-spin CoII ions;[11]
logue [CoCu
netically.[3b] Alternatively, Gatteschi et al. reported the man-
ganese(II) nitronyl nitroxide chain [Mn(hfac)2(NITiPr)]
(hfac=hexafluoroacetylacetonate, NITiPr=2-isopropyl-
A
G
ACHTREUNG
A
N
ACHTREUNG
4,4,5,5-tetramethylimidazolin-1-oxyl-3-oxide) as the first ex-
ample of a metalloradical chain which orders ferromagneti-
cally (TC =7.6 K).[4a] Despite the fact that the CuII ion (S=1/
2) had been replaced by a nitronyl nitroxide (NIT) radical
(S=1/2), a similar 1D ferrimagnetic behaviour was observed
in the MnIICuII bimetallic chains and the MnIINITradical
chains. The recent finding of slow magnetic relaxation and
hysteresis effects in the cobalt(II) nitronyl nitroxide radical
5) [{Fe(Tp)(CN)3}2Cu(CH3OH)]·2CH3OH (Tp=tris(pyrazo-
G
lyl)borate), which consists of chains with ferromagnetic cou-
pling between the high-spin FeIII ions and the CuII ions;[12]
and 6) (NEt4)
ACHTREUNG
N,N’-ethylenebis(5-methoxysalicylideneiminate)),
consists of chains of MnIIIFeIIIMnIII trinuclear units with al-
ternating intratrimer antiferro- and intertrimer ferromagnet-
ic interactions through cyanide and phenolate bridges, re-
spectively.[13]
chain analogue [Co
N
E
4’-methoxyphenyl-4,4,5,5-tetramethylimidazolin-1-oxyl-3-
oxide),[5] which are not associated with 3D ordering but
have a purely 1D origin, has renewed the field and opened
exciting new perspectives for storing information in low-di-
mensional magnetic materials.[6]
Here we present a rational synthetic strategy to obtain
SCMs which is based on the use of sterically hindered di-
A
gands towards fully solvated, divalent transition metal cat-
ions such as manganese(II) and cobalt(II) (Scheme 1). This
“complex-as-ligand” approach results in oxamato-bridged
heterobimetallic chains with three different possible struc-
tures. The linear chain structure[14a,b] is obtained when the
solvent molecules coordinate to the MII ions (M=Mn and
Co) in trans positions (Scheme 1a). When they coordinate
in cis positions, zigzag[14c] or helical structures result, de-
pending on the steric requirements of the CuII precursor
complex (Scheme 1b and c, respectively).
As early as 1963, Glauber[7] predicted that a single chain
with Ising-type magnetic anisotropy, either ferri- or ferro-
magnetic, should exhibit exponential divergence of the re-
laxation time t of the magnetisation as the temperature de-
creases. This situation is reminiscent of that recently ob-
served in some discrete polynuclear complexes, so-called
single-molecule magnets (SMMs).[8a,b] These molecules have
a large ground-state spin S and a large negative axial mag-
netic anisotropy D which result in a large activation energy
Ea for the magnetisation reversal, given by Ea =jDjS2. In
the case of single chains, the activation energy depends on
the intrachain coupling constant J according to the expres-
sion Ea =(4 jJj+jDj)S2.[8c] Although this theory has been
known for more than forty years, it is only recently that the
first example of a 1D compound exhibiting slow magnetic
relaxation below a blocking temperature TB, a so-called
single-chain magnet (SCM), has been reported. An easy axis
of magnetisation and a large ratio of intrachain to interchain
coupling constants (jJ/jj>104) are both required to observe
this unique magnetic behaviour experimentally. In spite of
these restrictive requirements, some additional examples of
SCMs were published in the last five years since the first ex-
ample was reported by Gatteschi et al., which still exhibits
the highest blocking temperature (TB ꢁ15 K).[5a] These in-
clude ferri- and ferromagnetic, homo- and heterobimetallic
chains with different bridges and metal ions: 1) [Mn2-
We report the preparation and structural and magnetic
characterisation of a novel series of heterobimetallic manga-
nese(II)–copper(II) and cobalt(II)–copper(II) chains pre-
pared from monomeric copper(II) complexes with the aro-
matically substituted oxamate ligands L1 =N-2-methylphe-
nyloxamate, L2 =N-2,6-dimethylphenyloxamate and L3 =N-
2,4,6-trimethylphenyloxamate. In a preliminary communica-
tion,[15] we have shown that the combination of an orbitally
degenerate octahedral high-spin CoII ion (4T1g) and
a
square-planar CuII precursor with the L3 ligand results in a
large Ising-type magnetic anisotropy in the chain and mini-
misation of the interchain interactions. Both conditions were
necessary to observe the phenomenon of slow relaxation of
the magnetisation in the cobalt(II)–copper(II) chain
[CoCu(L3)2
(H2O)2]·4H2O, where magnetic coupling between
A
the CoII and CuII ions through the oxamate bridge is large
enough to allow long-range intrachain magnetic correlation
(ꢀJ=35.0 cmꢀ1).[15] At present, our goal is to perform a sys-
tematic analysis of the effect of the steric requirements in
the CuII precursor complex (number of methyl substituents)
and the local magnetic anisotropy of the MII ion (M=Mn,
Co) on the structure and magnetic properties of the result-
ing MIICuII chain compound. This work provides an example
A
G
(saltmen=N,N’-(1,1,2,2-tetra-
ACHTREUNG
ACHTREUNG
Chem. Eur. J. 2007, 13, 2054 – 2066
ꢁ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2055

F. Lloret et al.
80%) by deprotonation and hy-
drolysis of the corresponding
ligand HEtLx with NaOH in
water, and subsequent addition
of the stoichiometric amount of
CuII as a nitrate salt. In the
third step, the heterobimetallic
chain compounds of general
formula [MnCu(Lx)2]·mDMSO
(1a–3a)
and
[CoCu(Lx)2]·
mDMSO (1b–3b) were ob-
tained by reaction of the corre-
sponding sodium salt of the
mononuclear CuII precursor
and MnII or CoII nitrate, re-
spectively, in hot DMSO
(Scheme 2c).
Alternatively,
single crystals of cobalt(II)–
copper(II) chain compounds
Scheme 1. Proposed structures for the oxamato-bridged heterobimetallic chain compounds: a) linear, b) zigzag
and c) helical.
[CoCu(L2)2
(H2O)2] (2b’) and
A
of successful design of a new family of oxamato-bridged het-
erobimetallic SCMs, which expand the range of the few re-
ported examples of slowly magnetically relaxing 1D materi-
als.
Results and Discussion
Scheme 2. Synthetic pathway to the oxamato-bridged heterobimetallic
Synthesis and general physical characterization: Neutral ox-
amato-bridged MIICuII chains (M=Mn, Co) were synthe-
chain compounds. a) C2O2ClOEt, THF (708C); b) NaOH/CuII
ACHTREUNG
H2O (RT); c) MII
H2O (RT ).
A
ACHTREUNG
A
synthesis of the N-substituted monooxamate ligands H2Lx by
straightforward condensation of ethyl chlorooxoacetate with
three different aniline derivatives in THF (Scheme 2a).
They were isolated as the ethyl ester derivatives HEtLx in
excellent yields (ca. 80–100%). The second step consists of
synthesising the mononuclear copper(II)–Lx complexes as
their sodium salts of general formula [Na2Cu(Lx)2]·nH2O
(Scheme 2b). They were obtained in good yields (ca. 75–
[CoCu(L3)2
(H2O)2]·4H2O (3b’) were grown by slow diffu-
U
sion in an H-shaped tube of aqueous solutions containing
stoichiometric amounts of CoII nitrate in one arm and the
corresponding sodium salt of the mononuclear CuII precur-
sor in the other (Scheme 2c’). All our attempts to get X-ray-
2056
www.chemeurj.org
ꢁ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 2054 – 2066

Heterobimetallic Single-Chain Magnets
FULL PAPER
Figure 1. a) Top and b) side views of the mononuclear complex anion of Na2[Cu(L3)2]·6H2O with atom labelling for the metal coordination environment.
3
¯
Hydrogen atoms are omitted for clarity. c) and d) Crystal packing of Na2[Cu(L )2]·6H2O along the [111] and [100] directions. Coordinative bonds to
sodium are represented by solid lines (symmetry codes: a=ꢀx, 1ꢀy, 1/2ꢀz; b=1/2ꢀx, ꢀy, 1/2ꢀz).
quality single crystals of the cobalt(II)–copper(II)–L1 chain
analogue were unsuccessful.
Table 1. Selected
bond
lengths
[Š]
and
angles
[8]
of
Na2[Cu(L3)2]·6H2O.[a,b]
ꢀ
ꢀ
Cu1 N1a
1.922(4)
Cu3 N3
1.920(4)
1.920(4)
1.955(3)
1.955(3)
1.917(4)
1.917(4)
1.954(3)
1.954(3)
180.0
84.14(15)
95.86(15)
95.86(15)
84.14(15)
180.0
The chemical identity of the ligands, the mononuclear
copper(II) complexes and the chain compounds was con-
firmed by elemental analysis (C, H, N, S) and IR and
1H NMR spectroscopy. Analytical and general spectroscopic
data are listed in the Supporting Information in Tables S1–
S3.
ꢀ
ꢀ
Cu1 N1
1.922(4)
1.955(3)
1.955(3)
1.923(4)
1.923(4)
1.953(3)
1.953(3)
180.0
95.77(15)
84.23(15)
84.22(15)
95.78(15)
180.0
Cu3 N3c
ꢀ
ꢀ
Cu1 O3a
Cu3 O9c
ꢀ
ꢀ
Cu1 O3
Cu3 O9
ꢀ
ꢀ
Cu2 N2
Cu4 N4
ꢀ
ꢀ
Cu2 N2b
Cu4 N4d
ꢀ
ꢀ
Cu2 O6b
Cu4 O12d
ꢀ
ꢀ
Cu2 O6
Cu4 O12
N1a-Cu1-N1
N1a-Cu1-O3
N1-Cu1-O3
N1a-Cu1-O3a
N1-Cu1-O3a
O3-Cu1-O3a
N2-Cu2-N2b
N2-Cu2-O6b
N2c-Cu2-O6
N2-Cu2-O6
N2b-Cu2-O6
O6b-Cu2-O6
N3-Cu3-N3c
N3-Cu3-O9
Description of the structures
N3c-Cu3-O9
N3-Cu3-O9c
N3c-Cu3-O9
O9-Cu3-O9c
N4-Cu4-N4d
N4-Cu4-O12c
N4d-Cu4-O4d
N4-Cu4-O12
N4d-Cu4-O12
O12d-Cu4-O12
Na2[Cu(L3)]·6H2O: Na2[Cu(L3)]·6H2O consists of centro-
symmetric mononuclear dianionic copper(II) units
[Cu(L3)2]2ꢀ (Figure 1a), coordinated sodium cations, and
both coordinated and crystallisation water molecules. Select-
ed bond lengths and angles are summarised in Table 1.
There are four crystallographically independent copper
atoms Cu1–Cu4, which are geometrically equivalent with
almost identical bond lengths and angles for the metal coor-
dination environment. The copper atoms are four-coordi-
nate with two amidate-nitrogen and two carboxylate-oxygen
atoms from the two oxamate ligands in a trans arrangement
180.0
180.0
95.80(15)
84.20(15)
84.20(15)
95.80(15)
180.0
95.52(15)
84.48(15)
84.48(15)
95.52(15)
180.0
[a] Estimated standard deviations are given in parentheses. [b] Symmetry
codes: a=ꢀx, ꢀy, ꢀz; b=ꢀx, ꢀy, ꢀzꢀ1; c=ꢀxꢀ1, ꢀy, ꢀz+1; d=ꢀx,
ꢀyꢀ1, ꢀzꢀ1.
Chem. Eur. J. 2007, 13, 2054 – 2066
ꢁ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2057

F. Lloret et al.
ꢀ
building a square-planar surrounding. The average Cu N
ꢀ
and Cu O bond lengths are 1.921(2) and 1.954(3) Š, respec-
tively, and they compare well with those reported earlier for
related square-planar copper(II) complexes with two oxa-
ꢀ
ꢀ
mate ligands in cis arrangement (average Cu N and Cu O
bond lengths of 1.910 and 1.945 Š, respectively).[16] The
average dihedral angle between the basal plane of the
copper atoms and the planes of the oxamato groups is
1.4(2)8. The phenyl rings are practically perpendicular to the
oxamate groups to prevent steric repulsion between the 2-
and 6-methyl substituents and the carbonyl oxygen atoms
(Figure 1b). The average dihedral angle between their mean
planes is 82.1(4)8. More likely, the alternative cis arrange-
ment of the oxamate ligands is precluded because of steric
hindrance between the trimethyl-substituted phenyl groups.
In this regard, it is important to point out that the trans ar-
rangement of the oxamate ligands in Na2[Cu(L3)2]·6H2O is
kept in the bimetallic chains as well (see below).
In the crystal lattice, the bis(oxamato)copper(II) entities
A
are coordinated to the sodium atoms through the carbonyl
oxygen atoms to form a corrugated 3D rhombic network
structure (Figure 1c). The four copper atoms Cu1–Cu4
occupy the centre of each edge of the rhombus, while the
four sodium atoms Na1–Na4 define the vertexes. Additional
water molecules, both bridging and terminal, complete the
coordination sphere of the sodium atoms and thus lead to
pillared 1D arrays (Figure 1d). There are both five- and six-
ꢀ
coordinate sodium atoms with Na O distances in the range
2.315(4)–2.439(4) Š. The shortest Cu1···Cu2 distance
through the coordinated sodium ions is 7.2667(8) Š.
Figure 2. a) View of a fragment of the chain of 2b’ with atom labelling
for the metal coordination environments. Hydrogen atoms are omitted
G
U
¯
for clarity. b) and c) Crystal packing of the chains of 2b’ along the [101]
and [101] directions. Hydrogen bonds are represented by dashed lines
(symmetry codes: a=1/2ꢀx, 1/2ꢀy, ꢀz; b=ꢀx, y, 1/2ꢀz; c=x, 1ꢀy, 1ꢀz;
d=x, y, 1/2ꢀz).
entity acts as a bis-bidentate ligand through the cis carbonyl
oxygen atoms towards cis-diaquacobalt(II) units. Because of
the cis conformation of the octahedral Co atoms, two differ-
ent isomers (D and L) exist that alternate regularly along
the chain and thus lead to an achiral zigzag chain structure.
A chiral helical chain structure would result in the case of
identical isomers (D or L). Selected bond lengths and angles
for 2b’ are summarised in Table 2.
The copper atom is four-coordinate with two amidate-ni-
trogen and two carboxylate-oxygen atoms from the two oxa-
mate ligands in a trans arrangement building a square-
planar surrounding (Cu1 N1 1.926(3), Cu1 O3 1.956(3) Š).
The dihedral angle between the basal plane of the copper
atom and the plane of the oxamato groups is 2.85(12)8. T he
cobalt atom is six-coordinate with two cis-coordinated water
molecules and four carbonyl oxygen atoms from two oxa-
mate ligands forming a distorted octahedral surrounding.
Table 2. Selected bond lengths [Š] and angles [8] of 2b’.[a,b]
ꢀ
ꢀ
Cu1 N1
1.923(4)
1.958(3)
2.083(3)
2.123(3)
2.076(4)
92.1(3)
Cu1 N1a
1.923(4)
1.958(3)
2.083(3)
2.123(3)
2.076(4)
170.09(14)
92.4(2)
ꢀ
ꢀ
Cu1 O3
Cu1 O3a
ꢀ
ꢀ
Co1 O1
Co1 O1b
ꢀ
ꢀ
Co1 O2
Co1 O2b
ꢀ
ꢀ
Co1 O1w
Co1 O1wb
O1w-Co1-O1wb
O1wb-Co1-O1
O1-Co1-O1w
O1w-Co1-O2
O1-Co1-O2
O1w-Co1-O1
O1w-Co1-O1b
O1-Co1-O1b
O1wb-Co1-O2
O1b-Co1-O2
O1wb-Co1-O2b
O1b-Co1-O2b
N1-Cu1-N1a
N1a-Cu1-O3a
N1a-Cu1-O3
92.4(2)
170.09(14)
91.15(14)
79.95(12)
91.03(14)
97.69(12)
176.8(2)
95.09(14)
84.91(14)
180.0
84.5(2)
91.03(14)
97.69(12)
91.15(14)
79.95(12)
180.0
O1wb-Co1-O2
O1-Co1-O2b
O2-Co1-O2b
N1-Cu1-O3a
N1-Cu1-O3
84.91(14)
95.09(14)
O3a-Cu1-O3
ꢀ
The bond lengths around the cobalt atom (Co1 O1w
[a] Estimated standard deviations are given in parentheses. [b] Symmetry
codes: a=ꢀx+1/2, ꢀy+1/2, ꢀz; b=ꢀx, y, ꢀz+1/2.
ꢀ
ꢀ
2.073(5), Co1 O1 2.118(3), Co1 O2 2.082(3) Š) are similar
to those observed for the high-spin CoII ion in the related
oxamato-bridged cobalt(II)–copper(II) chain [CoCu
(dmso)3] (opba=o-phenylenebis
(oxamate)).[14a] The intra-
chain Co1···Cu1 separation through the oxamato bridge in
A
G
U
2b’ is 5.296(1) Š, a value which compares well with that re-
ported for [CoCu
A
E
2058
www.chemeurj.org
ꢁ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 2054 – 2066

Heterobimetallic Single-Chain Magnets
FULL PAPER
¯
The chains of 2b’ grow parallel to the [101] direction and
they are rather well separated from each other (Figure 2b).
The phenyl rings are practically perpendicular to the oxa-
mate groups (the dihedral angle between their mean planes
is 85.6(2)8) and thus afford effective shielding between the
copper atoms of neighbouring chains in the ac plane. How-
ever, the cis conformation of the octahedral cobalt units pre-
vents the aromatic rings from arranging alternately in the
two perpendicular directions to the basal plane of the
copper atoms, and this facilitates approach between the
cobalt atoms of neighbouring chains (Figure 2c). This leads
to weak interchain hydrogen-bonding interactions involving
the coordinated water molecules and the carbonyl oxygen
atoms of the oxamate ligands (O1w···O3c 2.867(4) Š, Fig-
ure 2c). The shortest interchain Co1···Co1c and Cu1···Co1c
distances are 5.995(5) and 6.628(3) Š, respectively (symme-
try code: c=ꢀx, ꢀy, 1ꢀz).
A
N
neutral oxamato-bridged cobalt(II)–copper(II) linear chains
(Figure 3a) and water molecules of crystallisation. Within
each chain, the bis(oxamato)copper(II) entity acts as a bis-
E
bidentate ligand through the cis carbonyl oxygen atoms to-
wards trans-diaquacobalt(II) units. This situation contrasts
with that of 2b’, where cis coordination of the two water
molecules to cobalt atoms leads to a zigzag chain structure.
A precedent exists in the literature of two related oxamato-
bridged manganese(II)–copper(II) chains [MnCu(opba)-
G
A
G
A
linear and zigzag structures, respectively.[14b,c] Selected bond
lengths and angles for 3b’ are summarised in Table 3.
Table 3. Selected bond lengths [Š] and angles [8] of 3b’.[a,b]
ꢀ
ꢀ
Cu1 N1
1.949(3)
1.958(3)
2.066(3)
2.109(3)
2.158(5)
180.0
84.55(12)
95.45(12)
180.0
81.07(12)
98.93(12)
90.0
90.0
90.0
Cu1 N1a
1.949(3)
1.958(3)
2.066(3)
2.109(3)
2.158(5)
95.45(12)
84.55(12)
180.0
98.93(12)
81.07(12)
180.0
90.0
90.0
ꢀ
ꢀ
Cu1 O3
Cu1 O3a
ꢀ
ꢀ
Co1 O1
Co1 O1b
ꢀ
ꢀ
Co1 O2
Co1 O2b
ꢀ
ꢀ
Co1 O1w
Co1 O1wb
N1-Cu1-N1a
N1-Cu1-O3a
N1-Cu1-O3
N1a-Cu1-O3a
N1a-Cu1-O3
O3a-Cu1-O3
O1-Co1-O2b
O1-Co1-O2
O2b-Co1-O2
O1b-Co1-O1wb
O2-Co1-O1wb
O1b-Co1-O1w
O2-Co1-O1w
O1-Co1-O1b
O1b-Co1-O2b
O1b-Co1-O2
O1-Co1-O1wb
O2b-Co1-O1wb
O1-Co1-O1w
O2b-Co1-O1w
O1wb-Co1-O1w
90.0
90.0
90.0
180.0
[a] Estimated standard deviations are given in parentheses. [b] Symmetry
codes: a=ꢀx, y, ꢀz; b=ꢀ x, y, 1ꢀz.
The copper atom is four-coordinate with two amidate-ni-
trogen and two carboxylate-oxygen atoms from the two oxa-
mate ligands in a trans arrangement building a square-
planar environment. The bond lengths around the copper
ꢀ
ꢀ
atom (Cu1 N1 1.949(3), Cu1 O3 1.958(3) Š) are close to
those of the monomeric precursor Na2[Cu(L3)2]·6H2O. The
cobalt atom is six-coordinate with two trans-coordinated
water molecules and four carbonyl oxygen atoms from two
oxamate ligands forming a distorted octahedral surrounding.
ꢀ
The bond lengths around the cobalt atom (Co1 O1w
Figure 3. a) View of a fragment of the chain of 3b’ with atom labelling
for the metal coordination environments. Hydrogen atoms are omitted
for clarity. b) and c) Crystal packing of the chains of 3b’ along the [001]
and [111] directions. Hydrogen bonds are represented by dashed lines
(symmetry codes: a=ꢀx, y, ꢀz; b=ꢀx, y, 1ꢀz; c=x, 1+y, z; d=1+x, y,
z; e=1/2ꢀx, 1/2ꢀy, ꢀz).
ꢀ
ꢀ
2.158(5), Co1 O1 2.109(3), Co1 O2 2.066(3) Š) are similar
to those observed for 2b’.
The chains of 3b’ grow parallel to the [001] direction and
they are well separated from each other (Figure 3b). The
phenyl rings are practically perpendicular to the oxamate
Chem. Eur. J. 2007, 13, 2054 – 2066
ꢁ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2059

F. Lloret et al.
groups (the dihedral angle between their mean planes is
89.7(2)8) and thus afford effective shielding between neigh-
bouring chains in the ab plane. There are, however, some
weak interchain hydrogen-bonding interactions along the b
axis in 3b’ involving the coordinated and the crystallisation
water molecules (O1w···O2w 2.878(6) Š), which form a hy-
drogen-bonded square motif (Figure 3c). The intrachain
Co1···Cu1 distance across bridging oxamato ligand in 3b’ is
5.3010(15) Š, a value which is much shorter than the short-
est interchain metal–metal separation of 8.702(3) Š for
Co1···Co1c and Cu1···Cu1c (symmetry code: c=x, 1+y, z).
g=2.0) in magnetic isolation. This suggests relatively large
intrachain antiferromagnetic interaction between the CuII
and MnII ions through the oxamate bridge. On cooling, cMT
decreases and attains a minimum around 125 K for 1a and
2a and at 105 K for 3a (inset of Figure 4a). The presence of
a minimum is characteristic of ferrimagnetic manganese(II)–
copper(II) chain compounds.[3a,14b,c] The cMT value reaches a
maximum at about 2.0 K for 1a and 2a and at 6.8 K for 3a,
with cMT values in the range 10.5–24.0 cm3 molꢀ1 K, due to
saturation effects. The lack of a maximum in cM allows us to
rule out the occurrence of 3D long-range antiferromagnetic
order and thus suggests that the chains are magnetically well
isolated.
Magnetic properties
The M versus H plots for 1a–3a at 2.0 K (M is the mag-
netisation per MnCu pair and H the applied magnetic field)
are shown in Figure 4b. The magnetisation values at 5 T are
in the range 3.70–3.95Nb and they are consistent with the
predicted value (Ms =3.90Nb) for an S=2 state resulting
from antiferromagnetic coupling between a high-spin MnII
ion (SMn =5/2 with g=2.0) and a CuII ion (SCu =1/2 with g=
2.1). Moreover, the magnetisation isotherms show fast satu-
ration with about 90% of Ms being reached in a field of
1000 G. This reveals strong short-range correlation along the
chain favouring antiparallel alignment of the spins of CuII
and MnII ions.
Manganese(II)–copper(II) chains: The dc magnetic proper-
ties of 1a–3a in the form of cMT versus T plots (cM is the
magnetic susceptibility per MnCu pair) are shown in Fig-
ure 4a. The values of cMT at room temperature vary in the
The magnetic susceptibility data of 1a–3a were analysed
by using the one-dimensional model developed by Kahn
et al.[3a] In this model, the MnII and CuII ions are treated as
classic (SMn =5/2) and quantum (SCu =1/2) spins, respective-
P
ly. The spin Hamiltonian is H=
bH(gMn
[ꢀJSMn,iSCu,i
+
i
S
Mn,i+gCuSCu,i)], where i runs over the MnCu units, J
is the exchange coupling between neighbouring spins and
gMn and gCu are the LandØ factors. Least-squares fitting of
the experimental data through this model in the tempera-
ture range 30–300 K gave ꢀJ values of 27.9 (1a), 28.2 (2a)
and 24.7 cmꢀ1 (3a; Table 4). The theoretical curves repro-
II
Table 4. Selected dc magnetic data for the Mn CuII chain compounds.
[a]
Complex
ꢀJ [cmꢀ1
]
gMn
gCu
1a
2a
3a
27.9
28.2
24.7
2.00
2.00
2.00
2.09
2.07
2.06
[a] J is the exchange coupling parameter.
duce very well the observed minima in the cMT versus T
plots (solid lines in the inset of Figure 4a). The antiferro-
magnetic coupling between CuII and MnII ions through the
oxamate bridge in 1a and 2a is somewhat stronger than in
3a. This suggests that the copper atoms in 1a and 2a are
square-planar, whereas that in 3a is square-pyramidal with
an apical DMSO ligand, as previously reported for the relat-
*
Figure 4. a) Temperature dependence of cMT of 1a ( ), 2a (&) and 3a
~
(
) in an applied magnetic field of 1 T( Tꢂ50 K) and 250 G (T<50 K).
The inset shows the minimum of cMT. The solid lines are the best-fit
~
)
*
curves (see text). b) Field dependence of M of 1a ( ), 2a (&) and 3a (
at 2.0 K. The solid lines are guides for the eye.
narrow range 4.31–4.35 cm3 molꢀ1 K, and they are lower
ed chain [MnCu
N
N
than expected for the sum of a square-planar CuII ion (SCu
=
copper atom lies in the oxamato plane (CuN2O2 chromo-
phore), whereas it is out of this mean plane in the latter
case (CuN2O3 chromophore). This leads to better overlap of
1/2, cMT=0.40 cm3 molꢀ1 K with g=2.1) and an octahedral
high-spin MnII ion (SMn =5/2, cMT=4.38 cm3 molꢀ1 K with
2060
www.chemeurj.org
ꢁ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 2054 – 2066

Heterobimetallic Single-Chain Magnets
FULL PAPER
the magnetic orbitals of CuII and MnII ions through the s in-
plane exchange pathway of the oxamate bridge, and thus to
a larger antiferromagnetic coupling.
The ac magnetic measurements for 1a–3a showed no evi-
dence of slow magnetic-relaxation effects. In fact, no out-of-
phase magnetic susceptibility signals were observed above
2.0 K for these compounds (data not shown). This is likely
due to the magnetically isotropic character of the octahedral
high-spin MnII ion (6A1g, D value close to zero).
Cobalt(II)–copper(II) chains: The dc magnetic properties of
1b–3b in the form of cMT versus T plots (cM is the magnetic
susceptibility per CoCu pair) are shown in Figure 5a, and
*
Figure 6. a) Temperature dependence of cMT of 2b’ ( ) and 3b’ (&) in an
applied magnetic field of 1 T( Tꢂ50 K) and 250 G (T<50 K). The inset
shows the minimum of cMT. The solid lines are the best-fit curves (see
*
text). b) Field dependence of M of 2b’ ( ) and 3b’ (&) at 2.0 K. The
solid lines are guides for the eye.
minimum in the temperature range 90–115 K (insets of Fig-
ures 5a and 6a), which is typical of ferrimagnetic cobalt(II)–
copper(II) chain compounds.[3b,14a] The cMT value reaches a
maximum at about 2–9 K, with cMT values in the range 7.5–
40.0 cm3 molꢀ1 K, which is due to magnetic anisotropy and
saturation effects. As for their manganese(II)–copper(II) an-
alogues 1a–3a, the lack of a maximum in cM allows us to
rule out 3D long-range antiferromagnetic order and thus
suggests that the chains are magnetically well isolated from
each other.
*
Figure 5. a) Temperature dependence of cMT of 1b ( ), 2b (&) and 3b
~
(
) in an applied magnetic field of 1 T( Tꢂ50 K) and 250 G (T<50 K).
The inset shows the minimum of cMT. The solid lines are the best-fit
~
)
*
curves (see text). b) Field dependence of M of 1b ( ), 2b (&) and 3b (
The M versus H plots for 1b–3b at 2.0 K (M is the mag-
netisation per CoCu pair) are shown in Figure 5b, and those
of 2b’ and 3b’ in Figure 6b. The magnetisation values at 5 T
vary in the range 0.98–1.10Nb and they are consistent with
the predicted value (Ms =1.10Nb) for partial spin cancella-
tion resulting from antiferromagnetic coupling between a
high-spin CoII ion (Seff =1/2 with g=4.2) and a CuII ion (S=
1/2 with g=2.1). Moreover, the magnetisation isotherms
show fast saturation with about 80% of the value of MS
being reached in a field of 1000 G. This reveals a strong
short-range correlation along the chain favouring antiparal-
lel alignment of the spins of CuII and CoII ions.
at 2.0 K. The solid lines are guides to the eye.
those of 2b’ and 3b’ in Figure 6a. The cMT values at room
temperature vary in the range 2.42–2.70 cm3 molꢀ1 K and
they are smaller than expected for the sum of a square-
planar CuII ion (SCu =1/2, cMT=0.40 cm3 molꢀ1 K with g=
2.1) and an octahedral high-spin CoII ion (SCo =3/2) with an
4
orbitally degenerate T1g single-ion ground state (cMT=2.5–
3.0 cm3 molꢀ1 K)[17] in magnetic isolation. This suggests the
occurrence of a relatively large intrachain antiferromagnetic
interaction between the CuII and the CoII ions through the
oxamate bridge. On cooling, cMT decreases and it attains a
To analyse the magnetic susceptibility data of 1b–3b, 2b’
and 3b’, in which the octahedral high-spin CoII ions are orbi-
Chem. Eur. J. 2007, 13, 2054 – 2066
ꢁ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2061

F. Lloret et al.
tally degenerate (4T1g), we used the branch chain model pre-
viously developed by Kahn et al. for the related oxamato-
netic susceptibilities per CoCu pair, respectively) are shown
in Figures 7–10. The cM’’ value becomes nonzero below
3.5 K for 2b, and shows a maximum at 2.1 K for the highest
bridged chain [CoCu
N
E
assumes that only z components of spin and orbital momen-
ta are coupled and that the applied magnetic field is along
the quantisation axis. The corresponding Hamiltonian is
P
H=
A
S
Cu,i(z) +J’LCo,i(z)
S
Co,i(z) +DL2Co,iðzÞꢀbH(z)-
i
(gCoSCo,i(z) +gCuSCu,i(z) +kLCo,i(z))], where i runs over the CoCu
units, J and J’ are the exchange and spin-orbit coupling pa-
rameters, k and D the orbital reduction and local anisotropy
parameters of the cobalt(II) ion, and gCo and gCu the LandØ
factors. Least-squares fitting of the experimental data
through this model in the temperature range 30–300 K gave
ꢀJ values in the range 35.0–45.8 cmꢀ1 for 1b–3b, 2b’ and
3b’ (Table 5). The theoretical curves reproduce quite well
II
Table 5. Selected dc magnetic data for the Co CuII chain compounds.
[a]
[b]
[d]
Complex ꢀJ [cmꢀ1
]
gCo
gCu
ꢀl [cmꢀ1
]
k[c]
D [cmꢀ1
]
1b
2b
3b
2b’
3b’
40.3
40.5
44.3
45.8
35.0
2.24 2.08 110
2.30 2.07 107
0.98 538
0.97 719
0.95 710
0.94 692
0.80 610
2.21 2.08
2.24 2.09
99
82
2.38 2.05 110
[a] J is the exchange coupling parameter. [b] l is the spin–orbit coupling
parameter (A=3/2). [c] k is the orbital reduction parameter. [d] D is the
local anisotropy parameter.
the observed minima in the cMT versus T plots (solid lines
in the inset of Figures 5a and 6a). The antiferromagnetic
coupling between CuII and CoII ions through the oxamato
bridges in 1b–3b, 2b’ and 3b’ is somewhat stronger than
Figure 7. Temperature dependence of a) cM’ and b) cM’’ of 2b in zero ap-
plied static field and under 1 G oscillating field at different frequencies
~
*
*
^
^
of the oscillating field: 1 ( ), 10 ( ), 100 ( ), 333 ( ), 1000 Hz ( ). The
that reported for [CoCu
G
(H2O)3]·2H2O (ꢀJ=
solid lines are guides for the eye.
18.0 cmꢀ1).[3b] As mentioned above, this is likely due to the
fact that the copper atom in these chains lies in the oxamato
plane (CuN2O2 chromophore in a square-planar surround-
ing), as evidenced from the X-ray crystal structures of 2b’
and 3b’, whereas it lies out of this mean plane in [CoCu-
applied frequency (n=1000 Hz; Figure 7b). For the ana-
logue 2b’, cM’’ becomes nonzero below 3.0 K, but no cM’’
maxima are observed above 2.0 K (Figure 8b), which is the
lowest temperature in our susceptometer. On the contrary,
cM’’ shows maxima between 3.5 (n=1400 Hz) and 2.2 K (n=
300 Hz) for 3b (Figure 9b). For the analogue 3b’, the cM’’
maxima are located at lower temperatures between 2.3 (n=
1400 Hz) and 2.0 K (n=300 Hz; Figure 10b). Overall, the
temperature Tmax of the maxima of cM’’ along this series of
cobalt(II)–copper(II)–Lx chain compounds varies according
to the steric hindrance of the aromatically substituted oxa-
mate ligand in the series L1 <L2 <L3. It appears that the in-
terchain interactions, which are always present whatever the
compound, as revealed by the crystal structures of 2b’ and
3b’, may tune the dynamic magnetic properties by lowering
the blocking temperature with respect to that expected for a
perfectly isolated chain. Although this hypothesis has no
precedent in the relatively recent literature on SCMs,[5,9–13]
there are some studies that point out the influence of inter-
molecular interactions on the dynamic behaviour of related
SMMs.[18–20]
A
N
pyramidal environment). The effective spin–orbit coupling
can be related with the spin–orbit coupling parameter l
through the expression J’=ꢀAkl. This gives values of ꢀl in
the range 82–110 cmꢀ1 (with A=3/2) for 1b-3b, 2b’ and 3b’
(Table 5; cf. l=ꢀ180 cmꢀ1 for the free ion). The local mag-
netic anisotropy parameter D is related to the energy split-
4
ting of the orbital triplet ground state T1g resulting from
axial distortion of the octahedral geometry of the CoII ion.
The large values of D in the range 538–719 cmꢀ1 for 1b-3b,
2b’ and 3b’ (Table 5) are similar to those reported for other
axially distorted six-coordinate high-spin cobalt(II) com-
plexes.[17]
The ac magnetic properties of this series of cobalt(II)–
copper(II) chain compounds show evidence of slow magnet-
ic-relaxation effects which are typical of SCMs in all cases
except for 1b. T hecM’ and cM’’ versus T plots for 2b, 2b’, 3b
and 3b’ (cM’ and cM’’ are the in-phase and out-of-phase mag-
2062
www.chemeurj.org
ꢁ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 2054 – 2066

Heterobimetallic Single-Chain Magnets
FULL PAPER
Figure 8. Temperature dependence of a)cM’ and b) cM’’ of 2b’ in zero ap-
plied static field and under 1 G oscillating field at different frequencies
Figure 9. Temperature dependence of a) cM’ and b) cM’’ of 3b in zero ap-
plied static field and under 1 G oscillating field at different frequencies
~
*
*
^
^
of the oscillating field: 10 ( ), 30 ( ), 100 ( ), 333 ( ), 1000 Hz ( ). The
~
~
*
*
^
^
of the oscillating field: 0.1 ( ), 1 ( ), 10 ( ), 50 ( ), 75 ( ), 100 ( ), 200
solid lines are guides for the eye.
!
!
&
( ), 400 ( ), 700 ( ), 1000 ("), 1400 Hz ("). The solid lines are guides
for the eye.
The relaxation time t for 3b and 3b’ can be calculated
from the maximum of cM’’ at a given frequency n, whereby
it is assumed that switching of the oscillating ac field match-
es the relaxation rate of the magnetisation (1/t=2pn). The
calculated t values at Tmax follow the Arrhenius law charac-
relative variation of the temperature of the maximum of cM’’
with respect to the frequency, expressed by the so called F
parameter defined by Mydosh as F=(DTmax/Tmax)/D
(lgn),[22b]
A
is 0.25 for 3b and 0.12 for 3b’ (Table 6). These moderate
values of F are typical of SCMs and they are greater than
those expected for spin glasses (F values smaller than
0.01).[22c]
teristic of a thermally activated mechanism: t=t0 exp(Ea/
A
kBT) (Figure 11a). The activation energies Ea are 38.0 (3b)
and 16.3 cmꢀ1 (3b’), and the preexponential factors t0 2.3”
10ꢀ11 (3b) and 4.0”10ꢀ9 s (3b’; Table 6). As expected, the
larger the antiferromagnetic intrachain coupling [ꢀJ=44.3
(3b) and 35.0 cmꢀ1 (3b’)], the greater the activation energy
to reverse the magnetisation direction. However, these
values must be regarded with caution because of the very
narrow temperature range available for both cobalt(II)–cop-
per(II)–L3 chain compounds. Indeed, the cM’’ versus cM’ plot
at 2.0 K for 3b and 3b’ (Cole–Cole plot[21a]) in the frequency
range 1–1400 Hz gives a semicircle in both cases (Fig-
ure 11b). The least-squares fit of the experimental data
through the Cole–Cole expression[21], c=cS +(cTꢀcS)/[1+
(iwt)1ꢀa] gave very small values for the a parameter of 0.10
for 3b and 0.05 for 3b’ (a=0 for an ideal Debye model
with a single relaxation time); thus, spin-glass behaviour is
ruled out.[22] The adiabatic (cS) and isothermal (cT) suscepti-
bilities are 0.5 and 3.1 cm3 molꢀ1 for 3b and 0.1 and
4.7 cm3 molꢀ1 for 3b’, respectively (Table 6). Moreover, the
Conclusion
A new family of oxamato-bridged heterobimetallic chains
has been obtained through ligand design from the self-as-
sembly of potentially bis-bidentate square-planar copper(II)
complexes with sterically hindered aromatically substituted
oxamate ligands and transition metal ions such as mangane-
se(II) and cobalt(II). Both manganese(II)–copper(II) and
cobalt(II)–copper(II) chains behave as ferrimagnetic chains
with moderately strong intrachain antiferromagnetic cou-
pling between CuII and MII ions (M=Mn and Co) through
the oxamate bridge. In spite of this, only the CoIICuII chains
show slow magnetic relaxation effects which are typical of
SCMs. Together with the large magnetic anisotropy of the
orbitally degenerate octahedral high-spin CoII ion, however,
Chem. Eur. J. 2007, 13, 2054 – 2066
ꢁ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2063

F. Lloret et al.
Figure 10. Temperature dependence of a) cM’ and b) cM’’ of 3b’ in zero
Figure 11. a) Arrhenius plots and b) Cole–Cole plots at 2.0 K for 3b (&)
applied static field and under 1 G oscillating field at different frequencies
!
*
&
and 3b’ ( ). The solid lines are the best-fit curves (see text).
*
^
^
of the oscillating field: 50 ( ), 75 ( ), 100 ( ), 200 ( ), 300 (&), 400 ( ),
!
700 ( ), 1000 ("), 1400 Hz ("). The solid lines are guides for the eye.
ed hypothesis must still be supported by new experimental
and theoretical works, we guess that the implications that
can be derived from it may be relevant to some of the pro-
posed mechanisms for the slow magnetic relaxation of
SCMs in the near future. We are currently exploring the lim-
itations of this strategy by playing with the steric require-
ments of the substituents in the copper(II) precursor in
order to obtain new members of this family of oxamato-
bridged heterobimetallic SCMs with a higher blocking tem-
perature.
II
Table 6. Selected ac magnetic data for Co CuII chain compounds 3b and
3b’.
Complex t0 [s][a]
Ea
[cmꢀ1
a[c] cS
[cm3 molꢀ1
cT
F[f]
[b]
[d]
[e]
G
]
]
[cm3 molꢀ1
]
3b
3b’
2.3”10ꢀ11 38.0
4.0”10ꢀ9 16.3
0.10 0.5
0.05 0.1
3.1
4.7
0.25
0.12
[a] t0 is the preexponential factor. [b] Ea is the activation energy. [c] a is
the Cole–Cole parameter. [d] cS is the adiabatic susceptibility. [e] cT is the
isothermal susceptibility. [f] F is the Mydosh parameter.
certain steric requirements in the CuII precursor must be ful-
filled in order to observe SCM behaviour at sufficiently high
temperatures (above 2.0 K). Thus, the overall increase in
Tmax observed in the series of CoIICuII chains when increas-
ing the number of methyl substituents in the benzene ring
from one to three is related to the increased separation be-
tween chains and hence decreased interchain interactions.
Moreover, the nature of the coordinated solvent also influ-
ences the SCM behaviour, with higher values of Tmax for the
series of CoIICuII chains with DMSO compared to those
with H2O, probably due to a better isolation of the chains
because of the larger size of the former. At present, there
are no reports on SCMs showing that the dynamics of mag-
netic relaxation is affected by interchain interactions, as our
results suggest. Although the relevance of this unprecedent-
Experimental Section
Reagents: Metal nitrates, sodium hydroxide, ethyl chlorooxoacetate and
the corresponding aniline derivatives were purchased from commercial
sources and used as received. Elemental analyses (C, H, N, S) were car-
ried out by the Microanalytical Service of the Universitat de Val›ncia.
N-2-methylphenyloxamate (H2L1), N-2,6-dimethylphenyloxamate (H2L2)
and N-2,4,6-trimethylphenyloxamate (H2L3): The ethyl ester derivative of
the proligands H2Lx (x=1–3) were synthesised by following a standard
procedure.[16] In a typical preparation, the corresponding aniline deriva-
tive (60 mmol) was treated with ethyl chlorooxoacetate (7.0 mL,
60 mmol) in THF (150 mL) at 08C under continuous stirring. The result-
ing solution was refluxed for 1 h and the solvent was removed on a rota-
tory evaporator to afford an oil which solidified when left at room tem-
perature. The white solid obtained was filtered off, washed with a small
amount of diethyl ether and dried under vacuum.
2064
www.chemeurj.org
ꢁ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 2054 – 2066

Heterobimetallic Single-Chain Magnets
FULL PAPER
Na2[Cu(L1)2]·2H2O, Na2[Cu(L2)2]·2H2O and Na2[Cu(L3)2]·6H2O:
T
he
paper. These data can be obtained free of charge from the Cambridge
Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_request/cif.
sodium salts of the copper(II)–Lx (x=1–3) precursors were prepared in a
standardised manner.[16] In a typical experiment, the corresponding ethyl
ester ligand derivative of H2Lx (x=1–3) (10 mmol) was suspended in
water (25 mL) and allowed to react with an aqueous solution (25 mL) of
NaOH (1.0 g, 25 mmol). A solution of Cu(NO3)2·3H2O (1.21 g, 5 mmol)
A
Acknowledgements
in water (25 mL) was added dropwise to the resulting colourless solution
at room temperature under continuous stirring. The resulting deep green
solution was then filtered through paper to remove the small amount of
solid particles. The solvent was reduced to one-fourth volume on a rota-
tory evaporator and a solid was formed. The green polycrystalline solid
was filtered off, washed with acetone and diethyl ether and dried under
vacuum. Well-formed prisms of Na2[Cu(L3)2]·6H2O suitable for X-ray
diffraction were obtained by slow diffusion of methanol into the aqueous
mother liquor.
This work was supported by the MEC (Projects CTQU2004-03633 and
MAT2004-03112), the CNRS and the CAPES (Project COFECUB 460/
04), the European Union through the Network QuEMolNa (Project
MRTN-CT-2003-504880) and the Magmanet Network of Excellence
(Contract 515767-2), the Generalitat Valenciana (Grupos 03/197) and the
Gobierno Autónomo de Canarias (PI2002/175). E.P. and F.D. thank the
Spanish Ministry of Science and Education for a grant.
A
(1a),
[MnCu(L2)2]·2DMSO
(2a)
and
[1] a) O. Kahn, Molecular Magnetism, VCH, New York, 1993; b) O.
Kahn, Adv. Inorg. Chem. 1995, 43, 179.
[2] a) A. Caneschi, D. Gatteschi, R. Sessoli, P. Rey, Acc. Chem. Res.
1989, 22, 392; b) A. Caneschi, D. Gatteschi, P. Rey, Prog. Inorg.
Chem. 1991, 39, 331.
a
ACHTREUNG
(0.062 g, 0.25 mmol) was dissolved in hot DMSO (10 mL) (708C) and
added dropwise to a solution of the corresponding sodium salt of the cop-
per(II)–Lx (x=1–3) precursor (0.25 mmol) dissolved in hot DMSO
(10 mL). The resulting dark green solution was filtered while hot and the
filtrate was allowed to stand at room temperature. After several days, a
brownish green polycrystalline solid appeared which was collected by fil-
tration and air-dried.
[3] a) O. Kahn, Y. Pei, M. Verdaguer, J. P. Renard, J. Sletten, J. Am.
Chem. Soc. 1988, 110, 782; b) P. J. Van Kroningsbruggen, O. Kahn,
K. Nakatani, Y. Pei, J. P. Renard, M. Drillon, P. Legoll, Inorg. Chem.
1990, 29, 3325; c) F. Lloret, K. Nakatani, Y. Journaux, O. Kahn, Y.
Pei, J. P. Renard, J. Chem. Soc. Chem. Commun. 1988, 642; d) K.
Nakatani, J. Y. Carriat, Y. Journaux, O. Kahn, F. Lloret, J. P. Renard,
Y. Pei, J. Sletten, M. Verdaguer, J. Am. Chem. Soc. 1989, 111, 5739;
e) F. Lloret, M. Julve, R. Ruiz, Y. Journaux, K. Nakatani, O. Kahn,
J. Sletten, Inorg. Chem. 1993, 32, 27.
[4] a) A. Caneschi, D. Gatteschi, J. P. Renard, P. Rey, R. Sessoli, Inorg.
Chem. 1989, 28, 1976; b) A. Caneschi, D. Gatteschi, J. P. Renard, P.
Rey, R. Sessoli, Inorg. Chem. 1989, 28, 2940; c) A. Caneschi, D. Gat-
teschi, J. P. Renard, P. Rey, R. Sessoli, Inorg. Chem. 1989, 28, 3314.
[5] a) A. Caneschi, D. Gatteschi, N. Lalioti, C. Sangregorio, R. Sessoli,
G. Venturi, A. Vindigni, A. Rettori, M. G. Pini, M. Novak, Angew.
Chem. 2001, 113, 1810; Angew. Chem. Int. Ed. 2001, 40, 1760; b) A.
Caneschi, D. Gatteschi, N. Lalioti, R. Sessoli, L. Sorace, V. Tangou-
lis, A. Vindigni, Chem. Eur. J. 2002, 8, 286; c) A. Caneschi, D. Gatte-
schi, N. Lalioti, C. Sangregorio, R. Sessoli, G. Venturi, A. Vindigni,
A. Rettori, M. G. Pini, M. Novak, Europhys. Lett. 2001, 58, 771;
d) 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. 2004, 92, 207204.
[CoCu(L1)2]·3DMSO (1b), [CoCu(L2)2]·3DMSO (2b) and [Co-
Cu(L3)2]·3DMSO (3b): Complexes 1b–3b were prepared as indicated for
1a–3a but with manganese(II) nitrate instead of cobalt(II) nitrate. In a
typical experiment, Co(NO3)2·6H2O (0.073 g, 0.25 mmol) was dissolved
G
in hot DMSO (10 mL) (708C) and added dropwise to a solution of the
corresponding sodium salt of the copper(II)–Lx (x=1–3) precursor
(0.25 mmol) in hot DMSO (10 mL). The resulting dark green solution
was filtered while hot, and the filtrate was allowed to stand at room tem-
perature. After several days, a green polycrystalline solid appeared which
was collected by filtration and air-dried.
[CoCu(L2)2
N
G
formed deep green octahedral prisms of 2b’ and 3b’ suitable for X-ray
diffraction were obtained by slow diffusion in an H-shaped tube of aque-
ous solutions containing stoichiometric amounts of the corresponding
sodium salt of the copper(II)–Lx (x=2 and 3) precursor (0.25 mmol) in
one arm, and Co(NO3)2·6H2O (0.073 g, 0.25 mmol) in the other. They
were isolated by filtration on paper and air-dried.
Physical techniques: 1H NMR spectra were recorded at 250 MHz on a
Bruker AC 250 spectrometer. Chemical shifts are reported in d [ppm]
versus SiMe4 with the residual proton signals of deuterated DMSO sol-
vent as internal standard. IR spectra were recorded on a Perkin-Elmer
882 spectrophotometer with KBr pellets. Variable-temperature dc and ac
magnetic susceptibility measurements and variable-field magnetisation
measurements were carried out on polycrystalline samples with a Quan-
tum Design SQUID magnetometer. The susceptibility data were correct-
ed for the diamagnetism of the constituent atoms and the sample holder.
U
[6] R. Lescouzec, L. M. Toma, J. Vaissermann, M. Verdaguer, F. S.
Delgado, C. Ruiz-Perez, F. Lloret, M. Julve, Coord. Chem. Rev.
2005, 249, 2691.
[7] R. J. Glauber, J. Math. Phys. 1963, 4, 294.
[8] a) L. Thomas, F. Lionti, R. Ballou, D. Gatteschi, R. Sessoli, B. Bar-
bara, Nature 1996, 383, 145; b) J. R. Friedman, M. P. Sarachik, J.
Tejada, R. Ziolo, Phys. Rev. Lett. 1996, 76, 3830; c) C. Coulon, R.
ClØrac, L. Lecren, W. Wernsdorfer, H. Miyasaka, Phys. Rev. B 2004,
69, 132408.
[9] a) R. ClØrac, H. Miyasaka, M. Yamashita, C. Coulon, J. Am. Chem.
Soc. 2002, 124, 12837; b) H. Miyasaka, R. ClØrac, K. Mizushima, K.
Sugiura, M. Yamashita, W. Wernsdorfer, C. Coulon, Inorg. Chem.
2003, 42, 8203.
[10] a) R. Lescouzec, J. Vaissermann, C. Ruiz-PØrez, F. Lloret, R. Car-
rasco, M. Julve, M. Verdaguer, Y. Dromzee, D. Gatteschi, W. Werns-
dorfer, Angew. Chem. 2003, 115, 1521; Angew. Chem. Int. Ed. 2003,
42, 1483; b) L. M. Toma, R. Lescouzec, F. Lloret, M. Julve, J. Vais-
sermann, M. Verdaguer, Chem. Commun. 2003, 1850; c) L. M.
Toma, R. Lescouzec, J. Pasµn, C. Ruiz-PØrez, J. Vaissermann, J.
Cano, R. Carrasco, W. Wernsdorfer, F. Lloret, M. Julve, J. Am.
Chem. Soc. 2006, 128, 4842.
X-ray crystal structure determination and refinement: The structures of
Na2[Cu(L3)2]·6H2O, 2b’ and 3b’ were solved by direct methods and re-
fined with full-matrix least-squares technique on F2 using the SHELXS-
97 and SHELXL-97 programs.[23a] Data collection and data reduction
were done with the COLLECT[23b] and EVALCCD[23c] programs. Empiri-
cal absorption corrections were carried out using SADABS[23d] for all
compounds. All calculations for data reduction, structure solution and re-
finement were done by standard procedures (WINGX).[23e] The final geo-
metrical calculations and the graphical manipulations were carried out
with PARST97[23f] and CRYSTAL MAKER
programs, respectively.
[23g]
The hydrogen atoms of the organic ligands were calculated and refined
with isotropic thermal parameters, while those from the water molecules
were neither found nor calculated. Crystallographic data for
Na2[Cu(L3)2]·6H2O, 2b’ and 3b’ are listed in Table S4 in the Supporting
Information.
CCDC-611876 (Na2[Cu(L3)2]·6H2O), CCDC-611877 (2b’) and CCDC-
227520 (3b’) contain the supplementary crystallographic data for this
[11] T. Liu, D. Fu, S. Gao, Y. Zhang, H. Sun, G. Su, Y. Liu, J. Am. Chem.
Soc. 2003, 125, 13976.
[12] S. Wang, J. L. Zuo, S. Gao, Y. Song, H. C. Zhou, Y. Z. Zhang, X. Z.
You, J. Am. Chem. Soc. 2004, 126, 8900.
Chem. Eur. J. 2007, 13, 2054 – 2066
ꢁ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2065

F. Lloret et al.
[13] M. Ferbinteanu, H. Miyasaka, W. Wernsdorfer, K. Nakata, K. Su-
giura, M. Yamashita, C. Coulon, R. ClØrac, J. Am. Chem. Soc. 2005,
127, 3090.
[14] a) H. O. Stumpf, C. L. M. Pereira, A. C. Doriguetto, C. Konzen,
L. C. B. Meira, N. G. Fernandes, Y. P. Mascarenhas, J. Ellena, M.
Knobel, Eur. J. Inorg. Chem. 2005, 24, 5018; b) H. O. Stumpf, Y. Pei,
L. Ouahab, F. Leberre, E. Codjovi, O. Kahn, Inorg. Chem. 1993, 32,
5687; c) H. O. Stumpf, Y. Pei, O. Kahn, J. Sletten, J. P. Renard, J.
Am. Chem. Soc. 1993, 115, 6738.
[15] E. Pardo, R. Ruiz-García, F. Lloret, J. Faus, M. Julve, Y. Journaux,
F. Delgado, C. Ruiz-PØrez, M. C. MuÇoz, Adv. Mater. 2004, 16, 1597.
[16] B. Cervera, J. L. Sanz, M. J. IbµÇez, G. Villa, F. Lloret, M. Julve, R.
Ruiz, X. Ottenwaelder, A. Aukauloo, S. Poussereau, Y. Journaux,
M. C. MuÇoz, J. Chem. Soc. Dalton Trans. 1998, 781.
[17] a) J. M. Herrera, A. Bleuzen, Y. Dromzee, M. Julve, F. Lloret, M.
Verdaguer, Inorg. Chem. 2003, 42, 7052; b) D. Maspoch, N. Domi-
ngo, D. Ruiz-Molina, K. Wurst, J. M. Hernµndez, G. Vaughan, C.
Rovira, F. Lloret, J. Tejada, J. Veciana, Chem. Commun. 2005, 5035.
[18] a) W. Wernsdorfer, N. Aliaga-Alcalde, D. N. Hendrickson, G. Chris-
tou, Nature 2002, 416, 406; b) W. Wernsdorfer, S. Bhaduri, R. Tiron,
D. N. Hendrickson, G. Christou, Phys. Rev. Lett. 2002, 89, 197201;
c) K. Park, M. R. Pederson, S. L. Richardson, N. Aliaga-Alcalde, G.
Christou, Phys. Rev. B 2003, 68, 020405; d) S. Hill, R. S. Edwards, N.
Aliaga-Alcalde, G. Christou, Science 2003, 302, 1015.
[20] a) H. Miyasaka, K. Nakata, K. Sugiura, M. Yamashita, R. ClØrac,
Angew. Chem. 2004, 116, 725; Angew. Chem. Int. Ed. 2004, 43, 707;
b) J. Yoo, W. Wernsdorfer, E. C. Yang, M. Nakano, A. L. Rheingold,
D. N. Hendrickson, Inorg. Chem. 2005, 44, 3377; c) H. Miyasaka, K.
Nakata, L. Lecren, C. Coulon, Y. Nakazawa, T. Fujisaki, K. Sugiur,
M. Yamashita, R. ClØrac, J. Am. Chem. Soc. 2006, 128, 3770.
[21] a) K. S. Cole, R. H. Cole, J. Chem. Phys. 1941, 9, 341; b) C. J. F.
Boettcher, Theory of Electric Polarization, Elsevier, Amsterdam,
1952; c) S. M. Aubin, Z. Sun, L. Pardi, J. Krzysteck, K. Folting, L. J.
Brunel, A. L. Rheingold, G. Christou, D. N. Hendrickson, Inorg.
Chem. 1999, 38, 5329.
[22] a) M. A. Girtu, C. M. Wynn, W. Fujita, K. Awaga, A. Epstein, J.
Phys. Rev. B 1998, 57, R11058; b) J. A. Mydosh, Spin Glasses: An
Experimental Introduction, Taylor & Francis, London, 1993; c) D.
Chowdhury, Spin Glasses and Other Frustrated Systems, Princeton
University Press, New Jersey, 1986; d) K. Moorjani, J. M. Coey,
Magnetic Glasses, Elsevier, New York, 1984; e) K. Binder, A. P.
Young, Rev. Mod. Phys. 1986, 58, 801.
[23] a) G. M. Sheldrick, SHELX97, Programs for Crystal Structure Anal-
ysis (Release 97-2), Universität Gçttingen, Gçttingen, Germany,
1998; b) R. W. W. Hooft, COLLECT, Nonius BV, Delft, The Nether-
lands, 1999; c) A. J. M. Duisenberg, L. M. J. Kroon-Batenburg,
A. M. M. Schreurs, J. Appl. Crystallogr. 2003, 36, 220 (EVALCCD);
d) SADABS, version 2.03, Bruker AXS Inc., Madison, WI, 2000;
e) L. J. Farrugia, J. Appl. Cryst. 1999, 32, 837; f) M. Nardelli, J. Appl.
Crystallogr. 1995, 28, 659; g) D. Palmer, CRYSTAL MAKER, Cam-
bridge University Technical Services, Cambridge, UK, 1996.
[19] a) R. Tiron, W. Wernsdorfer, N. Aliaga-Alcalde, G. Christou, Phys.
Rev.
B 2003, 68, 140407(R); b) R. Tiron, W. Wernsdorfer, D.
Foguet-Albiol, N. Aliaga-Alcalde, G. Christou, Phys. Rev. Lett. 2003,
91, 227203; c) C. Boskovic, R. Bircher, P. L. W. Tregenna-Piggott,
H. U. Güdel, C. Paulsen, W. Wernsdorfer, A. L. Barra, E. Khatsko,
A. Neels, H. Stoeckli-Evans, J. Am. Chem. Soc. 2003, 125, 14046.
Received: July 11, 2006
Published online: December 5, 2006
2066
www.chemeurj.org
ꢁ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 2054 – 2066