Syntheses, structures and magnetic properties of layered metal(II) mandelates

Article abstract of DOI:10.1002/ejic.200400623

A series of Mn^II, Fe^II, Co^II, Ni ^II and Cu^II layered mandelates has been synthesised in the form of crystalline powders by a hydrothermal route. The five compounds are isostructural with the copper analogue, also obtained in a single crystal form, except that the latter shows a strong Jahn-Teller distortion of the coordination spheres around the Cu^II centres. The complexes crystallise in the monoclinic centrosymmetric space group P2₁/α and their structures consist of layers of six-coordinate metal(II) ions interconnected through carboxylato bridges in a diamond-like network. The magnetic data indicate very small antiferromagnetic in-plane couplings for 1, 2, 3 and 5, with J/k_B = -0.122 K for 1. The Ni^II compound 4 exhibits ferromagnetic coupling within the layers (J/k_B = +1.213 K) and ferromagnetic 3D ordering can be observed at T_c = 2.7 K. Wiley-VCH Verlag GmbH & Co. KGaA, 2005.

Full text of DOI:10.1002/ejic.200400623

FULL PAPER
Syntheses, Structures and Magnetic Properties of Layered
Metal(II) Mandelates
Adel Beghidja,[a,b] Sylvain Hallynck,[a] Richard Welter,*[b] and Pierre Rabu*[a]
Keywords: Carboxylate ligands / Layered compounds / Magnetic properties / Transition metals / X-ray powder diffraction
A series of MnII, FeII, CoII, NiII and CuII layered mandelates
has been synthesised in the form of crystalline powders by a
hydrothermal route. The five compounds are isostructural
with the copper analogue, also obtained in a single crystal
form, except that the latter shows a strong Jahn−Teller distor-
tion of the coordination spheres around the CuII centres. The
complexes crystallise in the monoclinic centrosymmetric
space group P21/a and their structures consist of layers of six-
coordinate metal(II) ions interconnected through carboxylato
bridges in a diamond-like network. The magnetic data indi-
cate very small antiferromagnetic in-plane couplings for 1, 2,
3 and 5, with J/kB = −0.122 K for 1. The NiII compound 4
exhibits ferromagnetic coupling within the layers (J/kB
=
+1.213 K) and ferromagnetic 3D ordering can be observed at
TC = 2.7 K.
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2005)
Introduction
metal based layers.[14] Striking results have been obtained in
the field of low dimensional magnetic materials exhibiting
unusual behaviour, by carrying out solvothermal
reactions.[15Ϫ19] In particular, recent reports have been con-
cerned with the use of carboxylate moieties as bridging li-
gands between paramagnetic transition metal ions with the
versatility of the carboxylate coordination group leading to
magnetic networks with various topologies.[20Ϫ23]
Our recent investigation of CoII alkanoates [Co(O2C-
(CH2)nCH3)2(H2O)2]n has shown that for long alkane
chains, layered structures are obtained when the magnetic
layers consist of chains of CoII ions interconnected through
carboxylate bridges in a syn-anti configuration. This leads
to weak antiferromagnetic coupling.[24] On the other hand,
results on transition metal phenoxyalkanoate hydrates
[M(O2C(CH2)OC6H4R)2(H2O)2]n (M ϭ Mn, Co; R ϭ H,
Cl, F, ...), the structures of which consist of stacked
metal(ii) carboxylate layers separated by an organic sub-
network of phenyl rings, provided suitable examples of
square planar arrays of magnetic centres.[25,26] The
diaquabis(phenoxyacetato)manganese(ii) derivative exhibits
typical behaviour for an antiferromagnetic 2D square-
planar S ϭ 5/2 system down to TC ϭ 1.633 K where a 3D
ordered spin state leads to weak ferromagnetism. Magnetic
data for the CoII analogue suggest the presence of ferro-
magnetic interactions.[27]
Correlations between the magnetic properties of mol-
ecule-based solids and their structures together with the na-
ture of the chemical bond are useful in the design of new
magnetic materials.[1,2] In this respect, low dimensional sys-
tems in which the magnetic centres interact within chains
(1D) or planes (2D) are of particular interest since their
physics may be approximated and analysed to a high degree.
This has been possible thanks to the development of a sub-
stantial amount of experimental and theoretical work in the
area over several decades[3Ϫ6] and to the efforts of synthetic
chemists in providing model compounds, particularly
through the molecular approach, i.e. by using molecular
building blocks for the design of new interconnected net-
works,[7,8] involving several kinds of metallic ions or rad-
icals, which exhibit predictable magnetic properties.[9Ϫ11]
Focusing on 2D systems, as found in layered transition me-
tal compounds, it has been shown that the occurrence of
3D magnetic ordering is mainly driven by the divergence of
the correlation length within the layers which depends on
the in-plane magnetic exchange coupling and on the top-
ology of the magnetic centres.[4,6,12,13] Recent results on
organic-inorganic layered magnets have shown that it is
possible to tune the coupling between spin layers by using
suitable organic ligands separating or connecting transition
As a continuation of our investigations of layered sys-
tems, the present work is concerned with the synthesis,
through a hydrothermal route, of a series of layered tran-
sition metal (M) carboxylates obtained using the chiral
mandelate ion C6H5CH(OH)CO2 (M ϭ Mn, Fe, Co, Ni,
Cu) and an analysis of their crystal structures and magnetic
[a]
´
Institut de Physique et Chimie des Materiaux de Strasbourg,
UMR 7504 du CNRS,
23 rue du Loess, 67034 Strasbourg, France
Fax: (internat.) ϩ 33-3-88107247
E-mail: pierre.rabu@ipcms.u-strasbg.fr
Laboratoire DECMET, ILB, Universite Louis Pasteur,
4 rue Blaise Pascal, 67000 Strasbourg, France
Ϫ
[b]
´
662
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/ejic.200400623
Eur. J. Inorg. Chem. 2005, 662Ϫ669

Layered Metal(II) Mandelates
FULL PAPER
behaviour. Some molecular complexes of transition metals
(Cu, Co, V) have been reported in the literature in which
the mandelate anion acts as a co-ligand with lactic acid de-
rivatives such as phenanthroline or bipyridine, resulting in
peculiar coordination geometries.[28Ϫ31] Noble metal com-
plexes were also prepared.[32] It is worth noticing, however,
that very few results have been reported in the literature on
compounds involving only the mandelate ion and most of
these studies have concerned the solubility, vibrational spec-
troscopic properties and optical activities of the isolated
metal complexes.[33] Only one very recent work reports the
crystal structure and magnetic susceptibility measurements
of a CoII derivative.[34]
Results and Discussion
Synthesis and Structural Analysis
Figure 1. ORTEP plot of the structure of 5 showing coordination
of the CuII atoms by the mandelate ligands and interconnection
through carboxylato bridges
A series of five transition metal mandelates has been syn-
thesised through a hydrothermal method by treating metal
chlorides with mandelic acid and NaOH in water (see Ex-
perimental Section). The complexes of manganese (1), iron
(2), cobalt (3) and nickel (4) were obtained as crystalline
powders consisting of very thin platelets. The FeII solid 2
contains a small amount of a magnetic impurity, namely
iron oxide, which was revealed by the magnetic measure-
ments. The copper(ii) analogue 5 crystallises from the
mother solution (resulting from hydrothermal treatment) as
thick single crystals suitable for X-ray analysis. It is worth
noticing also that reactions under standard conditions at
room temperature led, for the five metal ions, to the same
products but only as fine and less crystalline powders. This
highlights the importance of an interest in the hydrothermal
method. The structure of 5 has been solved by a single-
crystal X-ray diffraction analysis (see crystallographic sec-
tion) and is similar to that reported in the literature for the
CoII analogue[34] except that an accurate Fourier analysis
ascertained the presence of the hydrogen of the α-hydroxide
˚
Table 1. Bond lengths [A] and angles [°] in the copper dimandel-
ate compound
Atoms
Distances
Atoms
Angles
CuϪO1
CuϪO2
CuϪO3#1
O1ϪC1
C1ϪC2
O2ϪC2
C2ϪC3
1.924(2)
1.987(2)
2.418(2)
1.299(4)
1.530(5)
1.466(3)
1.506(4)
O1ϪCuϪO2
83.41(10)
96.59(10)
95.18(9)
84.82(9)
95.18(9)
O1ϪCuϪO2#2
O1ϪCuϪO3#1
O1#2ϪCuϪO3#1
O2ϪCuϪO3#1
O2#2ϪCuϪO3#1
C1ϪO1ϪCu
89.63(9)
115.72(19)
113.32(18)
123.5(3)
C2ϪO2ϪCu
O3ϪC1ϪO1
O1#3ϪO2#2
2.752(4)
(intermolecular)
O3ϪC1ϪC2
119.1(3)
#1: Ϫx ϩ 1/2, y ϩ 1/2, Ϫz ϩ 1; #2: Ϫ x ϩ 1, Ϫy, Ϫz ϩ 1; #3: Ϫx
ϩ 1, Ϫy ϩ 1, Ϫz ϩ 1
coordinated to the copper(ii) centres, whereas this was not (Figure 2). The CuϪCu distances across the carboxylato
˚
considered for the cobalt species. The compound can be bridges are 5.325(5) A which correspond to the exchange
described as a (diaqua)copper(ii)Ϫµ-mandelate polymer, pathways schematised in Figure 3. Moreover, intermolecu-
the asymmetric unit of which is displayed in Figure 1. Each lar hydrogen bonds can be identified [O1#3ϪO2#2 2.752
CuII atom is coordinated to six oxygen atoms forming elon- (4) A, O1ϪH1ϪO2 171°], the CuO6 octahedra thus being
˚
gated octahedra. In the equatorial plane, they are sur- connected by two O1ϪO2 hydrogen bonds along the b axis.
rounded by two mandelate ligands each chelating the metal All the mandelates are positioned quasi-perpendicular to
˚
through one oxygen atom of the carboxylate group [CuϪO1 the copper layers which are 15.12(1) A apart. In addition,
˚
˚
1.924(2) A] and the α-OH [CuϪO2 1.987(2) A]. Two other a detailed examination of the structural stacking prompts a
mandelate groups complete the octahedral site with mono- consideration of the existence of σϪπ interactions between
dentate oxygen atoms from their carboxylate functions in the phenyl rings related symmetrically by the crystallo-
apical positions [CuϪO3#1 2.418(2) A] resulting in a dis- graphic a mirror plane. (Figure 2).[35] These cycles form an
˚
torted environment (Figure 1) similar to the CoII case but angle of 87° and the centroidϪcentroid distance is 5.21 A.
˚
with strong additional distortion of the octahedral sites The shortest distances for σ-π interactions are: C4ϪC8 3.70
around metal ions due to the JahnϪTeller effect (see A, C4ϪC7 3.74 A, C5ϪC7 3.93 A and C5ϪC6 4.19 A.
Table 1). It is apparent that the structure is centrosymmetric. Ac-
˚
˚
˚
˚
Within the a,b planes, the copper atoms are connected to cordingly, the present compounds exhibit no circular
four neighbours by carboxylato bridges in a syn-anti con- dichroism and involve both (R) and (S) enantiomers of the
formation (Figure 1), thus forming square-planar magnetic mandelate ligand although the pure (R) enantiomer was
arrays separated by a double layer of mandelate anions used as a starting ligand. This was also observed in the
Eur. J. Inorg. Chem. 2005, 662Ϫ669
www.eurjic.org
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
663

A. Beghidja, S. Hallynck, R. Welter, P. Rabu
FULL PAPER
from the reaction under ambient rather than hydrothermal
conditions. It thus appears that significant racemisation
may also occur in an acidic or neutral medium, perhaps
with different kinetics however.[37]
The other MII compounds were obtained as powders or
very thin crystals which somewhat frustrated characteris-
ation by single-crystal diffraction methods. However, the
corresponding spectroscopic characterisations (Figure 4)
and elemental analyses confirm that all complexes are iso-
structural with the copper(ii) and cobalt(ii) dimandelates de-
scribed above. This was further confirmed by a comparison
between their powder X-ray patterns shown in Figure 5.
The powder diffraction patterns were fully indexed and the
crystallographic cell parameters for the series of compounds
Figure 2. Partial view along the b axis of the layered structure of 5
Figure 4. Infrared spectra of the transition metal mandelates; for
clarity, the spectra are shifted in intensity
Figure 3. Drawing of the diamond-like MII network in the (a,b)
planes; each metal ion interacts with its four nearest neighbours as
in a square planar system
cobalt(ii) analogue.[34] For the latter, it was pointed out that
this racemisation, on the basis of previous studies on α-
hydroxycarboxylic acids, is favoured by the use of an alka-
line medium.[36] It is worth noticing here that the present
copper(ii) analogue, contrary to all the other metal deriva-
tives, can be obtained without addition of an alkaline base.
Figure 5. Comparison between the powder X-ray diffraction pat-
terns of the five layered transition metal mandelates, shifted in
intensity for clarity; some characteristic diffraction lines are
Similarly, the same compound can be obtained as a powder labelled
Table 2. Crystallographic parameters of the first row transition metal dimandelates refined from X-ray powder patterns; all compounds
crystallise in the monoclinic space group P21/a
M(C6H5CH(OH)CO2]2
Mn
Fe
Co
Ni
Cu[a]
˚
a [A]
8.925(5)
4.933(1)
15.605(4)
101.31(5)
673.70(8)
9.511(1)
4.8370(8)
15.569(2)
101.37(1)
702.19(6)
9.3532(7)
4.8737(3)
15.495(1)
99.95(1)
9.239(2)
4.851(1)
15.510(1)
100.04(1)
684.49(6)
9.489(1)
4.9156(6)
15.551(1)
102.749(9)
707.48(2)
˚
b [A]
˚
c [A]
β [°]
3
˚
V [A ]
695.71(5)
[a]
3
˚
˚
˚
˚
Single crystal parameters are: a ϭ 9.454 A; b ϭ 4.905 A; c ϭ 15.495 A; β ϭ 102.56°; V ϭ 701.3 A .
664
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2005, 662Ϫ669

Layered Metal(II) Mandelates
FULL PAPER
have been refined (Table 2) by means of a full pattern susceptibility displayed in Figure 6 for this compound
matching procedure using the Fullprof software package.[38] corresponds to the data corrected using the constant ferro-
magnetic moment deduced from variations in room tem-
Spectroscopic Properties
perature magnetisation as a function of the field.
The IR and UV data collected for the five compounds
were easy to analyse on the basis of the crystal data (see
Exp. Sect.). The IR spectra exhibit characteristic bands of
the mandelate ligand. The frequency difference between the
symmetric and asymmetric carboxylate vibrations, ∆ν ϭ
150Ϫ165 cmϪ1, is much smaller than that observed in the
corresponding acid (∆νi ഠ 270 cmϪ1) in agreement with the
bidentate coordination mode of the carboxylate groups.[39]
The presence of broad νOϪH bands in the range
3200Ϫ3400 cmϪ1 is related to the mandelate OH moieties
bonded to the metal ions. Interestingly, whereas the IR
spectra of the MnII, FeII, CoII, NiII compounds are almost
identical, thus confirming very similar local structures, sig-
nificant shifts of the νOϪH and carboxylate bands can be
observed for CuII species. The latter is merely related to
the distortion of the coordination sphere around the copper
atoms due to the JahnϪTeller effect, as found from the
single-crystal structure analysis (Table 1). The UV spectra
are also consistent with the presence of octahedral geo-
metries for the MII ions.[40,41] The values of Dq and the
Racah parameter B have been calculated for the d7 and d8
ions giving Dq ϭ 985 cmϪ1, B ϭ 710 cmϪ1 and Dq/B ϭ 1.4
for the CoII and Dq ϭ 811 cmϪ1, B ϭ 977 cmϪ1 and
Dq/B ϭ 0.83 for the NiII compound in accordance with
weak crystal field and high-spin configurations.[42] For CoII,
admixture of forbidden d-d transitions at 480 nm (20833
cmϪ1) and 505 nm (19801 cmϪ1) results in a multiple struc-
tured band. A value of Dq ϭ 939 cmϪ1 can be deduced for
Figure 6. Temperature variation of the magnetic susceptibilities of
the compounds 1, 2, 3 and 5 as χT vs. T plots; the full line corre-
sponds to the fit of the data for the MnII compound 1, as detailed
in the text
The four compounds exhibit paramagnetic behaviour
corresponding to quasi-isolated metal centres. The inverse-
susceptibility versus T curves are linear above 200 K and
were fitted using the CurieϪWeiss law. The corresponding
values of the Curie constants and Weiss temperatures are
given in Table 3 and the corresponding values of the
effective moments µeff ϭ (8C)1/2 agree well with those
expected for MII ions in an octahedral crystal field.
Upon cooling, the χT product of the CoII compound 3
exhibits a regular decrease from 2.85 cm3·K·molϪ1 at 300 K
to 1.8 cm3·K·molϪ1 at 10 K. The decrease of the χT product
is well understood as being due to the spin-orbit coupling
effects for octahedral CoII ions.[43] Indeed, high-spin octa-
the iron(ii) ions and a shoulder at λ ϭ 1260 nm (7936 cmϪ1
)
can be attributed to splitting of the 5Eg state. The UV spec-
trum of the CuII derivative exhibits the usual broad band
(λ ϭ 745 nm/13422 cmϪ1) due to transitions from the t2g
components to the x2Ϫy2 level.
4
hedral CoII has a T1g ground state and, as a consequence,
exhibits unquenched spin-orbit coupling in addition to a
zero-field splitting which dominates at low temperature and
stabilises a doublet ground state. It should be pointed out
that for isolated octahedral CoII ions, a minimum value of
Magnetic Properties
The magnetic properties of 1, 2, 3, 4 and 5 were investi- χT ഠ 1.8 cm3·K·molϪ1 can be expected corresponding to a
gated in the temperature range 2Ϫ300 K. The temperature pseudo spin S ϭ 1/2 and g ഠ 4.4. Moreover, the magnetis-
dependence of the magnetic susceptibilities χ above 2 K is ation vs. field variation at 2 K (not shown) is consistent
shown in Figure 6 as χT vs. T plots for 1, 2, 3 and 4, after with paramagnetic-like behaviour, indicating that no signifi-
correction for diamagnetic contributions. The iron(ii) com- cant coupling takes place. The saturation moment, Ms ϭ
pound 2 contains small amounts of an impurity which is 2.25 µB·molϪ1, is in agreement with the expected value for
ferromagnetic at ambient temperature. Consequently, the octahedral high spin CoII ions (2Ϫ3 µB per cobalt atom).[9]
Table 3. Main magnetic data of the first row transition metal dimandelates
M[C6H5CH(OH)CO2]2
MnII
FeII
CoII
NiII
CuII
C [K·emu·molϪ1
]
4.4
5.9
ca. 3.9
ca. 5.6
Ϫ
isolated
Ϫ
2.85
4.8
Ϫ5
isolated
Ϫ
1.16
3.0
0.44
1.9
Ϫ3
isolated
Ϫ
µeff [µB.molϪ1
Θ [K]
]
Ϫ0.8
AF/J ϭ Ϫ0.122 K
Ϫ
ϩ11.36
F/J ϭ ϩ1.213 K
F/TC ϭ 2.7 K
In-plane interactions[a]
3D ordering[a]
[a]
AF: antiferromagnetic; F: ferromagnetic.
Eur. J. Inorg. Chem. 2005, 662Ϫ669
www.eurjic.org
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
665

A. Beghidja, S. Hallynck, R. Welter, P. Rabu
FULL PAPER
Although the corrected data for compound 2 are ap-
Figure 8 shows a magnetisation loop measured at 2 K ex-
proximate owing to the presence of the magnetic impurity, hibiting a sharp increase in the magnetisation at low field,
the general behaviour of the susceptibility is in agreement up to ca. the half of the saturation value. No hysteresis (at
with that expected for quasi-isolated FeII ions exhibiting least within a few Gauss) was observed, in accordance with
spin-orbit coupling effects similar to the CoII analogue the isotropic character of the Heisenberg NiII ions and the
above.
magnitude of TC. The saturation moment tends to 2µB per
The copper(ii) derivative show a quasi-constant χT prod- NiII at 5 T which is expected at full saturation for S ϭ 1
uct in the whole temperature range and is paramagnetic, spin moments. It is important to notice, however, that M
whereas the MnII compound 3 exhibits a significant de- vs. H shows a smooth and quasi-linear variation in the
crease below 30 K indicating short-range antiferromagnetic high-field region which suggests that small canting between
in-plane coupling. No long-range order is evident from moments might occur, thereby explaining such particular
either susceptibility or magnetisation measurements for this features and leading to rather small residual moments.
compound. The magnetic behaviour of 3 was modelled with
the 2D Heisenberg approach in an effort to evaluate the
magnitude of the nearest-neighbour exchange interaction.
The susceptibility data were fit between 2 and 300 K with
the high-temperature series expansion reported by Rush-
brooke and Wood[44,45,46] for an antiferromagnetic 2D Hei-
senberg square planar system [Equation (1)],[47] giving g ϭ
2.01(1) and J/kB ϭ Ϫ0.122(4) K, Figure 6. The coupling is
characteristic of very weak antiferromagnetic interactions
across the three-atom dicarboxylate bridge.
(1)
Figure 8. Magnetisation vs. field hysteresis loop for the NiII
mandelate 4, recorded at 2 K
The NiII compound 4 exhibits very different behaviour.
Ferromagnetic interactions are implied by the inverse sus-
A fit of the magnetic susceptibility above TC was carried
out using the high temperature series expansion for a ferro-
magnetic square planar array of S ϭ 1 spins, the expression
of which is similar to Equation (1) above.[48] In this ex-
pression, the zero-field splitting (D) for NiII ions is neglected
although it is known to be effective in the low temperature
range, typically below 10 K.[9] Within this approximation,
the data are very well fit in the full temperature range with
g ϭ 2.15(1) and J/kB ϭ ϩ1.213(4) K, Figure 7. Finally, the
occurrence of 3D magnetic ordering can be undoubtedly
established by the temperature variation of the specific heat
recorded at low temperature, Figure 9. The curve shows a
typical lambda anomaly, the maximum of which enables a
ceptibility 1/χ vs. T variation (not shown) with a positive
CurieϪWeiss temperature (Table 3). As shown in Figure 7,
the temperature variation of the dc magnetic susceptibility
implies the existence of a significant ferromagnetic coupling
between neighbouring nickel(ii) ions within the plane, with
an increase in χT from 1.21 K·emu·molϪ1 at 295 K to
47.5 K·emu·molϪ1 at 2 K. Ferromagnetic long range order
between the layers is demonstrated by the occurrence of a
peak in the out-of-phase signal χЈЈ of the ac susceptibility
shown in the inset of Figure 7. The Curie temperature value
TC ഠ 2.5 K can be deduced from the maximum of χЈ where
an onset of the imaginary part of the susceptibility χЈЈ
occurs.
Figure 7. Temperature variation of the magnetic susceptibility of
the NiII mandelate 4 as a χT vs. T plot; the full line corresponds
to the fit of the data, as detailed in the text; the inset shows the
variation of the real and imaginary parts of the ac susceptibility
recorded in a 3.5 Oe/30 Hz magnetic field
Figure 9. Low-temperature variation of the specific heat of the NiII
mandelate 4
666
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2005, 662Ϫ669

Layered Metal(II) Mandelates
FULL PAPER
(SHELXS-97) and refined against F2 using the SHELXL-97
software.[50] The absorption was corrected empirically with Sor-
tav.[51] All non-hydrogen atoms were refined anisotropically. Hydro-
gen atoms were generated according to stereochemistry and refined
using a riding model in SHELXL-97. CCDC-236368 contains the
supplementary crystallographic data for this paper. These data can
be obtained free of charge at www.ccdc.cam.ac.uk/conts/retriev-
ing.html [or from the Cambridge Crystallographic Data Centre, 12
Union Road, Cambridge CB2 1EZ, UK; Fax: (internat.) ϩ 44-
1223-336-033; E-mail: deposit@ccdc.cam.ac.uk]. Green-blue crys-
tal, crystal dimension: 0.13 ϫ 0.10 ϫ 0.02 mm, C16H14CuO6, M ϭ
precise determination of the Curie temperature TC
2.7(1) K.
ϭ
Conclusion
A series of MnII, FeII, CoII, NiII and CuII mandelates has
been crystallised and structurally and magnetically charac-
terised. Although the starting mandelate was enantiomer-
ically pure, the ligand was found to racemise during the
synthetic process. The five compounds are isostructural
with the copper analogue, the latter having been obtained
as single crystals and showing a strong JahnϪTeller distor-
tion of the coordination spheres around the CuII centres.
Their solid-state structures consist of layers of six-coordi-
nate metal(ii) ions, thus forming planar arrays of octahedra
interconnected through carboxylato bridges in a diamond-
like network. The planes are separated by double layers of
phenyl groups from the mandelate ligands resulting in a 2D
365.81 g·molϪ1, monoclinic space group P21/a, a ϭ 9.454(5) A, b ϭ
˚
˚
˚
4.905(5) A, c ϭ 15.495(5) A, β ϭ 102.564(5)°, Z ϭ 2, Dcalcd.
ϭ
1.732 g·cmϪ3, µ (Mo-Kα) ϭ 1.588 mmϪ1, Tmin. ϭ 0.890, Tmax.
ϭ
0.970 mm, a total of 2406 reflections, 2.69° Ͻ θ Ͻ 31.96°, 2406
independent reflections with 1833 having I Ͼ 2σ(I), 106 param-
eters, Final results: R1 ϭ 0.0817, wR2 ϭ 0.1349, Gof ϭ 1.16, maxi-
Ϫ3
˚
mum residual electronic density ϭ 1.152 e·A
.
Manganese(II) Dimandelate, C16H14O6Mn (1): C16H14O6Mn (1)
was synthesised by a hydrothermal method from MnCl2·4H2O
system. From a magnetic point of view, it has been shown (99% Aldrich), (R)-mandelic acid (98% Acros) and NaOH (Merck).
The starting mixture, which consisted of [MnCl2·4H2O], C8H8O3,
NaOH and H2O in a ratio of 2:4:4:20, was homogenised and trans-
ferred to a sealed Teflon lined hydrothermal bomb (120-mL bomb
volume). After heating at 170 °C for 120 h under autogenous press-
ure, a white, slightly pink crystalline powder was obtained which
was washed and rinsed with distilled water and absolute ethyl al-
cohol (yield: 0.43 g, 60% on the basis of MnCl2·6H2O). The purity
was confirmed by elemental analysis [C16H14O6Mn (361.21): calcd.
C 53.80, H 3.95; found C 53.21, H 3.85] and oxidising pyrolysis
[heating in air from 20 to 700 °C at 3 °C/min, at 700 °C for 30 min
to transform manganese into Mn2O3 and cooling to 20 °C at 10
°C/min: calcd. Mn 15.38%; found 17.16%]. IR (KBr pellet): ν˜ ϭ
3237 cmϪ1 (νOH), 1560 cmϪ1 (νasymCO) and 1413 cmϪ1 (νsymCO).
that the in-plane exchange pathways between neighbouring
magnetic centres form a square-like planar array. The ex-
change coupling involving bridging carboxylate ligands in
a syn-anti conformation is known to promote antiferromag-
netic interactions between neighbouring MII ions[49] and to
be more favourable than the anti-anti conformation.[24] This
agrees with the observed behaviour of the χT product for
the series of transition metal mandelates except for the
nickel derivative for which ferromagnetic behaviour is ob-
served with ferromagnetic 3D ordering at TC ϭ 2.7 K.
Experimental Section
Iron(II) Dimandelate C16H14O6Fe (2): C16H14O6Fe (2) was syn-
thesised by a hydrothermal method according to the same process
as for the manganese analogue 1 but using FeCl2·6H2O instead of
MnII dichloride. A very pale green crystalline powder was obtained,
washed and rinsed with distilled water and absolute ethyl alcohol
(yield: 0.36 g, 50% on the basis of FeCl2·6H2O). The purity was
checked by elemental analysis [C16H14O6Fe (358.12): calcd. C
53.66, H 3.94; found C 53.63, H 3.91] and oxidising pyrolysis [heat-
ing in air from 20 to 700 °C at 2 °C/min, at 700 °C for 30 min to
transform into Fe3O4 and cooling to 20 °C at 10 °C/min: calcd. Fe
15.59%; found 18.85%]. IR (KBr pellet): ν˜ ϭ 3247 cmϪ1 (νOH),
1565 cmϪ1 (νasymCO) and 1417 cmϪ1 (νsymCO). UV/Vis (reflec-
General Remarks: Thermogravimetric experiments were performed
using a Setaram TG92 instrument (heating rate of 3 °C/min, air
stream). FT-IR studies were performed with an ATI Mattson Gen-
esis computer-driven instrument (0.1 mm thick powder samples in
KBr). UV/Vis/NIR studies were performed with a PerkinϪElmer
Lambda 19 instrument (spectra recorded by reflection with a reso-
lution of 4 nm and a sampling rate of 480 nm/min). Elemental
analyses were performed using a Thermo-Finnigan EA 1112 setup
(Service d’analyses, ICS, Strasbourg, France). Magnetic studies
were carried out using a Quantum Design SQUID MPMS-XL
magnetometer (Ϯ5 T, 2Ϫ300 K) on several sets of crystalline or
powdered samples. Low temperature specific heat measurements
were carried out using a MagLab HC Oxford Instrument. Powder
X-ray diffraction patterns for the whole series where obtained using
tance for octahedral coordination of FeII):
λ ϭ 1065 nm
(5EgǞ5T2g).
Cobalt(II) Dimandelate C16H14O6Co (3): C16H14O6Co (3) was syn-
thesised similarly to the above procedure but starting from
CoCl2·6H2O (98% Acros). Very thin pink platelet-like crystals were
obtained, washed and rinsed with distilled water and absolute ethyl
alcohol (yield: 0.47 g, 65% on the basis of CoCl2·6H2O). The purity
was confirmed by elemental analysis [C16H14CoO6 (361.21): calcd.
C 53.20, H 3.91; found C 52.84, H 3.85] and oxidising pyrolysis
[heating in air from 20 to 700 °C at 3 °C/min, at 700 °C for 30 min
˚
BraggϪBrentano Siemens D500 (Co-Kα1, λ ϭ 1.78897 A) and
D5000 (Cu-Kα1, λ ϭ 1.5406 A) diffractometers equipped with pri-
mary beam monochromators. All the powder X-ray patterns exhi-
bit large preferential orientations of the crystallites parallel to the
stacking axis c.
˚
X-ray Crystallographic Study: A single crystal of the copper man-
delate 5 was mounted on a Nonius Kappa-CCD area detector dif-
fractometer (Mo-Kα, λ ϭ 0.71073 A). The complete conditions of to transform into Co3O4 and cooling to 20 °C at 10 °C/min: calcd.
˚
data collection (Denzo software) and structure refinements are
given below. The cell parameters were determined from reflections
Co 16.32%; found 16.78%]. IR (KBr pellet): ν˜ ϭ 3255 cmϪ1 (νOH),
1573 cmϪ1 (νasymCO) and 1408 cmϪ1 (νsymCO). UV/Vis (reflec-
taken from one set of 10 frames (1.0° steps in ϕ angle), each with an tance for octahedral coordination of CoII):
λ
ϭ
1140 nm
exposure time of 20 s. The structure was solved by direct methods
(4T1gǞ4T2g), 645 nm (4T1gǞ4A2g), 545 nm [4T1gǞ4T1g (P)].
Eur. J. Inorg. Chem. 2005, 662Ϫ669
www.eurjic.org
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
667

A. Beghidja, S. Hallynck, R. Welter, P. Rabu
FULL PAPER
Commun. 1996, 10, 1107Ϫ1108; W. Fujita, K. Awaga, Inorg.
Chem. 1996, 35, 1915Ϫ1917; P. Rabu, J. M. Rueff, Z. L. Huang,
S. Angelov, J. Souletie, M. Drillon, Polyhedron 2001, 20,
1677Ϫ1685; P. Rabu, M. Drillon, K. Awaga, W. Fujita, T. Sek-
ine, in Magnetism: Molecules to Materials (Eds.: J. S. Miller, M.
Drillon), Wiley-VCH, Weinheim, Germany, 2001, p. 357Ϫ395.
P. Rabu, P. Janvier, B. Bujoli, J. Mater. Chem. 1999, 9,
1323Ϫ1326.
Z.-L. Huang, M. Drillon, N. Masciocchi, A. Sironi, J.-T. Zhao,
P. Rabu, P. Panissod, Chem. Mater 2000, 12, 2805Ϫ2812.
S. O. H. Gutschke, D. J. Price, A. K. Powell, P. T. Wood, Angew.
Chem. Int. Ed. 2001, 40, 1920Ϫ1923.
D. J. Price, S. Tripp, A. K. Powell, P. T. Wood, Chem. Eur. J.
2001, 7, 200Ϫ208.
R. Feyerherm, A. Loose, P. Rabu, M. Drillon, Solid State Sci-
ence 2003, 5, 321Ϫ326.
Y. Rodriguez-Martin, C. Ruiz-Perez, J. Sanchiz, F. Lloret, M.
Nickel(II) Dimandelate C16H14O6Ni (4): C16H14O6Ni (4) was syn-
thesised in the same manner as above from NiCl2·6H2O (99.99%
Sigma). A green powder was obtained, washed and rinsed with
distilled water and absolute ethyl alcohol (yield: 0.5 g, 70% on the
basis of NiCl2·6H2O). The purity was confirmed by elemental
analysis [C16H14NiO6 (360.97): calcd. C 53.24, H 3.91; found C
53.15, H 3.87] and oxidising pyrolysis [heating in air from 20 to
700 °C at 3 °C/min, at 700 °C for 30 min to transform into NiO,
and cooling to 20 °C at 10 °C/min: calcd. Ni 16.26%; found
17.48%]. IR (KBr pellet): ν˜ ϭ 3245 cmϪ1 (νOH), 1577 cmϪ1
(νasymCO) and 1411 cmϪ1 (νsymCO). UV/Vis (reflectance for octa-
hedral coordination of NiII): λ ϭ 410 nm (3A2gǞ3T2g), 685 nm
(3A2gǞ3T1g), 745 nm (3A2gǞ1Eg), 1235 nm [3A2gǞ3T1g(P)].
[15]
[16]
[17]
[18]
[19]
[20]
Copper(II) Dimandelate C16H14O6Cu (5): C16H14O6Cu (5) was syn-
thesised from CuCl2·2H2O (99% Aldrich) and (R)-mandelic acid
(98% Acros). The starting mixture, corresponding to CuCl2·2H2O,
C8H8O3 and H2O in a molar ratio of 2:4:20, was homogenised and
transferred to a sealed, Teflon-lined hydrothermal bomb (120-mL
bomb volume). Hydrothermal treatment (170 °C, 72 h, autogenous
pressure) led to white-yellow crystals and a blue-green solution.
These crystals were left to dissolve in the mother liquor. After 6
d, large green-blue crystals of 5 suitable for single-crystal X-ray
diffraction analysis were obtained. The crystals were washed with
distilled water and absolute ethyl alcohol. (yield: 0.38 g, 52% on
the basis of CuCl2·2H2O). The purity was confirmed by elemental
analysis [C16H14O6Cu (365.82): calcd. C 53.66, H 3.94; found C
53.63, H 3.91] and oxidising pyrolysis [heating in air from 20 to
700 °C at 3 °C/min, at 700 °C for 30 min to transform into CuO
and cooling to 20 °C at 10 °C/min: calcd. Cu 17.37%; found
16.99%]. IR (KBr pellet): ν˜ ϭ 3133 cmϪ1 (νOH), 1635 cmϪ1
(νasymCO) and 1456 cmϪ1 (νsymCO). UV/Vis (reflectance for octa-
hedral coordination of CuII): λ ϭ 745 nm (transition from t2g com-
ponents to x2Ϫy2).
´
Julve, Inorg. Chim. Acta 2001, 318, 159165; C. Ruiz-Perez, J.
`
Sanchiz, M. Hernadez-Molina, F. Lloret, M. Julve, Inorg.
Chim. Acta 2000, 298, 202Ϫ208.
M. Riou-Cavellec, C. Albinet, C. Livage, N. Guillou, M. No-
[21]
`
`
´
gues, J. M. Greneche, G. Ferey, Solid State Sciences 2002, 4,
267Ϫ270.
[22]
[23]
M. J. Plater, M. R. S. J. Foremen, R. A. Howie, J. M. S. Skakle,
A. M. Z. Slawin, Inorg. Chim. Acta 2001, 315, 126Ϫ132.
N. Snejko, E. Guttierez-Puebla, J. L. G. Smith, D. E. Lynch,
T. C. W. Mak, W.-H. Yip, C. H. L. Kennard Martinez, M. A.
Monge, C. Ruiz-Valero, Chem. Mater 2002, 14, 1879Ϫ1883.
J. M. Rueff, N. Masciocchi, P. Rabu, A. Sironi, A. Skoulios,
Chem. Eur. J. 2002, 2813Ϫ2820; J. M. Rueff, N. Masciocchi, P.
Rabu, A. Sironi, A. Skoulios, Eur. J. Inorg. Chem. 2001,
2843Ϫ2848.
G. Smith, D. E. Lynch, T. C. W. Mak, W.-H. Yip, C. H. L.
Kennard, Polyhedron 1993, 12, 203Ϫ208.
G. Smith, O. J. O’Reilly, C. H. L. Kennard, J. Chem. Soc.,
Dalton Trans. 1980, 2462Ϫ2466.
T. Asaji, K. Aoki, D. Nakamura, Z. Naturforsch., Teil A 1984,
39, 371Ϫ375.
R. Carballo, A. Castinieras, S. Balboa, B. Covelo, J. Niclos,
Polyhedron 2002, 21, 2811.
R. Carballo, B. Covelo, S. Balboa, A. Castinieras, J. Niclos, Z.
Anorg. Allg. Chem. 2001, 627, 948.
D. A. House, J. Browning, J. R. Ripal, W. T. Robinson, Inorg.
Chim. Acta 1999, 292, 73.
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[1]
Research Frontiers in Magnetochemistry (Ed.: C. O’Connor),
World Scientific Publishers, Singapore, 1993.
Magnetic Molecular Materials (Eds.: D. Gatteschi, O. Kahn, J.
[2]
S. Miller, F. Palacio), Kluwer, Dordrecht, 1991.
L. P. Kadanoff, W. Götze, D. Hamblen, R. Hecht, E. A. S.
[3]
Lewis, V. V. Palciauskas, M. Rayl, J. Swift, Rev. Mod. Phys.
1967, 39, 395Ϫ431.
I. K. Smatanova, J. Marek, P. Svancarek, P. Schwendt, Acta
Crystallogr., Sect. C 2000, 56, 154.
[4]
L. J. de Jongh, A. R. Miedema, Adv. Phys. 1974, 23, 1Ϫ260.
G. Meister, G. Rheinwald, H. Stoeckli-Evans, G. Suss-Fink, J.
Organomet. Chem. 1995, 496, 197.
[5]
See for instance the articles in Magnetism: Molecules to Mate-
rials-Models and Experiments (Eds.: J. S. Miller, M. Drillon),
Wiley-VCH, Weinheim, 2001.
J.-P. Renard, in Organic and inorganic Low-dimensional crystal-
[33] [33a] K. N. Kovalenko, D. V. Kazachenko, V. P. Kurbatov, L. G.
Kovaleva, Zh. Neorg. Khim. 1971, 16, 2444. [33b] A. Ranade, Z.
[6]
[33c]
Anorg. Allg. Chem. 1972, 388, 105.
America, Ind. J. Chem. 1975, 13, 525.
Kapoor, J. Prakt. Chem. 1963, 20, 237.
P. V. Khadikar, R. L.
line materials (Eds.: P. Delhaes, M. Drillon), NATO-ASI Series,
Plenum Press New York and London, 1987, p. 125Ϫ140.
[33d]
A. B. Sen, S. N.
[33e]
R. Larsson, B.
[7]
C. Janiak, Dalton Trans. 2003, 2781.
[33f]
Folkesson, Acta Chem. Scand. 1968, 22, 1970Ϫ1980.
A. J.
[8]
S. L. James, Chem. Soc. Rev. 2003, 32, 276.
R. L. Carlin, Magnetochemistry, Springer-Verlag, Berlin, Heid-
Fischinger, A. Sarapu, A. Companion, Can. J. Chem. 1969, 47,
[9]
2629Ϫ2637. [33g] P. Spacu, S. Plostinaru, L. Patron, A. Contes-
elberg, 1986; O. Kahn, Molecular Magnetism, Wiley-VCH,
[33h]
cuand, N. Stanica, Rev. Roum. Chim. 1985, 30, 837Ϫ842.
Weinheim, 1993.
Book review: Molecular Magnets Ϫ Recent Highlights (Eds.:
R. P. Rajguru, C. Panda, J. Inst. Chem. (India) 1991, 63,
37Ϫ39.
H. Kumagai, Y. Oka, K. Inoue, M. Kurmoo, J. Phys. Chem.
Sol. 2004, 65, 55Ϫ60.
C. Janiak, J. Chem. Soc., Dalton Trans. 2000, 3885Ϫ3896.
I. L. Finar, Organic Chemistry, 4th ed., Longmans, London,
1968, vol. 2, p. 69.
Winther, Z. Physik Chem. 1906, 56, 465Ϫ511; A. N. Campbell,
A. J. R. Campbell J. Am. Chem. Soc. 1932, 54, 3834Ϫ3841; A.
N. Campbell, A. J. R. Campbell Nature 1932, 129, 281.
Fullprof: J. Rodriguez-Carvajal, Physica 1993, B192, 55.
K. Nakamoto, Infrared and Raman Spectra of Organic and Co-
[10]
W. Linert, M. Verdaguer), Springer, Wien, New York, 2003.
Molecular Magnetism (Eds.: K. Itoh, M. Kinoshita), Kodansha
[34]
[11]
Ltd. and Gordon and Breach Science Publishers, Amsterdam,
[35]
[36]
London, 2000.
[12]
M. Drillon, P. Panissod, J. Magn. Magn. Mater. 1998, 188,
93Ϫ99.
[37]
[13]
Magnetic Properties of Layered Transition Metal Compounds
(Ed.: L. J. de Jongh), vol. 9A (‘‘Physics and Chemistry of Mate-
rials with Low-dimensional Structures’’), Kluwer Academic
Publishers, Dordrecht, 1990.
[38]
[39]
[14]
P. Rabu, S. Rouba, V. Laget, C. Hornick, M. Drillon, Chem.
668
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2005, 662Ϫ669

Layered Metal(II) Mandelates
FULL PAPER
[47]
[48]
ordination Compounds, 4th ed., Wiley Ϫ Interscience Publi-
cation, New York, 1986, 231.
A. B. P. Lever, Inorganic Electronic Spectroscopy, 2nd ed., El-
sevier, Amsterdam, 1984.
L. Banci, A. Bencini, C. Benelli, D. Gatteschi, C. Zanchini,
Struct. Bonding 1982, 52, 37Ϫ86.
Y.-S. Dou, J. Chem. Educ. 1990, 67, 134.
F. E. Mabbs, D. J. Machin, Magnetism and transition metal
complexes, Chapman and Hall Ltd., London, 1973.
G. S. Rushbrooke, G. A. Baker, P. J. Wood, in Phase Transitions
and Critical Phenomena (Eds.: C. Domb, M. S. Green), Aca-
demic Press, London, New York, 1974, vol. III, chapter 5.
K. Yamaji, J. J. Kondo, Phys. Soc. Jpn. 1973, 35, 25Ϫ32.
R. Navarro, in Magnetic Properties of Layered Transition Metal
x ϭ ϪJ/(kBT) for antiferromagnetic coupling, N ϭ Avogadro’s
number, µB ϭ Bohr magneton, g ϭ Lande g value, a1
23.33333, a2 396.66666, a3 6212.888888, a4
1161923.54664125, a6
180849416.031247, and a8
ϭ
ϭ
ϭ
ϭ
[40]
ϭ
ϭ
87638.8119558377, a5
14796191.6938613, a7
2142650986.87194.
ϭ
[41]
ϭ
x ϭ J/(kBT) for ferromagnetic coupling, a1 ϭ 5.333333333,
a2 ϭ 18.66666667, a3 ϭ 56.88888889, a4 ϭ 152.3950631,
a5 ϭ 376.9415553, a6 ϭ 918.8538407, a7 ϭ 2134.805724, and
a8 ϭ 4572.414518.
E. Coronado, M. Drillon, D. Beltran, J. C. Bernier, Inorg.
Chem. 1984, 23, 4000Ϫ4004.
G. M. Sheldrick, SHELXL-97, University of Göttingen, Ger-
many, 1997.
Sortav: R. H. Blessing, J. Appl. Crystallogr. 1989, 22, 396Ϫ397.
Received July 17, 2004
[42]
[43]
[44]
[49]
[50]
[51]
[45]
[46]
Compounds (Ed.: L. J. de Jongh), Kluwer Academic Publishers,
Dordrecht, Boston, London, 1990, p. 105Ϫ190.
Published Online January 3, 2005
Eur. J. Inorg. Chem. 2005, 662Ϫ669
www.eurjic.org
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
669