Ground spin state variation in carboxylate-bridged tetranuclear [Fe2Mn2O2]8+ cores and a comparison with their [Fe4O2]8+ and [Mn4O2]18+ congeners

Article abstract of DOI:10.1002/ejic.200390076

Syntheses of ten new tetranuclear complexes containing the [M_A(μ₃-O)₂M_B]⁸⁺ butterfly core [M_A = M_B = Fe^III (1 and 2), M_A = M_B = Mn^III (3 and 4) and M_A = Fe^III, M_B = Mn^III (5-10)] using 1,4,7-trimethyl-1,4,7-triazacyclononane (L) as the capping ligand, salicylaldoximato dianion (salox^2-) and six different carboxylate anions as bridging ligands, are described. The crystal structures of [L₂Fe₂(μ₃-O)₂ (salox)₂(diphenylglycolate)₃Fe₂] (ClO₄) (1), [L₂Mn₂(μ₃-O)₂ (salox)₂(diphenylglycolate)₃Mn₂] (ClO₄) (3), [L₂Fe₂(μ₃-O)₂ (salox)₂(acetate)₃Mn₂]- (ClO₄) (5), [L₂Fe₂(μ₃-O)₂ (salox)₂(diphenylglycolate)₃Mn₂]- (ClO₄) (6), and [L₂Fe₂(μ₃-O)₂ (salox)₂(benzoate)₃Mn₂]- (ClO₄) (8) were determined by X-ray crystallography. The complexes are isostructural, with M^III ions disposed in a butterfly core where bridging between the M^III ions occurs via two μ₃-oxo anions. In complexes 5, 6 and 8 the "wing-tip" positions of the butterfly are occupied by the LFe³⁺ units, whereas the "body" metal ions are d⁴ high-spin Mn^III ions. Complexes 1-10 have been characterised by variable-temperature (2-290 K) magnetic susceptibility measurements (1 T) and by Moessbauer spectroscopy. A "2J" model has been applied to simulate the experimental μ_eff vs. T plots. Overall exchange interactions are antiferromagnetic in nature. Complexes with the Fe₄O₂ core (1 and 2) and the Mn₄O₂ core (3 and 4) possess an S_t = 0 and S_t = 3 ground state, respectively, regardless of the nature of the bridging carboxylate ligand. In contrast, changing the bridging carboxylates in the [Fe₂Mn₂O₂] core from acetate (5) to diphenylglycolate (6), triphenyl acetate (7), benzoate (8), chloroacetate (9) and propionate (10), results in a variation of the ground state from S_t = 1 to S_t = 3, due to spin frustration of the "body" manganese centres. The spin-correlation diagram for the Fe₂Mn₂O₂ butterfly arrangement clearly demonstrates the ground state variation as a function of the ratio of two competing coupling interactions. Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim, Germany, 2003.

Full text of DOI:10.1002/ejic.200390076

FULL PAPER
Ground Spin State Variation in Carboxylate-Bridged Tetranuclear
[Fe2Mn2O2]8؉ Cores and a Comparison with Their [Fe4O2]8؉ and [Mn4O2]8؉
Congeners
Phalguni Chaudhuri,*[a] Eva Rentschler,[a] Frank Birkelbach,[a] Carsten Krebs,[a]
Eckhard Bill,[a] Thomas Weyhermüller,[a] and Ulrich Flörke[b]
Keywords: Oxo ligands / Carboxylate ligands / Iron / Manganese / Magnetic properties / Tetranuclear complexes
Syntheses of ten new tetranuclear complexes containing the
[MA(µ3-O)2MB]8+ butterfly core [MA = MB = FeIII (1 and 2),
MA = MB = MnIII (3 and 4) and MA = FeIII, MB = MnIII (5−10)]
perature (2−290 K) magnetic susceptibility measurements (1
T) and by Mössbauer spectroscopy. A ‘‘2J’’ model has been
applied to simulate the experimental µeff vs. T plots. Overall
using 1,4,7-trimethyl-1,4,7-triazacyclononane (L) as the cap- exchange interactions are antiferromagnetic in nature. Com-
ping ligand, salicylaldoximato dianion (salox2−) and six dif-
ferent carboxylate anions as bridging ligands, are described.
The crystal structures of [L2Fe2(µ3-O)2(salox)2(diphenylgly-
colate)3Fe2](ClO4) (1), [L2Mn2(µ3-O)2(salox)2(diphenylgly-
plexes with the Fe4O2 core (1 and 2) and the Mn4O2 core (3
and 4) possess an St = 0 and St = 3 ground state, respectively,
regardless of the nature of the bridging carboxylate ligand.
In contrast, changing the bridging carboxylates in the
colate)3Mn2](ClO4) (3), [L2Fe2(µ3-O)2(salox)2(acetate)3Mn2]- [Fe2Mn2O2] core from acetate (5) to diphenylglycolate (6),
(ClO4) (5), [L2Fe2(µ3-O)2(salox)2(diphenylglycolate)3Mn2]-
(ClO4) (6), and [L2Fe2(µ3-O)2(salox)2(benzoate)3Mn2]-
(ClO4) (8) were determined by X-ray crystallography. The
complexes are isostructural, with MIII ions disposed in a but-
triphenyl acetate (7), benzoate (8), chloroacetate (9) and pro-
pionate (10), results in a variation of the ground state from
St = 1 to St = 3, due to spin frustration of the ‘‘body’’ mangan-
ese centres. The spin-correlation diagram for the Fe2Mn2O2
terfly core where bridging between the MIII ions occurs via butterfly arrangement clearly demonstrates the ground state
two µ3-oxo anions. In complexes 5, 6 and 8 the ‘‘wing-tip’’
positions of the butterfly are occupied by the LFe3+ units,
whereas the ‘‘body’’ metal ions are d4 high-spin MnIII ions.
Complexes 1−10 have been characterised by variable-tem-
variation as a function of the ratio of two competing coup-
ling interactions.
( Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2003)
Introduction
made, and there are relatively few studies of the analogous
heterometallic butterfly complexes,[13,14] despite the fact
that such studies might be more informative in comparison
to those of homometal complexes. New exchange pathways
can be expected for heteronuclear complexes,[15Ϫ18] where
unusual sets of magnetic orbitals can be brought into
close proximity.
There are several intriguing features associated with oxo-
bridged carboxylate clusters. Firstly, these complexes can
have unusual electronic structures and may serve as sources
of fundamental information about exchange coupling in
multinuclear assemblies. A second general reason to study
polynuclear metal complexes is that they may be building
blocks for molecular-based magnetic materials.[19] Because
of their topology, molecules that have large numbers of un-
paired electrons should serve as good starting points for
constructing molecular magnetic materials. Even though
the pairwise exchange interactions in these complexes are
found almost always to be antiferromagnetic, spin frustra-
tion[20,21] in the general sense of the term, or more accur-
ately, competing spin interactions, can result in polynuclear
complexes having a relatively large number of unpaired
Tetranuclear, oxide-bridged metal carboxylate assemblies
are relevant to several areas including bioinorganic model-
ling and magnetochemistry. In the field of iron chemistry,
the pursuit of model complexes for the iron proteins has
yielded several interesting tetranuclear complexes[1Ϫ6]
amongst other polynuclear iron complexes with unusual
electronic structures.[7] The knowledge that in PSII the site
of water oxidation is a tetranuclear manganese aggregate[8]
in which four manganese atoms are essential for activity
and appear to be in close proximity to one another, has
augmented the search for tetranuclear manganese com-
plexes. Thus, tetramanganese butterfly[6,9Ϫ12] complexes
with the [Mn4O2]8ϩ core have emerged as the most intensely
studied oxide-bridged carboxylate clusters. Comparatively
fewer studies of the butterfly [Fe4O2]8ϩ core have been
[a]
Max-Planck-Institut für Strahlenchemie,
Stiftstraße 34Ϫ36, 45470 Mülheim an der Ruhr, Germany
E-mail: Chaudh@mpi-muelheim.mpg.de
[b]
Anorganische und Analytische Chemie, Universität Paderborn,
33095 Paderborn, Germany
Eur. J. Inorg. Chem. 2003, 541Ϫ555  2003 WILEY-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim 1434Ϫ1948/03/0203Ϫ0541 $ 20.00ϩ.50/0
541

P. Chaudhuri et al.
FULL PAPER
electrons in their ground states. Although spin frustration of [L2Fe2(µ3-O)2(µ-carboxylate)3(salox)2Mn2]ClO4 in reas-
is a well-known magnetic phenomenon for extended lat- onable yield. Complex 5 was prepared by using MnIII acet-
tices, its application to the magnetochemistry of discrete po- ate.
lynuclear complexes is not widely recognised. Competing
The IR spectra of the tetranuclear complexes 1Ϫ10 ex-
spin interactions may give rise to unpredictable ground state hibit, besides ligand and ClO4 absorptions, a sharp me-
spins and peculiar spin state spectra. Thus, the situation of dium-intensity band in the region 670Ϫ660 cmϪ1, probably
ground state degeneracy induced by competing spin interac- due to the νasym vibration of the four modes of the
tions is worth investigating.
FeMn2(µ3-O) triangle.[29] The ν(CN) vibration is assigned
We have been exploring the feasibility of using salicylal- to the medium-intensity band at ca. 1600 cmϪ1. One of the
doxime to influence the nuclearity and metal topology of characteristics in the IR spectra is a strong carboxyl stretch-
metal carboxylate complexes. Although salicylaldoxime[22] ing vibration found in the region 1550Ϫ1580 cmϪ1. The
has been known as a metal-chelating agent since 1930, re- complexes show a strong band, whose intensity is similar to
ports on the ligating properties of salicylaldoxime towards that of ν(COO), at ca. 1090 cmϪ1 (antisymmetric stretch)
trivalent transition metal ions are rare.[23,24] Our approach and a sharp band at ca. 620 cmϪ1 (antisymmetric bend),
of using oximes as ligands has yielded several interesting 3d indicative of uncoordinated perchlorate anions.
metal complexes with the cores [MA(oxime)3MB]nϩ [24]
,
[M
Interestingly, the mass spectra of complexes 1Ϫ7 differ
III(µ3-O)(µ2-OPh)]6ϩ (M ϭ Ti, V, Cr, Mn, Fe and Co),[25,26] widely. Although the all-iron complexes 1 and 2 show sig-
3
[MI6II(µ3-O)2(µ2-OOCR)2]12ϩ (M ϭ V, Cr, Mn, Fe)[27] and nals due to the respective monocations [Mϩ] with the ex-
[(MIAII)2(µ3-O)2(MIBII)2]8ϩ (MA ϭ MB ϭ FeIII [4]
;
M
A ϭ CrIII
,
pected isotope distributions in the ESI positive mode, the
MB ϭ FeIII [13]
;
MA ϭ CrIII, MB ϭ MnIII).[14] In this paper manganese compound 3 does not show the corresponding
we report the extension of our studies to complexes having molecular ion. However, in the case of the mixed-metal
tetranuclear cores, and include the preparation and charac- complexes 5Ϫ7, mass spectrometry in the ESI-positive
terisation of [L2Fe4(µ3-O)2(µ2-OOCC(OH)Ph2]3(salox)2]- mode has proved to be a very useful tool for characteris-
ClO4 (1), [L2Fe4(µ3-O)2(µ2-OOCCPh3)3(salox)2]ClO4 (2), ation.
[L2Mn4(µ3-O)2(µ2-OOCC(OH)Ph2]3(salox)2]ClO4
(3),
and
(5Ϫ10)
[L2Mn4(µ3-O)2(µ2-OOCCPh3)3(salox)2]ClO4
[L2Fe2Mn2(µ3-O)2(µ2-carboxylate)3(salox)2]ClO4
(4)
Structures of Complexs 1 and 3
[carboxylate ϭ acetate (5), benzilate (diphenylglycolate) (6),
triphenylacetate (7), benzoate (8), chloroacetate (9) and
propionate (10)], where L is the cyclic amine 1,4,7-trime-
thyl-1,4,7-triazacyclononane and salox is the dianion of sal-
icylaldoxime (H2salox). Molecules 1Ϫ10 contain MIII ions
disposed in a butterfly [M4O2]8ϩ core, where bridging be-
tween the MIII ions occurs via two µ3-oxo anions.
Figure 1 shows a perspective view of the cation of 1. Al-
though this complex contains an [Fe4(µ-diphenyl-
glycolate)3(µ3-oxo)2] core very similar to that in the previ-
ously reported [Fe4(µ-acetato)3(µ3-oxo)2] complex,[4] we
have included the description of the structure of 1 in this
paper to show that the OH groups of the diphenylglycolate
ligands are not coordinated to the oxophilic FeIII centres. It
should also be noted that there are very few tetranuclear
FeIII complexes with the [Fe4O2]8ϩ butterfly core reported
Results
Our previous investigations of the use of solvolysis react-
ivity of [LFeCl3][28] in the synthesis of homo- and heteropo-
lynuclear complexes led us to anticipate compounds with a
butterfly structure, in which the ‘‘body’’ metal ions differ
from the ‘‘wing-tip’’ (Fe3ϩ) ions. The fact that the iron
centres remain bonded to the macrocyclic ligand L during
the formation of the polynuclear complexes was recognised
and subsequently utilised as a synthetic strategy for build-
ing Fe2(O)2Mn2 cores with Mn3ϩ as the ‘‘body’’ ions. Ac-
cordingly, the dark brown solution obtained from mangan-
ese() chloride, a carboxylic acid, salicylaldehyde oxime and
triethylamine in methanol reacted with [LFeCl3] to afford,
Figure 1. Structure of the monocation of complex 1; the bridging
diphenylglycolate (benzilate) anions are denoted by RCOOϪ, i.e.
after addition of a perchlorate solution, dark brown crystals ‘‘Ph2C(OH)’’ has been replaced by ‘‘R’’ for clarity
542
Eur. J. Inorg. Chem. 2003, 541Ϫ555

Ground Spin State Variation in Carboxylate-Bridged Tetranuclear [Fe2Mn2O2]8ϩ Cores
FULL PAPER
˚
˚
˚
to 1.988 A, the two shortest being Fe(4)ϪO(2) [1.872(7) A]
Table 1. Selected bond lengths [A] and angles [°] for 1
˚
and Fe(3)ϪO(1) [1.874(7) A].
Fe(1)ϪO(22)
Fe(1)ϪO(1)
Fe(1)ϪO(10)
1.906(7)
1.988(6)
2.111(9)
Fe(1)ϪO(2)
Fe(1)ϪO(3)
Fe(1)ϪN(21)
Fe(2)ϪO(32)
Fe(2)ϪO(2)
Fe(2)ϪO(9)
Fe(3)ϪO(1)
Fe(3)ϪO(4)
Fe(3)ϪN(1)
Fe(4)ϪO(2)
Fe(4)ϪO(7)
Fe(4)ϪN(5)
1.949(7)
2.109(6)
2.115(10)
1.910(8)
1.981(7)
2.130(6)
1.874(7)
2.031(10)
2.204(11)
1.872(7)
2.039(8)
2.217(10)
The
dihedral
angle
between
the
planes
Fe(1)Fe(2)Fe(3)O(1) and Fe(1)Fe(2)Fe(4)O(2) is 145.9°,
which is slightly smaller than that in the corresponding
acetate complex (153.9°). Selected bond lengths and angles
for the Fe4(µ3-O)2 core and for the rest of the cation are
given in Table 1.
In addition to the two µ3-oxo groups, there are three
bridging carboxylate groups in the tetranuclear structure
originating from the diphenylglycolate ion, one each be-
tween iron atom pairs Fe(4)Fe(2), Fe(1)Fe(2) and
Fe(1)Fe(3). Additionally two NϪO groups of the depro-
tonated salicylaldoximes are coordinated to iron atom pairs
Fe(4)Fe(2) and Fe(3)Fe(1) through their oxygen and nitro-
gen donor atoms, respectively.
It should be noted that the enolic OH group of di-
phenylglycolic acid is not coordinated to any iron centre.
It seems that the nonplanarity of the [Fe4O2]8ϩ unit is a
consequence of the bridging carboxylate (diphenylglycolate)
groups. Specifically, the carboxylate-bridged iron atoms are
apparently ‘‘pulled’’ closer to one another [Fe(1)···Fe(3) ϭ
3.197(2) A, Fe(2)···Fe(4) ϭ 3.192(2) A] than those without
carboxylate bridging [Fe(1)···Fe(4)
Fe(2)···Fe(3) ϭ 3.738(2) A]. Just as there are two long and
two short Fe···Fe separations, there are also two long
[Fe(1)ϪO(1) ϭ 1.988(6) A, Fe(2)ϪO(2) ϭ 1.981(7) A] and
two short [Fe(2)ϪO(1) ϭ 1.949(7) A, Fe(1)ϪO(2) ϭ
1.949(7) A] ironϪµ3-oxo bonds in the inner
Fe(1)Fe(2)O(1)O(2) core.
Fe(2)ϪO(1)
Fe(2)ϪN(31)
Fe(2)ϪO(6)
Fe(3)ϪO(21)
Fe(3)ϪN(3)
Fe(3)ϪN(2)
Fe(4)ϪO(31)
Fe(4)ϪN(6)
1.949(7)
2.112(10)
2.140(7)
1.926(6)
2.167(12)
2.267(11)
1.928(8)
2.171(10)
2.294(10)
Fe(4)ϪN(4)
O(22)-Fe(1)ϪO(2)
O(2)ϪFe(1)ϪO(1)
O(2)ϪFe(1)ϪO(3)
O(22)ϪFe(1)ϪO(10)
O(1)ϪFe(1)ϪO(10)
O(22)ϪFe(1)ϪN(21)
O(1)ϪFe(1)ϪN(21)
O(10)ϪFe(1)ϪN(21)
O(2)ϪFe(1)ϪFe(2)
O(3)ϪFe(1)ϪFe(2)
N(21)ϪFe(1)ϪFe(2)
O(32)ϪFe(2)ϪO(2)
O(32)ϪFe(2)ϪN(31)
O(2)ϪFe(2)ϪN(31)
O(1)ϪFe(2)ϪO(9)
N(31)ϪFe(2)ϪO(9)
O(1)ϪFe(2)ϪO(6)
N(31)ϪFe(2)ϪO(6)
O(32)ϪFe(2)ϪFe(1)
O(2)ϪFe(2)ϪFe(1)
O(9)ϪFe(2)ϪFe(1)
O(1)ϪFe(3)ϪO(21)
O(21)ϪFe(3)ϪO(4)
O(21)ϪFe(3)ϪN(3)
O(1)ϪFe(3)ϪN(4)
O(4)ϪFe(3)ϪN(1)
O(1)ϪFe(3)ϪN(2)
O(4)ϪFe(3)ϪN(2)
N(1)ϪFe(3)ϪN(2)
O(2)ϪFe(4)ϪO(7)
O(2)ϪFe(4)ϪN(6)
O(7)ϪFe(4)ϪN(6)
O(31)ϪFe(4)ϪN(5)
N(6)ϪFe(4)ϪN(5)
O(31)ϪFe(4)ϪN(4)
N(6)ϪFe(4)ϪN(4)
Fe(3)ϪO(1)ϪFe(2)
Fe(2)ϪO(1)ϪFe(1)
Fe(4)ϪO(2)ϪFe(2)
Fe(1)···Fe(2)
104.2(3)
O(22)-Fe(1)ϪO(1)
O(22)-Fe(1)ϪO(3)
O(1)ϪFe(1)ϪO(3)
O(2)ϪFe(1)ϪO(10)
O(3)ϪFe(1)ϪO(10)
O(2)ϪFe(1)ϪN(21)
O(3)ϪFe(1)ϪN(21)
O(22)-Fe(1)ϪFe(2)
O(1)ϪFe(1)ϪFe(2)
O(10)ϪFe(1)ϪFe(2)
O(32)ϪFe(2)ϪO(1)
O(1)ϪFe(2)ϪO(2)
O(1)ϪFe(2)ϪN(31)
O(32)ϪFe(2)ϪO(9)
O(2)ϪFe(2)ϪO(9)
O(32)ϪFe(2)ϪO(6)
O(2)ϪFe(2)ϪO(6)
O(9)ϪFe(2)ϪO(6)
O(1)ϪFe(2)ϪFe(1)
N(31)ϪFe(2)ϪFe(1)
O(6)ϪFe(2)ϪFe(1)
O(1)ϪFe(3)ϪO(4)
O(1)ϪFe(3)ϪN(3)
O(4)ϪFe(3)ϪN(3)
O(21)ϪFe(3)ϪN(1)
N(3)ϪFe(3)ϪN(1)
O(21)ϪFe(3)ϪN(2)
N(3)ϪFe(3)ϪN(2)
O(2)ϪFe(4)ϪO(31)
O(31)ϪFe(4)ϪO(7)
O(31)ϪFe(4)ϪN(6)
O(2)ϪFe(4)ϪN(5)
O(7)ϪFe(4)ϪN(5)
O(2)ϪFe(4)ϪN(4)
O(7)ϪFe(4)ϪN(4)
N(5)ϪFe(4)ϪN(4)
Fe(3)ϪO(1)ϪFe(1)
Fe(4)ϪO(2)ϪFe(1)
Fe(1)ϪO(2)ϪFe(2)
Fe(2)···Fe(3)
172.6(3)
83.0(3)
96.8(3)
84.8(3)
96.2(3)
85.3(4)
87.6(3)
81.4(4)
44.1(2)
109.7(2)
125.4(3)
172.8(3)
86.3(4)
86.5(3)
92.3(3)
79.9(4)
99.8(3)
89.1(3)
143.1(2)
43.2(2)
80.0(2)
96.4(3)
92.4(4)
92.7(4)
100.5(4)
89.6(5)
176.0(4)
86.1(5)
77.5(4)
98.2(3)
96.4(3)
163.0(4)
163.7(4)
81.3(4)
86.3(4)
80.1(4)
155.8(4)
92.4(3)
111.8(3)
87.4(3)
90.3(3)
94.7(3)
167.4(3)
169.5(3)
88.1(4)
144.0(3)
43.3(2)
82.2(2)
104.0(3)
33.2(3)
166.5(3)
83.9(4)
95.3(3)
88.0(3)
91.3(3)
166.8(3)
44.4(2)
122.7(2)
112.4(2)
97.4(3)
96.8(4)
164.3(4)
162.6(4)
81.2(5)
85.5(4)
79.6(5)
95.7(3)
91.8(3)
95.3(4)
100.5(3)
87.5(4)
176.2(3)
85.0(4)
77.4(4)
111.7(3)
155.5(4)
92.6(3)
˚
˚
˚
ϭ
3.734(2) A,
˚
˚
˚
˚
˚
Each iron atom is in a distorted octahedral environment,
with Fe(1) and Fe(2) having an FeNO5 and Fe(3) and Fe(4)
an FeN3O3 coordination sphere, respectively. FeϪN and
FeϪO bond lengths are typical of high-spin ferric com-
plexes. The FeϪN bonds trans to the short FeϪµ3-O bonds
are noticeably longer than those cis to them, as observed
2.841(2)
3.738(2)
Fe(1)···Fe(3)
Fe(1)···Fe(4)
3.197(2)
3.734(2)
Fe(2)···Fe(4)
Fe(3)···Fe(4)
3.192(2)
6.077(2)
in the literature.[1Ϫ6] The metal geometry of the cluster may
be described as ‘‘butterfly’’ based on two edge-sharing
Fe3(µ3-O) triangular units, with the oxygen atoms O(1)
˚
˚
(0.029 A) and O(2) (0.004 A) above the corresponding Fe3
˚
triangles. The Fe(1)···Fe(2) distance [2.841(2) A] is at the
˚
low end of the range (2.829Ϫ3.108 A) observed in other
Figure 2. Perspective view of the cation of 3; ‘‘Ph2C(OH)’’ -units
of the bridging diphenylglycolate (benzilate) anions are denoted
[Fe4O2]-containing compounds.[1Ϫ6] There are three types
of independent FeϪO(oxo) distances, ranging from 1.872 only by ‘‘R’’ for the sake of clarity
Eur. J. Inorg. Chem. 2003, 541Ϫ555
543

P. Chaudhuri et al.
FULL PAPER
earlier.[4] The structural parameters for 1 lie in the range of the butterfly are 0.04 and 0.008 A, respectively, above
˚
reported for the other [Fe4O2]8ϩ complexes.[1Ϫ6]
their Mn3 least-squares planes. The dihedral angle between
A perspective view of the cation of complex 3 is shown the planes comprising Mn(1)Mn(2)Mn(3)O(1) and
in Figure 2, and selected structural parameters are listed in Mn(1)Mn(2)Mn(4)O(2) is 145.8°, a value identical with that
Table 2. The cation, like that in 1, possesses an M4O2 core in the Fe4O2 core of 1. The shortest Mn···Mn contact is
˚
which can be considered as two edge-sharing Mn3O units, 2.844(2) A, between Mn(1) and Mn(2). The coordination
the metal dispositions being a butterfly structure with spheres of ‘‘wing-tip’’ [Mn(4) and Mn(3)], and ‘‘body’’
Mn(1) and Mn(2) occupying the ‘‘body’’ sites and Mn(3) [Mn(1) and Mn(2)] manganese ions are MnN3O3 and
and Mn(4) occupying ‘‘wing-tip’’ sites. Each body-wing-tip MnNO5, respectively, in which the distortions along the
and body-body Mn2 pair is bridged by a carboxylate group axes O(3)Mn(1)O(10), O(6)Mn(2)O(9), N(2)Mn(3)O(1) and
of the diphenylglycolate anion in its common syn,syn bind- N(4)Mn(4)O(2) are as expected for JahnϪTeller-distorted
ing mode. The µ3-O atoms O(1) and O(2) on each Mn3 wing high-spin d4 ions in an octahedral geometry. The MnϪN
and MnϪO bond lengths are consistent with a d4 high-spin
electronic configuration for the MnIII centres. The essential
features of the structure of the cation of 3 are very similar
to those of 1.
˚
Table 2. Selected bond lengths [A] and angles [°] for 3
Mn(1)ϪO(22)
Mn(1)ϪO(2)
Mn(1)ϪO(1)
Mn(1)ϪO(10)
1.913(8)
1.961(8)
1.971(7)
2.107(9)
Mn(2)ϪO(32)
Mn(2)ϪO(1)
Mn(2)ϪO(2)
Mn(2)ϪN(31)
1.904(8)
1.959(8)
1.966(8)
2.101(10)
2.116(9)
2.125(9)
1.869(8)
1.926(8)
2.048(9)
2.167(11)
2.221(11)
2.267(11)
Structures of Complexes 5, 6 and 8
Mn(1)ϪN(21)
2.109(10) Mn(2)ϪO(9)
Mn(1)ϪO(3)
Mn(3)ϪO(1)
Mn(3)ϪO(21)
2.117(9)
1.872(8)
1.938(9)
Mn(2)ϪO(6)
Mn(4)ϪO(2)
Mn(4)ϪO(31)
The molecular geometry and the atom-labelling scheme
of the cation of 5 are shown in Figure 3. The structure con-
sists of a discrete monocationic tetranuclear unit having a
twofold rotation axis. Selected interatomic distances and
angles are listed in Table 3. The cation possesses a ‘‘butter-
fly’’ [Fe2(µ3-O)2Mn2] core. Mn(1) and Mn(1a) occupy
‘‘body’’ positions, and Fe(1) and Fe(1a) occupy the ‘‘wing-
tip’’ positions, with O(1) and O(1a) triply bridging each
FeMn2 wing. The structure can thus be considered as two
edge-sharing FeMn2O triangular units with a dihedral
angle of 145.1°. In contrast to the Mn4O2 complex 3, O(1)
Mn(3)ϪO(4)
Mn(3)ϪN(3)
Mn(3)ϪN(1)
Mn(3)ϪN(2)
2.048(11) Mn(4)ϪO(7)
2.170(13) Mn(4)ϪN(6)
2.198(13) Mn(4)ϪN(5)
2.281(12) Mn(4)ϪN(4)
O(22)ϪMn(1)ϪO(2)
O(22)ϪMn(1)ϪO(1)
O(2)ϪMn(1)ϪO(1)
O(22)ϪMn(1)ϪO(10)
O(2)ϪMn(1)ϪO(10)
O(1)ϪMn(1)ϪO(10)
O(22)ϪMn(1)ϪN(21)
O(2)ϪMn(1)ϪN(21)
O(1)ϪMn(1)ϪN(21)
O(10)ϪMn(1)ϪN(21)
O(22)ϪMn(1)ϪO(3)
O(2)ϪMn(1)ϪO(3)
O(1)ϪMn(1)ϪO(3)
O(10)ϪMn(1)ϪO(3)
N(21)ϪMn(1)ϪO(3)
O(1)ϪMn(3)ϪMn(1)
O(21)ϪMn(3)ϪMn(1)
O(4)ϪMn(3)ϪMn(1)
N(3)ϪMn(3)ϪMn(1)
N(1)ϪMn(3)ϪMn(1)
N(2)ϪMn(3)ϪMn(1)
O(1)ϪMn(3)ϪO(21)
O(1)ϪMn(3)ϪO(4)
O(21)ϪMn(3)ϪO(4)
O(1)ϪMn(3)ϪN(3)
O(21)ϪMn(3)ϪN(3)
O(4)ϪMn(3)ϪN(3)
O(1)ϪMn(3)ϪN(1)
O(21)ϪMn(3)ϪN(1)
O(4)ϪMn(3)ϪN(1)
N(3)ϪMn(3)ϪN(1)
O(1)ϪMn(3)ϪN(2)
O(21)ϪMn(3)ϪN(2)
O(4)ϪMn(3)ϪN(2)
N(3)ϪMn(3)ϪN(2)
N(1)ϪMn(3)ϪN(2)
Mn(1)···Mn(2)
103.9(3)
O(32)ϪMn(2)ϪO(1)
O(32)ϪMn(2)ϪO(2)
O(1)ϪMn(2)ϪO(2)
O(32)ϪMn(2)ϪN(31)
O(1)ϪMn(2)ϪO(31)
O(2)ϪMn(2)ϪN(31)
O(32)ϪMn(2)ϪO(9)
O(1)ϪMn(2)ϪO(9)
O(2)ϪMn(2)ϪO(9)
N(31)ϪMn(2)ϪO(9)
O(32)ϪMn(2)ϪO(6)
O(1)ϪMn(2)ϪO(6)
O(2)ϪMn(2)ϪO(6)
N(31)ϪMn(2)ϪO(6)
O(9)ϪMn(2)ϪO(6)
O(2)ϪMn(4)ϪO(31)
O(2)ϪMn(4)ϪO(7)
O(31)ϪMn(4)ϪO(7)
O(2)ϪMn(4)ϪN(6)
O(31)ϪMn(4)ϪN(6)
O(7)ϪMn(4)ϪN(6)
O(2)ϪMn(4)ϪN(5)
O(31)ϪMn(4)ϪN(5)
O(7)ϪMn(4)ϪN(5)
N(6)ϪMn(4)ϪN(5)
O(2)ϪMn(4)ϪN(4)
O(31)ϪMn(4)ϪN(4)
O(7)ϪMn(4)ϪN(4)
N(6)ϪMn(4)ϪN(4)
N(5)ϪMn(4)ϪN(4)
Mn(3)ϪO(1)ϪMn(2)
Mn(3)ϪO(1)ϪMn(1)
Mn(2)ϪO(1)ϪMn(1)
Mn(4)ϪO(2)ϪMn(1)
Mn(4)ϪO(2)ϪMn(2)
Mn(1)ϪO(2)ϪMn(2)
Mn(2)···Mn(3)
104.5(4)
173.0(3)
82.9(3)
85.1(4)
94.8(3)
96.5(4)
86.1(4)
169.1(4)
87.4(4)
81.5(4)
86.4(3)
96.8(3)
90.9(3)
166.9(3)
88.1(4)
34.8(2)
67.5(3)
80.4(3)
117.6(5)
129.8(4)
147.3(3)
95.9(4)
98.6(4)
92.7(4)
97.8(5)
93.0(5)
161.9(5)
101.1(4)
162.5(4)
88.5(6)
81.1(6)
177.4(5)
85.0(4)
83.7(5)
79.7(6)
77.8(5)
172.4(4)
83.1(3)
86.0(4)
166.0(4)
86.5(4)
83.8(4)
91.6(4)
95.5(4)
80.2(4)
87.6(4)
100.6(3)
91.6(3)
88.9(4)
166.5(4)
95.4(4)
98.2(3)
91.7(4)
96.7(4)
95.5(4)
162.7(4)
100.7(4)
163.8(4)
87.7(4)
81.0(5)
177.3(4)
85.5(4)
84.3(4)
80.7(4)
78.4(4)
154.8(4)
112.3(4)
92.7(3)
154.8(4)
112.3(4)
92.8(3)
˚
lies significantly out (0.14 A) of the FeMn2 plane. In addi-
tion to the two µ3-oxo groups, there are three bridging acet-
ate groups in the tetranuclear structure, one each between
the metal atom pairs Fe(1)Mn(1), Fe(1a)Mn(1a) and
Mn(1)Mn(1a). Additionally, a salicylaldoximate dianion is
2.844(2)
3.738(2)
Mn(1)···Mn(3)
Mn(1)···Mn(4)
3.191(3)
3.737(2)
Mn(2)···Mn(4)
Mn(3)···Mn(4)
3.186(2)
6.070(2)
Figure 3. Molecular structure of the heterometal cation of 5, show-
ing the atom-numbering scheme
544
Eur. J. Inorg. Chem. 2003, 541Ϫ555

Ground Spin State Variation in Carboxylate-Bridged Tetranuclear [Fe2Mn2O2]8ϩ Cores
FULL PAPER
˚
ponding bond length in the analogous cation of 6, but is
similar to that in 8.
Table 3. Selected bond lengths [A] and angles [°] for 5
Fe(1)ϪO(1)
1.8795(13)
The coordination geometry of the ‘‘wing-tip’’ ferric ion
Fe(1) is distorted octahedral, with three nitrogen atoms
from the facially coordinated tridentate macrocyclic amine
(L) and three oxygen atoms [O(2) from the deprotonated
oxime group, O(4) from the bridging acetate and O(1) from
the µ3-bridging oxo group], resulting in a fac-FeN3O3 coor-
Fe(1)ϪO(2)
Fe(1)ϪO(4)
Fe(1)ϪN(4)
Fe(1)ϪN(3)
Fe(1)ϪN(2)
Mn(1)ϪO(3)
Mn(1)ϪO(1)#1
Mn(1)ϪO(1)
1.974(2)
1.989(2)
2.184(2)
2.230(2)
2.279(2)
1.897(2)
1.9343(13)
1.9397(14)
2.047(2)
2.160(2)
2.212(2)
94.35(6)
97.72(6)
98.21(7)
95.86(7)
90.36(7)
163.32(7)
104.06(6)
160.08(6)
86.95(7)
80.37(8)
174.17(7)
83.35(6)
87.93(7)
78.83(7)
77.60(7)
98.03(6)
175.73(6)
84.90(6)
87.31(7)
173.94(6)
89.94(6)
89.96(7)
91.35(6)
93.09(6)
85.76(6)
89.86(6)
95.11(6)
86.78(6)
87.74(6)
173.50(6)
140.19(5)
43.57(4)
43.41(4)
130.59(5)
82.49(4)
101.73(4)
152.41(8)
112.39(7)
93.02(6)
2.811(1)
3.174(1)
6.022(1)
3.704(1)
˚
dination sphere. The Fe(1)ϪN(2) distance [2.279(2) A] is
significantly longer than the other FeϪN distances, due to
Mn(1)ϪN(1)
Mn(1)ϪO(6)
Mn(1)ϪO(5)
˚
the trans effect. The FeϪN (average 2.23 A) and FeϪO (av-
5
˚
erage 1.95 A) distances are in agreement with a d high-
spin electronic configuration for the iron centres. The
O(1)ϪFe(1)ϪO(2)
O(1)ϪFe(1)ϪO(4)
O(2)ϪFe(1)ϪO(4)
O(1)ϪFe(1)ϪN(4)
O(2)ϪFe(1)ϪN(4)
O(4)ϪFe(1)ϪN(4)
O(1)ϪFe(1)ϪN(3)
O(2)ϪFe(1)ϪN(3)
O(4)ϪFe(1)ϪN(3)
N(4)ϪFe(1)ϪN(3)
O(1)ϪFe(1)ϪN(2)
O(2)ϪFe(1)ϪN(2)
O(4)ϪFe(1)ϪN(2)
N(4)ϪFe(1)ϪN(2)
N(3)ϪFe(1)ϪN(2)
O(3)ϪMn(1)ϪO(1)#1
O(3)ϪMn(1)ϪO(1)
O(1)#1ϪMn(1)ϪO(1)
O(3)ϪMn(1)ϪN(1)
O(1)#1ϪMn(1)ϪN(1)
O(1)ϪMn(1)ϪN(1)
O(3)ϪMn(1)ϪO(6)
O(1)#1ϪMn(1)ϪO(6)
O(1)ϪMn(1)ϪO(6)
N(1)ϪMn(1)ϪO(6)
O(3)ϪMn(1)ϪO(5)
O(1)#1ϪMn(1)ϪO(5)
O(1)ϪMn(1)ϪO(5)
N(1)ϪMn(1)ϪO(5)
O(6)ϪMn(1)ϪO(5)
O(3)ϪMn(1)ϪMn(1)#1
O(1)#1ϪMn(1)ϪMn(1)#1
O(1)ϪMn(1)ϪMn(1)#1
N(1)ϪMn(1)ϪMn(1)#1
O(6)ϪMn(1)ϪMn(1)#1
O(5)ϪMn(1)ϪMn(1)#1
Fe(1)ϪO(1)ϪMn(1)#1
Fe(1)ϪO(1)ϪMn(1)
Mn(1)#1ϪO(1)ϪMn(1)
Mn(1)···Mn(1a)
˚
Fe(1)Ϫµ3-oxo distance [1.880(1) A] is, as expected, the
shortest amongst the metalϪligand bond lengths. The iron
˚
atom Fe(1) is displaced by 0.28 A from the mean basal
plane comprising N(3)N(4)O(2)O(4) toward the apical µ3-
oxygen atom O(1).
The structure of the cation of complex 6 is presented in
Figure 4. The cation possesses an FeI2IIMnI2II(µ3-O)2 core
very similar to those present in the Fe4O2 (1), Mn4O2 (3)
and Fe2Mn2O2 (5 and 8) core structures. The metal centres
are thus disposed in a ‘‘butterfly’’ arrangement based on
two edge-sharing FeMn2(µ3-O) triangular units, with the
˚
oxygen atoms O(1) and O(2) slightly above (0.07 A) and
˚
below (0.02 A) the Mn2Fe triangle, as depicted below; note
that the drawing is not meant to suggest a stepwise assem-
bly process.
Mn(1)···Fe(1)
Fe(1)···Fe(1a)
Fe(1)···Mn(1a)
coordinated to Mn(1) through its phenolate oxygen atom
O(3) and oxime nitrogen atom N(1), thus making Mn(1)
six-coordinate. The X-ray structure clearly shows that the
‘‘body’’ metal ions are JahnϪTeller-distorted, high-spin d4
MnIII ions; the axially elongated sites are occupied by the
carboxylate oxygen atoms of acetate, with Mn(1)ϪO(6)
˚
˚
2.160(2)
A,
Mn(1)ϪO(5)
2.212(2)
A
and
Figure 4. Structure of the tetranuclear monocation of 6; phenyl
O(5)ϪMn(1)ϪO(6) 173.5°. Interestingly, the Mn(1)ϪO(1)
rings (Ph) of the bridging diphenylglycolate (benzilate)
distance [1.940(1) A] is significantly shorter than the corres- Ph2C(OH)COOϪ are denoted by C atoms
˚
Eur. J. Inorg. Chem. 2003, 541Ϫ555
545

P. Chaudhuri et al.
Table 4. Selected bond lengths [A] and angles [°] for 6
FULL PAPER
˚
Mn(1)ϪO(22)
Mn(1)ϪO(2)
Mn(1)ϪO(1)
Mn(1)ϪO(3)
Mn(1)ϪO(10)
Mn(1)ϪN(21)
Mn(1)ϪMn(2)
Mn(2)ϪO(32)
Mn(2)ϪO(1)
Mn(2)ϪO(2)
Mn(2)ϪN(31)
Mn(2)ϪO(9)
1.906(2)
1.930(2)
1.972(2)
2.097(2)
2.103(2)
2.113(2)
2.8310(6)
1.907(2)
1.944(2)
1.966(2)
2.098(2)
2.114(2)
2.127(2)
Fe(1)ϪO(1)
Fe(1)ϪO(21)
Fe(1)ϪO(4)
Fe(1)ϪN(3)
Fe(1)ϪN(1)
Fe(1)ϪN(2)
Fe(2)ϪO(2)
Fe(2)ϪO(31)
Fe(2)ϪO(7)
Fe(2)ϪN(6)
Fe(2)ϪN(5)
Fe(2)ϪN(4)
1.863(2)
1.933(2)
2.041(2)
2.167(3)
2.188(2)
2.268(3)
1.876(2)
1.931(2)
2.045(2)
2.149(2)
2.205(2)
2.270(2)
Mn(2)ϪO(6)
The dispositions of the µ3-oxo groups are not completely
symmetrical, but are consistent with all metal atoms being
in the 3ϩ oxidation state. The dihedral angle between the
planes Mn(1)Mn(2)Fe(2) and Mn(1)Mn(2)Fe(1) is 145.9°.
Selected bond lengths and angles for the FeI2IIMnI2IIO2 core
and for the rest of the cation are given in Table 4. The
‘‘body’’ metal centres [Mn(1) and Mn(2)] show clear evid-
ence of JahnϪTeller distortions, as expected for high-spin
octahedral d4 MnIII, and the axially elongated sites are all
occupied by carboxylate oxygen atoms of diphenylglycolic
acid [O(3), O(10) for Mn(1) and O(6), O(9) for Mn(2)]. The
hydroxy oxygen atoms of (C6H5)2C(OH)COOϪ O(8), O(5)
O(22)ϪMn(1)ϪO(2)
O(22)ϪMn(1)ϪO(1)
O(2)ϪMn(1)ϪO(1)
O(22)ϪMn(1)ϪO(3)
O(2)ϪMn(1)ϪO(3)
O(1)ϪMn(1)ϪO(3)
O(22)ϪMn(1)ϪO(10)
O(2)ϪMn(1)ϪO(10)
O(1)ϪMn(1)ϪO(10)
O(3)ϪMn(1)ϪO(10)
O(22)ϪMn(1)ϪN(21)
O(2)ϪMn(1)ϪN(21)
O(1)ϪMn(1)ϪN(21)
O(3)ϪMn(1)ϪN(21)
103.93(8)
O(1)ϪFe(1)ϪO(21)
O(1)ϪFe(1)ϪO(4)
O(21)ϪFe(1)ϪO(4)
O(1)ϪFe(1)ϪN(3)
O(21)ϪFe(1)ϪN(3)
O(4)ϪFe(1)ϪN(3)
O(1)ϪFe(1)ϪN(1)
O(21)ϪFe(1)ϪN(1)
95.68(8)
97.72(8)
93.00(9)
96.82(10)
92.73(10)
163.75(10)
100.61(9)
163.08(9)
172.69(8)
83.21(7)
86.86(8)
96.82(8)
90.74(8)
85.02(8)
95.00(8)
96.05(8)
166.98(8)
85.82(9)
169.12(8)
87.21(8)
88.42(8)
80.86(9)
143.91(6)
43.89(5)
43.30(5)
109.38(6)
83.00(5)
125.31(6)
103.88(8)
172.94(8)
83.02(7)
86.14(9)
166.16(8)
86.81(8)
83.73(8)
91.94(8)
94.70(8)
79.53(8)
88.20(8)
100.22(7)
92.03(7)
89.47(8)
166.74(8)
142.84(6)
44.08(5)
42.90(5)
122.91(6)
80.11(5)
112.36(5)
2.831(1)
3.185(1)
3.716(1)
O(4)ϪFe(1)ϪN(1)
N(3)ϪFe(1)ϪN(1)
O(1)ϪFe(1)ϪN(2)
O(21)ϪFe(1)ϪN(2)
O(4)ϪFe(1)ϪN(2)
N(3)ϪFe(1)ϪN(2)
N(1)ϪFe(1)ϪN(2)
O(2)ϪFe(2)ϪO(31)
O(2)ϪFe(2)ϪO(7)
O(31)ϪFe(2)ϪO(7)
O(2)ϪFe(2)ϪN(6)
O(31)ϪFe(2)ϪN(6)
O(7)ϪFe(2)ϪN(6)
O(2)ϪFe(2)ϪN(5)
O(31)ϪFe(2)ϪN(5)
O(7)ϪFe(2)ϪN(5)
O(6)ϪFe(2)ϪN(5)
O(2)ϪFe(2)ϪN(4)
O(31)ϪFe(2)ϪN(4)
O(7)ϪFe(2)ϪN(4)
N(6)ϪFe(2)ϪN(4)
N(5)ϪFe(2)ϪN(4)
89.38(10)
80.85(11)
176.44(9)
85.53(9)
85.54(10)
79.77(11)
77.95(10)
95.65(8)
98.24(8)
91.73(8)
95.98(8)
95.43(8)
163.37(8)
101.08(8)
163.17(8)
87.65(8)
81.16(9)
175.81(8)
85.60(8)
85.71(8)
79.91(9)
77.58(8)
and O(11) are not coordinated to any of the metal centres. O(10)ϪMn(1)ϪN(21)
O(22)ϪMn(1)ϪMn(2)
O(2)ϪMn(1)ϪMn(2)
O(1)ϪMn(1)ϪMn(2)
O(3)ϪMn(1)ϪMn(2)
There are three types of independent FeϪO distances, ran-
˚
ging from 1.863(2) to 2.045(2) A, with the shortest, as ex-
pected, being the FeϪµ3-oxo distances Fe(1)ϪO(1)
˚
˚
[1.863(2) A] and Fe(2)ϪO(2) [1.876(2) A]. The FeϪN dis-
tances trans to the µ3-oxo groups [Fe(1)ϪN(2) ϭ 2.268(3)
O(10)ϪMn(1)ϪMn(2)
N(21)ϪMn(1)ϪMn(2)
O(32)ϪMn(2)ϪO(1)
O(32)ϪMn(2)ϪO(2)
O(1)ϪMn(2)ϪO(2)
O(32)ϪMn(2)ϪN(31)
O(1)ϪMn(2)ϪN(31)
O(2)ϪMn(2)ϪN(31)
O(32)ϪMn(2)ϪO(9)
O(1)ϪMn(2)ϪO(9)
O(2)ϪMn(2)ϪO(9)
N(31)ϪMn(2)ϪO(9)
O(32)ϪMn(2)ϪO(6)
O(1)ϪMn(2)ϪO(6)
O(2)ϪMn(2)ϪO(6)
N(31)ϪMn(2)ϪO(6)
O(9)ϪMn(2)ϪO(6)
O(32)ϪMn(2)ϪMn(1)
O(1)ϪMn(2)ϪMn(1)
O(2)ϪMn(2)ϪMn(1)
N(31)ϪMn(2)ϪMn(1)
O(9)ϪMn(2)ϪMn(1)
O(6)ϪMn(2)ϪMn(1)
Mn(1)···Mn(2)
˚
˚
A, Fe(2)ϪN(4) ϭ 2.270(2) A] are longer than the other
FeϪN distances, due to the trans effect. The coordination
geometry of the ‘‘wing-tip’’ ferric ions, Fe(1) and Fe(2), is
distorted octahedral, with three nitrogen atoms from the
facially coordinated tridentate macrocyclic amine (L) and
three oxygen atoms [O(31) or O(21) of the deprotonated
oxime group, O(4) or O(7) of the bridging carboxylate and
O(1) or O(2) of the µ3-bridging oxo group] resulting in a
fac-FeN3O3 coordination sphere. The FeϪN [average
˚
˚
2.20(6) A] and FeϪO [average 1.95(6) A] distances are sim-
ilar to literature values[28] for high-spin FeIII complexes with
this macrocyclic amine. A deviation from the idealised octa-
hedral geometry at the metal centre is found for the ligand
L; the NϪFeϪN angles range from 77.6 to 81.2°, whereas
the OϪFeϪO angles fall between 91.7 and 98.2°. The iron
Fe(1)ϪO(1)ϪMn(2)
Fe(1)ϪO(1)ϪMn(1)
Mn(2)ϪO(1)ϪMn(1)
Fe(2)ϪO(2)ϪMn(1)
Fe(2)ϪO(2)ϪMn(2)
Mn(1)ϪO(2)ϪMn(2)
154.88(10)
112.26(9)
92.62(8)
154.95(10)
111.80(9)
93.21(8)
˚
ions Fe(1) and Fe(2) are displaced by 0.28 A from the mean
basal planes comprising O(21)O(4)N(3)N(1) and
N(5)N(6)O(7)O(31), respectively, toward the apical µ3-oxy-
gen atoms O(1) and O(2).
The ‘‘body’’ manganese ions, Mn(1) and Mn(2), are
bridged by two oxo groups, O(2) and O(1). Each Mn is
coordinated to three oxygen atoms [O(1), O(2) and
Mn(2)···Fe(1)
Mn(2)···Fe(2)
Fe(2)···Fe(1)
3.715(1)
3.182(1)
6.048(1)
Mn(1)···Fe(1)
Mn(1)···Fe(2)
phenolate O(22) or O(32)] and an imine nitrogen atom ese centres Mn(1) and Mn(2) lie out of the best basal
[N(21) or N(31)] of an oxime group in the basal plane. The planes comprising Mn(1)N(21)O(22)O(1)O(2) and
˚
˚
carboxylate oxygen atoms O(3)/O(10) or O(6)/O(9) occupy Mn(2)N(31)O(32)O(1)O(2) by 0.028 A and 0.09 A in the
the axially elongated apical positions. The MnϪN and direction of the bridging O(3) or O(6), respectively.
MnϪO bond lengths are consistent with a d4 high-spin
The structure of 8·2CH3CN is very similar to that of 5
electronic configuration for the MnIII centres. The mangan- and 6 and hence only the core structure is shown in Fig-
546
Eur. J. Inorg. Chem. 2003, 541Ϫ555

Ground Spin State Variation in Carboxylate-Bridged Tetranuclear [Fe2Mn2O2]8ϩ Cores
FULL PAPER
˚
ure 5. Table 5 gives selected bond lengths and angles for 8.
Remarkably, the MnϪOϪMn angles in 5, 6 and 8 (93°) are
nearly identical. On the other hand, the MnϪµ3-O dis-
Table 5. Selected bond lengths [A] and angles [°] for 8
Fe(1)ϪO(1)
Fe(1)ϪO(50)
Fe(1)ϪO(61)
Fe(1)ϪN(7)
Fe(1)ϪN(4)
Fe(1)ϪN(1)
Fe(1)ϪO(2)
Fe(1)ϪO(60)
Fe(1)ϪO(71)
Fe(1)ϪN(27)
Fe(1)ϪN(24)
Fe(1)ϪN(21)
1.892(2)
1.959(2)
2.013(2)
2.190(2)
2.226(2)
2.265(2)
1.895(2)
1.961(2)
1.985(2)
2.196(2)
2.213(2)
2.268(2)
Mn(1)ϪO(41)
Mn(1)ϪO(2)
Mn(1)ϪO(1)
Mn(1)ϪN(49)
Mn(1)ϪO(81)
Mn(1)ϪO(63)
Mn(1)ϪMn(2)
Mn(2)ϪO(51)
Mn(2)ϪO(1)
Mn(2)ϪO(2)
Mn(2)ϪN(59)
Mn(2)ϪO(83)
1.890(2)
1.927(2)
1.933(2)
2.022(2)
2.187(2)
2.222(2)
2.8034(6)
1.890(2)
1.917(2)
1.918(2)
2.010(2)
2.220(2)
2.294(2)
96.86(8)
176.12(8)
84.03(7)
88.70(8)
171.68(8)
90.84(8)
89.73(8)
91.01(7)
94.04(7)
82.80(8)
87.97(7)
96.89(7)
88.17(7)
89.47(8)
171.98(7)
97.49(7)
177.81(8)
84.70(7)
87.71(8)
174.32(8)
90.10(8)
89.21(7)
93.75(7)
90.68(7)
84.02(8)
91.50(7)
95.73(7)
88.24(7)
86.36(8)
170.32(7)
˚
tances in 8 (average 1.92 A) are shorter than those in 5 or
˚
6 (average 1.94 A), although the FeϪµ3-O distances for 5,
˚
6 and 8 are nearly identical (average 1.88 A). This difference
in bond lengths presumably has significant effects on the
ground state St values, as will be shown later.
Mn(2)ϪO(73)
O(1)ϪFe(1)ϪO(50)
O(1)ϪFe(1)ϪO(61)
O(50)ϪFe(1)ϪO(61)
O(1)ϪFe(1)ϪN(7)
O(50)ϪFe(1)ϪN(7)
O(61)ϪFe(1)ϪN(7)
O(1)ϪFe(1)ϪN(4)
O(50)ϪFe(1)ϪN(4)
O(61)ϪFe(1)ϪN(4)
N(7)ϪFe(1)ϪN(4)
O(1)ϪFe(1)ϪN(1)
O(50)ϪFe(1)ϪN(1)
O(61)ϪFe(1)ϪN(1)
N(7)ϪFe(1)ϪN(1)
N(4)ϪFe(1)ϪN(1)
O(2)ϪFe(2)ϪO(60)
O(2)ϪFe(2)ϪO(71)
O(60)ϪFe(2)ϪO(71)
O(2)ϪFe(2)ϪN(27)
O(60)ϪFe(2)ϪN(27)
O(71)ϪFe(2)ϪN(27)
O(2)ϪFe(2)ϪN(24)
O(60)ϪFe(2)ϪN(24)
O(71)ϪFe(2)ϪN(24)
N(27)ϪFe(2)ϪN(24)
93.52(8)
96.42(7)
100.55(8)
97.14(8)
92.00(8)
160.89(8)
106.76(8)
158.64(8)
84.10(8)
79.13(8)
173.57(8)
81.35(8)
88.36(8)
79.27(8)
77.94(8)
93.89(8)
96.96(8)
98.69(8)
96.72(8)
90.30(8)
163.06(8)
105.74(8)
158.95(8)
86.61(8)
80.09(8)
173.58(8)
81.18(8)
87.89(8)
79.29(8)
78.66(8)
154.22(10)
111.25(8)
93.49(7)
112.25(9)
150.98(10)
93.64(8)
2.803(1)
3.157(2)
3.700(2)
O(41)ϪMn(1)ϪO(2)
O(41)ϪMn(1)ϪO(1)
O(2)ϪMn(1)ϪO(1)
O(41)ϪMn(1)ϪN(49)
O(2)ϪMn(1)ϪN(49)
O(1)ϪMn(1)ϪN(49)
O(41)ϪMn(1)ϪO(81)
O(2)ϪMn(1)ϪO(81)
O(1)ϪMn(1)ϪO(81)
N(49)ϪMn(1)ϪO(81)
O(41)ϪMn(1)ϪO(63)
O(2)ϪMn(1)ϪO(63)
O(1)ϪMn(1)ϪO(63)
N(49)ϪMn(1)ϪO(63)
O(81)ϪMn(1)ϪO(63)
O(51)ϪMn(2)ϪO(1)
O(51)ϪMn(2)ϪO(2)
O(1)ϪMn(2)ϪO(2)
O(51)ϪMn(2)ϪN(59)
O(1)ϪMn(2)ϪN(59)
O(2)ϪMn(2)ϪN(59)
O(51)ϪMn(2)ϪO(83)
O(1)ϪMn(2)ϪO(83)
O(2)ϪMn(2)ϪO(83)
N(59)ϪMn(2)ϪO(83)
O(51)ϪMn(2)ϪO(73)
O(1)ϪMn(2)ϪO(73)
O(2)ϪMn(2)ϪO(73)
N(59)ϪMn(2)ϪO(73)
O(83)ϪMn(2)ϪO(73)
Figure 5. Atom connectivity of the Fe2Mn2O2 core in the cation
of 8
Mössbauer Spectroscopy
The Mössbauer spectra of 1 (Figure 6, curve A) and 2
were recorded without an applied field at 80 K. Each spec-
trum is best fit to two equal-area doublets, whose isomer
shifts (δFe) and quadrupole splitting (∆EQ) values are listed
in Table 6. The two doublets imply two Fe sites with differ- O(2)ϪFe(2)ϪN(21)
O(60)ϪFe(2)ϪN(21)
ent coordination spheres, in agreement with the molecular
O(71)ϪFe(2)ϪN(21)
structure of 1, and are thus assigned to the ‘‘body’’ and
N(27)ϪFe(2)ϪN(21)
‘‘wing-tip’’ pairs of FeIII atoms. The isomer shift δFe of 0.5
N(24)ϪFe(2)ϪN(21)
Ϯ 0.05 mm/s is of the magnitude expected for the high-spin
ferric state and is close to the values reported for similar
compounds.[30] The ∆EQ values for both sites in 1 and 2
reflect deviations from octahedral geometry, since the val-
ence electron contribution to ∆EQ is negligible for high-spin
FeIII. The two ∆EQ values are significantly different, and
larger than the 0.4Ϫ0.7 mm/s range observed for basic iron
carboxylates.[31] By comparing the bond lengths for the
NO5 coordination sphere around Fe(1) and Fe(2) with
those of the more symmetrical N3O3 coordination sphere
Fe(1)ϪO(1)ϪMn(2)
Fe(1)ϪO(1)ϪMn(1)
Mn(2)ϪO(1)ϪMn(1)
Fe(2)ϪO(2)ϪMn(2)
Fe(2)ϪO(2)ϪMn(1)
Mn(2)ϪO(2)ϪMn(1)
Mn(1)···Mn(2)
Mn(2)···Fe(1)
Mn(2)···Fe(2)
Fe(2)···Fe(1)
3.713(2)
3.166(2)
6.024(2)
Mn(1)···Fe(1)
Mn(1)···Fe(2)
about Fe(3) and Fe(4), the doublet with ∆EQ ഠ 1.37 mm/s in the solid-state structure, namely the ‘‘wing-tip’’ iron sites
can be assigned to the ‘‘body’’ iron atoms Fe(1) and Fe(2), Fe(1) and Fe(2). The δFe values of ca. 0.45 mm/s are consist-
whereas the inner doublet is assignable to the ‘‘wing-tip’’ ent with a d5 high-spin electron configuration for the iron
iron atoms Fe(3) and Fe(4). This assignment is also corrob- centres Fe(1) and Fe(2) in complexes 5Ϫ10. The isomer
orated by the Mössbauer data for complexes 5Ϫ10.
shift values are surprisingly constant within these com-
The powder-state Mössbauer spectrum of 6 at 80 K is plexes, and indicate that a change in the carboxylate group
displayed in Figure 6, curve B. Essentially similar spectra Ϫ with its varying electron-withdrawing power Ϫ has little
were obtained for the remaining complexes of the group. effect upon the s-electron density at the iron nucleus. The
The δFe and ∆EQ values for 5Ϫ10 are listed in Table 6. ∆EQ values (0.53Ϫ0.61 mm/s) for 5Ϫ10 lie within the
Mössbauer results indicate a single quadrupole splitting, in 0.4Ϫ0.7 mm/s range found for basic iron carboxylates, and
accordance with only one type of iron coordination sphere are similar to those for analogous structures with a dis-
Eur. J. Inorg. Chem. 2003, 541Ϫ555
547

P. Chaudhuri et al.
FULL PAPER
Scheme 1
ent for the ‘‘body’’ site [Fe(1) of complexes 1 and 2] origin-
2
2
ates predominantly from strong covalency of the dx Ϫy or-
bital and the resulting anisotropy in the distribution of val-
ence electrons. In contrast, the short bonds in the ‘‘wing-
tip’’ sites are distributed in a more balanced manner among
the three major axes. In addition, the covalency due to the
particularly short Fe4ϪO2 and Fe2ϪO2 bonds in com-
pounds 1 and 6, respectively, appears to be partly compens-
ated for by a pronounced trans effect in the opposite bonds.
Thus, the covalent delocalisation of 3d electrons seems to
be more isotropic, as it affects both eg orbitals. Since these
Figure 6. Mössbauer spectra of 1 (A) and 6 (B) at 80 K
Table 6. Mössbauer parameters at 80 K
Complex
δFe [mm·sϪ1 [a]
]
∆EQ [mm·sϪ1
]
1
2
0.47
0.53
0.47
0.53
0.45
0.46
0.45
0.46
0.46
0.46
0.95
1.37
0.95
1.38
0.61
0.54
0.53
0.58
0.54
0.54
2
2
2
orbitals, dx Ϫy and dz , have equally strong but opposite
contributions to the total electric field gradient, the smaller
quadrupole splittings observed for the ‘‘wing-tip’’ iron sites
appears to be plausible.
5
6
7
8
9
10
Magnetic Properties
[a]
Isomer shift relative to iron foil at room temperature. Typical
The magnetic susceptibility data for polycrystalline
samples of 1Ϫ10 were collected in the temperature range
2Ϫ290 K in an applied magnetic field of 1 T and are dis-
played in Figures 7Ϫ9 as plots of the effective magnetic mo-
ment (µeff) versus temperature (T). Table 7 summarises the
intracluster exchange parameters, together with the ob-
served spin ground states, and selected µeff values for 1Ϫ10
are listed in Table 8. The solid lines in Figures 7Ϫ9 repres-
errors are Ϯ0.01 for δFe and Ϯ0.02 for ∆EQ.
torted octahedral symmetry. Comparison of the Mössbauer
data for 5Ϫ10 with those for 1 and 2 makes it clear that the
‘‘wing-tip positions in complexes 5Ϫ10 are occupied by Fe
atoms in all cases, and hence the cyclic amine 1,4,7-trime-
thyl-1,4,7-triazacyclononane is coordinated to iron centres
only.This is confirmed by the X-ray structures of 5, 6 and 8.
The remarkable difference in quadrupole splitting of the
‘‘wing-tip’’ and ‘‘body’’ iron sites in complexes 1, 2 and 6
can be rationalised by comparing the distributions of bond
lengths for the different coordination polyhedra. To this
end, the ironϪligand bonds were classified into ‘‘short’’ and
‘‘long’’ bonds as sketched in Scheme 1.
It is striking that for the ‘‘wing-tip’’ iron atoms, the three
shortest FeϪO bonds lie in a fac configuration, whereas for
the ‘‘body’’ site, they have a mer arrangement. Thus, the
‘‘body’’ site can be regarded as a distorted octahedron with
bond compression in the equatorial plane (if the z axis is
taken in the O3ϪFe1ϪO10 direction). Since large quadru-
pole splittings of high-spin FeIII (3d5 configuration) are the
result of strong anisotropy in the covalent delocalisation of
valence electrons, as suggested for the FeϪOϪFe unit in
Figure 7. Temperature dependence of the magnetic moments of
complexes 1 and 2; the solid lines represent the best fit of data to
other systems,[32] we infer that the large electric field gradi- the HeisenbergϪDiracϪvan Vleck model (see text)
548
Eur. J. Inorg. Chem. 2003, 541Ϫ555

Ground Spin State Variation in Carboxylate-Bridged Tetranuclear [Fe2Mn2O2]8ϩ Cores
FULL PAPER
Table 8. Selected µeff values as a function of temperature for com-
plexes 1Ϫ10
Com- Complex
pound
µeff, µB (T [K])
1
2
Fe4 benzilate
0.56 (2), 1.26 (40), 2.46 (110),
4.32 (290)
0.52 (5), 1.30 (40), 2.64 (110),
4.66 (290), 4.76 (300)
Fe4 triphenyl acetate
3
4
Mn4 benzilate
4.56 (2), 6.93 (10), 8.00 (40),
9.07 (110), 9.79 (290)
4.50 (2), 6.87 (10), 7.95 (40),
9.27 (110), 10.22 (290)
Mn4 triphenyl acetate
5
Fe2Mn2 acetate
2.67 (2), 3.14 (5), 3.38 (10),
3.49 (15), 4.05 (40),
5.37 (110), 7.13 (290),
7.16 (295)
Figure 8. Plots of µeff vs. T for solid 3 and 4; the solid lines are the
best least-squares fits of the experimental data to the theoretical
equation
6
7
8
Fe2Mn2 benzilate
2.60 (2), 2.98 (5), 3.23 (15),
4.20 (40), 5.58 (110)
7.05 (290)
Fe2Mn2 triphenyl acetate 3.63 (2), 5.20 (5), 5.95 (15),
6.79 (40), 8.01 (110)
9.27 (290)
3.62 (2), 4.30 (5), 4.44 (10),
4.50 (15), 4.96 (40)
Fe2Mn2 benzoate
6.10 (110), 7.75 (290)
3.63 (2), 5.06 (15), 5.82 (40),
6.97 (110), 8.42 (290)
2.72 (2), 3.62 (15), 4.73 (40),
6.49 (110), 8.39 (290)
9
Fe2Mn2 chloroacetate
Fe2Mn2 propionate
10
Figure 9. Temperature dependence of the magnetic moment, µeff
,
netic behaviour of the tetranuclear complexes. Con-
sequently, the other conceivable ‘‘3J’’ model has not been
considered to avoid overparametrisation.
for 5Ϫ10; the solid lines represent the simulation with the spin
Hamiltonian (see text)
ent the simulations. The model used for the simulations in-
The resulting spin Hamiltonian used to describe the iso-
volves two pairwise exchange interactions, Jwb for the wing- tropic exchange interactions is given by
ˆ ˆ ˆ ˆ ˆ ˆ
ˆ
H ϭ Ϫ2Jwb (S1•S3 ϩ S2•S3 ϩ S1•S4 ϩ S2•S4) Ϫ
ˆ ˆ
tip-body and Jbb for the body-body interactions. The oxide
bridges have been shown to be more important mediators 2Jbb(S1•S2)
ˆ
ˆ
5
for the exchange interactions than the carboxylate bridges
in the structurally related complexes [Fe4(µ3-O)2]8ϩ [3,4]
[Mn4(µ3-O)2]8ϩ [9,35] [Cr4(µ3-O)2]8ϩ [33] [V4(µ3-O)2]8ϩ [34]
[Cr2(µ3-O)2Fe2]8ϩ [13] and [Cr2(µ3-O)2Mn2]8ϩ [14]
in which a
where Si (i ϭ 1, 2, 3, 4) ϭ /2 for 1 and 2, Si (i ϭ 1, 2,
5
,
3, 4) ϭ 2 for 3 and 4, and S1 ϭ S2 ϭ 2; S3 ϭ S4 ϭ /2
,
,
,
for 5Ϫ10.
,
The magnetic moment for 1 and 2 decreases monotonic-
similar ‘‘2J’’ model is able to adequately describe the mag- ally with decreasing temperature to an essentially diamag-
Table 7. Magnetic parameters for complexes 1Ϫ10
Compound
Complex
Jwb [cmϪ1
(J13 ϭ J24 ϭ J23 ϭ J14)
]
Jbb [cmϪ1] g
D [cmϪ1
]
Ground state
(J12)
(g1 ϭ g2 ϭ g3 ϭ g4) (D1 ϭ D2 ϭ D3 ϭ D4) St
1
2
Fe4 benzilate
Fe4 triphenyl acetate
Ϫ41.4
Ϫ38.2
0 (fixed)
0 (fixed)
2.0 (fixed)
2.0 (fixed)
Ϫ
Ϫ
0
0
3
4
Mn4 benzilate
Mn4 triphenyl acetate
Ϫ0.47
Ϫ1.63
Ϫ7.73
Ϫ6.71
2.099 (fixed)
2.215 (fixed)
0 (fixed)
0 (fixed)
3
3
5
Fe2Mn2 acetate
Fe2Mn2 benzilate
Fe2Mn2 triphenyl acetate
Fe2Mn2 benzoate
Fe2Mn2 chloroacetate
Fe2Mn2 propionate
Ϫ25.2
Ϫ22.3
Ϫ5.5
Ϫ19.4
Ϫ10.6
Ϫ14.4
Ϫ35.1
Ϫ31.6
Ϫ10.2
Ϫ28.9
Ϫ17.8
Ϫ19.4
2.05 (fixed)
2.00 (fixed)
2.00 (fixed)
2.09
2.01 (fixed)
2.13 (fixed)
Ϫ0.38
0 (fixed)
0 (fixed)
0
2.07
0 (fixed)
1
1
2, 3
2, 1
2
6
7
8
9
10
1
Eur. J. Inorg. Chem. 2003, 541Ϫ555
549

P. Chaudhuri et al.
FULL PAPER
netic value of 0.55 µB for 1 at 2 K and 0.52 µB at 5 K for 2; value of µeff ϭ 2.67 µB at 2 K, indicating a ground state of
the residual nonzero moment is due primarily to an amount St ϭ 1 (Figure 9). A full-matrix diagonalisation approach
of a paramagnetic impurity PI (assumed S ϭ 5/2). Thus, the including spin exchange (Ϫ2J S1·S2), Zeeman interactions,
2
temperature-dependent behaviours of the magnetic mo- and single-ion zero-field interaction (DSz ), was employed
ments of 1 and 2 clearly indicate an St ϭ 0 electronic to fit the data. The solid line in Figure 9 shows the good
ground state, in agreement with previous observations[1Ϫ6] quality of the fit above T ϭ 15 K with the following para-
for other molecules of similar structure. The theoretical van meters:
Vleck equation used for simulation of the susceptibility data
J
FeMn ϭ Jwb ϭ Ϫ25·2 cmϪ1; JMnMn ϭ Jbb ϭ Ϫ35.1 cmϪ1
;
has been described earlier.[4]
D(MnIII) ϭ Ϫ0.38 cmϪ1 and g ϭ 2.05 (fixed). As it is diffi-
The Zeeman interaction of the high-spin ferric ion in a cult, if not impossible, to unambiguously determine the sign
6A1 ground state with practically no contribution from or- of D from powder magnetic susceptibility measurements,
bital angular momentum is isotropic and the observed g one should not put too much weight on the absolute value
values are thus very close to the free-electron spin value of of D, although inclusion of D as a parameter slightly im-
2.0. Hence, we have used a fixed value of g ϭ 2.0 for 1 and proved the quality of the fit. We addressed the point of
2 during simulation. As the quality of the fit is negligibly global and local minima by consideration of the two-dimen-
dependent on Jbb,[3,4] we have kept Jbb ϭ 0. Previous work sional contour projection of the error surface. The error
has shown that Jbb cannot be precisely determined from the surface shows that the fitting procedure has correctly identi-
susceptibility data in the temperature range ca 4Ϫ300 K for fied the global minimum and we conclude that 5 has an
such complexes, but Jwb can be evaluated with reasonable St ϭ 1 ground state lying 22.3 cmϪ1 below the first excited
accuracy.[3,4] Additionally, an error-surface plot for 1 as a state with St ϭ 2. For a system of two high-spin FeIII and
function of Jwb and Jbb shows that, while the value of Jwb two high-spin MnIII ions disposed in a butterfly-type ar-
is well determined by the fitting procedure, the value of Jbb rangement as found in complexes 5Ϫ10, the overall spin
is not. The best fits are shown as solid lines in Figure 7 and degeneracy (900) is distributed over 110 levels with St values
the fitting parameters are Jwb ϭ Ϫ41.4 cmϪ1, Jbb ϭ 0 ranging from 0 to 9.
(fixed), g ϭ 2.0 (fixed) and PI ϭ 0.008 for 1, and Jwb
ϭ
The effective magnetic moment (µeff/molecule) for 6 de-
Ϫ38.2 cmϪ1, Jbb ϭ 0 (fixed), g ϭ 2.00 (fixed) and PI ϭ creases monotonically with decreasing temperature from
0.009 for 2. The Jwb values obtained compare closely with 7.05 µB at 290 K to 2.60 µB at 2 K (Figure 9). The two ex-
those previously determined by us for the Fe4 acetate sys- change coupling constants, Jwb and Jbb, were allowed to
tem[4], and also those reported in the literature.[3,6]
vary independently in the full-matrix diagonalisation pro-
The effective magnetic moment of complexes 3 and 4 as a gram to optimise the agreement. A good fit was found with
function of temperature is shown in Figure 8. The magnetic the following parameters: Jwb ϭ JFeMn ϭ Ϫ22.3 cmϪ1
,
moment gradually decreases to 6.93 µB at 10 K for 3 and Jbb ϭ JMnMn ϭ Ϫ31.6 cmϪ1, g ϭ 2.0 (fixed) and D ϭ 0
6.87 µB at 10 K for 4, whereupon the moment falls sharply (fixed). The ground state of St ϭ 1 for 6 lies about 14 cmϪ1
to 4.56 µB for 3 and 4.50 µB for 4 by 2 K. Thus, the temper- below the first excited state with St ϭ 2.
ature-dependent behaviours of the magnetic moments of 3
A steady decrease of the magnetic moment with decreas-
and 4 suggest an St ϭ 3 ground state. In both cases, only ing temperature is observed for 7. The value of µeff at 2 K
data for temperatures above 20 K were considered in the fit, (3.63 µB) clearly indicates a different ground state than that
to avoid complications at low temperature from zero-field for 5 or 6 with St ϭ 1. The solid line in Figure 9 shows the
splitting. The solid lines in Figure 8 represent the best fits good quality of the fit with the following parameters: Jwb ϭ
with the following parameters: Jwb ϭ Ϫ0.47 cmϪ1, Jbb
ϭ
JFeMn ϭ Ϫ5.5 cmϪ1; Jbb ϭ JMnMn ϭ Ϫ10.16 cmϪ1; g ϭ
Ϫ7.73 cmϪ1, g ϭ 2.099 (fixed) and D ϭ 0 (fixed) for 3, and 2.00 (fixed) and D ϭ 0 (fixed). The relative error surface
Jwb ϭ Ϫ1.63 cmϪ1, Jbb ϭ Ϫ6.71 cmϪ1, g ϭ 2.215 (fixed) for fitting the magnetic data of 7 as a function of both
and D ϭ 0 cmϪ1 (fixed) for 4. Field-dependent magnetis- JFeMn and JMnMn, representing a well-defined global min-
ation measurements (not shown) at 2 K up to 6 T also con- imum in the parameter space, has allowed us to estimate
firm that the ground state for 3 is St ϭ 3 [simulation with the error bars on the quoted JFeMn and JMnMn values as
S ϭ 3.0 (fixed), g ϭ 2.02, D ϭ 4.13 cmϪ1]. The ground state approximately Ϯ0.2 and Ϯ1.0 cmϪ1, respectively. The
is within 5 cmϪ1 of the first excited state with St ϭ 4.
We note that the g values employed for the simulation coupling constants is separated from the first excited state
are high for MnIII. Nevertheless, they do not affect our con- with St ϭ 3 by an energy gap of only 5 cmϪ1
ground state of St ϭ 2 for 7 obtained from the exchange
.
clusion that 3 and 4 have an St ϭ 3 ground state. The
A good fit of the magnetic data for 8 yields Jwb ϭ
ground state of St ϭ 3 has been found for all carboxylate- JFeMn ϭ Ϫ19.38 cmϪ1, Jbb ϭ JMnMn ϭ Ϫ28.91 cmϪ1, g ϭ
bridged clusters with an [Mn4O2]8ϩ core except one with a 2.09 and D ϭ 0 (Figure 9). The ground state of St ϭ 1 for
quintuply degenerate ground state.[35]
8 is not well separated (∆E ഠ 1 cmϪ1) from the first excited
At 295 K the magnetic moment (µeff/molecule) for 5 is state with St ϭ 2.
7.16 µB, which is significantly lower than the value of µeff ϭ
The magnetic behaviour of 9 can be easily interpreted in
10.86 µB expected for an uncoupled system containing two terms of the competing influence of Jwb and Jbb upon spin
FeIII (S ϭ 5/2) and two MnIII (S ϭ 2) ions. On lowering the coupling, yielding a ground state of St ϭ 2. An excellent fit
temperature, µeff decreases monotonically until it reaches a representing a well-defined global minimum for 9, together
550
Eur. J. Inorg. Chem. 2003, 541Ϫ555

Ground Spin State Variation in Carboxylate-Bridged Tetranuclear [Fe2Mn2O2]8ϩ Cores
FULL PAPER
with the experimental magnetic moments, is shown in Fig- for 10 (propionate). Figure 10 gives a selective spin-correla-
ure 9. The parameters used for the least-squares fitting (fit tion diagram for the FeI2IIMnI2II(O)2 butterfly arrangement
shown as a solid line in Figure 9) are Jwb ϭ JFeMn ϭ Ϫ10.6 in which the energy (in units of Jwb) of a spin state denoted
cmϪ1, Jbb ϭ JMnMn ϭ Ϫ17.8 cmϪ1, g ϭ 2.01 (fixed) and by (St, SMn, SFe) is plotted as a function of Jbb/Jwb. The
D ϭ 2.07 cmϪ1 (fixed). The relative error surface calcula- correlation diagram clearly demonstrates the ground state
tion has allowed us to estimate the error bars on the JFeMn variation as a function of the ratio of Jbb/Jwb, which is also
and JMnMn values as approximately Ϯ0.6 cmϪ1 and Ϯ1.5 consistent with the experimental findings for the different
cmϪ1, respectively. The relatively large uncertainties in the carboxylate bridging ligands within the heterometallic
absolute magnitudes of JFeMn and JMnMn do not affect our Fe2Mn2 core. As an example, the ground state St ϭ 3 has
conclusion that 9 has an St ϭ 2 ground state lying 15.8 component SFe and SMn values of 5 and 2, respectively.
cmϪ1 below the first excited state of St ϭ 1.
These values result from different relative orientations of
For 10, a monotonous decrease of the magnetic moment
SFe(3), SFe(4), SMn(1), SMn(2), which is represented below as
with decreasing temperature, reaching a value of µeff ϭ 2.72 an overall spin alignment due to spin frustration.
µB at 2 K, is observed. This clearly indicates a ground state
with St Ͼ 1, which is attained by competing antiferromag-
netic coupling between the wing-body (Jwb ϭ Ϫ14.4 cmϪ1
)
and the body-body (Jbb ϭ Ϫ19.4 cmϪ1) metal centres. The
g value used for the simulation of the experimental data
(data shown as squares in Figure 9) is g ϭ 2.13. The solid
line in Figure 9 is the simulation of the magnetic data with
the abovementioned parameters. The evaluated lowest lying
state of St ϭ 1 lies 18 cmϪ1 below the second lowest state
with St ϭ 2.
Discussion
In the preceding section it has been experimentally estab-
lished that 1 and 2, with the Fe4O2 core, possess an St ϭ
0 ground state. All thermally populated excited states are
effectively only determined by Jwb, and the intrinsic charac-
ter (i.e. antiparallel coupling) of the Fe(2)···Fe(1) interac-
tion, Jbb ϭ J12, can be totally negated due to spin frustra-
tion,[20] making the Jbb interaction indeterminate in the
[Fe4(µ3-O)2]8ϩ core. On the other hand, 3 and 4, with the
Mn4O2 core, possess the ground state St ϭ 3, irrespective
of the nature of the bridging carboxylate, as has been found
earlier.[9] In contrast, variation of the six bridging carb-
oxylates from acetate (5), to benzilate (diphenylglycolate)
(6), triphenyl acetate (7), benzoate (8), chloroacetate (9),
and propionate (10), causes changes in the ground states
Figure 10. A selected energy-level diagram for the [Fe2Mn2O2]8ϩ
core emphasising the ground state variation as a function of the
ratio Jbb/Jwb; spin states are denoted by St, SMn and SFe, where St ϭ
SMn ϩ SFe
Concluding Remarks
The results described in the present paper show that not
for 5Ϫ10, varying from St ϭ 1 to 3, although the replace- only homometal tetranuclear metal clusters of FeIII and
ment of acetate in 5 by benzilate in 6 or benzoate in 8 has MnIII with a butterfly disposition, [M4(µ3-O)2]8ϩ, can be
little structural effect within the [Fe2Mn2O2]8ϩ core, as is isolated, but also their mixed-metal analogues
evident from the X-ray structures. These ground states have [FeI2IIMnI2II(µ3-O)2]8ϩ are easily accessible. The tetranuclear
been confirmed by the field-dependent (0.5Ϫ6.5 T) mag- cores are amenable to carboxylate bridging ligand variation
netisation measurements. This indicates that the ‘‘R’’ groups but the structural parameters show insignificant variation.
of the different carboxylic acids (RCOOH) might not be that Although the spin ground states resulting from the ex-
innocent in mediating exchange interactions, as is commonly change interactions within the [Fe4O2]8ϩ and [Mn4O2]8ϩ
taken for granted. The ground state does not represent the cores are little affected by the peripheral carboxylate li-
smallest spin state possible for the electronic dn configura- gands, variation of the six carboxylates (acetate, benzilate,
tion when both interactions are antiferromagnetic. It is triphenyl acetate, benzoate, chloroacetate propionate)
noteworthy in this respect that the ratio of competing coup- causes significant changes in the electronic ground state
ling pathways, and not so much their absolute magnitudes, structure (St ϭ 1Ϫ3) of the [Fe2Mn2O2]8ϩ core. As the crys-
determines the electronic structures of these butterfly com- tallographic data exclude steric effects as the cause of the
plexes. The ratio JMnMn/JFeMn varies from 1.39 for 5 (acet- ground state variation, the electronic (donor) properties
ate), to 1.42 for 6 (benzilate), 1.85 for 7 (triphenyl acetate), seem to be responsible for the dependence of spin ground
1.49 for 8 (benzoate), 1.68 for 9 (chloroacetate), and 1.35 states on the nature of the bridging carboxylates. Presum-
Eur. J. Inorg. Chem. 2003, 541Ϫ555
551

P. Chaudhuri et al.
FULL PAPER
Table 9. Crystal data and structure refinement for 1, 3, 5, 6 and 8
1
3
5
6
8
Empirical formula
Formula mass
Temperature [K]
Crystal system
Space group
C74H85ClFe4N8O19 C74H85ClMn4N8O19 C38H61ClFe2Mn2N8O16 C74H85ClFe2Mn2N8O19 C53H67ClFe2Mn2N8O16·2CH3CN
1649.35
1645.71
1142.98
1647.53
1411.28
293(2)
293(2)
100(2)
100(2)
100(2)
orthorhombic
Pna21
orthorhombic
Pna21
tetragonal
P4(1)2(1)2
a ϭ 11.6230(9)
b ϭ 11.6230(9)
c ϭ 35.827(2)
α ϭ β ϭ γ ϭ 90°
4840.0(6); 4
1.569
orthorhombic
Pna21
monoclinic
P21/n
˚
Unit cell dimensions [A]
a ϭ 16.329(4)
b ϭ 18.644(4)
c ϭ 25.009(6)
α ϭ β ϭ γ ϭ 90°
7613.7(3); 4
1.439
a ϭ 16.317(3)
b ϭ 18.655(4)
c ϭ 25.001(5)
α ϭ β ϭ γ ϭ 90°
7610.1(3); 4
1.436
a ϭ 16.144(2)
b ϭ 18.396(2)
c ϭ 24.759(3)
α ϭ β ϭ γ ϭ 90°
7353(2); 4
1.488
a ϭ 16.0785(6)
b ϭ 14.6897(5)
c ϭ 27.5133(10)
β ϭ 106.23°
6239.3(4); 4
1.502
3
˚
Volume [A ]; Z
Density (calcd.) [Mg/m3]
Absorpt. coeff. [mmϪ1
F(000)
]
0.857
0.759
1.227
0.836
0.969
3432
3416
2368
3424
2928
Crystal size [mm]
Reflections collected
Independent reflections
0.40 ϫ 0.37 ϫ 0.12 0.50 ϫ 0.45 ϫ 0.07 0.31 ϫ 0.28 ϫ 0.12
0.42 ϫ 0.24 ϫ 0.09
59215
20348
0.30 ϫ 0.25 ϫ 0.25
53728
14147
8991
8991
9056
8648
54594
9541
Data/restraints/parameters 8972/214/931
8634/321/954
Ϫ0.01(3)
9536/0/316
0.012(12)
20331/1/964
Ϫ0.016(8)
14124/0/801
Ϫ
Absolute structure
parameter
0.00(3)
Final R1 [I Ͼ 2σ(I)]
R2 (all data)
0.0770
0.178
0.0827
0.168
0.0396
0.0965
0.0439
0.0961
0.0436
0.1021
meter 2000. The magnetochemical measurements of the powdered
samples were performed with a Quantum Design SQUID-Magne-
tometer, MPMS generally in a field of 1 T. The samples were put
in gelatine capsules and the response function was measured four
times for each of 32 measured temperature points. Diamagnetic
contributions were estimated for each compound by making use
of Pascal’s constants. The Mössbauer spectrometer worked in the
conventional constant-acceleration mode with a 57Co/Rh source.
Isomer shifts are given relative to α-Fe at room temperature.
ably, spin coupling through carboxylate is not negligible, in
contrast to the prevailing notion. Thus, future low molecu-
lar weight structural modelling of metalloproteins is ex-
pected to be guided by proper choice of the carboxylate
ligands, in order to mimic the amino acids in the biomolec-
ules.
The [M4O2]8ϩ tetranuclear cores are structurally incon-
gruent with currently known biological metal sites and
hence might be irrelevant from a biological view point.
However, being readily prepared and amenable to carb-
oxylate variation, 5Ϫ10 have proven useful starting points
to increase our knowledge on fine-tuning the spin proper-
ties of molecular aggregates. This fine-tuning is a result of
spin frustration. Both the lowest and the intermediate spin
X-ray Crystallography: The crystallographic data for 1, 3, 5, 6 and
8 are summarised in Table 9. Graphite-monochromated Mo-Kα ra-
˚
diation (λ ϭ 0.71073 A) was used throughout. X-ray diffraction
data were collected with a Siemens R3m/V diffractometer for 1
and 3. Dark brown single crystals of 5, 6 and 8 were fixed with
perfluoropolyether on glass fibres and mounted on a Siemens
ground states have been observed in the isostructural com- SMART diffractometer equipped with a cryogenic nitrogen cold
stream, and intensity data were collected at 100(2) K. Final cell
constants were obtained from a least-squares fit of the setting
angles of several thousand strong reflections. Intensity data were
corrected for Lorentz and polarisation effects. An empirical ab-
sorption correction by Ψ scans was applied for 1 and 3; intensity
data of 5 and 8 were corrected for absorption using the program
SADABS (G. M. Sheldrick, University of Göttingen, 1994) giving
maximum and minimum transmission factors of 0.928 and 0.736
for 5 and 0.853 and 0.768 for 8. Intensities of 6 were left uncorrec-
ted. The Siemens ShelXTL, Vers. 5.0 (Siemens Industrial Automa-
tion, Inc., 1994) software package was used for solution, refinement
and artwork of the structures. Neutral atom scattering factors of
the program were used. All structures were solved and refined by
direct methods and difference Fourier techniques. Non-hydrogen
atoms were refined anisotropically and hydrogen atoms were placed
at calculated positions and refined as riding atoms with isotropic
displacement parameters. A crystallographic twofold axis passes
through the central acetate bridge of the complex cation of 5 and
plexes 5Ϫ10 which vary only in the nature of their carb-
oxylate bridging ligands. As is seen from Figure 10 ‘‘spin-
frustration degeneracy’’ of the ground state leading to un-
usual electronic properties is expected for the [Fe2Mn2O2]8ϩ
core if Jbb/Jwb ഠ 1.52 and 2.0.
Experimental Section
General: CAUTION: Although we experienced no difficulties, the
unpredictable behaviour of perchlorate salts necessitates extreme
caution in their handling. The complex LFeCl3 was prepared as
described previously.[36] All other reagents were used as received.
Elemental microanalyses were performed by the microanalytical
laboratory Kolbe, Mülheim. Iron and manganese were determined
spectrophotometrically by using pyridine-2,6-dicarboxylic acid as
described in the literature.[37] Perchlorate ion was determined gravi-
metrically as tetraphenylarsonium perchlorate. Infrared spectra the perchlorate anion is disordered since this axis is not coincident
Ϫ
were measured as KBr disks with a PerkinϪElmer FT-IR spectro-
with the twofold axis of the ClO4 moiety. The disorder was satis-
552
Eur. J. Inorg. Chem. 2003, 541Ϫ555

Ground Spin State Variation in Carboxylate-Bridged Tetranuclear [Fe2Mn2O2]8ϩ Cores
factorily modelled by refining the complete anion slightly displaced
FULL PAPER
[L2MnIII4(salox)2(µ3-O)2(Ph3CCOO)3]ClO4 (4): Complex 4 was
prepared similarly to 3 with triphenylacetic acid as the carboxylate
source. Deep brown crystals were obtained in moderately good
yield (45%). C92H97ClMn4N8O16 (1825.99): calcd. C 60.51, H 5.35,
N 6.14, Mn 12.03, ClO4 5.45; found C 58.9, H 5.5, N 6.2, Mn 12.2,
ClO4 5.1. MS(ESI positive in CH3CN): m/z (%) ϭ 1727 (80) [M]ϩ,
1484 (100) [M Ϫ L(Mn)(O)]ϩ, 1439 (60) [M Ϫ (C6H5)3CCOO]ϩ.
IR (KBr): ν˜ ϭ 3441, 1587, 1541, 1152 cmϪ1. Complex 4 was also
prepared using CH3CN as solvent.
from the axis with an occupation factor of 0.5. A phenyl group of
a benzilate ligand in 6 was found to be disordered and therefore a
split atom model was used. Two positions of the phenyl ring were
refined with occupancies of 0.5 each. CCDC-187680 (1), -187681
(3), -187705 (5), -187706 (6) and -187707 (8) contain the supple-
mentary 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].
[L2FeIII2(µ3-O)2(salox)2(µ-OOC·CH3)3MnIII2]ClO4 (5): ‘‘Basic
manganese() acetate’’[39] (0.27 g, 1 mmol) was added to a vigor-
ously stirred solution of salicylaldoxime (0.135 g, 1 mmol) in meth-
anol (60 mL). The resulting suspension was stirred at room temper-
ature for 0.5 h to yield a dark brown solution which was then
charged with a solid sample of LFeCl3 (0.33 g, 1 mmol), sodium
acetate (0.24 g, 3.0 mmol) and triethylamine (2 mL). The suspen-
sion was refluxed for 1 h and then filtered to remove some green
solid (presumably [Mn(Hsalox)3]). Sodium perchlorate monohydr-
ate (0.4 g) was added to the clear solution and the red-brown solu-
tion was kept at ambient temperature in a closed vessel. After 12 h,
the dark brown crystals were filtered off and air-dried. Yield:
400 mg (70%). C38H61ClFe2Mn2N8O16 (1142.98): calcd. C 39.93, H
5.38, N 9.80, Fe 9.77, Mn 9.61, ClO4 8.70; found C 39.7, H 5.4, N
9.8, Fe 9.9, Mn 9.5, ClO4 8.9. IR (KBr): ν˜ ϭ 1594, 1544, 1155,
Preparation of Complexes: Complexes 1 and 2 were prepared sim-
ilarly by using a modified procedure for the corresponding acetate-
bridged tetranuclear FeIII compound.[4]
[L2FeIII4(salox)2(µ3-O)2(Ph2C(OH)COO)3]ClO4 (1): All operations
were carried out under argon. Fe(ClO4)2·6H2O (0.30 g, 1 mmol),
sodium diphenylglycolate (benzilate) (0.50 g, 2 mmol) and salicylal-
doxime (0.27 g, 2 mmol) were added to vigorously stirred, degassed
methanol (60 mL). The resulting suspension was stirred at room
temperature for 1 h to yield a dark violet solution. This solution
was charged with a solid sample of [LFeCl3][36] (0.33 g, 1 mmol),
sodium diphenylglycolate (benzilate) (0.75 g, 3 mmol) and triethyl-
amine (1 mL). The suspension was refluxed for 1 h and then filtered
to remove some dark solid, identified as [LFeIII(salox)3FeIII] by IR
as described earlier.[24b] Sodium perchlorate monohydrate (0.4 g)
was added to the filtrate and the red-brown solution was kept in
air at ambient temperature. After 1 d, the red-brown crystals were
filtered off and air-dried. Yield: 390 mg (47%). C74H85ClFe4N8O19
(1648.35): calcd. C 53.89, H 5.19, N 6.79, Fe 13.54, ClO4 6.03;
found C 53.9, H 5.1, N 6.8, Fe 13.7, ClO4 5.8. MS (ESI positive in
CH3CN/H2O): m/z (%) ϭ 1549 (100) [M]ϩ. IR (KBr): ν˜ ϭ 3472,
755, 743, 663 cmϪ1
.
[L2FeIII2(µ3-O)2(salox)2(µ-OOC·C(OH)Ph2]3MnIII2]ClO4 (6): A so-
lution containing MnCl2·4H2O (0.20 g, 1 mmol), salicylaldoxime
(0.27 g, 2 mmol) diphenylglycolic acid (0.63 g, 3 mmol) and triethy-
lamine (1 mL) in methanol (60 mL) was heated to reflux for 0.5 h.
LFeCl3 (0.33 g, 1 mmol) and sodium diphenylglycolate (0.25 g,
1 mmol) were added to the resulting suspension. The mixture was
refluxed for 1 h and then filtered to remove some green solid. The
red-brown solution was kept at room temperature following addi-
tion of NaClO4·H2O (0.8 g). After 24 h, the deep brown crystals
were filtered off and air-dried. Yield: 510 mg (ca. 62%).
C74H85ClFe2Mn2N8O19 (1647.53): calcd. C 53.95, H 5.20, N 6.80,
Fe 6.78, Mn 6.67, ClO4 6.04; found C 53.80, H 5.1, N 6.8, Fe 6.5,
Mn 6.4, ClO4 5.5. MS (ESI positive in CH3CN): m/z (%) ϭ 1547
(100), 1548 (90) [M]ϩ. IR (KBr): ν˜ ϭ 3742, 1617, 1592, 1540, 1158,
1600, 1569, 1157, 700, 669, 648, 638 cmϪ1
.
[L2FeIII4(salox)2(µ3-O)2(Ph3CCOO)3]ClO4 (2): Complex 2 was ob-
tained similarly to complex 1 with sodium triphenylacetate used
instead of sodium diphenylglycolate as the carboxylate source.
Dark brown crystals were obtained. Yield: 410 mg (42%).
C92H97ClFe4N8O16 (1829.63): calcd. C 60.39, H 5.34, N 6.12, Fe
12.21, ClO4 5.43; found C 58.4, H 5.4, N 6.0, Fe 12.3, ClO4 5.5.
MS (ESI positive in CH3CN/H2O): m/z (%) ϭ 1730 (100) [M]ϩ. IR
704, 670, 638, 603 cmϪ1
[L2FeIII2(µ3-O)2(salox)2(µ-OOC·CPh3)3MnIII2]ClO4 (7), [L2FeIII
.
(KBr): ν˜ ϭ 1594, 1155, 1567, 699, 671, 652, 623 cmϪ1
.
-
2
(µ3-O)2(salox)2(µ-OOC·Ph)3MnIII2]ClO4 (8), [L2FeIII2(µ3-O)2-
(salox)2(µ-OOCCH2Cl)3MnIII2]ClO4 (9), [L2FeIII2(µ3-O)2(salox)2-
(µ-OOC·CH2CH3)3MnIII2]ClO4 (10): Complexes 7Ϫ10 were pre-
pared according to the same protocol for 6 using triphenylacetic
acid (7), benzoic acid (8), chloroacetic acid (9) or propionic acid
(10) as the bridging carboxylate sources. Red-brown crystals were
obtained with yields in the range 40Ϫ50%.
[L2MnIII4(salox)2(µ3-O)2(Ph2C(OH)COO)3]ClO4 (3): Step 1: A sus-
pension of diphenylglycolic acid (0.46 g, 2 mmol) and sodium hy-
droxide (0.08 g, 2 mmol) in water (30 mL) was stirred vigorously
for 15 min. MnCl2·4H2O (0.20 g, 1 mmol) was added and stirring
was continued for a further 15 min. The precipitated solid was fil-
tered off, washed thoroughly with water, and dried for 24 h under
vacuum. Step 2: A suspension of the dried solid obtained from
step 1 and salicylaldoxime (0.27 g, 2 mmol) in methanol (40 mL)
containing triethylamine (1 mL) was refluxed in air for 0.5 h to
7: C92H97ClFe2Mn2N8O16 (1827.81): calcd. C 60.45, H 5.35, N
6.13, Fe 6.11, ClO4 5.44; found C 58.9, H 5.4, N 6.3, Fe 6.0, Mn
6.01, ClO4 5.8. MS (ESI positive in CH3CN/H2O): m/z (%) ϭ 1727
obtain a dark green suspension. The green suspension was charged
[38]
(100) [M]ϩ. IR (KBr): ν˜ ϭ 1591, 1540, 701, 670, 645, 612 cmϪ1
.
with LMnCl3
(0.33 g, 1 mmol) under argon and treated with
triethylamine (0.5 mL). The resulting suspension was refluxed un-
der argon for 0.5 h and then filtered to remove some green solid
(presumably [Mn(Hsalox)3]). Sodium perchlorate monohydrate
(0.8 g) was added, and after 1Ϫ2 d the dark brown solution yielded
deep brown crystals of 3, which were collected by filtration and air-
dried. Yield: 275 g (ca. 33%). C74H85ClMn4N8O19 (1645.71): calcd.
C 54.00, H 5.21, N 6.81, Mn 13.35, ClO4 6.04; found C 53.8, H
5.1, N 6.8, Mn 13.1, ClO4 5.9. IR (KBr): ν˜ ϭ 3472, 1617, 1592,
8: C53H67ClFe2Mn2N8O16 (1329.18): calcd. C 47.89, H 5.08, N
8.43, Fe 8.40, Mn 8.27, ClO4 7.48; found C 48.1, H 5.0, N 8.5, Fe,
8.4, Mn, 8.2, ClO4 7.9. IR (KBr): ν˜ ϭ 1594, 1541, 1157, 738, 723,
663, 636 cmϪ1
.
9: C38H58Cl4Fe2Mn2N8O16 (1246.30): calcd. C 36.62, H 4.69, N
8.99, Fe 8.96, Mn 8.82, ClO4 8.00; found C 35.0, H 5.0, N 9.0, Fe
9.2, Mn 8.8, ClO4 8.1. IR (KBr): ν˜ ϭ 1594, 1543, 1156, 757, 745,
1574, 1158 cmϪ1
.
663 cmϪ1
.
Eur. J. Inorg. Chem. 2003, 541Ϫ555
553

P. Chaudhuri et al.
FULL PAPER
Fallah, X. Solans, M. Forit-Bardia, Eur. J. Inorg. Chem. 2000,
10: C41H67ClFe2Mn2N8O16 (1185.06): calcd. C 41.55, H 5.70, N
9.46, Fe, 9.43, Mn 9.27, ClO4 8.39; found C 42.0, H 5.6, N 9.6, Fe
9.1, Mn 9.5, ClO4 8.4. IR (KBr): ν˜ ϭ 1592, 1538, 1158, 704, 672,
675. [17c] A. Aukauloo, X. Ottenwaelder, R. Ruiz, Y. Journaux,
Y. Pei, E. Riviere, M. C. Munoz, Eur. J. Inorg. Chem. 2000,
[17d]
951.
B. Cervera, R. Ruiz, F. Lloret, M. Julve, J. Faus, C.
640, 602 cmϪ1
.
[17e]
M. Munoz, Y. Journaux, Inorg. Chim. Acta 1999, 288, 57.
E. Colacio, J. M. Dominguez-Vera, M. Ghazi, R. Kivekäs, M.
Klinga, J. M. Moreno, Inorg. Chem. 1998, 37, 3040. [17f] S. Mo-
hanta, K. K. Nand, L. K. Thompson, U. Flörke, K. Nag, In-
org. Chem. 1998, 37, 1465.
Acknowledgments
O. Kahn, Adv. Inorg. Chem. 1995, 43, 179. [18b] K. S. Mur-
[18] [18a]
This work was initiated in Bochum. Our thanks are due to Mrs.
M. Hess, H. Schucht (Mülheim) and M. Winter (Bochum) for their
technical assistance. Support from the Max Planck Society and
Fonds der Chemischen Industrie is gratefully acknowledged.
ray, Adv. Inorg. Chem. 1995, 43, 261.
[19] [19a]
P. D. W. Boyd, Q. Li, J. B. Vincent, K. Folting, H. R.
Chang, W. E. Streib, J. C. Huffman, G. Christou, D. N. Hend-
rickson, J. Am. Chem. Soc. 1988, 110, 8537. [19b] D. Sessoli, H.
L. Tsai, A. R. Schake, S. Wang, J. B. Vincent, K. Folting, D.
Gatteschi, G. Christou, D. N. Hendrickson, J. Am. Chem. Soc.
[1]
[19c]
´
V. I. Ponomarev, L. O. Atovmyan, S. A. Bobkova, K. I. Turte,
Dokl. Akad. Nauk. SSSR. 1984, 274, 368.
1993, 115, 1804.
D. Gatteschi, A. Caneschi, L. Pardi, R.
[19d]
Sessoli, Science 1994, 265, 1054.
S. M. J. Aubin, M. W.
[2]
[3]
W. H. Armstrong, M. E. Roth, S. J. Lippard, J. Am. Chem.
Soc. 1987, 109, 6318.
J. M. McCusker, J. B. Vincent, E. A. Schmitt, M. L. Mino, K.
Shin, D. K. Coggin, P. M. Hagen, J. C. Huffman, G. Christou,
D. N. Hendrickson, J. Am. Chem. Soc. 1991, 113, 3012.
P. Chaudhuri, M. Winter, P. Fleischhauer, W. Haase, U. Flörke,
H.-J. Haupt, Inorg. Chim. Acta 1993, 212, 241.
Wemple, D. M. Adams, H. L. Tsai, G. Christou, D. N. Hend-
rickson, J. Am. Chem. Soc. 1996, 118, 7746.
A. Caneschi, A. Cornia, D. Fabrizi de Biani, D. Gatteschi, C.
[19e]
A. L. Barra,
Sanggregorio, R. Sessoli, L. Sorace, J. Am. Chem. Soc. 1999,
[19f]
121, 5302.
G. Aromi, S. M. J. Aubin, M. A. Bolcar, G.
[4]
[5]
[6]
Christou, H. J. Eppley, K. Folting, D. N. Hendrickson, J. C.
Huffman, R. C. Squire, H.-L. Tsai, S. Wang, M. W. Wemple,
[19g]
L. Wu, M. Pressprich, P. Coppens, M. J. DeMarco, Acta Crys-
tallogr., Sect. C 1993, 49, 1255.
Polyhedron 1998, 17, 3005 and references therein.
J. Lari-
nova, M. Gross, M. Pilkington, H. Andres, H. Stoeckli-Evans,
H. U. Güdel, S. Decurtins, Angew. Chem. Int. Ed. 2000, 1605.
[19h] D. Gatteschi, R. Sessoli, A. Cornia, Chem. Commun. 2000,
725 and references therein.
D. N. Hendrickson, Research Frontiers in Magnetochemistry
(Ed.: C. J. O’Connor), World Scientific Singapore, 1993, p. 87.
O. Kahn, Chem. Phys. Lett. 1997, 265, 109.
M. W. Wemple, D. K. Coggin, J. B. Vincent, J. K. McCusker,
W. E. Streib, J. C. Huffman, D. N. Hendrickson, G. Christou,
J. Chem. Soc., Dalton Trans. 1998, 719.
[7] [7a]
[20]
K. Wieghardt, K. Pohl, I. Jibril, G. Huttner, Angew. Chem.
Int. Ed. Engl. 1984, 23, 77. [7b] C. Delfs, D. Gatteschi, L. Pardi,
[21]
[22]
R. Sessoli, K. Wieghardt, D. Hanke, Inorg. Chem. 1993, 32,
[7c]
3099.
A.-L. Barra, D. Gatteschi, R. Sessoli, Chem. Eur. J.
F. Ephraim, Ber. Dtsch. Chem. Ges. 1930, 63, 1928.
2000, 6, 1608.
[23] [23a]
E. Toyota, K. Umakoshi, Y. Yamamoto, Bull. Chem. Soc.
[8] [8a]
V. L. Pecoraro, Ed. Manganese Redox Enzymes, VCH Ver-
H. Sigel, A. Sigel, Ed.
Metal Ions in Biological Systems, Marcel Dekker, New York,
Jpn. 1995, 68, 858. [23b] J. M. Thorpe, R. L. Beddoes, D. Colli-
[8b]
lagsgesellschaft, Weinheim, 1992.
son, C. D. Garner, M. Helliwell, J. M. Holmes, P. A. Tasker,
Angew. Chem. Int. Ed. 1999, 38, 1119. [23c] S. J. Rettig, A. STorr,
[8c]
2000, vol. 37.
R. H. Holm, E. I. Solomon (Guest Eds.),
[23d]
J. Trotter, Acta Crystallogr., Sect. C 1992, 48, 1587.
K. R.
Chem. Rev. 1996, no. 7.
Manolov, Russ. J. Inorg. Chem. 1967, 12, 1431.
[9]
J. B. Vincent, C. Christmas, H.-R. Chang, Q. Li, P. D. Boyd, J.
C. Huffman, D. N. Hendrickson, G. Christou, J. Am. Chem.
Soc. 1989, 111, 2086.
E. Libby, J. K. McCusker, E. A. Schmitt, K. Folting, D. N.
Hendrickson, G. Christou, Inorg. Chem. 1991, 30, 3486.
E. Bouwman, M. A. Bolcar, E. Libby, J. C. Huffman, K. Folt-
ing, G. Christou, Inorg. Chem. 1992, 31, 5185.
M. W. Wemple, H.-L. Tsai, S. Wang, J. B. Claude, W. E. Streib,
J. C. Huffman, D. N. Hendrickson, G. Christou, Inorg. Chem.
1996, 35, 6437.
P. Chaudhuri, M. Winter, P. Fleischhauer, W. Haase, U. Flörke,
H.-J. Haupt, J. Chem. Soc., Chem. Commun. 1993, 566.
P. Chaudhuri, F. Birkelbach, M. Winter, V. Staemmler, P. Fle-
ischhauer, W. Haase, U. Flörke, H.-J. Haupt, J. Chem. Soc.,
Dalton Trans. 1994, 2313.
[24] [24a]
P. Chaudhuri, Proc. Indian Acad. Sci. (Chem. Sci) 1999,
[24b]
111, 397.
C. N. Verani, E. Bothe, D. Burdinski, T.
Weyhermüller, U. Flörke, P. Chaudhuri, Eur. J. Inorg. Chem.
2001, 2161.
[10]
[11]
[12]
[25]
[26]
[27]
[28]
E. Bill, C. Krebs, M. Winter, M. Gerdan, A. X. Trautwein, U.
Flörke, H.-J. Haupt, P. Chaudhuri, Chem. Eur. J. 1997, 3, 193.
P. Chaudhuri, M. Hess, T. Weyhermüller, E. Bill, H.-J. Haupt,
U. Flörke, Inorg. Chem. Commun. 1998, 1, 39.
P. Chaudhuri, M. Hess, E. Rentschler, T. Weyhermüller, U.
Flörke, New J. Chem. 1998, 533.
[13]
[14]
´
P. Chaudhuri, M. Winter, B. P. C. Della Vedova, P. Fleisch-
hauer, W. Haase, U. Flörke, H.-J. Haupt, Inorg. Chem. 1991,
30, 4777.
[29]
R. D. Cannon, R. P. White, Progr, Inorg. Chem. 1988, 36, 195.
[30] [30a]
[15]
P. G ütlich, in Mössbauer Spectroscopy (Ed.: U. Gonser),
O. Kahn, Molecular Magnetism, VCH Publishers, Weinheim,
1993.
[30b]
Springer-Verlag, Berlin, Heidelberg, New York, 1975.
N.
[16] [16a]
N. Greenwood, T. C. Gibb, Mössbauer Spectroscopy, Chapman
and Hall, London, 1971.
F. Birkelbach, U. Flörke, H.-J. Haupt, C. Butzlaff, A. X.
Trautwein, K. Wieghardt, P. Chaudhuri, Inorg. Chem. 1998, 37,
[31]
2000. [16b] F. Birkelbach, M. Winter, U. Flörke, H.-J. Haupt, C.
G. J. Long, W. T. Robinson, W. P. Tappmeyer, D. L. Bridges,
J. Chem. Soc., Dalton Trans. 1973, 573.
Butzlaff, M. Lengen, E. Bill, A. X. Trautwein, K. Wieghardt,
[32] [32a] J. H. Rodriguez, H. Ok, Y.-M. Xia, P. G. Debrunner, B. E.
[16c]
P. Chaudhuri, Inorg. Chem. 1994, 33, 3990.
C. Krebs, M.
Hinrichs, T. Meyer, N. H. Packard, J. Phys. Chem. 1996, 100,
Winter, T. Weyhermüller, E. Bill, K. Wieghardt, P. Chaudhuri,
[32b]
[16d]
6849.
J. H. Rodriguez, Y.-M. Xia, P. G. Debrunner, P.
J. Chem. Soc., Chem. Commun. 1995, 1913.
P. Chaudhuri,
Chaudhuri, K. Wieghardt, J. Am. Chem. Soc. 1996, 118, 7542.
A. Bino, R. Chayat, E. Pedersen, A. Schneider, Inorg. Chem.
1991, 30, 856.
S. L. Castro, Z. Sun, C. M. Grant, J. C. Bollinger, D. N. Hend-
rickson, G. Christou, J. Am. Chem. Soc. 1988, 120, 2365.
C. Canada-Vilalta, J. C. Huffman, G. Christou, Polyhedron
2001, 20, 1785.
M. Winter, P. Fleischhauer, W. Haase, U. Flörke, H.-J. Haupt,
[33]
[34]
[35]
[16e]
J. Chem. Soc., Chem. Commun. 1990, 1728.
C. N. Verani,
E. Rentschler, T. Weyhermüller, E. Bill, P. Chaudhuri, J. Chem.
Soc., Dalton Trans. 2000, 4263.
[17]
[17a]
A few selected examples:
K. E. Vostrikova, D. Luneau, W.
Wernsdorfer, P. Rey, M. Verdagner, J. Am. Chem. Soc. 2000,
[17b]
122, 718.
C. Diaz, J. Ribas, R. Costa, J. Tercero, M. S. E.
554
Eur. J. Inorg. Chem. 2003, 541Ϫ555

Ground Spin State Variation in Carboxylate-Bridged Tetranuclear [Fe2Mn2O2]8ϩ Cores
FULL PAPER
[36]
[39]
P. Chaudhuri, M. Winter, K. Wieghardt, S. Gehring, W. Haase,
B. Nuber, J. Weiss, Inorg. Chem. 1988, 27, 1564.
H. Hartkamp, Z. Anal. Chem. 1964, 199, 183.
K. Wieghardt, U. Bossek, B. Nuber, J. Weiss, J. Bonvoisin, M.
Corbella, S. E. Vitols, J. J. Girerd, J. Am. Chem. Soc. 1988,
110, 7398.
F. A. Cotton, G. Wilkinson, Advanced Inorganic Chemistry, 5th
ed., John Wiley & Sons, New York, 1988, p. 704.
Received July 3, 2002
[37]
[38]
[I02363]
Eur. J. Inorg. Chem. 2003, 541Ϫ555
555