Homoleptic 1-D iron selenolate complexes - Synthesis, structure, magnetic and thermal behaviour of ∞1[Fe(SeR)2] (R = Ph, Mes)

Article abstract of DOI:10.1039/c1dt10089k

The first examples of polymeric homoleptic iron chalcogenolato complexes 1∞[Fe(SePh)₂] and 1∞[Fe(SeMes)₂] (Ph = phenyl = C₆H₅, Mes = mesityl = C₆H₂-2,4,6- (CH₃)₃) have been both prepared by reaction of [Fe(N(SiMe₃)₂)₂] with two equivalents of HSeR (R = Ph, Mes) while 1∞[Fe(SePh)₂] was found to be also easily accessible through reactions of either FeCl₂, Fe(OOCCH ₃)₂ or FeCl₃ with PhSeSiMe₃ in THF. In the crystal, the two compounds form one-dimensional chains with bridging selenolate ligands comprising distinctly different Fe-Se-Fe bridging angles, namely 71.15-72.57° in 1∞[Fe(SePh)₂] and 91.80° in 1∞[Fe(SeMes)₂]. Magnetic measurements supported by DFT calculations reveal that this geometrical change has a pronounced influence on the antiferromagnetic exchange interactions of the unpaired electrons along the chains in the two different compounds with a calculated magnetic exchange coupling constant of J = -137 cm^-1 in 1∞[Fe(SePh)₂] and J = -20 cm^-1 in 1∞[Fe(SeMes)₂]. In addition we were able to show that the ring molecule [Fe(SePh)₂]₁₂ which is a structural isomer of 1∞[Fe(SePh)₂] behaves magnetically similar to the latter one. Investigations by powder XRD reveal that the ring molecule is only a metastable intermediate which converts in THF completely to form 1∞[Fe(SePh)₂]. Thermal gravimetric analysis of 1∞[Fe(SePh)₂] under vacuum conditions shows that the compound is thermally labile and already starts to decompose above 30 °C in a two step process under cleavage of SePh₂ to finally form at 250 °C tetragonal PbO-type FeSe. The reaction of 1∞[Fe(SePh)₂] with the Lewis base 1,10-phenanthroline yielded, depending on the conditions, the octahedral monomeric complexes [Fe(SePh)₂(1,10-phen)₂] and [Fe(1,10-phen)₃][Fe(SePh)₄].

Full text of DOI:10.1039/c1dt10089k

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
Dalton
Transactions
Cite this: Dalton Trans., 2011, 40, 7022
www.rsc.org/dalton
PAPER
Homoleptic 1-D iron selenolate complexes—synthesis, structure, magnetic
and thermal behaviour of 1•[Fe(SeR)2] (R = Ph, Mes)†
Andreas Eichho¨fer,*a Gernot Buth,b Francesco Dolci,c Karin Fink,a Richard. A. Moled and Paul T. Woode
Received 17th January 2011, Accepted 12th April 2011
DOI: 10.1039/c1dt10089k
The first examples of polymeric homoleptic iron chalcogenolato complexes 1•[Fe(SePh)2] and
1•[Fe(SeMes)2] (Ph = phenyl = C6H5, Mes = mesityl = C6H2-2,4,6-(CH3)3) have been both prepared by
reaction of [Fe(N(SiMe3)2)2] with two equivalents of HSeR (R = Ph, Mes) while 1•[Fe(SePh)2] was found
to be also easily accessible through reactions of either FeCl2, Fe(OOCCH3)2 or FeCl3 with PhSeSiMe3
in THF. In the crystal, the two compounds form one-dimensional chains with bridging selenolate
ligands comprising distinctly different Fe–Se–Fe bridging angles, namely 71.15–72.57◦ in •1 [Fe(SePh)2]
and 91.80◦ in •1 [Fe(SeMes)2]. Magnetic measurements supported by DFT calculations reveal that this
geometrical change has a pronounced influence on the antiferromagnetic exchange interactions of the
unpaired electrons along the chains in the two different compounds with a calculated magnetic
exchange coupling constant of J = -137 cm-1 in •1 [Fe(SePh)2] and J = -20 cm-1 in •1 [Fe(SeMes)2]. In
addition we were able to show that the ring molecule [Fe(SePh)2]12 which is a structural isomer of
1•[Fe(SePh)2] behaves magnetically similar to the latter one. Investigations by powder XRD reveal that
the ring molecule is only a metastable intermediate which converts in THF completely to form
1•[Fe(SePh)2]. Thermal gravimetric analysis of •1 [Fe(SePh)2] under vacuum conditions shows that the
compound is thermally labile and already starts to decompose above 30 ◦C in a two step process under
cleavage of SePh2 to finally form at 250 ◦C tetragonal PbO-type FeSe. The reaction of •1 [Fe(SePh)2] with
the Lewis base 1,10-phenanthroline yielded, depending on the conditions, the octahedral monomeric
complexes [Fe(SePh)2(1,10-phen)2] and [Fe(1,10-phen)3][Fe(SePh)4].
Introduction
There exists a long standing interest in the synthesis of polynuclear
iron sulphur species, due to their potential biological relevance.
aInstitut fu¨r Nanotechnologie, Karlsruhe Institut fu¨r Technologie (KIT),
Postfach 3640, Karlsruhe, 76021, Germany. E-mail: andreas.eichhoefer@
kit.edu; Fax: +721-608-26368; Tel: +721-608-26371
Many examples of different cluster types have been synthesized
and thoroughly characterized by a number of groups.1 Recent
reports on dinuclear biomimetic [2Fe–2S] complexes,2,3 as well as
cyclic oligomers of type [(tBu3SiS)MX]n (M = Fe, Co, Ni; X =
halide),4 and the synthesis of new binuclear and trinuclear iron
clusters5 prove the actual interest in this research area.
The build up of the structures of homoleptic and neutral
complexes of the general type [M(ER)n] (n = 1–3; M = 3d transition
metal; E = S, Se, Te; R = org. group) is mostly determined
by the interplay of the steric demand of the organic ligands
versus the tendency of the low coordinated metal ions to realize
higher coordination modes together with a minor influence of
the kind of chalcogen element. The recently published series
of quasi-two-coordinate transition metal dithiolates [M(SR)2]
(M = Cr, Mn, Fe, Co, Ni, Zn)6 with the very bulky organic
ligand R = C6H3-2,6(C6H2-2,4,6-iPr3)2 represents the lowest limit
with respect to the coordination number in homoleptic neutral
chalcogenolato complexes of these elements. Decrease of the steric
demand of the organic ligands results either in the formation of
bInstitut fu¨r Synchrotronstrahlung, Karlsruhe Institut fu¨r Technologie (KIT),
Postfach 3640, 76021, Karlsruhe, Germany
cEuropean Commission - JRC Institute for Energy, Cleaner Energy Unit,
Westerduinweg 3, NL-1755, ZG Petten, The Netherlands
dZWE FRM-II, Lichtenbergstrasse 1, 85747, Garching, Germany
eUniversity Chemical Laboratory, Lensfield Rd, Cambridge, UK, CB2 1EW
† Electronic supplementary information (ESI) available. Figure S1: Mea-
sured and simulated X-ray powder patterns of [Fe(SePh)2]12 as a suspension
of fresh crystals in CH2Cl2 and as a suspension of crystals after one
night in THF compared with the calculated pattern of 1, Figure S2:
measured and simulated X-ray powder patterns for 1 and 2, Figure
S3: Rietveld refinement results for 1 after thermal treatment, Table S1:
Refined parameters for 1 after heat treatment, Figure S4: Crystal structure
of 1 before and after heating, Figure S5: Susceptibility behaviour for
[Fe(SePh)2]12, Figure S6: Isothermal magnetization curves of 1 and 2 at 5
K, Figure S7 and S8: UV-Vis spectra of 1–4 in solution, Figure S9: Powder
XRD patterns of the residue of 1 after thermal treatment at different
heating temperatures. CCDC reference numbers 737597–737600. For ESI
and crystallographic data in CIF or other electronic format see DOI:
10.1039/c1dt10089k
7022 | Dalton Trans., 2011, 40, 7022–7032
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
mostly ring-like oligomeric molecules (e.g. M = Fe,7 Ni,8 Pd,9
Cu, Ag, Au10,11) or in the formation of insoluble polymers12
which could in most cases not be characterized by single crystal
X-ray diffraction. There exists only a limited number of crystal-
lographically characterized polymeric [M(ER)n] species namely
monomeric complexes [Fe(SePh)2(1,10-phen)2] (3) or dark red
crystals of the ionic compound [Fe(1,10-phen)3][Fe(SePh)4] (4)
(Scheme 4).
1•[Mn(SePh)2] (Ph = C6H5),13 1 [Mn(SeMes)2] (Mes = C6H2-2,4,6-
•
(CH3)3),14 2•[AgSePh],15 1 [Cd(TeMes)2],16 3 [Zn4(SPh)8CH3OH],17
•
•
3•[Cd4(SR)8] (R = Ph, 4-CH3-C6H4)18 and •3 [Cd(SePh)2].19
The synthesis and structural characterization of first examples
of polymeric homoleptic 1-dimensional iron-selenolate chain
1
•
1
•
compounds namely [Fe(SePh)2] and [Fe(SeMes)2] are reported
herein along with an investigation of their magnetic properties.
Results and discussion
Synthesis and structure
1
The iron selenolate complexes [Fe(SeR)2] (R = Ph (1), R = Mes
•
Scheme 4
(2)) can be prepared by reaction of the well known precursor
compound bis[(bistrimethylsilyl)amido]iron(II) with two equiva-
lents of RSeH in the prescence of PnPr3 (Scheme 1). In addition, 1
is also accessible by the reaction of anhydrous iron(II) chloride
as well as iron(III) chloride with three or four equivalents of
PhSeSiMe3 in THF (Scheme 2). Crystals suitable for X-ray analysis
were synthesized from reaction solutions of iron(II)acetate with
2.5 eq. of PhSeSiMe3 in the presence of 2 eq. of PnBu3 by layering
(Scheme 3).
1 crystallizes in the tetragonal space group I41/a (Table 1). In the
crystal structure the iron atoms are m2-bridged in one dimension by
the phenylselenolato ligands to form infinite chains (Fig. 1). Four
selenium atoms (Se(1), Se(2), Se(1¢) and Se(2¢)) of the phenylse-
lenolato ligands build a distorted tetrahedral coordination envi-
ronment around the iron atom with five nearly similar Se–Fe–Se
bond angles of about 105.03(4)◦ (Se(1¢¢)–Fe(1)–Se(1)), 106.39(4)◦
(Se(2¢)–Fe(1)–(Se(1)), 106.16(4)◦ (Se(1¢¢)–Fe(1)–Se(2)), 108.62(4)◦
(Se(2¢)–Fe(1)–Se(2)) and 108.20(4)◦ (Se(1)–Fe(1)–Se(2)) and only
one distinctly larger angle of 121.82(5)◦ for Se(1¢¢)–Fe(1)–Se(2¢).
The Fe2Se2 four-membered rings display a butterfly type shape
with Fe–Se–Fe bridging angles of 72.12(3)◦ for Fe(1)–Se(1)–Fe(1¢)
and 72.00(3)◦ for Fe(1)–Se(2)–Fe(1¢¢) and the rhomb folded by
22.61◦ from ideal planarity along the Se–Se edge. The selenium
atoms and thus the phenyl rings of the phenylselenolato ligands
are both oriented above a virtual mean planar rhomb while the
iron atoms lie below ( 0.1433(5) pm). Fe–Se distances range from
244.0(1) to 245.9(1) pm and the non-bonding Fe ◊ ◊ ◊ Fe distance
was found to be 287.9(2) pm which is distinctly smaller than
those found in planar Fe2Se2 dinuclear units for [Fe(SeC6H2-2,4,6-
Ph3)(N(SiMe3)2)]2 (317.0 pm),20 [Fe(SeC6H2-2,4,6-Ph3)2]2 (320.7
pm)20 and (iPrNMe3)2[Fe2(SeiPr)6] (302.6 pm),21 but larger than
Fe–Fe bonds in [(CO)3FeSepTol]2 (253.6 pm).22 Interestingly the
structure of the related polynuclear ring molecule [Fe(SePh)2]12,
which is a structural isomer of 1, displays similar bond distances
(Fe–Se 244.1 pm, Fe ◊ ◊ ◊ Fe 287.9 pm)7 and also similar Fe–Se–
Fe bridging angles (71.15–72.57◦). Only the tetrahedral Se–Fe–Se
bond angles show values different from those observed in 1 (e.g.
99.97, 105.81, 105.50, 114.26, 114.64, 115.75◦). The formation of
a certain structure by oligomerisation of the same monomeric
‘Fe(SePh)2’ unit seems to be determined by the orientation of the
phenyl groups and thus the direction of the bending of the Fe2Se2
rhomb. In a slightly simplified picture, one can realize that in 1
viewed along c (Fig. 1b) an alternating orientation of the phenyl
rings (up, left, down and right) is characteristic for the formation
of a helical chain structure while in the polynuclear ring molecule
[Fe(SePh)2]127 the phenyl rings in every second rhomb are oriented
in the same direction (‘outside down’ or ‘outside up’). Practically
the polarity of the reaction solvent (thf for 1 and CH2Cl2 for
[Fe(SePh)2]12) together with the presence of PPh3 in the synthesis
Scheme 1
Scheme 2
Scheme 3
Reaction of 1 with the Lewis base 1,10-phenanthroline yielded,
depending on the ratio, either dark green crystals of the octahedral
This journal is
The Royal Society of Chemistry 2011
Dalton Trans., 2011, 40, 7022–7032 | 7023
©

View Article Online
Table 1 Crystallographic Data for 1•[Fe(SePh)2] (1), 1•[Fe(Se-C6H2-2,4,6-(CH3)3)2] (2), [Fe(SePh)2(1,10-phen)2] (3) and [Fe(phen)3][Fe(SePh)4] (4)
1
2
3·C4H8O
4·(CH3)2NCHO
Sum formula
fw (g mol-1)
C12H10FeSe2
367.97
Tetragonal
I41/a
22.092(3)
C18H22FeSe2
452.13
Tetragonal
P42/mbc
15.710(2)
C40H34FeN4OSe2
800.5
Monoclinic
C63H51Fe2N7OSe4
1349.7
Crystal system
Triclinic
¯
Space group
P21/n
P1
˚
Cell a (A)
14.849(3)
11.434(3)
20.168(3)
13.290(3)
13.461(3)
16.475(3)
88.85(3)
77.74(3)
89.44(3)
2879.6(10)
2
b
c
10.887(2)
7.0097(14)
a (◦)
b
94.30(3)
g
3
˚
V (A )
5313.5(15)
16
150(2)
0.0800
1.840
8.866
2816
52
8899
1752
0.0791
1314
1730.0 (5)
4
190(2)
Mo-Ka
1.736
5.071
896
3414.6(12)
4
190(2)
Mo-Ka
2.613
2.547
1616
51
22983
6414
0.0259
5545
433
0.0325
0.0854
Z
T (K)
190(2)
Mo-Ka
1.557
3.080
1348
50
26364
9954
0.0394
8231
669
˚
l (A)
dc (g cm-3)
m (l) (mm-1)
F[000]
]
2qmax (◦)
54
Meas. reflns
7739
1001
0.0662
926
63
0.0320
0.0906
Unique reflns
Rint
Reflns with I > 2s(I).
Refined params
R1(I > 2s(I))a
wR2 (all data)b
159
0.0383
0.0689
0.0474
0.1558
a R1 = RꢀFo| – |Fcꢀ/R|Fo|. b wR2 = {R[w(Fo - Fc2)2] / R[w(Fo2)2]}
2
1/2
of [Fe(SePh)2]12 play, most probably, the important role for the
formation of a certain structure. Furthermore we were able to
prove by powder XRD that the ring molecule [Fe(SePh)2]12 is only
a metastable intermediate and converts in CH2Cl2 afters several
days and in THF over night completely to form a microcrystalline
powder of the chain compound 1 (Fig. S1 in the supporting
information†).
2 crystallises in the tetragonal space group P42/mbc (Table 1).
The complex exists in the solid state also as a one dimensional
infinite chain, with symmetrically bridging selenolate ligands
(Fig. 2). The Fe–Se bond distances of 251.81(4) pm are slightly
longer than those in 1. In contrast to 1 the Fe2Se2 rings are planar
and close to square shape with the Se–Fe–Se angles (88.20(2)◦)
being less acute and Fe–Se–Fe angles (91.80(2)◦) being more acute
than in 1. As a result the distance of the iron atoms along the chain
is significantly increased to 350.5(1) pm. The mesityl groups are
oriented around the chain so as to minimize steric interactions.
than in 1 and slightly longer than in 2. The Fe–N distances
(217.8–226.3(2) pm) are slightly longer than those found in
[FeCl2(phen)2]2- (213.0–222.1 pm)23 and distinctly longer than
those found in [(CO)2Fe(SePh)2(phen)] (198.9 pm).24 The N–Fe–N
bite angles (N(1)–Fe(1)–N(2): 74.52(8), N(3)–Fe(1)–N(4): 74,69(8)
◦) are comparable with those found in [FeCl2(phen)2]2- (75.87 and
76.28◦) and smaller than that found in [(CO)2Fe(SePh)2(phen)]
(81.31◦).
Ionic 4 crystallizes like 3 also as a racemate of the D and Kisomer
of the [Fe(phen)3]2+ dication in the triclinic spacegroup P1 (Table
¯
1). The crystal structure consists of well separated [Fe(phen)3]2+
dications and [Fe(SePh)4]2- dianions (Fig. 4). The well known
[Fe(phen)3]2+ dication comprises a ferrous atom which is in the
distorted octahedral coordination of six nitrogen atoms of the
three phenanthroline ligands. Bond angles and bond lengths are
comparable to those found in previously characterized compounds
like [Fe(phen)3]Cl2 for example.25 The FeSe4 core of the anion is
slightly more distorted (Se(1)–Fe(1)–Se(3) 98.76(4)–Se(3)–Fe(1)–
Se(4) 118.25(4)◦) than the one observed in (NEt4)2[Fe(SePh)4]
(103.6–114.9◦)26 while the Fe–Se bond lengths are
similar.
The structure of
1 is strongly reminiscent of that of
1•[Mn(SePh)2]13 while 2 is similar to [Mn(SeC6H2-2,4,6-CH3)2].14
The bond angles for each isostructural couple of iron and
manganese structures are quite similar while the bond distances
(Table 1) are smaller in the iron compounds as expected due to the
smaller radii of the iron ions (atom radii [pm]: Mn(125), Fe(126);
ionic radii [pm] M2+, hs, CN = 4: Mn(80), Fe(77)).
Powder X-ray diffraction studies
Monomeric 3 crystallizes as a racemate of the D and K isomer
in the monoclinic space group P21/n (Table 1) and exhibits a
distorted octahedral coordination sphere around the iron atom
(Fig. 3). The coordination sphere involves four nitrogen atoms
(N(1), N(2), N(3), N(4)) from the two bidentate phenanthroline
ligands and two selenium atoms (Se(1), Se(2)) from the phenylse-
lenolato ligands which coordinate in a cis geometry. The Fe–Se
distances of 255.0(1) pm and 259.3(1) pm are significantly longer
The calculated (150 K) and measured X-ray diffraction powder
patterns (293 K) for the carefully dried powders of 1–4 show a
good agreement, taking into account the temperature difference
(Fig. S2 in the supporting information†). However, powders of
1 taken from different reaction batches, sometimes displayed a
certain disagreement between the calculated and the experimental
profiles: a second diffraction peak with varying intensity appeared
at 2H = 8.18◦, close to the most intense peak at 2H = 7.93◦ of
7024 | Dalton Trans., 2011, 40, 7022–7032
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
Fig. 1 a) Section of the one-dimensional chain structure of 1•[Fe(SePh)2]
(1) viewed down a (ellipsoids drawn with 50% probability). Symmetry
transformation for generation of equivalent atoms: ¢ y + 1/4, -x +
5/4, z + 1/4 ¢¢ -y + 5/4, x - 1/4, z - 1/4 ¢¢¢ -x + 3/2, -y + 1, z -
1/2. Selected bond length (pm) and angles (◦): Se(1)–Fe(1¢) 243.97(13),
Se(1)–Fe(1) 245.13(12), Se(2)–Fe(1) 245.86(12), Fe(1)–Se(2¢) 243.96(13),
Fe(1) ◊ ◊ ◊ Fe(1¢) 287.92(15). Se(1¢¢)–Fe(1)–Se(1) 105.03(4), Se(2¢)–Fe(1)–
Se(1) 106.39(4), Se(1¢¢)–Fe(1)–Se(2) 106.16(4), Se(2¢)–Fe(1)–Se(2)
108.62(4), Se(1)–Fe(1)–Se(2),108.20(4), Se(1¢¢)–Fe(1)–Se(2¢) 121.82(5),
Fe(1¢)–Se(1)–Fe(1) 72.12(3), Fe(1¢¢)–Se(2)–Fe(1) 72.00(3), b) view along c.
Fig.
2
Section of the one-dimensional chain structure of
1•[Fe(SeC6H2-2,4,6-(CH3)3)2] (2) viewed down
a (ellipsoids drawn
with 50% probability). Symmetry transformation for generation of
equivalent atoms: ¢ -x + 2, -y, -z ¢¢ -y + 1, x - 1, z + 1/2 ¢¢¢ y + 1, -x +
the original structure. In order to reproduce the occurrence of the
additional peak we succeeded by gentle heating of 1 to 110 ◦C for
5 h to completely transform the original powder pattern of 1 to a
new but quite similar pattern (Fig. 5).
Rietveld analysis was used to investigate the changes which
occurred after the heating procedure (Fig. S3 and S4 and Table
S1 in the supporting information†).27 Cell parameters refinement
resulted in slightly changed unit cell edges with about 2.3% shorter
IV
V
1, -z + 1/2 y + 1, -x + 1, z + 0.5 y + 1, -x + 1, z - 0.5. Selected
bond length (pm) and angles (◦): Se(1)–Fe(1) 251.81(4), Fe(1) ◊ ◊ ◊ Fe(1¢)
350.5(1). Fe(1)–Se(1)–Fe(1¢) 88.20(2), Se(1¢¢)–Fe(1)–Se(1) 118.97(1),
Se(1)–Fe(1)–Se(1¢) 91.80(2), Se(1)–Fe(1)–Se(1¢¢¢) 118.97(1).
formed by the edge sharing FeSe4 tetrahedra (Fig. S4a and b
in the supporting information†). Most probably this irreversible
structural rearrangement is induced by loss of THF molecules
upon drying. Lattice THF molecules were identified within the
structure of 1, but these were located on a fourfold axis and
therefore badly disordered and could not be adequately refined
(see experimental part). Although elemental analysis and TGA of
1 after vacuum drying without heating suggest a loss of all THF
molecules, a complete phase change could only be achieved by
˚
˚
a and b axes (21.5799(4) A against 22.092(3) A), while the c axis
˚
˚
increased slightly from 10.887(2) A to 10.9024(5) A. As a base
for refinement the atom positions for iron and selenium were
taken from the structural model obtained from single crystal
diffraction. The positions and thermal factors of carbon and
hydrogen atoms of the phenyl rings have not been optimized
during the Rietveld refinement process and the quality of the
fit did not improve significantly when they were refined. The
refined structural model is given in table S1 of the supporting
information section.† The main structural changes involve a slight
shrinkage of the unit cell accompanied by a twist of the 1D chains
additional heating. In this respect it is also interesting to note that
3
˚
the elemental cell of the heated material still comprises 564 A of
3
˚
solvent accessible voids in comparison to 615 A of the original
compound.
This journal is
The Royal Society of Chemistry 2011
Dalton Trans., 2011, 40, 7022–7032 | 7025
©

View Article Online
Fig. 3 Molecular structure of [Fe(SePh)2(1,10–phen)2] (3) (ellipsoids
drawn with 50% probability). Selected bond length (pm) and angles (◦):
Fe(1)–N(4) 217.8(2), Fe(1)–N(2) 217.9(2), Fe(1)–N(3) 224.3(2), Fe(1)–N(1)
226.3(2), Fe(1)–Se(1) 255.0(1), Fe(1)–Se(2) 259.3(1), N(4)–Fe(1)–N(2)
157.27(8), N(4)–Fe(1)–N(3) 74.69(8), N(2)–Fe(1)–N(3) 88.97(8),
N(4)–Fe(1)–N(1) 88.01(8), N(2)–Fe(1)–N(1) 74.52(8), N(3)–Fe(1)–N(1)
84.34(8), N(4)–Fe(1)–Se(1) 99.11(6), N(3)–Fe(1)–Se(1) 173.77(6),
N(1)–Fe(1)–Se(1) 96.15(5), N(4)–Fe(1)–Se(2) 97.22(6), N(2)–Fe(1)–Se(2)
98.60(6), N(3)–Fe(1)–Se(2) 90.38(6), N(1)–Fe(1)–Se(2) 171.35(5),
Se(1)–Fe(1)–Se(2) 89.846(16).
Fig. 5 Measured X-ray diffraction powder patterns of 1 [Fe(SePh)2] (1)
•
before and after heating to 110 ◦C.
may become significant producing a change in dimensionality
leading to long range order. Secondly, many spin systems exhibit
Ising (1-D) or XY (2-D) anisotropy resulting in zero-field splitting
of the spin states which is not accounted for in simple models.
Finally, systems with integer spins will have a gap in their energy
level manifold leading to a more rapid decrease in susceptibility
at low temperatures.32 Two approaches have been taken in order
to derive mathematical models for idealized chains. For large
moments, such as in our previous work on 1•[Mn(SeR)2] (R = Ph,
Mes),13 the classical approximation used by Fisher29 is estimated
to be very good, while for smaller, quantum spin systems,
the approach of Bonner and Fisher,30 expanded by Weng,31 is
more appropriate, though neither of these models introduce an
anisotropy parameter.
The magnetic behavior of 1 and 2 (Fig. 6a and b) is very similar
1
to that of [Mn(SeR)2] (R = Ph, Mes).13 As with the manganese
•
analogue, 1, has a very small magnitude of susceptibility, well
below that expected at room temperature for a system of non-
interacting iron ions and it decreases smoothly down to around
40 K. This is indicative of very strong antiferromagnetic superex-
change as the broad maximum must occur at a temperature far
higher than 300 K. At very low temperatures, a rapid rise in
susceptibility has been observed. We ascribe this to paramagnetic
impurities arising from a very small degree of decomposition.
Interestingly the magnetic behaviour of 1 is similar to related solid
state compounds like KFe(III)E2 (E = S, Se) which also consists of
edge-linked chains of FeE4 tetrahedra.33
Fig. 4 Molecular structure of [Fe(phen)3][Fe(SePh)4] (4) (ellipsoids
drawn with 50% probability). Selected bond length (pm) and angles
(◦): Se(1)–Fe(1) 244.43(11), Se(2)–Fe(1) 244.59(11), Se(3)–Fe(1)
246.08(10), Se(4)–Fe(1) 247.41(12), Fe(2)–N(6) 195.6(4), Fe(2)–N(2)
196.7(4), Fe(2)–N(3) 196.9(4), Fe(2)–N(5) 197.4(4), Fe(2)–N(4) 197.7(4),
Fe(2)–N(1) 198.3(4). Se(1)–Fe(1)–Se(2) 113.76(4), Se(1)–Fe(1)–Se(3)
98.76(4), Se(2)–Fe(1)–Se(3) 114.06(4), Se(1)–Fe(1)–Se(4) 104.07(4),
Se(2)–Fe(1)–Se(4)
107.44(4),
Se(3)–Fe(1)–Se(4)
118.25(4),
N(2)–Fe(2)–N(1) 82.19(19), N(3)–Fe(2)–N(4) 82.74(16), N(6)–Fe(2)–N(5)
83.04(16), N(3)–Fe(2)–N(5) 175.42(16), N(2)–Fe(2)–N(4) 176.51(16),
N(6)–Fe(2)–N(1) 172.35(18).
For comparison, we resynthesized the ring molecule
7
[Fe(SePh)2]12 for which only the room temperature magnetic
moment was originally determined. The temperature dependence
of the magnetic susceptibility of this compound (Fig. S5 in the
supporting information†) is broadly similar to that measured for
1, showing decreasing susceptibility with decreasing temperature
down to low temperatures. This is indicative for very strong
antiferromagnetic coupling within the ring as the maximum in
c must occur far above 300 K. The effective magnetic moment meff
of the ring compound was originally found to be 1.51 mB/Fe at
293 K8 which agrees with our value of 1.57 mB/Fe at 300 K derived
from c. The increase in the susceptibility below 32 K probably
results from paramagnetic impurities. The magnetic behaviour
is thus very similar to that observed in 1 a similarity which is
Magnetic properties
The magnetic behavior of antiferromagnetic chain compounds
has been extensively studied for many years.28 Ideal systems
show a broad maximum in the temperature dependence of the
magnetic susceptibility as a result of short range ordering. Long
range order is not possible in a one-dimensional system. Various
models have been derived for the temperature dependence of the
susceptibility for chains of isotropic spins29–31 which depend on
the magnitude of the exchange interaction and the spin quantum
number. In real systems, there are a number of factors which
produce deviations from ideal behavior. Interchain interactions
7026 | Dalton Trans., 2011, 40, 7022–7032
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
⎛
⎜
⎜
⎞
⎟
⎟
⎟
⎟
⎠
2JS (S +1)
kBT
u = coth
−
⎜
⎜
⎝
kBT
2JS S +1
(
)
and cM is molar susceptibility, NA is Avogadro constant, mB is
Bohr magneton, S is total spin quantum number, kB is Boltzmann
constant, J is the coupling constant and T is temperature.
It should be noted that this value is approximately double that
obtained for the manganese analogues.
Deviation from the model at low temperatures can have a
number of possible causes. Firstly, a more rapid drop in c could
be associated with a Haldane gap, however a similar behavior
was observed in the half integer manganese analogue. The second
possibility is that interchain interactions become dominant and
there is a second order phase transition to a long range ordered
state. This is also implausible as the decrease is actually too rapid
for an isotropic 3-D system. Finally, the deviation may be due to
zero field splitting where states with low Jeff may be preferentially
populated as the temperature decreases. As with 1, a paramagnetic
tail is observed below ca. 20 K.
Isothermal magnetization curves of 1 and 2 at 5 K show no satu-
ration up to 4.5 T (1) and 9 T (2) consistent with antiferromagnetic
behaviour (Fig. S6 in the supporting information†).
Density Functional Theory (DFT) calculations were performed
to obtain the exchange coupling constants for the dinuclear
segments [Fe2(SeR)6]2- (R = Ph (5), R = Mes (6)) of the chain
compounds 1 and 2 by the broken symmetry approach35–37 (for
further details see experimental section). The calculated exchange
coupling constants of the two compounds are significantly dif-
ferent. For 5 a coupling constant of J = -137 cm-1 was obtained
while a much lower negative value of J = -20 cm-1 was obtained for
compound 6. It is known for a long time that the superexchange
coupling is a function of the geometrical structure of the bridging
unit B in M–B–M between two magnetic ions M.38–40 Important
structural parameters are the M–B distance and the M–B–M
angle. Furthermore, the role of direct exchange between the metal
atoms compared to superexchange is of interest. The influence of
these parameters on the magnetic exchange coupling have been the
subject of several quantum chemical calculations (see for example
references [41–46]).
We constructed model clusters to analyze which of the geomet-
rical differences between 5 and 6 is responsible for the change in
the coupling, the Fe–Se distance, the Fe–Se–Fe angle or a direct
coupling of the two metal atoms. In [Fe2(–)2(SeCH3)4]2– (model
1) (Fig. 7a), each Se–R bridge was substituted by one negative
point charge. This allows for the analysis of the direct Fe–Fe
coupling. We obtained a coupling constant of -16 cm-1 for the
short Fe–Fe distance in 5 and -4 cm-1 for the long Fe–Fe distance
in 6. There is also a difference in the occupation of the d-orbitals.
For the long distance the dxy orbital is doubly occupied, while
Fig. 6 The temperature dependence of the magnetic susceptibility of a)
1•[Fe(SePh)2] (1) and b) 1 [Fe(SeC6H2-2,4,6-(CH3)3)2] (2). The solid line
•
represents a fit to Fisher’s equation, described in the text.
expected in view of similar geometry in the Fe–Se–Fe bridges
(see DFT calculations below). Slight differences in the distortion
of the essentially tetrahedral ligand fields in 1 and [Fe(SePh)2]12
might lead to slight differences in magnetic behavior. Much weaker
magnetic coupling is observed in the Fe(II) rings with mixed halide
and thiolate bridges; [FeCl(SSitBu3)]12 and [FeBr(SSitBu3)]12.34
This was assigned to rather long intermetallic distances (Fe–Fe
312.7 and 317.1 pm, respectively). In view of our theoretical
calculations (see below) this weaker coupling might be better
explained by the larger Fe–S–Fe bridging angles (83.9 and 85◦
respectively) which lead to a weaker superexchange through the
ligands.
There are also distinct similarities between 2 and its manganese
analogue. The presence of the broad maximum centered close to
125 K shows that the antiferromagnetic exchange is strong, but
not as strong as in 1. Fits to the classical Fisher method (eqn (1))
2
a combination of dz and dxy is occupied for the short Fe–Fe
yield J = -16.2(2) cm-1.
For the spin Hamiltonian
distance (for the orientation of the molecules see Fig. 7). This
shows that direct exchange plays a minor role in coupling between
the iron atoms. In [Fe2(SeCH3)6]2– (model 2) (Fig. 7b), the Fe–Se–
Fe angle was varied from 70◦ to 110◦ for Fe–Se distances of 246
pm and 252 pm. The results are given in Table 3. As expected,
the coupling is for each angle in average by 32% weaker overall
in 6 (less negative J values), where the Fe–Se distance is by 6
pm longer. More importantly, the size of the exchange interaction
ˆ
H = −2J
Si ⋅Si+1
∑
these equations
i
have the form:
2
2
⎛
⎜
⎝
⎞
⎟
⎟
NAg mBS(S +1) 1+u
⎜
⎜
cM
=
(1)
⎟
⎟
⎠
3kBT
1−u
where
This journal is
The Royal Society of Chemistry 2011
Dalton Trans., 2011, 40, 7022–7032 | 7027
©

View Article Online
Table 2 Comparison of bond distances (pm) in 1 and 2 with those in
coordination geometry around the iron atoms do obviously not
have a strong influence on the coupling constant.
1•[Mn(SePh)2]13 and [Mn(SeC6H2-2,4,6-(CH3)3)2].14
1
1•[Mn(SePh)2]
2
[Mn(SeMes)2]
Optical properties
M–Se
243.9(1)–245.9(1) 254.7(1)–257.8(1) 251.8 (4) 259.0 (3)
M ◊ ◊ ◊ M 287.9(2)
304.6(1)
350.5 (1) 359.3 (3)
UV–Vis spectra of the crystalline powders of 1–4 were measured
as a nujol mull pressed between quartz plates. For 1 and 2 (Fig. 8a),
approximately three broad and nearly featureless absorption
maxima can be identified between 650 and 300 nm (1) and
350 nm (2). According to recent observations in dinuclear iron
thiophenolate clusters, these transitions should be mainly assigned
to LMCT bands from –SeR- (R = Ph, Mes) to Fe2+.2,3 The strong
absorption features at higher energies (1: 250 nm; 2: 250 and
295 nm) can be assigned in accordance to spectra of related
manganese selenolato complexes to p–p* transitions of the –SeR-
(R = Ph, Mes) ligands.13 As expected 3 (dark green) and 4 (dark red)
display different absorption spectra (Fig. 8b). Due to the different
types of ligands, absorption features above 350 nm could originate
from –SePh- to Fe2+ LMCT bands as observed in 1 as well as metal
to ligand charge transfer transitions (MLCT) into low lying empty
p* orbitals of phenanthroline ligands. A noticeable very broad and
weak absorption shoulder in 3 ranges from approximately 1300
down to 800 nm, while no absorption above 800 nm is present in
4. Similar strong absorption features below 350 nm can be assigned
to p–p* transitions of either the phenanthroline ligands or of the
PhSe- ligands or of a mixture of both.
*Mes = SeC6H2-2,4,6-(CH3)3
Table 3 Magneto-structural correlation of the exchange coupling in
[Fe2(SeCH3)6]2– (model 2). The coupling constants are given in cm-1
Fe–Se–Fe angle
J (dFe–Se = 246 pm)
J (dFe–Se = 252 pm)
110
100
90
80
70
-14.7
-5.2
-28.4
-77.0
-120.4
-9.3
-2.9
-19.9
-57.5
-91.4
Fig.
7
Model clusters for a) the direct Fe–Fe interaction
[Fe2(–)2(SeCH3)4]2– (model 1) and b) the magneto-structural correlation
[Fe2(SeCH3)6]2– (model 2).
decreases by 76.4% for dFe–Se = 246 pm and 78.2% for dFe–Se = 252
pm as the Fe–Se–Fe angle increases for 70 to 90◦. This illustrates
that the source of the difference in behaviour between 1 and 2 is
related to the steric bulk of the aromatic groups; phenyl is less
bulky and allows more efficient packing with smaller Fe–Fe and
Fe–Se distances and more important smaller Fe–Se–Fe angles and
therefore strong antiferromagnetic coupling. Equivalent reasoning
shows why more bulky mesityl yields weaker coupling.
Calculations on a binuclear model for the [Fe(SePh)2]12 ring
molecule yielded a coupling constant of J = -135 cm-1 similar
to that obtained for the model unit 5 of the chain complex 1.
This result again confirms the importance of the geometry of
the Fe2Se2 rhombs for the magnetic properties, which is similar
for both complexes. In comparison, differences in the tetrahedral
Fig. 8 UV-Vis spectra of a) 1 [Fe(SePh)2] (1) and 1 [Fe(SeC6H2-2,4,6-
(CH3)3)2] (2) and b) [Fe(SePh)2(1,10-phen)2] (3) and [Fe(phen)3][Fe(SePh)4]
(4) in solid state (powder in mineral oil between quartz plates).
•
•
7028 | Dalton Trans., 2011, 40, 7022–7032
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
Absorption spectra of 1–4 in solution display different features
than in solid state (Fig. S7 and S8 in the supporting information†).
The differences in both types of the spectra, which are on the
one hand due to the non validity of Lambert Beer’s law for
the solid state spectra, might also indicate different coordination
geometries of the Fe2+ ions in solution and in solid state. In
the crystal structure of 1 and 2 the Fe2+ ions have a tetrahedral
distorted coordination by four selenium atoms of the selenolato
ligands. In solution, the one-dimensional polymeric structures are
most probably destroyed along with an additional coordination
of solvent molecules which can either result in a tetrahedral or
octahedral geometry. Interestingly the solution spectra of 3 and
4 in DMF are very similar as already indicated by the formation
of a red solution suggesting the formation of identical complex
fragments in solution.
Thermal decomposition
Thermogravimetric analysis (TGA) of 1 in a helium gas flow shows
that the thermal decomposition occurs from 150 ◦C up to 250 ◦C in
a one step process (Fig. 9). In vacuum (~10-6 mbar), 1 already starts
to decompose above 30 ◦C in two successive steps up to 175 ◦C. The
mass change of both decomposition reactions corresponds to the
calculated cleavage of one SePh2 (calcd 63.3%). Its formation and
purity was confirmed by 1H and 13C-NMR of the collected liquid
colourless cleavage product. The black solid residue of the TGA
of 1 thus has the formal composition FeSe (C < 0.1%). The XRD
powder pattern of a sample heated to 550 ◦C in vacuum (Fig. 10)
reveals that in agreement with the phase diagram of the Fe–Se
system,47 the powder consists of a tetragonal Fe1+xSe (0.012 < x
< 0.02)48 and a hexagonal Fe7Se8 (with 52 to 53 at.% Se)49 phase
(Tables 2–4). In samples heated to 250 ◦C, only broad peaks of the
tetragonal phase could be identified which sharpen accompanied
by the evolution of the peaks for the hexagonal Fe7Se8 phase by
heating to 350 and 450 ◦C (Fig. S9 in the supporting information†).
This could be interesting with respect to the fact that recently
superconductivity was observed in tetragonal PbO-type FeSe with
a critical temperature (Tc) of 8 K.50
Fig. 10 Powder XRD pattern of t◦he residue of the TGA experiment under
vacuum (3 ¥ 10-6 mbar, up to 550 C) of •1 [Fe(SePh)2] (1) compared to the
indexed reflection patterns of tetragonal Fe1+xSe (0.012 < x < 0.02) 48 and
hexagonal Fe7Se8 (with 52 to 53 at.% Se).49
(CH3)3) have been prepared in good yields. In the crystal, the
compounds form one-dimensional chains with bridging selenolate
ligands comprising Fe–Se–Fe bridging angles which differ by
approximately 20◦. Magnetic measurements supported by DFT
calculations reveal that the different bridging angles have a pro-
nounced influence on the antiferromagnetic exchange interactions
of the unpaired electrons along the chains in the two different
compounds. In principle, these investigations on magnetostruc-
tural correlations in one-dimensional chain compounds should be
extendable to related sulphur and tellurium bridged compounds as
well as chalcogenolato complexes of cobalt and nickel. Preliminary
data from single crystals X-ray analysis and powder diffraction
measurements indicate that related sulphur bridged compounds
of manganese and iron are accessible via similar reactions.
Experimental section
Synthesis
Standard Schlenk techniques were employed throughout the
syntheses using a double manifold vacuum line with high pu-
rity dry nitrogen (99.999990%) and a MBraun Glovebox with
high purity dry argon (99.9999990%). The solvents THF and
diethylether were dried over sodium-benzophenone, and distilled
under nitrogen. Anhydrous dimethylformamide (DMF) (H2O <
0.005%) obtained from Aldrich was degassed, freshly distilled
and stored over molecular sieves under nitrogen. Fe(OOCCH3)2,
FeCl3 and 1,10-phenanthroline were purchased from Aldrich.
PhSeSiMe3,51 PhSeH,51 MesSeH (Mes = C6H2-2,4,6-(CH3)3)52 and
Fig. 9 Thermogravimetric analysis of 1 [Fe(SePh)2] (1) under a) He gas
flow and b) in vacuum.
•
Conclusion
The first examples of polymeric homoleptic iron chalcogenolato
complexes 1 [Fe(SePh)2] and 1 [Fe(SeMes)2] (Mes = C6H2-2,4,6-
•
•
This journal is
The Royal Society of Chemistry 2011
Dalton Trans., 2011, 40, 7022–7032 | 7029
©

View Article Online
53
7
Crystallography
Fe{N(Si(CH3)3)2}2 and [Fe(SePh)2]12 were prepared according
to literature procedures.
Crystals suitable for single crystal X-ray diffraction were taken
directly from the reaction solution of the compound and then
selected in perfluoroalkylether oil. Single-crystal X-ray diffraction
[Fe(SePh)2] (1): a) Fe{N(Si(CH3)3)2}2 (0.185 g (0.49 mmol))
was dissolved in 30 mL THF and 0.08 mL (0.4 mmol) of PnPr3
was added to give a light yellow solution. Addition of HSePh
(0.11 mL (1.03 mmol)) at -30 ◦C resulted in the immediate
formation of a deep red solution. Standing over night at 2 ◦C
allowed the formation of tiny dark red crystals of 1 which were
filtered after 2 additional days of rest at room temperature and
washed subsequently twice with THF to give a total yield of 68%
(0.123 g). C12H10FeSe2 (368.0): calcd C 39.2, H 2.7; found C 39.3,
H 3.1%.
b) FeCl3 (0.26 g (1.6 mmol)) was dissolved in 35 mL THF.
PhSeSiMe3 (1.23 mL (6.41 mmol)) was then added, followed by
a quick color change from light green to dark green then yellow
orange to dark red. When the stirring of the solution is stopped
soon a dark red microcrystalline precipitate of 1 starts to form.
After one night, the solution is warmed up without stirring for 1 h
at 60 ◦C and then, after cooling to room temperature, filtered and
washed with THF to give 0.35 g (59%). C12H10FeSe2 (368.0): calcd
C 39.2, H 2.7; found C 39.0, H 2.9%.
c) Crystals of 1 for single crystal diffraction: Fe(OOCCH3)2
(0.175 g (1.01 mmol)) was suspended with PnBu3 (0.5 mL, 2.01
mmol) in 25 mL of THF. PhSeSiMe3 (0.48 mL (2.53 mmol)) was
then added and the mixture stirred over night to give a deep red
solution. Careful layering of 10 mL of the reaction solution with
10 mL THF and 20 mL of diethylether in Schlenk tubes resulted
after three weeks in the crystallisation of tiny dark red needles
of Fe(SePh)2 (1) some of which being suitable for single crystal
diffraction at a synchrotron source. Filtration and washing with
THF resulted in a total yield of 43% (0.159 g) for 1. C12H10FeSe2
(368.0): calcd C 39.2, H 2.7; found C 39.2, H 3.0%.
˚
data of 1 were collected using synchrotron radiation (l = 0.80 A) on
a STOE IPDS II (Imaging Plate Diffraction System) at the ANKA
synchrotron source in Karlsruhe. The quality and intensity of the
reflections of the weakly diffracting crystals of 1 with a resolution
smaller than d = 0.92 decreases significantly indicated by a decrease
of the mean I/s ratio below 6.04 and an increase of the R(int) value
above 0.2526. Therefore data were measured only up to 2q 52◦ at
l = 0.80000. Single-crystal X-ray diffraction data of 2, 3 and 4
were collected with graphite-monochromatized Mo-Ka radiation
˚
(l = 0.71073 A) on a STOE IPDS II (Imaging Plate Diffraction
System). Raw intensity data were collected and treated with the
STOE X-Area software Version 1.27. Data for all compounds
were corrected for Lorentz and polarisation effects. Based on an
optimized crystal description numerical absorption corrections
were applied.54 The structures were solved with the direct methods
program SHELXS of the SHELXTL PC suite programs,55 and
were refined with the use of the full-matrix least-squares program
SHELXL.55 Molecular diagrams were prepared using Diamond.56
Lattice THF molecules were identified within the structure of 1,
but these were located on a fourfold axis and badly disordered and
could not be adequately refined. The data were therefore corrected
for these using the SQUEEZE option within the PLATON57
program package finding a total of 130 electrons (~2.5 THF) in
˚
a potential solvent accessible area of 615 A. All Fe, Se, N, O and
C atoms were refined with anisotropic displacement parameters
except those of the solvent DMF molecule in 4 whilst all H atoms
were calculated in fixed positions.
CCDC 737597(1), 737598 (2), 737599 (3) and 747600 (4)
contain the supplementary crystallographic data for this pa-
per. These data can be obtained free of charge at www.ccdc.
cam.ac.uk/conts/retrieving.html (or from the Cambridge Crys-
tallographic Data Centre, 12 Union Road, Cambridge CB2
1EZ, UK; fax: (internat.) +44-1223/336-033; Email: deposit@
ccdc.cam.ac.uk).
X-ray powder diffraction patterns (XRD) were measured on
a STOE STADI P diffractometer (Germanium monochromator,
Debye–Scherrer geometry) with Cu-Ka1 radiation (1–4) and
with Co-Ka1 radiation (bulk-FeSe) in sealed glass capillaries.
Theoretical powder diffraction pattern for 1–4 were calculated
on the basis of the atom coordinates obtained from single crystal
X-ray analysis by using the program package STOE WinXPOW.58
Rietveld analysis was performed with the program MAUD.27 The
position for iron and selenium atoms were refined together with
their Debye–Waller factors.
[Fe{SeC6H2-2,4,6-(CH3)3}2] (2): Fe{N(Si(CH3)3)2}2 (0.30 g
(0.81 mmol)) was dissolved in 40 mL toluene and 0.08 mL (0.4
mmol) of PnPr3 was added to give a light green solution. Addition
of HSe{C6H2-2,4,6-(CH3)3}2 (0.34 mL (1.69 mmol)) resulted in
the immediate formation of a deep red solution. Standing over
night in the dark afforded the formation of red-brown crystals of
2 which were filtered after 5 days and washed subsequently twice
with toluene and once with pentane to give a total yield of 61%
(0.22 g). C18H22FeSe (452.14): calcd C 47.8, H 4.9; found C 47.9,
H 5.1%.
[Fe(SePh)2(1,10-phen)2] (3): Fe(SePh)2 (0.123 g (0.33 mmol))
was dissolved in a mixture of 10 mL THF and 1 mL of DMF to
give a red solution. A solution of 1,10-phenanthroline (0.120 g
(0.67 mmol)) in 5 mL DMF was then added, followed, after
approximately 10 min, by the formation of dark green crystals
of 3 which were filtered and washed with THF after 4 days to give
a total yield of 78% (0.19 g). C36H26FeN4Se2 (728.4): calcd C 59.4,
H 3.6, N 7.7; found C 59.6, H 4.1, N 7.2%.
[Fe(1,10-phen)3][Fe(SePh)4]·(CH3)2NCHO
(4):
Fe(SePh)2
Physical measurements
(0.100 g (0.27 mmol)) was dissolved in 10 mL DMF to give
a dark-green suspension. A solution of 1,10-phenanthroline
(0.049 g (0.27 mmol)) in 3 mL DMF was then added to give a
deep red solution. Addition of 10 mL of THF and standing for
three days afforded dark red crystals of 4 which were filtered
and washed with THF to give a total yield of 75% (0.27 g).
C60H44Fe2N6Se4·C3H7NO (1349.7): calcd C 56.1, H 3.8, N 7.3;
found C 56.1, H 3.9, N 7.1%.
Zero-Field-Cooled temperature dependent susceptibilities were
recorded for 1 and 2 in RSO mode using a MPMS-5S (Quantum
Design) SQUID magnetometer over a temperature range from 5
to 300 K in a homogeneous 100 Oe (1) and 1000 Oe (2) external
magnetic field. The magnetization curve of 1 was measured on
the MPMS-5S (Quantum Design) SQUID magnetometer while
the magnetization curve of 2 was measured on a PPMS Quantum
7030 | Dalton Trans., 2011, 40, 7022–7032
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
Design magnetometer. The samples were contained in gelatine
capsules filled in a glove box under argon atmosphere owing
to the high degree of moisture and oxygen sensitivity of the
compounds. The data were corrected for the sample holder and
for diamagnetism using Pascal’s constants.28,59,60
UV-Vis absorption spectra of cluster molecules in solution
were measured on a Varian Cary 500 spectrophotometer in
quartz cuvettes. Solid state spectra were measured as micron
sized crystalline powders in nujol oil between quartz plates with a
Labsphere integrating sphere.
9 N. R. Kunchur, Acta Crystallogr., Sect. B: Struct. Crystallogr. Cryst.
Chem., 1968, B24, 1623–1633.
10 B. Krebs and G. Henkel, Angew. Chem., Int. Ed. Engl., 1991, 30(7),
769–788.
11 B. Krebs and G. Henkel, Chem. Rev., 2004, 104(2), 801–824.
12 M. E. Peach, J. Inorg. Nucl. Chem., 1979, 41, 1390–1392.
13 A. Eichho¨fer, P. T. Wood, R. N. Viswanath and R. A. Mole, Eur. J.
Inorg. Chem., 2007, 4794–4799.
14 M. Bochmann, A. K. Powell and X. Song, J. Chem. Soc., Dalton Trans.,
1995, 1645–1648.
15 H. L. Cuthbert, A. I. Wallbank, N. J. Taylor and J. F. Corrigan, Z.
Anorg. Allg. Chem., 2002, 628, 2483–2488.
16 M. Bochmann, A. P. Coleman, K. J. Webb, M. B. Hursthouse and M.
Mazid, Angew. Chem., Int. Ed. Engl., 1991, 30, 973–975.
17 I. G. Dance, R. G. Garbutt and D. C. Craig, Inorg. Chem., 1987, 26,
3732–3740.
Quantum chemical calculations
18 I. G. Dance, R. G. Garbutt, D. C. Craig and M. L. Scudder, Inorg.
All calculations were performed at DFT level using the RI-
approximation with the program package Turbomole.61,62 Seg-
ments with two iron atoms were cut out of the one-dimensional
chain compounds 1 and 2. All [SeR]- groups bound to the
two metal atoms are taken into account. The positions of the
Fe, Se, and C atoms were taken from X-ray structure analysis,
the positions of the H atoms were optimized using the BP86
functional.63,64 For the calculation of the exchange coupling
constant, the Spin-Hamiltonian (eqn (2)) was used.
Chem., 1987, 26, 4057–4064.
19 K. S. Anjali and J. J. Vittal, Inorg. Chem. Commun., 2000, 3, 708–
710.
20 R. Hauptmann, R. Kliß and G. Henkel, Angew. Chem., Int. Ed., 1999,
38, 377–379.
21 R. Hauptmann, J. Lackmann, C. Chen and G. Henkel, Acta Crystal-
logr., Sect. C: Cryst. Struct. Commun., 1999, 55, 1084–1087.
22 R. Wang, Q. Xu, Y. Souma, L-Ch. Song, J. Sun and J. Chun,
Organometallics, 2001, 20, 2226–2232.
23 B. N. Figgis, J. M. Patrick, P. A. Reynolds, B. W. Skelton, A. H. White
and P. C. Healy, Aust. J. Chem., 1983, 36, 2043–2055.
24 W.-F. Liaw, C.-H. Chen, C.-M. Lee, G.-Y. Lin, C.-Y. Ching, G.-H. Lee
and S.-M. Peng, J. Chem. Soc., Dalton Trans., 1998, 353–358.
25 M. Dela Varga, W. Petz, B. Neumu¨ller and R. Costa, Inorg. Chem.,
2006, 45, 9053–9063.
26 J. M. McConnachie and J. A. Ibers, Inorg. Chem., 1991, 30, 1170–1173.
27 L. Lutterotti and P. Scardi, J. Appl. Crystallogr., 1990, 23, 246–252 and
http://www.ing.unitn.it/~maud/index.html.
28 O. Kahn, Molecular Magnetism, Wiley-VCH, Weinheim, 1993.
29 M. E. Fisher, Am. J. Phys., 1964, 32, 343–346.
30 J. C. Bonner and M. E. Fisher, Phys. Rev., 1964, 135, 640–658.
31 C. Y. Weng, PhD Thesis, Carnegie Mellon University, Pittsburgh PA,
USA, 1969.
32 F. D. M. Haldane, Phys. Rev. Lett., 1983, 50, 1153–1156.
33 W. Bronger, A. Kayas and P. Mu¨ller, J. Solid State Chem., 1987, 70,
262–270.
34 O. L. Sydora, T. P. Henry, P. T. Wolczanski, E. B. Lobovsky, E.
Rumberger and D. N. Hendrickson, Inorg. Chem., 2006, 45, 609–626.
35 L. Noodleman, J. Chem. Phys., 1981, 74, 5737–5743.
36 K. Yamaguchi, H. Fukui and T. Fueno, Chem. Lett., 1986, 15, 625–628.
37 M. Nishino, S. Yamanaka, Y. Yoshioka and K. Yamaguchi, J. Phys.
Chem. A, 1997, 101, 705–712.
38 J. B. Goodenough, J. Phys. Chem. Solids, 1958, 6, 287–297.
39 J. Kanamori, J. Phys. Chem. Solids, 1959, 10, 87–98.
40 A. P. Ginsberg, Inorg. Chim. Acta, Rev., 1971, 5, 45–68.
41 C. Wang, K. Fink and V. Staemmler, Chem. Phys., 1995, 201, 87–94.
42 K. Fink, C. Wang and V. Staemmler, Int. J. Quantum Chem., 1997, 65,
633–641.
ˆ
(2)
ˆ ˆ
H = -2JS1S2
J was obtained from unrestricted DFT calculations with the
B3-LYP functional63,65,66 using the broken symmetry approach35–37
and applying the scheme of Nishino et al.37
E(BS)−E(HS)
J =
(3)
S2
− S2
HS
BS
E(HS) is the energy of a state where all unpaired electrons have
the same spin. In the BS state, the spin of the unpaired electrons at
one iron center is switched to opposite spin. The values obtained
with eqn (3) are almost identical to those obtained with eqn (4)
given by Noodleman.35
E(BS)−E(HS)
J =
(4)
SH2S
The difference in the denominator was 0.1% only.
Acknowledgements
The authors are grateful to Prof. Dr D. Fenske and Prof. A. K.
Powell for helpful discussions and generous support, E. Tro¨ster
for her valuable assistance in the practical work.
43 J. Casanovas and F. Illas, J. Chem. Phys., 1994, 101, 7683–7685.
44 S. A. Fiethen, V. Staemmler, N. N. Nair, J. Ribas-Arino, E. Schreiner
and D. Marx, J. Phys. Chem. B, 2010, 114(35), 11612–11619.
45 E. Ruiz, P. Alemany, S. Alvarez and J. Cano, Inorg. Chem., 1997, 36(17),
3683–3688.
References
46 D. A. Pantazis, V. Krewald, M. Orio and F. Neese, Dalton Trans., 2010,
39, 4959–4967.
1 P. R. Venkateswara and R. H. Holm, Chem. Rev., 2004, 104(2), 527–559.
2 J. Ballmann, S. Dechert, S. Demeshko and F. Meyer, Eur. J. Inorg.
Chem., 2009, 3219–3225.
3 J. Ballmann, S. Dechert, E. Bill, U. Ryde and F. Meyer, Inorg. Chem.,
2008, 47, 1586–1596.
4 O. L. Sydora, T. P. Henry, P. T. Wolczanski, E. B. Lobovsky, E.
Rumberger and D. N. Hendrickson, Inorg. Chem., 2006, 45, 609–626.
5 R. Pryadun and R. H. Holm, Inorg. Chem., 2008, 47, 3366–3370.
6 T. Nguyen, A. Panda, M. M. Olmstead, A. F. Richards, M. Stender,
M. Brynda and P. P. Power, J. Am. Chem. Soc., 2005, 127, 8545–8552.
7 D. Fenske and A. Fischer, Angew. Chem., Int. Ed. Engl., 1995, 34,
307–309.
47 H. Okamoto, J. Phase Equilibria, 1991, 12(3), 383–389.
48 G. Hagg and A. L. Kindstroem, Z. Phys. Chem. B, 1933, 22(5-6), 453–
464.
49 J. Parise, A. Nakano, M. Tokonami and N. Morimoto, Acta Crystal-
logr., Sect. B: Struct. Crystallogr. Cryst. Chem., 1979, 35, 1210–1212.
50 T. M. McQueen, Q. Huang, V. Ksenofontov, C. Felser, Q. Xu, H.
Zandbergen, Y. S. Hor, J. Allred, A. J. Williams, D. Qu, J. Checkelsky,
N. P. Ong and R. J. Cava, Phys. Rev. B: Condens. Matter Mater. Phys.,
2009, 79(1), 014522–014529.
51 N. Miyoshi, H. Ishii, K. Kondo, S. Mui and N. Sonoda, Synthesis,
1979, 301–304.
52 M. Bochmann, K. J. Webb, M. B. Hursthouse and M. Mazid, J. Chem.
Soc., Dalton Trans., 1991, 2317–2323.
8 A. A. Ivanov, M. A. Kozee, W. A. Merrill, S. Agarwal and L. Dahl, J.
Chem. Soc., Dalton Trans., 2002, 4105–4115.
This journal is
The Royal Society of Chemistry 2011
Dalton Trans., 2011, 40, 7022–7032 | 7031
©

View Article Online
53 R. A. Andersen, K. Faegri, J. C. Green, A. Haaland, M. F. Lappert,
A.-P. Leung and K. Rypdal, Inorg. Chem., 1988, 27, 1782–1786.
54 (a) X-SHAPE 1.06, Crystal Optimisation for Numerical Ab-
sorption Correction Program, Stoe & Cie GmbH, Darmstadt, 1999;
(b) X-RED32 1.01, Data Reduction Program, Stoe & Cie GmbH,
Darmstadt, Germany, 2001.
55 G. M. Sheldrick, SHELXTL PC version 5.1 An Integrated System for
Solving, Refining, and Displaying Crystal Structures from Diffraction
Data, Bruker Analytical X-ray Systems, Karlsruhe, 2000.
56 Diamond Version 2.1d, K. Brandenburg, Crystal Impact GbR, 1996-
2000.
58 STOE, WinXPOW, STOE & Cie GmbH, Darmstadt, 2000.
59 H. Lueken, Magnetochemie, B. G. Teubner, Stuttgart, Leibzig, 1999 S.
426.
60 W. Haberditzl, Angew. Chem., 1966, 78, 277–288, (Angew. Chem., Int.
Ed. Engl., 1966, 5, 288–298).
61 R. Ahlrichs, M. Ba¨r, M. Ha¨ser, H. Horn and C. Ko¨lmel, Chem. Phys.
Lett., 1989, 162, 165–169.
62 O. Treutler and R. Ahlrichs, J. Chem. Phys., 1995, 102, 346–354.
63 A. D. Becke, Phys. Rev. A: At., Mol., Opt. Phys., 1988, 38, 3098–3100.
64 J. P. Perdew, Phys. Rev. B, 1986, 33, 8822–8824.
65 C. Lee, W. Wang and R. G. Parr, Phys. Rev. B, 1988, 37, 785–789.
66 A. D. Becke, J. Chem. Phys., 1993, 98, 5648–5652.
57 A. L. Spek, Acta Cryst, 1990, A46, C–34.
7032 | Dalton Trans., 2011, 40, 7022–7032
This journal is
The Royal Society of Chemistry 2011
©