Amphiphilic iron(ii) complexes with short alkyl chains-crystal packing and spin transition properties

Article abstract of DOI:10.1039/c3nj00991b

An amphiphilic iron(ii) spin crossover complex with relatively short octyl chains was synthesised and the crystal structures of the high spin and the low spin state could be determined. In further reactions, a second modification of the hexa-coordinated c

Full text of DOI:10.1039/c3nj00991b

NJC
View Article Online
View Journal | View Issue
PAPER
Amphiphilic iron(II) complexes with short alkyl
chains – crystal packing and spin transition
properties†
Cite this: New J. Chem., 2014,
38, 1965
Stephan Schlamp, Katja Dankhoff and Birgit Weber*
An amphiphilic iron(II) spin crossover complex with relatively short octyl chains was synthesised and the
crystal structures of the high spin and the low spin state could be determined. In further reactions, a
second modification of the hexa-coordinated complex and two different penta-coordinated complexes
could be obtained and characterised by X-ray structure analysis. The examples demonstrate an influence
of the alkyl chains on the stoichiometry of the final product. Different arrangements of the alkyl chains
were observed in the crystal packing. Despite those differences, the spin transition of the hexa-
coordinated complexes is always gradual and comparable to that observed in solution.
Received (in Victoria, Australia)
23rd August 2013,
Accepted 15th October 2013
DOI: 10.1039/c3nj00991b
www.rsc.org/njc
with Schiff base-like ligands can self-assemble into lipid layer like
arrangements.11,12 Additionally, an influence of the alkyl
Introduction
Since the discovery of the spin crossover (SCO) phenomenon in chains on the spin transition behaviour is possible. A rearran-
1931,1 a vast amount of compounds showing spin transition gement of the alkyl chains could trigger the spin transition or
with a huge variety of ligands and metal centres have been vice versa.13 Thus the modification of the ligand could help to
synthesised.2,3 In recent times, the interest in adding multi- increase the cooperativity of spin crossover complexes based on
functionality to these compounds has increased so that they the Schiff base-like ligand system used in our group. In a crystal
exhibit not only spin crossover but also additional functionalities engineering approach we want to study the influence of these
like liquid crystallinity,4 gel formation5 or magnetic exchange alkyloxy substituents as a structure determining element on the
interactions,6 just to mention three examples. Also the nano- packing of the complexes in the crystal and therefore on the
structuring of spin crossover compounds, either by the synthesis SCO behaviour. So far, lipid layer like structures are the only
of nanoparticles7 or through patterning methods,8 is a recent motif observed for such complexes with dodecyl or hexadecyl
field of research that is an important step towards possible alkyl chains.11,12 Thus the question arises if other structural
future applications.
motifs are possible and to what extent the alkyl chains influence
For cooperative SCO materials intermolecular interactions the crystal packing and by this the magnetic properties. Here we
like hydrogen bonds, p–p interactions or also van der Waals present X-ray structures and magnetic properties of complexes
(vdW) interactions are of utmost importance. Studies on compounds with comparatively short octyl chains.
exhibiting huge hysteresis showed that particularly hydrogen bonds
are related to strong cooperative effects leading to the hysteresis
phenomenon.9 This work is focused on the influence of vdW
interactions on the spin crossover behaviour. The addition of long
Synthesis and general characterisation
alkyl chains to the periphery of the ligand adds a new functionality The synthesis of the new amphiphilic ligand, visualised in
to the spin crossover system by the generation of amphiphilic Scheme 1, is realised in a four step reaction. In the ESI,†
molecules.10 In previous work we did show that such complexes Fig. S1, the NMR spectrum of the free ligand H2L is displayed
with the signal assignment. Conversion with iron(II)-acetate14
results in the hexa-coordinated complex [Fe(L)(MeOH)2] with two
methanol as axial ligands. Replacement of the methanol mole-
cules with 4-dimethylaminopyridine (dmap) leads, depending
on the exact reaction conditions, to the desired compound
[Fe(L)(dmap)2] (1) (with varying amounts of included methanol
molecules) or the penta-coordinated complex [Fe(L)(dmap)] (2). In
a first approach a 30-fold excess of dmap was used with the aim of
¨
¨
Inorganic Chemistry II, Universitat Bayreuth, Universitatsstrabe 30, NW 1,
95440 Bayreuth, Germany. E-mail: weber@uni-bayreuth.de; Fax: +49-92155-2157;
Tel: +49-92155-2555
† Electronic supplementary information (ESI) available: 1H NMR spectrum of
H2L, magnetic susceptibility data for the iron complexes discussed in the work
and parameters used in the X-ray structure determination. CCDC 944434–944438.
For ESI and crystallographic data in CIF or other electronic format see DOI:
10.1039/c3nj00991b
This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2014
New J. Chem., 2014, 38, 1965--1972 | 1965

View Article Online
Paper
NJC
Scheme 1 General procedure for the synthesis of the new amphiphilic
Schiff base-like ligand and its iron(II) complexes.
obtaining an octahedral complex. Indeed, the desired complex
[Fe(L)(dmap)2]ꢀMeOH (1a) could be isolated. However, slight
variations in the reaction conditions (reaction time, temperature
of precipitation), keeping the stoichiometry constant, resulted in
the isolation of the penta-coordinated complex [Fe(L)(dmap)] (2).
Here two different samples (2a and 2b), both with the same
composition but differences in the relative orientation of the alkyl
chains, were obtained. This was unexpected as the characteristic
colour change of the solution upon cooling with liquid nitrogen
suggested the presence of a hexa-coordinated species in solution.
Due to the contrasting results further syntheses with a different
excess of the axial ligand were carried out to obtain a clear
synthetic protocol for the compounds 1 and 2. The excess of the
axial dmap ligand was systematically varied and stoichiometries
of 1: 20, 30, 50, 70, 90 and 110 were used. It turned out that the
penta-coordinated product (2) was obtained selectively with a
20 fold excess. By taking 30 equivalents of the axial ligand, it is
difficult to predict if the penta- or hexa-coordinated complex
will be obtained. In our synthetic approaches a 50 : 50 ratio
between the two possibilities was reached. But the system can
be forced to precipitate hexa-coordinated when a 50 fold excess
or higher is used. In this frame, another modification of the
hexa-coordinated complex, 1b ([Fe(L)(dmap)2]ꢀ1.5MeOH), was
obtained, that could be characterised by single crystal X-ray
structure analysis. For the approaches with a higher excess of
dmap, fine crystalline samples with additionally included
dmap/MeOH molecules were obtained.
Fig. 1 ORTEP drawing of 1a at 273 K (HS) (top) and 133 K (LS) (bottom).
Ellipsoids are drawn at the 50% probability level. Hydrogen atoms and
methanol molecules are omitted for clarity.
was low (high Rint) thus only the conformation of the complex
and the relative orientation of the molecules in the crystal
packing can be presented. In the case of 2a the disorder of the
alkyl chains could not be solved satisfactorily (due to the low
quality of the dataset and twinning of the crystal) thus only
the conformation of the molecule is presented. Attempts to
reproduce the crystals to obtain diffraction data of higher quality
led to the sample 2b. In the ESI,† Table S1, the crystallographic
data are given.
1a precipitated in the form of platelet-like crystals out of a
black solution with a 30 fold excess of dmap. The crystal
structure was determined at 273 K and 133 K, which correspond
to the high spin (HS) and the low spin (LS) state of the system
(see magnetic measurements). In Fig. 1 the asymmetric unit of
1a in the HS and the LS state is displayed. Selected bond lengths
and angles are summarised in Table 1.
The compounds were characterised by elemental analysis,
IR and mass spectrometry, magnetic measurements and, if
possible, X-ray structure analysis.
Description of the X-ray structures
%
Single crystals suitable for X-ray structure analysis were
1a crystallises in the triclinic space group P1 that does not
obtained for two hexa-coordinated (1a and 1b) and two penta- change upon spin transition. The bond lengths and angles within
coordinated (2a and 2b) samples. The quality of the data of 1b the first coordination sphere, Fe–Neq/Fe–Oeq, have an average
1966 | New J. Chem., 2014, 38, 1965--1972
This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2014

View Article Online
NJC
Paper
Table 1 Selected bond lengths/Å and angles/1 within the inner coordination sphere of 1a (LS) and 1a (HS), 1b, 2a and 2b
Fe–Neq/Å
Fe–Oeq/Å
Fe–Nax/Å
O–Fe–O/1
1a (LS)
1a (HS)
1b
1.909(2)/1.915(2)
2.065(3)/2.080(3)
1.9
1.948(1)/1.957(1)
2.002(3)/2.012(3)
2.0
2.013(2)/2.018(2)
2.216(3)/2.217(3)
2.0
2.1
91.68(6)
107.16(11)
92
2a
2.1
2.0
100
2b
2.073(4)/2.085(4)
2.002(3)/1.980(3)
2.127(4)
101.62(14)
of 2.07/2.01 Å (HS) and 1.91/1.95 Å (LS) in the region expected
for this ligand system with a bond length change of about
5% upon spin transition.15,18 The average Fe–N distances to the
axially attached ligands (2.22 Å (HS) and 2.02 Å (LS)) change by
about 10% (see Table 1) due to the higher flexibility of the axial
ligands. The O–Fe–O angle is 91.71 (LS) and 107.21 (HS) in the
expected region for complexes of this type of Schiff base-like
ligands.15,18 The change of the unit cell volume is DV/V = 4.0% at
the lower limit of what is expected for an iron(II) spin crossover
complex,3 but the value is higher than the one previously
reported for an SCO complex with C16-alkyl chains.12
One of the alkyl chains and the methanol molecule are
disordered in the LS state. Due to increase of thermal motion
of the atoms at higher temperatures, this disorder cannot be
solved in the HS state. The relative orientation of the aromatic
planes of the dmap to each other changes marginally from
89.31 to 88.11, and the Nax–Fe–Nax angle from 175.71 to 174.91
upon switching from the LS to the HS state, thus the axial
ligands are nearly perpendicular in both spin states.
1b precipitated as spicular crystals, space group P21/c, from
the synthetic approach with 50 fold excess of dmap and the
X-ray structure was determined at 133 K. Due to insufficient
quality of the data one can only talk about a motif. Average
values of selected bond lengths and angles are summarised in
Table 1.
Fig. 3 Molecular packing of 1a (LS) along [100]. Hydrogen bonds are
drawn as dashed lines. Disorder omitted for clarity.
should be in the LS state according to the magnetic measure-
ments. Indeed, the bond lengths and angles within the first
coordination sphere are very similar to those of 1a in the LS state.
The average values are 1.9 Å/2.0 Å for Fe–Neq/Fe–Oeq and 2.0 Å
for Fe–Nax. The O–Fe–O angle is about 921, the Nax–Fe–Nax angle
is 1741 and the dmap rings are twisted towards each other by an
angle of around 941 and are therefore also nearly perpendi-
cular. The main difference between the two samples lies in the
additional half of the solvent molecule, the conformation
(orientation of the alkyloxy chains) and the packing of the
molecules in the crystal.
Fig. 2 shows the asymmetric unit of 1b. At the temperature
used for the determination of the X-ray structure the complex
In the molecular packing of 1a, displayed in Fig. 3 for the LS
state, the disordered methanol molecule forms a hydrogen
bond with the O5 atom in the outer periphery of the equatorial
ligand. This hydrogen bond is weakened in the HS state. A few
further short contacts (more than 0.2 Å shorter than the sum of
the vdW radii) are observed, which are given in Table 2. The
major distinction between the molecular packing of 1a com-
pared to the previously discussed structures is the absence of a
lipid layer like structure. Instead, the axial dmap ligand of one
complex is ‘‘embraced’’ from the C8 chains of the neighbouring
complex. For the previously published structures the distance
between the alkyl chains are between 0.3 and 0.4 Å longer than
Fig. 2 ORTEP drawing of 1b. Ellipsoids are shown at the 50% probability
level. Hydrogen atoms and methanol molecules are omitted for clarity.
This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2014
New J. Chem., 2014, 38, 1965--1972 | 1967

View Article Online
Paper
NJC
Table 2 Short contacts and hydrogen bonds/Å and the corresponding
angle/1 of the obtained crystal structures
D–Hꢀ ꢀ ꢀA
D–H Hꢀ ꢀ ꢀA Dꢀ ꢀ ꢀA
D–Hꢀ ꢀ ꢀA
1a (LS)
C47–H47Bꢀ ꢀ ꢀO7a
0.98
2.60
2.637
2.57
1.92
2.606
2.04
3.436(3)
3.519(8)
3.393(3)
2.760(3)
3.550(6)
2.785(14) 151
3.347(7)
143
151
142
174
165
C29A–H29Bꢀ ꢀ ꢀC42b 0.97
C41–H41Cꢀ ꢀ ꢀO2c
O31B–H31Fꢀ ꢀ ꢀO5d
0.98
0.84
0.96
0.82
0.98
1a (HS) C32–H32Bꢀ ꢀ ꢀC6e
O9A–H9Aꢀ ꢀ ꢀO5 f
2b
C23–H23Cꢀ ꢀ ꢀO4g
2.400
162
a
b
c
d
e
1 + x, y, z. 1 ꢁ x, 1 – y, ꢁ z. 1 ꢁ x, 2 ꢁ y, 1 ꢁ z. x, y, z. 1 + x, y, z.
f
g
1 ꢁ x, 1 ꢁ y, 1 ꢁ z. –1 + x, y, z.
the sum of the vdW radii, indicative of stabilising interactions
between the chains.11,12 For 1a in the LS state several short
contacts are observed between the alkyl chains, and the alkyl
chains and the embraced dmap, respectively. However, the
contacts are too short to be considered for a stabilising effect.
In the HS state they are a bit longer. The strength of such
stabilising interactions (London dispersion forces) depends on
the length of the alkyl chains. Obviously, for the complex
discussed here, the C8 alkyl chains are not long enough to
form lipid layer like structures.
The packing of the molecules in the crystal of 1b is very
different to that of 1a (Fig. 4). Two molecules form pairs where the
alkyl chains are arranged such that stabilising vdW interactions can
be considered. The next pair is rotated by 901 with the two axial
dmap ligands pointing together. As only a structural motif is
obtained, no intermolecular contacts can be discussed. It can,
however, be pointed out that, as for 1a, no lipid layer like arrange-
ment of the complex molecules is obtained.
Fig. 4 Molecular packing of 1b along [100]. Methanol molecules are
omitted for clarity.
The two penta-coordinated complexes crystallise in the
%
space group P21/c (2a) and P1 (2b), respectively. In Fig. 5 ORTEP
drawings of the asymmetric units of 2a and 2b are displayed. In
Table 1 selected bond lengths and angles within the first
coordination sphere are given. For these two samples, no addi-
tional methanol molecules are included in the crystal packing.
Due to insufficient quality of the data of the spicular crystals
of 2a, only the conformation of the molecule is discussed in
comparison to that of 2b.
The average bond lengths of the inner coordination spheres
of the two penta-coordinated species are about 2.1 Å (Fe–Neq
)
and 2.0 Å (Fe–Oeq) very similar to the lengths of the HS
structure of 1a and in the same order of magnitude as observed
for other penta-coordinated complexes of this ligand system.16,17
The Fe–Nax bond lengths are about 0.1 Å shorter than in 1a (HS),
due to the penta-coordination. The O–Fe–O angles are 1001 in
the region expected for complexes with this kind of Schiff base-
like ligands and are between the values of the HS and the LS
state, respectively.16,17
Fig. 5 ORTEP drawing of 2a (left) and 2b (right). Ellipsoids are shown at
the 50% probability level. Hydrogen atoms are omitted for clarity.
In Fig. 6 the molecular packing in the unit cell of 2b is displayed.
In this case, pairs are built where the almost planar planes of the
In the case of 2a the C8 alkyl chains are spread widely out equatorial ligand including the iron(II) (the iron-N1N2O1O2 plane
with an angle of almost 901 between the two chains. In contrast to distance is 0.36 Å) are stacked above each other and the C8 alkyl
this, for 2b they are arranged parallel to each other, similar to 1b. chains are bent in the direction of the axial dmap of the
However, differently to 1b, in 2b the alkyl chains are not in plane neighbouring molecule. A short contact with a distance of
with the axial ligand but bent by almost 901 in the direction of the 0.2 Å smaller than the sum of the vdW radii connects one
axial dmap ligand.
molecule at the atom H23C of the dmap ligand with the oxygen
1968 | New J. Chem., 2014, 38, 1965--1972
This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2014

View Article Online
NJC
Paper
Fig. 7 Magnetic measurement of 1a (circles), 1b (triangles) and a methanol
solution of 1 with a 50 fold excess of dmap (open squares) in the
temperature range of 325–10 K.
again after complete removal of the methanol. This is confirmed
by a measurement on freshly prepared crystals (1c) where in the
first cycle an apparent hysteresis is observed (see ESI,† Fig. S3)
that is lost in a second cycle. The transition curve of sample 1a
can be reproduced by displaying a combined curve of the
weighted contributions (50: 50) of 1b and 1c.
The magnetic properties of 1b are very similar to that of 1a.
The spin transition is very gradual and complete and takes
place in the same temperature region. The wMT value at 325 K is
with 3.10 cm3 K mol–1 comparable to that of 1a. Decreasing of
the temperature leads to a gradual decrease of the wMT product
until 120 K (0.31 cm3 K mol–1) where a small plateau can be
observed. At 100 K the compound is completely in the LS state
(wMT = 0.06 cm3 K mol–1). Although methanol molecules are
included in the crystal packing of 1b, they are strongly disordered
and are, apparently, not involved in cooperative interactions as a
very gradual spin transition is observed.
Fig. 6 Molecular packing of 2b along [100].
atom O4 of the outer periphery of the ligand of the neighbouring
molecule. The dmap ligands are arranged parallel to each other
and also almost parallel to the alkyl chains. The same is observed
for the alkyl chains that are themselves arranged parallel to
each other. That means that the alkyl chains of one molecule
are again interacting with the chains of another molecule whose
axial ligand points to the opposite direction so that they form a
lipid layer like arrangement. The distances between the alkyl chains
support the idea of stabilising vdW interactions. Additionally, the
distances between the stacked planes of the equatorial ligand
(about 3.5 Å) suggest stabilising p–p-interactions.
Magnetic measurements
Temperature dependent magnetic measurements in the 325–10 K
In Fig. S2, ESI,† the spin transition curves of the products of
range were performed for all complexes discussed in this work. the different synthetic approaches are compared. The penta-
Additionally, the temperature dependent magnetic susceptibility coordinated samples (2a, 2b and 2 from the approach with
was determined in a methanol solution of the iron complex 1 with 20 fold excess of dmap) remain as expected in the HS state. The
a 50 fold excess of dmap. The concentration of the complex in the wMT vs. T plot of 2 is shown as a typical example. All samples,
solution is ca. 14.5 mg mLꢁ1. The presence of the octahedral where 50 equivalents of dmap and more was used, show the
complex [Fe(L)(dmap)2] in solution was confirmed by the colour same kind of spin transition as 1b, independent of additional
change upon cooling due to the spin transition. This is illustrated solvent or dmap molecules in the crystal packing. This is
in the ESI,† Fig. S4. In Fig. 7 the results for 1a, 1b and for the unexpected as for spin transition compounds often significant
methanol solution of the complex are given. At 325 K 1a is in the changes in the spin transition behaviour are observed, if the
high spin state, with a wMT value of 3.15 cm3 K molꢁ1
.
crystal packing is slightly modified or additional solvent mole-
Upon cooling a gradual decrease of the wMT product down to cules are included. The behaviour observed for the different
2.10 cm3 K molꢁ1 at 255 K is observed, where a small plateau is samples of 1 indicates the total absence of cooperative interactions,
visible. Further cooling causes a slow drop of the wMT product and thus the spin transition should be comparable to that of the
until the compound is in the low spin state at about 100 K complex in solution. As can be seen in Fig. 7, indeed the transition
(0.17 cm3 K molꢁ1). The plateau is due to a mixture of powder curves are almost identical. Only the transition temperature is
and crystalline parts in the sample used for the magnetic shifted to slightly higher temperatures in solution.
measurements. According to X-ray structure analysis the crystals
It was already shown that abrupt ST is realisable with
contain disordered methanol molecules. Results from CHN compounds that bear lipid layer like arrangements.12 The
analysis indicate the absence of additional methanol molecules increase of vdW interactions influences the packing in the
in the fine crystalline bulk material. The step and the small crystal through the formation of lipid layer like arrangements
hysteresis disappear completely if the compound is measured and by this cooperative effects like a network of hydrogen
This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2014
New J. Chem., 2014, 38, 1965--1972 | 1969

View Article Online
Paper
NJC
bonds between the polar groups can be enhanced. For the 90 min. After storing the reaction mixture at 5 1C overnight, the
complexes described here, no layered structure can be achieved precipitate was collected, washed with ethanol and recrystallised
because of the relatively short C8 alkyl chains. Additionally, the from 35 mL of ethanol. The bright yellow fine crystalline ligand was
molecule is very bulky and the volume change upon spin dried in air. Yield: 3.1 g (87%). Elem. anal. calcd for C34H52N2O8
transition relative to the overall volume is relatively small and (616.79 g molꢁ1): C 66.21, H 8.50, N 4.54; found: C 66.64, H 9.28,
almost in the region of the thermal contraction. Therefore, N 4.64. MS (DEI+): m/z (%) = 616 (100) [M]+, 584 (62), 552 (61) 501
cooperativity is decreased in the crystal and only a gradual SCO (54). 1H NMR (299.86 MHz, CDCl3, 298 K): d = 0.90 (t, J = 6.8 Hz, 6H,
as in solution is observed. This is the reason why the spin CH3), 1.30–1.40 (m, 16H, CH2), 1.45–1.55 (m, 4H, CH2), 1.80–1.89 (m,
transition is independent of the conformation of the molecule 4H, CH2), 2.57 (s, 6H, CH3), 3.79 (s, 6H, CH3), 4.03 (t, 4H, J = 6.5 Hz,
and the crystal packing. A similar effect was recently described CH2O), 6.76 (s, 2H, Har), 8.28 (d, 2H, J = 12.5 Hz, CHQ), 12.90 (d, 2H,
for the nanostructuring of mononuclear complexes.19
J = 12.5 Hz, NH) ppm. 13C NMR (75.40 MHz, CDCl3, 298 K): d = 14.3
(CH3), 22.9 (CH2), 26.2 (CH2), 29.4 (CH2), 29.5 (CH2), 29.6 (CH2) 31.3
(CH3), 32.1 (CH2), 45.0 (Cq), 51.5 (CH3), 70.3 (CH2, CH2O), 103.7 (Cq),
106.7 (CH, Car), 125.0 (Cq, Car–N), 148.8 (Cq, Car–O), 154.3 (CH), 167.4
(O–CQO), 200.4 (CQO) ppm. IR: n~ = 1699 s, 1612 vs, 1521 m,
1420 m, 1243 vs, 1189 vs, 1081 vs.
[Fe(L)(MeOH)2]. 1.5 g (2.43 mmol) of H2L and 0.85 g of
(4.86 mmol, 2 eq.) iron(II) acetate were dissolved in 100 mL of
methanol and the brown solution was heated to reflux for 1 h.
After cooling to room temperature, the brown precipitate was
collected, washed twice with 10 mL of methanol and dried in
vacuum. Yield: 1.35 g (76%). Elem. anal. calcd for C36H58FeN2O10
(734.80 g molꢁ1): C 58.85, H 7.96, N 3.81; found: C 59.29, H 7.86,
N 4.16. IR: n~ = 2925 m, 2854 w, 1706 m, 1577 s, 1506 w, 1429 s,
1384 s, 1258 vs, 1214 s, 1069 vs, 998 m, 841 m, 769 m.
[Fe(L)(dmap)2] (1/1a). 0.27 g (0.37 mmol) of [Fe(L)(MeOH)2]
and 1.35 g of (11.05 mmol, 30 eq.) dmap were dissolved in
15 mL of methanol and heated to reflux for 70 min. After storing
the solution at 5 1C for 14 d, greenish-black crystals (1a) were filtered
and dried in vacuum. Yield: 0.01 g (3%, crystals), 0.15 g (44%, fine
crystalline powder, 1/1a) Elem. anal. calcd for C48H70FeN6O8 (fine
crystalline powder, no methanol included) (914.95 g molꢁ1):
C 63.01, H 7.71, N 9.19; found: C 62.67, H 7.78, N 9.59.
[Fe(L)(dmap)] (2a). 0.25 g of (0.34 mmol) [Fe(L)(MeOH)2] and
1.25 g of (10.21 mmol, 30 eq.) dmap were dissolved in 10 mL of
methanol and heated to reflux for 60 min. After 1 d at room
temperature the reaction mixture was stored at ꢁ30 1C for 3 d.
Black crystals that precipitated were filtered, washed with
2.5 mL of methanol and dried in vacuum. Yield: 0.13 g (48%).
Elem. anal. calcd for C41H60FeN4O8 (792.78 g molꢁ1): C 62.12,
H 7.63, N 7.07; found: C 61.79, H 7.44, N 7.30.
Conclusions
In this article, several molecular setups and arrangements of
hexa- and penta-coordinated amphiphilic iron(II) complexes 1a
(HS), 1a (LS), 1b, 2a and 2b with dimethylaminopyridine as
axial ligands were investigated with the help of X-ray structure
analysis and magnetic measurements. The complex can precipitate
hexa-coordinated as well as penta-coordinated depending on the
excess of the axial ligand consumed. With 30 equivalents of dmap
the system can crystallise in both modifications, below this value it
is penta-coordinated and above, it can be forced to crystallise hexa-
coordinated. In the case of the octahedral complexes 1a and 1b, a
gradual spin crossover can be observed starting at about 325 K in
the high spin state and ending at about 125 K in the low spin state.
It should be pointed out that the spin transition is always the same
despite the significant differences in the composition and crystal
packing of the different samples. The crystal structures show no
lipid layer like arrangement due to the relatively short C8 alkyl
chains. In the crystal packing of the penta-coordinated compound
vdW interactions between the C8 alkyl chains are observed.
This in combination with the p–p-interactions could be the
reason for the complex to precipitate penta-coordinated up to a
relatively high excess of the used dmap ligand. The gradual
spin transition behaviour can be explained with missing lipid
layer like arrangement and the absence of other factors that are
responsible for cooperative interactions. Thus the same spin
transition as in solution is observed.
[Fe(L)(dmap)] (2b). 0.18 g (0.25 mmol) of [Fe(L)(MeOH)2]
and 0.90 g of (7.37 mmol, 30 eq.) dmap were dissolved in 15 mL
of methanol and heated to reflux for 60 min. After 5 weeks
Experimental section
Synthesis
The synthesis of the iron complexes was carried out under an at room temperature, black needles that precipitated were
argon atmosphere using Schlenk tube techniques. The solvents filtrated and dried in vacuum. Yield: 0.05 g (22%). C41H60FeN4O8
therefore were purified as described in the literature20 and distilled (792.78 g molꢁ1). Elem. anal. was not possible due to the insuffi-
under an atmosphere of argon. The precursors methoxymethylene- cient amount of product. IR: n~ = 2923 m, 2853 w, 1693 m, 1579 s,
methylacetoacetate,21 iron(II) acetate,14 1,2-dioctyloxybenzene, 1433 s, 1387 s, 1255 vs, 1212 vs, 1065 vs, 1008 s, 804 m, 768 m.
1,2-dinitro-4,5-dioctyloxybenzene, 1,2-diamino-4,5-dioctyloxy-
benzene22 were synthesised as described.
[Fe(L)(dmap)]. (20 eq.) 0.25 g (0.34 mmol) of [Fe(L)(MeOH)2]
and 0.83 g of (6.8 mmol, 20 eq.) dmap were dissolved in 15 mL
(E,E)[{dimethyl-2,20[4,5-dioctyloxy-1,2-phenylenebis(iminomethyl- of methanol and heated to reflux for 60 min. After 1 d at room
idyne)]bis-3-oxobutanato}] (H2L). Under argon, 2.1 g (5.76 mmol) of temperature, black crystals and brown powder were filtered and
1,2-diamino-4,5-dioctyloxybenzene and 2.4 g of (15.18 mmol, 2.6 eq.) dried in vacuum. Yield: 0.12 g (44%). Elem. anal. calcd for
methoxymethylenemethylacetoacetate were dissolved in 60 mL of
degassed ethanol and the yellow solution was heated to reflux for C 61.80, H 7.81, N 7.21.
C
41H60FeN4O8 (792.78 g molꢁ1): C 62.12, H 7.63, N 7.07; found:
1970 | New J. Chem., 2014, 38, 1965--1972
This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2014

View Article Online
NJC
Paper
[Fe(L)(dmap)2]. (30 eq.) 0.25 g of (0.34 mmol) [Fe(L)(MeOH)2] Mo-Ka radiation. The data were corrected for Lorentz and
and 1.25 g (10.2 mmol, 30 eq.) of dmap were dissolved in 15 mL polarisation effects. 1a (LS) 1b, 2a and 2b (Sir97),23 1a (HS)
of methanol and heated to reflux for 60 min. After 1 d at room (SHELXS-97)24 were solved by direct methods and refined by full-
2
temperature and 10 d at 6 1C, black crystals and brown powder matrix least-squares techniques against F0 (SHELXL-97).24 The
were filtered and dried in vacuum. Yield: 0.16 g (51%). Elem. hydrogen atoms were included at calculated positions with fixed
anal. calcd for C48H70FeN6O8 (914.95 g molꢁ1): C 63.01, H 7.71, displacement parameters, allowed to ride on their parent atoms.
N 9.19; found: C 62.77, H 8.15, N 9.30.
If not noted differently, for methyl groups and hydroxyl groups
[Fe(L)(dmap)2] (1b). (50 eq.) 0.27
g
of (0.37 mmol) the torsion angles were allowed to be refined according to the
[Fe(L)(MeOH)2] and 2.25 g (18.4 mmol, 50 eq.) of dmap were electron density. For the hydroxyl groups O99–H9A (1a), O98–
dissolved in 15 mL of methanol and heated to reflux for 70 min. H98 and O99–H99 (both 1b) no stable refinement was achieved
After 1 d at room temperature, 16 d at 6 1C and 1 d at ꢁ30 1C, thus idealised torsion angles were used. All non-hydrogen atoms
black crystals were filtrated, washed with 3 mL of methanol and were refined anisotropically. Due to bad quality of the data of 1b
dried in vacuum. Yield: 0.05 g (15%). Elem. anal. calcd for (bad Rint) and 2a only the general molecular setup could be
C
48H70FeN6O8 (no methanol included) (914.95 g molꢁ1): investigated. For 2a, twin refinement was conducted based on
C 63.01, H 7.71, N 9.19; found: C 62.87, H 8.03, N 9.66.
[Fe(L)(dmap)2]. (70 eq.) 0.25 g of (0.34 mmol) [Fe(L)(MeOH)2]
and 2.91 g (23.8 mmol, 70 eq.) of dmap were dissolved in 15 mL
of methanol and heated to reflux for 60 min. After 1 d at room
temperature, 14 d at 6 1C and 13 d at ꢁ30 1C, black crystals were
filtrated and dried in vacuum. Yield: 0.15 g (48%). Elem. anal.
calcd for C48H70FeN6O8 ꢂ dmap (1037.12 g molꢁ1): C 63.69,
H 7.77, N 10.80; found: C 63.79, H 8.39, N 10.87.
[Fe(L)(dmap)2]. (90 eq.) 0.25 g of (0.34 mmol) [Fe(L)(MeOH)2]
and 3.74 g (30.6 mmol, 90 eq.) of dmap were dissolved in 15 mL
of methanol and heated to reflux for 60 min. After 1 d at room
temperature, 14 d at 6 1C and 5 d at ꢁ30 1C, black crystals were
filtered and dried in vacuum. Yield: 0.16 g (51%). Elem. anal.
calcd for C48H70FeN6O8 ꢂ 2dmap ꢂ 0.5MeOH (1175.31 g molꢁ1):
C 63.87, H 7.89, N 11.92; found: C 64.06, H 8.24, N 11.68.
the twin law
0
1
ꢁ1:000 0:000 0:000
0:000 ꢁ1:000 0:000
B
B
C
C
B
@
C
A
0:369
0:000 1:000
found by PLATON.25 ORTEP-III26 was used for the structure
representation, Schakal-9927 and Mercury28 for the representa-
tion of the molecule packing.
Acknowledgements
We thank C. Lochenie, P. Thoma and J. Obenauf (University of
Bayreuth) for the collection of the X-ray data and W. Milius
for professional support. For financial support we thank the
University of Bayreuth and the Deutsche Forschungsgemeinschaft
(SFB 840/A10).
[Fe(L)(dmap)2]. (110 eq.) 0.25
g
of (0.34 mmol)
[Fe(L)(MeOH)2] and 4.57 g (37.4 mmol, 110 eq.) of dmap were
dissolved in 15 mL of methanol and heated to reflux for 60 min.
After 1 d at room temperature, 10 d at 6 1C and 4 d at ꢁ30 1C,
black fine crystalline powder was filtered and dried in vacuum.
Elem. anal. calcd for C48H70FeN6O8 ꢂ 2dmap ꢂ 0.5MeOH
(1175.31 g molꢁ1): C 63.87, H 7.89, N 11.92; found C 64.05,
H 8.63, N 11.82.
Notes and references
¨
1 L. Cambi and L. Szego, Ber. Dtsch. Chem. Ges. A/B, 1933, 66,
656–661.
Magnetic measurements
2 Spin-Crossover Materials, ed. M. A. Halcrow, John Wiley &
Sons Ltd, Oxford, UK, 2013.
3 Spin Crossover in Transition Metal Compounds I-III, ed.
P. Gu¨tlich and H. Goodwin, Springer Berlin/Heidelberg,
2004, pp. 233–235.
4 (a) S. Hayami, Y. Komatsu, T. Shimizu, H. Kamihata and
Y. H. Lee, Coord. Chem. Rev., 2011, 255, 1981–1990;
(b) C. Gandolfi, T. Cotting, P. N. Martinho, O. Sereda,
A. Neels, G. G. Morgan and M. Albrecht, Dalton Trans.,
2011, 40, 1855–1865; (c) P. N. Martinho, C. J. Harding,
H. Mu¨ller-Bunz, M. Albrecht and G. G. Morgan, Eur. J. Inorg.
Chem., 2010, 675–679; (d) Y. Komatsu, K. Kato, Y. Yamamoto,
H. Kamihata, Y. H. Lee, A. Fuyuhiro, S. Kawata and
S. Hayami, Eur. J. Inorg. Chem., 2012, 2769–2775;
Magnetic measurements on the bulk materials were carried out
using a SQUID MPMS-XL5 from Quantum Design with an
applied field of 1000, 2000 and 5000 G, respectively, and in
the temperature range from 325 to 10 K in the sweep and settle
modes. The sample was prepared in a gelatine capsule held in a
plastic straw. The raw data were corrected for the diamagnetic
part of the sample holder and the diamagnetism of the organic
ligand using tabulated Pascal’s constants.
For the measurements in solution the sample was prepared
in the plastic straw and measured in the settle mode with an
applied field of 20 000 G. The raw data were corrected for the
diamagnetism of the solution and the diamagnetism of the
organic ligand using tabulated Pascal’s constants.
¨
(e) Y. Bodenthin, U. Pietsch, H. Mohwald and D. G. Kurth,
X-ray diffraction
J. Am. Chem. Soc., 2005, 127, 3110–3114; ( f ) A. B. Gaspar,
M. Seredyuk and P. Gu¨tlich, Coord. Chem. Rev., 2009, 253,
2399–2413.
The intensity data of 1a (LS), 1a (HS), 1b, 2a and 2b were collected
using a Stoe IPDS II diffractometer using graphite-monochromated
This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2014
New J. Chem., 2014, 38, 1965--1972 | 1971

View Article Online
Paper
NJC
5 P. Grondin, O. Roubeau, M. Castro, H. Saadaoui, A. Colin 15 B. Weber, Coord. Chem. Rev., 2009, 253, 2432–2449.
´
¨
16 B. Weber and E.-G. Jager, Eur. J. Inorg. Chem., 2009, 465–477.
and R. Clerac, Langmuir, 2010, 26, 5184–5195.
6 A. Gaspar, M. Seredyuk and P. Gu¨tlich, J. Mol. Struct., 2009, 17 S. Schlamp, J. Schulten, R. Betz, T. Bauch, A. V. Mudring
924–926, 9–19. and B. Weber, Z. Anorg. Allg. Chem., 2012, 638, 1093–1102.
7 (a) S. M. Neville, C. Etrillard, S. Asthana and J.-F. Letard, Eur. 18 (a) B. Weber, E. Kaps, J. Weigand, C. Carbonera, J.-F. Letard,
´
´
´
J. Inorg. Chem., 2010, 282–288; (b) V. Martınez, I. Boldog,
K. Achterhold and F. G. Parak, Inorg. Chem., 2008, 47,
487–496; (b) B. Weber, J. Obel, D. Henner-Vasquez and
W. Bauer, Eur. J. Inorg. Chem., 2009, 5527–5534.
A. B. Gaspar, V. Ksenofontov, A. Bhattacharjee, P. Gu¨tlich
and J. A. Real, Chem. Mater., 2010, 22, 4271–4281;
´
´
(c) J. R. Galan-Mascaros, E. Coronado, A. Forment-Aliaga, 19 A. Tissot, L. Rechignat, A. Bousseksou and M.-L. Boillot,
M. Monrabal-Capilla, E. Pinilla-Cienfuegos and M. Ceolin,
J. Mater. Chem., 2012, 22, 3411.
Inorg. Chem., 2010, 49, 5706–5714.
20 H. G. O. Becker, Organikum. Organisch-chemisches Grund-
praktikum, Johann Ambrosius Barth, Berlin, 19th edn, 1993.
´
8 (a) C. Thibault, G. Molnar, L. Salmon, A. Bousseksou and
C. Vieu, Langmuir, 2010, 26, 1557–1560; (b) M. Cavallini, 21 W. Bauer, T. Ossiander and B. Weber, Z. Naturforsch., B:
I. Bergenti, S. Milita, G. Ruani, I. Salitros, Z.-R. Qu, Chem. Sci., 2010, 323–328.
R. Chandrasekar and M. Ruben, Angew. Chem., 2008, 120, 22 (a) M. J. Howard, F. R. Heirtzler and S. I. G. Dias, J. Org.
8724–8728; (c) C. M. Quintero, I. A. Gural’skiy, L. Salmon,
G. Molnar, C. Bergaud and A. Bousseksou, J. Mater. Chem.,
2012, 22, 3745–3751; (d) A. D. Naik, L. Stappers, J. Snauwaert,
J. Fransaer and Y. Garcia, Small, 2010, 6, 2842–2846.
Chem., 2008, 73, 2548–2553; (b) W. M. Lauer, C. Rondestvedt,
R. T. Arnold, N. L. Drake, J. van Hook and J. Tinker, J. Am.
¨
Chem. Soc., 1946, 68, 1546–1548; (c) M. Schlogl and B. Rieger,
Z. Naturforsch., B: Chem. Sci., 2004, 233–240; (d) K. Tahara,
S. Furukawa, H. Uji-i, T. Uchino, T. Ichikawa, J. Zhang,
W. Mamdouh, M. Sonoda, F. C. de Schryver, S. de Feyter
and Y. Tobe, J. Am. Chem. Soc., 2006, 128, 16613–16625.
23 A. Altomare, M. C. Burla, M. Camalli, G. L. Cascarano,
C. Giacovazzo, A. Guagliardi, A. G. G. Moliterni, G. Polidori
and R. Spagna, J. Appl. Crystallogr., 1999, 32, 115–119.
ˆ
9 (a) B. Weber, W. Bauer, T. Pfaffeneder, M. M. Dırtu,
A. D. Naik, A. Rotaru and Y. Garcia, Eur. J. Inorg. Chem.,
2011, 3193–3206; (b) B. Weber, W. Bauer and J. Obel, Angew.
Chem., Int. Ed., 2008, 47, 10098–10101.
10 (a) M. Seredyuk, A. B. Gaspar, V. Ksenofontov,
Y. Galyametdinov, J. Kusz and P. Gu¨tlich, J. Am. Chem.
Soc., 2008, 130, 1431–1439; (b) M. Seredyuk, A. B. Gaspar, 24 G. Sheldrick, Acta Crystallogr., Sect. A: Fundam. Crystallogr.,
V. Ksenofontov, Y. Galyametdinov, J. Kusz and P. Gu¨tlich,
Adv. Funct. Mater., 2008, 18, 2089–2101.
11 S. Schlamp, P. Thoma, B. Weber and Y. Garcia, Eur. J. Inorg.
Chem., 2012, 2759–2768.
12 S. Schlamp, B. Weber, A. D. Naik and Y. Garcia, Chem.
Commun., 2011, 47, 7152–7154.
2008, 64, 112–122.
25 A. L. Spek, PLATON - A Multipurpose Crystallographic Tool,
Utrecht University, Utrecht, The Netherlands, 2008.
26 (a) C. K. Johnson and M. N. Burnett, ORTEP-III, Oak-Ridge
National Laboratory, Oak-Ridge, TN, 1996; (b) L. Farrugia,
J. Appl. Crystallogr., 1997, 30, 565.
13 S. Hayami, Y. Kojima, D. Urakami, K. Ohta and K. Inoue, 27 E. Keller, Schakal-99, University of Freiburg, Germany, 1999.
Polyhedron, 2009, 28, 2053–2057.
14 B. Weber, R. Betz, W. Bauer and S. Schlamp, Z. Anorg. Allg.
Chem., 2011, 637, 102–107.
28 C. F. Macrae, P. R. Edgington, P. McCabe, E. Pidcock,
G. P. Shilds, R. Taylor, M. Towler and J. van de Streek,
J. Appl. Crystallogr., 2006, 39, 453–457.
1972 | New J. Chem., 2014, 38, 1965--1972
This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2014