The spin-states and spin-crossover behaviour of iron(II) complexes of 2,6-dipyrazol-1-ylpyrazine derivatives

Article abstract of DOI:10.1039/b210368k

The syntheses of [FeL₂]X₂ (L = 2,6-dipyrazol-1- ylpyrazine [L²H], 2,6-bis{3-methylpyrazol-1-yl}pyrazine [L ²Me], 2,6-bis{3,5-dimethylpyrazol-1-yl}pyrazine [L²Me ₂] or 2,6-bis{3-[2,4,6-trimethylphenyl]pyrazol-1-yl}pyrazine [L ²Mes]; X^- = BF₄^- or ClO ₄^-) are described. Solvent-free [Fe(L²H) ₂][BF₄]₂ and [Fe(L²H) ₂][ClO₄]₂ exhibit very similar abrupt spin-state transitions at 223 K and 208 K respectively, which show hysteresis loops of 3-5 K. Powder diffraction measurements afforded related, but not identical, unit cells for these two compounds, and imply that [Fe(L ²H)₂][ClO₄]₂ is isomorphous with [Fe(L¹H)₂][BF₄]₂ (L¹H = 2,6-dipyrazol-1-ylpyridine). The single crystalline solvate [Fe(L ²H)₂][BF₄]₂·3CH ₃NO₂ undergoes a similarly abrupt spin-state transition at 198 K. Polycrystalline [Fe(L²Me)₂][BF₄] ₂ and [Fe(L²Me)₂][ClO₄]₂ are isomorphous with each other and also exhibit spin-state transitions at low temperature, although these are very different in form. In contrast, both salts of [Fe(L²Me₂)₂]²⁺ and [Fe(L ²Mes)₂]²⁺ are fully low-spin at 295 K. Single crystal structures of [Fe(L²Me₂)₂][BF ₄]2·0.5{CH₃}2CO·0.1H₂O and [Fe(L²Mes)₂][BF₄]₂·5CH ₃NO₂ show low-spin complex dications, and imply that [Fe(L²Me₂)₂][BF₄]₂ is low-spin as a result of intra-ligand steric repulsions involving the pyrazole 5-methyl substituents. NMR and UV/vis data in MeCN and MeNO₂ show that the spin states of all four complex dications are similar in solution and the solid state except for [Fe(L²Me₂)₂] ²⁺, which exists as a mixture of high- and low-spin species in these solvents. The Royal Society of Chemistry 2003.

Full text of DOI:10.1039/b210368k

View Article Online / Journal Homepage / Table of Contents for this issue
The spin-states and spin-crossover behaviour of iron(II) complexes
of 2,6-dipyrazol-1-ylpyrazine derivatives†
Jérôme Elhaïk,a Victoria A. Money,b Simon A. Barrett,a Colin A. Kilner,a
Ivana Radosavljevic Evansb and Malcolm A. Halcrow*a
a Department of Chemistry, University of Leeds, Woodhouse Lane, Leeds, UK LS2 9JT.
E-mail: M.A.Halcrow@chem.leeds.ac.uk
b Department of Chemistry, University of Durham, South Road, Durham, UK DH1 3LE
Received 22nd October 2002, Accepted 6th January 2003
First published as an Advance Article on the web 23rd April 2003
The syntheses of [FeL2]X2 (L = 2,6-dipyrazol-1-ylpyrazine [L2H], 2,6-bis{3-methylpyrazol-1-yl}pyrazine [L2Me],
2,6-bis{3,5-dimethylpyrazol-1-yl}pyrazine [L2Me2] or 2,6-bis{3-[2,4,6-trimethylphenyl]pyrazol-1-yl}pyrazine
[L2Mes]; XϪ = BF4Ϫ or ClO4Ϫ) are described. Solvent-free [Fe(L2H)2][BF4]2 and [Fe(L2H)2][ClO4]2 exhibit very
similar abrupt spin-state transitions at 223 K and 208 K respectively, which show hysteresis loops of 3–5 K. Powder
diffraction measurements afforded related, but not identical, unit cells for these two compounds, and imply that
[Fe(L2H)2][ClO4]2 is isomorphous with [Fe(L1H)2][BF4]2 (L1H = 2,6-dipyrazol-1-ylpyridine). The single crystalline
solvate [Fe(L2H)2][BF4]2ؒ3CH3NO2 undergoes a similarly abrupt spin-state transition at 198 K. Polycrystalline
[Fe(L2Me)2][BF4]2 and [Fe(L2Me)2][ClO4]2 are isomorphous with each other and also exhibit spin-state transitions
at low temperature, although these are very different in form. In contrast, both salts of [Fe(L2Me2)2]2ϩ and
[Fe(L2Mes)2]2ϩ are fully low-spin at 295 K. Single crystal structures of [Fe(L2Me2)2][BF4]2ؒ0.5{CH3}2COؒ0.1H2O
and [Fe(L2Mes)2][BF4]2ؒ5CH3NO2 show low-spin complex dications, and imply that [Fe(L2Me2)2][BF4]2 is low-spin
as a result of intra-ligand steric repulsions involving the pyrazole 5-methyl substituents. NMR and UV/vis data in
MeCN and MeNO2 show that the spin states of all four complex dications are similar in solution and the solid
state except for [Fe(L2Me2)2]2ϩ, which exists as a mixture of high- and low-spin species in these solvents.
while three Fe() complexes of L3 represent the only other
Introduction
known Fe() compounds with ligands containing both pyrazine
Although the phenomenon of spin-state transitions in d–d
transition ion complexes has been known since the 1930’s, there
is continuing interest in the discovery of new spin-crossover
compounds and their solid-state physics.1,2 Much of this inter-
est is driven by the potential technological applications for
spin-crossover compounds, in display devices and for data
storage.3 To be useful in this regard, a spin-crossover material
must be genuinely bistable. That is, its spin-state transition must
show a hysteresis loop, so that within the hysteresis loop the
compound can be either high-spin or low-spin, depending on its
history. In addition, the transition should be abrupt, and occur
as close to room temperature as possible. These properties
depend on both the intra- and inter-molecular interactions
within the solid compound,1,2 and are difficult to engineer into
a material. The most common structural motif in known spin-
crossover compounds is of an octahedral Fe() centre with six
N-donor ligands, and a large number of such compounds have
been reported to date.
and pyrazole functionalities.15
We have previously reported that [Fe(L1H)2][BF4]2 under-
goes an abrupt spin-state transition at 261 K, exhibiting a
small hysteresis loop.4–6 The spin states of other complexes
[Fe(L1R)2]2ϩ that we have prepared vary in a predictable way
depending on the inductive properties of the ligand ‘R’ substit-
uents, when those substituents are sterically flat.7,8 We there-
fore decided to introduce hydrogen-bonding functionality into
the [Fe(L1R)2]2ϩ framework, with the aim of increasing the
cooperativity of any spin-state transitions we discovered. With
this in mind, we report here the synthesis and characterisation
of four complexes of type [Fe(L2R)2]2ϩ. The ligand L2H and its
Ag() complex have been reported previously,9 as have some
multi-dentate lanthanide chelators based on the L2R frame-
work.10 We are aware of only a small number of other spin-
crossover Fe() compounds containing pyrazine donors,11–14
Results and discussion
The ligand L2H was prepared by the published method.9 The
three new ligands in this study, L2Me, L2Me2 and L2Mes were
synthesised by an analogous procedure employing the corre-
sponding substituted pyrazole as starting material. The com-
pounds [Fe(L2R)2][BF4]2 (R = H, 1[BF4]2; R = Me, 2[BF4]2;
R = Me2, 3[BF4]2; R = Mes, 4[BF4]2) and [Fe(L2R) ][ClO4]2
2
(R = H, 1[ClO4]2; R = Me, 2[ClO4]2; R = Me2, 3[ClO4]2;
R = Mes, 4[ClO4]2) can be prepared, by complexation of the
relevant hydrated Fe() salt with 2 molar equivalents of L2R
in acetone. Once isolated, all these compounds are insoluble
in acetone, but moderately soluble in MeCN or MeNO2. For
4[BF4]2 and 4[ClO4]2 only, microanalysis consistently form-
ulated these products as monohydrates (hereafter written as
4[BF4]2ؒH2O and 4[ClO4]2ؒH2O). All of the other complexes
were obtained as solvent free, anhydrous solids. We were unable
† Based on the presentation given at Dalton Discussion No. 5, 10–12th
April 2003, Noordwijkerhout, The Netherlands.
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 3
D a l t o n T r a n s . , 2 0 0 3 , 2 0 5 3 – 2 0 6 0
2053

View Article Online
to prepare other hydrated crystals of salts of these compounds,
since they undergo rapid hydrolysis in water or MeOH solution,
precipitating free L2R. A large number of co-crystallisations
from aprotic solvents were also attempted of our [Fe(L2R)2]2ϩ
salts with aliphatic primary diamines or dialcohols, dihydroxy-
benzenes, or aliphatic or aromatic dicarboxylic acids. These
were intended to exploit the hydrogen bond-acceptor capability
of coordinated L2R, to form infinite hydrogen-bonded solid
lattices. However, all of these procedures yielded either
additive-free [Fe(L2R)2]X2 (XϪ = BF4Ϫ, ClO4Ϫ), the pure addi-
tive, or an intractable mixture of species.
Despite being centred at lower temperatures, the forms of the
thermal spin-state transitions shown by 1[BF4]2 and 1[ClO4]2 in
Fig. 1 are strikingly similar to that shown by [Fe(L1H)2][BF4]2,4
which suggested to us that these three compounds might be
isomorphous. Solvent-free single crystals of 1[BF4]2 and
1[ClO4]2 can be grown from acetone. Upon cooling on the
diffractometer, these crystals undergo a sharp colour change
from yellow to brown, at a temperature that is within 2 K of
the spin-state transition temperature derived from powder
magnetic measurements of these compounds (see above).
Unfortunately, however, they diffract X-rays so poorly that we
were unable to measure a unit cell from either compound.
Therefore, we undertook powder diffraction measurements of
these two solids at 295 K. Both compounds gave monoclinic
P unit cells at this temperature, with the following parameters.
For 1[BF4]2: a = 11.886(2), b = 11.878(2), c = 19.093(3) Å,
β = 92.291(9)Њ and V = 2693.4(8) Å3. For 1[ClO4]2: a = 8.694(1),
b = 8.668(1), c = 19.993(3) Å, β = 90.517(1)Њ and V = 1506.5(1)
Å3. For comparison, single crystalline [Fe(L1H)2][BF4]2 has the
following unit cell at 290 K: a = 8.4947(2), b = 8.5070(2), c =
19.0535(6) Å, β = 95.7050(18)Њ and V = 1370.07(6) Å3.4 The unit
cell of 1[ClO4]2 is similar enough to that of [Fe(L1H)2][BF4]2
that the two compounds are almost certainly isomorphous.
However, although the cell dimensions for 1[BF4]2 are super-
ficially similar to these other two compounds (a ≈ b, c ≈ 19 Å,
β ≈ 90Њ), its unit cell volume is larger and corresponds approx-
imately to Z = 4. These unit cells show that 1[BF4]2 and 1[ClO4]2
are likely to show similar, but not identical, packing arrange-
ments in the solid that are closely related to that shown by
[Fe(L2H)2][BF4]2.4–6 Variable temperature powder diffraction
measurements on these compounds are in progress, as are
attempts to refine the structures of 1[BF4]2 and 1[ClO4]2 from
their powder diffraction data. These will be reported separately.
Although useful single crystals of 1[BF4]2 and 1[ClO4]2 could
not be grown, crystals of formula 1[BF4]2ؒ3CH3NO2 are good
diffractors of X-rays. Interestingly, this compound is not
isomorphous with the CH3NO2 solvate of [Fe(L1H)2][BF4]2.5
The crystals are orange-yellow at room temperature, but
undergo a sharp change to a dark brown colour upon cooling
that is indicative of a spin-state transition. Variable temperature
unit cell measurements established that this transition occurred
abruptly at 198(1) K, and does not involve a crystallographic
phase transition (Fig. 2). In cooling mode, the transition is
characterised by a sharp increase of 1.18(10)Њ in the unit cell
angle (in the setting P21/n), a 0.18(4) Å decrease in the unit cell
dimension a and an increase of 0.064(13) Å in b; there is no
significant discontinuity in the c dimension. These changes
correspond to a decrease in the unit cell volume upon cooling
across the transition of 30(4) Å3, or 8(1) Å3 per molecule which
is much smaller than the volume change associated with spin-
crossover in crystals of [Fe(L1H)2][BF4]2, of 18(3) Å3 per mole-
cule.4–6 This presumably reflects the presence of additional
solvent molecules in 1[BF4]2ؒ3CH3NO2, compared to the liter-
ature compound. Crystals of 1[BF4]2ؒ3CH3NO2 could be
cycled across their transition several times, with only a small
degradation in crystal quality.
Solid state structures and magnetochemistry of [Fe(L2H)2]2؉ salts
Both 1[BF4]2 and 1[ClO4]2 are bright yellow microcrystalline
powders, that show XMT = 3.5 cm3 molϪ1 K at room-temper-
ature. Both these observations are consistent with their being
fully populated in a high-spin, S = 2 spin state.16 Upon cooling,
the samples undergo very similar abrupt spin-state transitions
to a fully populated S = 0 low-spin form. For 1[BF4]2, this
transition is centred at 223 K and shows a hysteresis loop of
3 K, while for 1[ClO4]2, the transition is centred at 206 K with a
5 K hysteresis loop (Fig. 1). These measurements were con-
firmed by DSC data, which showed one first-order transition
for each material. For 1[BF4]2, the DSC transition was centred
at 224 K with a hysteresis width of 4 K, while for 1[ClO4]2 the
transition was centred at 205 K with a 3 K hysteresis loop.
Interestingly, the thermodynamic parameters for the transition
in the two materials showed some differences. For 1[BF4]2, ∆H =
29.2(1) kJ molϪ1 and ∆S = 130.4(4) J molϪ1
K
Ϫ1; while, for
1[ClO4]2 ∆H = 19.8(4) kJ molϪ1 and ∆S = 97(2) J molϪ1 KϪ1. ∆H
and ∆S for both of these compounds are somewhat greater
than for [Fe(L1H)2][BF4]2,4 however.
Two X-ray datasets were collected from the same crystal of
1[BF4]2ؒ3CH3NO2 at 150 K and 300 K, which are temperatures
at which the crystal should be fully low-spin and fully high-spin,
respectively. At both temperatures, the complex dications show
only small deviations from local D2d symmetry (Fig. 3). The
Fe–N bond lengths in 1[BF4]2ؒ3CH3NO2 at 150 K and 300 K
are crystallographically indistinguishable from those in the
low-spin and high-spin forms of [Fe(L1H)2][BF4]2,4–6 respect-
ively, and hence are consistent with the compound being in
the expected, pure spin-state (Table 1). At both temperatures,
Ϫ
all the BF4 and CH3NO2 moieties in the crystal are badly
disordered. There are some differences in the disorder models
refined at 150 K and 300 K, but it is uncertain how significant
these changes are; or, if they are real, whether they result from
Fig. 1 Plots of XMT vs. T for polycrystalline samples of: (a) 1[BF4]2;
and (b) 1[ClO4]2. The inset graphs show the hysteresis loops associated
with these two spin-state transitions.
D a l t o n T r a n s . , 2 0 0 3 , 2 0 5 3 – 2 0 6 0
2054

View Article Online
Table 1 Selected bond lengths (Å) and angles (Њ) for [Fe(L2H)2][BF4]2ؒ
3CH3NO2 (1[BF4]2ؒ3CH3NO2)
150 K
300 K
Fe(1)–N(2)
Fe(1)–N(9)
Fe(1)–N(14)
Fe(1)–N(18)
Fe(1)–N(25)
Fe(1)–N(30)
1.896(3)
1.982(3)
1.981(3)
1.891(3)
1.998(3)
1.978(3)
2.114(3)
2.191(3)
2.179(3)
2.123(3)
2.184(3)
2.185(3)
N(2)–Fe(1)–N(9)
N(2)–Fe(1)–N(14)
N(2)–Fe(1)–N(18)
N(2)–Fe(1)–N(25)
N(2)–Fe(1)–N(30)
N(9)–Fe(1)–N(14)
N(9)–Fe(1)–N(18)
N(9)–Fe(1)–N(25)
N(9)–Fe(1)–N(30)
N(14)–Fe(1)–N(18)
N(14)–Fe(1)–N(25)
N(14)–Fe(1)–N(30)
N(18)–Fe(1)–N(25)
N(18)–Fe(1)–N(30)
N(25)–Fe(1)–N(30)
79.85(10)
79.53(10)
177.40(11)
102.60(10)
98.23(11)
159.37(11)
98.65(11)
91.18(10)
94.07(11)
101.98(11)
92.79(10)
89.38(11)
79.50(11)
79.73(11)
159.11(11)
73.21(11)
73.45(11)
173.18(11)
110.94(11)
102.97(11)
146.55(11)
101.54(11)
93.07(12)
97.97(12)
111.91(11)
96.25(12)
91.98(12)
73.27(12)
73.11(12)
146.07(12)
Fig. 2 Plot of the variation of the unit cell parameters of 1[BF4]2ؒ
3CH3NO2 with temperature.
Fig. 3 View of the complex dication in the crystal structure of
[Fe(L2H)2][BF4]2ؒ3CH3NO2 (1[BF4]2ؒ3CH3NO2) at 150 K, showing the
atom numbering scheme employed. For clarity, all H atoms have been
omitted. Thermal ellipsoids are at the 50% probability level. The
molecular structure of the complex dication in this crystal at 300 K is
visually almost indistinguishable from that shown here, and uses an
identical atom numbering scheme.
the spin state change, or simply from the effects of cooling the
crystal. Hence, we have been unable to determine whether or
not the abruptness of the spin-state change in the crystals is
mediated by changes in the disorder regime of the crystal across
the transition.4,17
Fig. 4 Plots of XMT vs. T for polycrystalline samples of: (a) 2[BF4]2
and (b) 2[ClO4]2.
Solid state structures and magnetochemistry of the other
[Fe(L2R)2]2؉ complexes
shows a two-stage transition: an abrupt step at 196 K, that does
not show detectable hysteresis; and a more gradual step centred
at 133 K (Fig. 4B). The boundary between these two steps
occurs when half of the Fe centres in the sample are low-spin,
to within experimental error. From precedent, this behaviour
could be interpreted in three ways. First, 2[ClO4]2 may contain
two independent [Fe(L2Me)2]2ϩ environments that undergo
spin-state transitions at different temperatures.18 Second, the
Both of the salts 2[BF4]2 and 2[ClO4]2 are yellow at room tem-
perature, showing that they are also high-spin under ambient
conditions. Upon cooling they also undergo transitions to a
low-spin state, although the shape of the XMT vs. T curve is
now very different for the two salts. The transition for 2[BF4]2 is
very gradual, occurring over the approximate range of 130–280
K and being centred at 235 K (Fig. 4A). However, 2[ClO4]2
D a l t o n T r a n s . , 2 0 0 3 , 2 0 5 3 – 2 0 6 0
2055

View Article Online
Table 2 Selected bond lengths (Å) and angles (Њ) for [Fe(L2Me2)2][BF4]2ؒ0.5{CH3}2COؒ0.1H2O (3[BF4]2ؒ0.5{CH3}2COؒ0.1H2O)
Molecule 1
Molecule 2
Fe(1)–N(2)
Fe(1)–N(9)
Fe(1)–N(16)
Fe(1)–N(22)
Fe(1)–N(29)
Fe(1)–N(36)
1.880(3)
1.997(3)
1.976(3)
1.882(3)
1.981(3)
1.986(3)
Fe(42)–N(43)
Fe(42)–N(50)
Fe(42)–N(57)
Fe(42)–N(63)
Fe(42)–N(70)
Fe(42)–N(77)
1.885(3)
1.967(3)
1.989(3)
1.884(3)
1.979(3)
1.984(3)
N(2)–Fe(1)–N(9)
N(2)–Fe(1)–N(16)
N(2)–Fe(1)–N(22)
N(2)–Fe(1)–N(29)
N(2)–Fe(1)–N(36)
N(9)–Fe(1)–N(16)
N(9)–Fe(1)–N(22)
N(9)–Fe(1)–N(29)
N(9)–Fe(1)–N(36)
N(16)–Fe(1)–N(22)
N(16)–Fe(1)–N(29)
N(16)–Fe(1)–N(36)
N(22)–Fe(1)–N(29)
N(22)–Fe(1)–N(36)
N(29)–Fe(1)–N(36)
80.30(12)
80.12(12)
176.10(12)
101.53(13)
98.59(12)
160.41(12)
96.27(12)
92.24(12)
93.67(12)
103.32(12)
91.27(12)
89.63(12)
80.37(13)
79.70(12)
159.71(12)
N(43)–Fe(42)–N(50)
N(43)–Fe(42)–N(57)
N(43)–Fe(42)–N(63)
N(43)–Fe(42)–N(70)
N(43)–Fe(42)–N(77)
N(50)–Fe(42)–N(57)
N(50)–Fe(42)–N(63)
N(50)–Fe(42)–N(70)
N(50)–Fe(42)–N(77)
N(57)–Fe(42)–N(63)
N(57)–Fe(42)–N(70)
N(57)–Fe(42)–N(77)
N(63)–Fe(42)–N(70)
N(63)–Fe(42)–N(77)
N(70)–Fe(42)–N(77)
80.21(12)
80.33(12)
179.50(14)
99.30(13)
100.21(13)
160.54(12)
100.14(12)
90.60(12)
93.37(13)
99.32(12)
92.45(12)
90.15(13)
80.35(13)
80.13(13)
160.47(12)
spin-state transition in this compound could occur concomi-
tantly with a phase transition when the fraction of high-spin
molecules in the solid (‘γ’) is 0.5, with the dynamics of the
spin-crossover being different in the two phases.19,20 Finally, a
balance of inter- and intra-molecular forces within the solid
may give rise to a local ordering of high- and low-spin
molecules in the sample when γ = 0.5, without causing a crystal-
lographic phase change.21
Single crystals of 2[BF4]2 and 2[ClO4]2 yield very similar unit
cells at 150 K and at 300 K, all with 4/m Laue symmetry. At
150 K the parameters are as follows: for 2[BF4]2, a = 9.04(1),
c = 17.59(1) Å, V = 1437.48(2) Å3; and for 2[ClO4]2, a = 9.16(1),
c = 17.71(1) Å, V = 1485.97(2) Å3. Unfortunately, we were
unable to solve crystallographic datasets obtained from either
compound, and it is unclear whether or not the crystals were
twinned. None-the-less, the unit cell volumes are consistent
with the crystals being solvent-free (for Z = 2), and it seems
clear that the two compounds are isomorphous. Hence, the very
different spin-state transitions shown by the two compounds
(Fig. 4) are intriguing. Unfortunately, in the absence of better
crystallographic data we are unable to distinguish between the
three potential origins of this behaviour listed in the previous
paragraph.
In contrast, 3[BF4]2, 4[BF4]2ؒH2O and their corresponding
Ϫ
ClO4 salts are all dark brown in colour at room temperature.
Magnetic measurements for all four compounds at 295 K
showed XMT <0.02 cm3 molϪ1 K, consistent with their being
completely low spin under ambient conditions.16 In order to
understand the large difference in the magnetic properties
between the complexes [Fe(L2Me)2]2ϩ and [Fe(L2Me2)2]2ϩ, a
single crystal X-ray analysis was carried out on crystals of
3[BF4]2ؒ0.5{CH3}2COؒ0.1H2O at 150 K. The asymmetric unit
of the crystals contains two crystallographically independent
complex dications, which show almost identical molecular
structures with molecular symmetry close to D2d (Fig. 5). The
Fe–N bond lengths in this crystal structure (Table 2) are crystal-
lographically indistinguishable from those in low-spin 1[BF4]2ؒ
3CH3NO2 (Table 1) confirming that this compound is com-
pletely in a low-spin state at 150 K. Examination of a space-
filling model shows that the 5-methyl groups on each pyrazole
ring are in van-der-Waals contact with a 3{5}-H atom of the
pyrazine ring. The relevant H3C ؒ ؒ ؒ H distances lie in the range
2.56–2.65 Å, compared to the sum of the Pauling radii of a H
atom and a ‘spherical’ methyl group, of 3.2 Å.22 Steric repulsion
between these two groups should force the pyrazole N donors
Fig.
5
Views of the crystallographically independent complex
dications in the structure of [Fe(L2Me2)2][BF4]2ؒ0.5{CH3}2COؒ0.1H2O
(3[BF4]2ؒ0.5{CH3}2COؒ0.1H2O), showing the atom numbering scheme
employed. For clarity, all H atoms have been omitted. Thermal
ellipsoids are at the 50% probability level.
closer to the Fe ions, favouring a low-spin state. This presum-
ably explains why 3[BF4]2 and 3[ClO4]2 are low-spin at room
temperature.
D a l t o n T r a n s . , 2 0 0 3 , 2 0 5 3 – 2 0 6 0
2056

View Article Online
Table 3 Selected bond lengths (Å) and angles (Њ) for [Fe(L2Mes)2]-
[BF4]2ؒ5CH3NO2 (4[BF4]2ؒ5CH3NO2)
spectrum of 3[BF4]2 was unexpected, since this compound is
fully low-spin in the solid at this temperature. 1H NMR
spectra obtained in CD3NO2 showed the same number of peaks
as in CD3CN, at chemical shifts that were identical for each
compound to within 2 ppm for the paramagnetic spectra, or
0.2 ppm for the diamagnetic spectra. However, for 1[BF4]2
only the spectrum in CD3CN was substantially broader than in
CD3NO2. This could indicate the presence of solvent-promoted
ligand exchange for this compound in CD3CN, which would be
consistent with the sensitivity of the [Fe(L2R)2]2ϩ complexes to
hydrolysis mentioned above.
Fe(1)–N(2)
Fe(1)–N(9)
1.885(2)
2.009(2)
Fe(1)–N(23)
2.009(2)
N(2)–Fe(1)–N(2Ј)
N(2)–Fe(1)–N(9)
N(2)–Fe(1)–N(9Ј)
N(2)–Fe(1)–N(23)
N(2)–Fe(1)–N(23Ј)
179.80(17)
80.01(9)
100.13(10)
80.08(9)
N(9)–Fe(1)–N(9Ј)
N(9)–Fe(1)–N(23)
N(9)–Fe(1)–N(23Ј)
N(23)–Fe(1)–N23Ј)
87.80(13)
160.09(9)
95.79(9)
87.49(13)
99.77(9)
Primed atoms are related to unprimed atoms by the relation y, x, Ϫz.
Low temperature solution-phase magnetic data were not
measured for 1[BF4]2 and 2[BF4]2, owing to their very poor
solubility in solvents that have freezing points near their solid
state spin-state transitions, such as {CD3}2CO. However, the
magnetic moment of 3[BF4]2 was measured in CD3NO2
between 253 K and 333 K, in order to gain further insight into
its unexpected NMR and UV/vis behaviour (see below). Under
these conditions XMT rose continuously from 0.5(1) cm3 molϪ1
K at 253 K to 1.0(1) cm3 molϪ1 K at 333 K. A spin-state transi-
tion in solution would yield an S-shaped XMT vs. T curve, and
would normally be complete over a temperature range of ca.
150 K.5,26 So, the extremely slow monotonic rise of XMT with
increasing temperature shown by 3[BF4]2 is unlikely to corre-
spond to a single Fe() centre in a spin equilibrium. Rather, it
implies a solution sample containing a mixture of high- and
low-spin Fe() complex species in rapid temperature-dependent
chemical exchange. This is further evidence for solution lability
of another of our [Fe(L2R)2]2ϩ centres.
A crystal structure was also achieved on crystals of 4[BF4]2ؒ
5CH3NO2, to probe the apparent lattice water found in this
compound. The structure shows a fully low-spin Fe() centre,
with crystallographic C2 symmetry and approximate D2d
symmetry (Fig. 6, Table 3). The Fe–N bond lengths in this
structure are up to 0.042(4) Å longer than for low-spin 1[BF4]2ؒ
3CH3NO2 or 3[BF4]2, but are similar to those in [Fe(L1Mes)2]-
[PF6]2.7 Importantly, even though dried material from the
same solvent analyses as 4[BF4]2ؒH2O, there is no trace of
lattice water in the solvated crystals analysed here. This shows
that the water content of the dried material does not play an
intrinsic role in its structure, but is simply atmospheric moisture
that has partially replaced the nitromethane content of the
freshly prepared crystals upon exposure to air. We have
observed this previously in salts of [M(L1Mes)2]2ϩ (M = Fe,7
Co,23 Ni,23 Cu,24 Zn25), which always form solvated single
crystals from CH3NO2 or CH3CN and are very difficult to
obtain as solvent-free solids. Hence, it seems unlikely that the
water content in powdered 4[BF4]2ؒH2O interacts specifically
with this complex’s non-coordinated pyrazine N donors.
The UV/vis/NIR spectra of the compounds in CH3CN and
CH3NO2 at 295 K are listed in the Experimental section. Very
similar spectra were obtained in the visible and near-UV ranges
in both solvents for each compound, which confirm the conclu-
sions drawn from the NMR studies. Hence, 1[BF4]2 and 2[BF4]2
show spectra arising from a pure high-spin Fe() chromophore,
being characterised by two overlapping d–d absorptions
centred near 9.3 and 11.8 × 103 cmϪ1, both with εmax ≈ 10 MϪ1
cmϪ1 27
and, a weak metal-to-ligand charge transfer (MLCT)
;
band near 23.5 × 103 cmϪ1 (sh, ε ≈ 250 MϪ1 cmϪ1). Compound
4[BF4]2ؒH2O, in contrast, exhibits spectra typical of a low-spin
Fe() centre,27 that is dominated by a much stronger MLCT
envelope centred at 22.4 × 103 cmϪ1 (εmax ≈ 3,200 MϪ1 cmϪ1).
However, the spectrum of 3[BF4]2 shows both a d–d absorption
characteristic of a high-spin compound at 12.0 × 103 cmϪ1
(11.4); and, a strong MLCT band at 21.2 × 103 cmϪ1 (3,000)
that is suggestive of a low-spin Fe() centre. This is consistent
with the NMR data mentioned above, which show that 3[BF4]2
exists as a mixture of high-spin and low-spin compounds under
these conditions.
Fig. 6 View of the complex dication in the crystal structure of
[Fe(L2Mes)2][BF4]2ؒ5CH3NO2 (4[BF4]2ؒ5CH3NO2), showing the atom
numbering scheme employed. For clarity, all H atoms have been
omitted. Thermal ellipsoids are at the 50% probability level. Primed
atoms are related to unprimed atoms by the relation y, x, Ϫz.
Conclusion
The compounds in this study represent a continuation of our
investigations of Fe() complexes of pyrazine-containing
chelates. Unlike our earlier system [Fe(L3)3]2ϩ 15
salts of
,
[Fe(L2H)2]2ϩ and [Fe(L2Me)2]2ϩ undergo spin-state transitions
at moderately low temperatures in the solid state and in
solution. Comparison of these data with our earlier results4–6
shows that substitution of the pyridine rings in [Fe(L1R)2]2ϩ by
pyrazine rings in [Fe(L2R)2]2ϩ results in only a small lowering of
the spin-crossover temperature, by ca. 50 K. This small effect of
introducing an additional heteroatom into [Fe(L2R)2]2ϩ com-
pared to [Fe(L1R)2]2ϩ is also reflected in the low-spin character
of [Fe(L2Mes)2]2ϩ. Since we have previously shown that
[Fe(L1Mes)2]2ϩ is low-spin,7 and because [Fe(L2H)2]2ϩ shows a
lower spin-crossover temperature than [Fe(L1H)2]2ϩ, we had
hoped that substitution of a pyrazine ring for the pyridine ring
in the ligand framework may lower the ligand field at Fe
sufficiently for [Fe(L2Mes)2]2ϩ to show a spin-state transition at
or below room temperature. Clearly, this is not the case. It
Solution measurements
All solution studies were carried out using the BF4Ϫ salts of the
complexes. The 1H NMR spectra of all these complexes show a
single C2 or m-symmetric L2R environment, consistent with the
D2d-symmetric molecular structures shown by, or anticipated
for, these compounds. The spectra of 1[BF4]2, 2[BF4]2 and
3[BF4]2 in CD3CN at 295 K were all broadened and strongly
contact shifted, consistent with these compounds being at least
partly high-spin in solution. In contrast, the spectrum of
4[BF4]2ؒH2O is well resolved, and lies completely within the
diamagnetic region of the spectrum. This demonstrates that
this complex is completely low-spin in solution. These data
are consistent with the solid-state properties of 1[BF4]2, 2[BF4]2
and 4[BF4]2ؒH2O (see above). However, the contact-shifted
D a l t o n T r a n s . , 2 0 0 3 , 2 0 5 3 – 2 0 6 0
2057

View Article Online
is also of interest that 1[BF4]2, 1[BF4]2ؒ3CH3NO2, 1[ClO4]2,
2[BF4]2 and 2[ClO4]2 all undergo spin-state transitions at
Synthesis of 2,6-bis(3,5-dimethylpyrazol-1-yl)pyrazine (L2Me2)
Method as for L2Me, using 3,5-dimethylpyrazole (2.3 g, 2.4 ×
10Ϫ2 mol). The freshly precipitated white solid product was ana-
lytically pure after drying over P2O5, and was not purified fur-
ther. Yield 4.6 g, 70%. Found: C, 62.8; H, 6.2; N, 31.3; Calcd.
for C14H16N6: C, 62.7; H, 6.0; N, 31.3%. Mp 167–169 ЊC. EI
mass spectrum: m/z 268 [M]ϩ. NMR spectra (CDCl3, 295 K):
1H; δ 9.05 (s, 2H, Py H3/5), 6.05 (d, 2.0 Hz, 2H, Pz H4), 2.61
(s, 6H, 5-CH3), 2.32 (s, 6H, 3-CH3) ppm. 13C; δ 151.3 (Py C2/6),
146.4 (Pz C3), 141.6 (Pz C5), 134.5 (Py C3/5), 109.9 (Pz C4),
14.1, 13.6 (2 × CH3) ppm.
broadly similar temperatures, with T = 196–235 K. Studies on
½
related tridentate ligand systems have implied that T for
½
[Fe(L2Me)2]2ϩ should be lower for than [Fe(L2H)2]2ϩ, because
of the inductive effect of the ligand methyl substituents;28
or, that it should be higher than for [Fe(L2H)2]2ϩ, owing to
steric repulsions involving these methyl groups.29 Clearly, these
effects balance out in our system, so that these two com-
plex dications undergo spin-state transitions at very similar
temperatures.
The apparent solution lability of 1[BF4]2 and 3[BF4]2, and
the sensitivity of all of our compounds to hydrolysis in protic
solvents, contrasts with [Fe(L1R)2]2ϩ (R = H, Pri, Ph, Mes),
which exhibit static structures by NMR5,7 and which are much
more chemically robust. Presumably, the increased lability of
[Fe(L2R)2]2ϩ reflects the lower basicity of the pyrazine donor in
L2R compared to the pyridine donor in L1R,30 combined with
the relatively poor basicity of the pyrazole donors in both
ligands.31
Synthesis of 2,6-bis(3-{2,4,6-trimethylphenyl}pyrazol-1-yl)-
pyrazine (L2Mes)
Method as for L2Me, using 3{5}-(2,4,6-trimethylphenyl)-
pyrazole (4.2 g, 2.4 × 10Ϫ2 mol). The freshly precipitated white
solid product solid contained L2Mes, and its regioisomer 2-(3-
{2,4,6-trimethylphenyl}pyrazolyl)-6-(5-{2,4,6-trimethylphenyl}-
1
pyrazolyl)pyrazine in approximately 9 : 1 molar ratio by H
It is interesting that, although we have synthesised and
crystallised our compounds in air, using non-predried solvents,
few of the crystalline or powder solids we have isolated contain
lattice water, or any other protic additives that might hydrogen
bond to the non-coordinated pyrazine N-donors. In addition,
while the salts of [Fe(L2Mes)2]2ϩ that we prepared appear to
contain lattice water by microanalysis, this is not present in
single crystals of 4[BF4]2 grown from the same solvent. The
apparent reluctance of [Fe(L2R)2]2ϩ to engage in hydrogen
bonding contrasts with the known Fe() complexes of L3, all of
which contain extensive hydrogen bonding to the L3 pyrazine
acceptors.15 We suggest that the incorporation of a second
electron-withdrawing pyrazole substituent into the L2R ring
may reduce the basicity of the non-coordinated L2R N-atom
to such an extent that it is only a poor hydrogen-bond
acceptor. Future work will investigate the synthesis of pyrazine-
based ligands containing other donor functions with more
favourable inductive properties, in order to ameliorate this
problem.
NMR. Recrystallisation of the mixture from acetone allowed
the isolation of pure L2Mes as a white solid, albeit with
substantial solubility losses. Yield 1.6 g, 31%. Found: C, 74.2;
H, 6.2; N, 18.7; Calcd. for C28H28N6: C, 75.0; H, 6.3; N, 18.7%.
Mp 189–191 ЊC. EI mass spectrum: m/z 448 [M]ϩ. NMR spectra
(CDCl3, 293 K): 1H; δ 9.20 (s, 2H, Py H3/5), 8.60 (d, 2.8 Hz, 2H,
Pz H5), 6.98 (s, 4H, Ph H3/5), 6.54 (d, 2.8 Hz, 2H, Pz H4),
2.34 (s, 6H, 4–CH3), 2.20 (s, 12H, 2,6–CH3) ppm. 13C; δ 154.9
(Py C2/6), 145.1 (Pz C3), 138.2 (Ph C4), 137.3 (Ph C2/6), 131.4
(Py C3/5), 129.7 (Ph C1), 128.4 (Ph C3/5), 127.7 (Pz C5), 110.4
(Pz C4), 21.1 (4–CH3), 20.5 (2,6–CH3) ppm.
Syntheses of the complexes
The syntheses of all of these complexes followed the same basic
procedure, as described here for [Fe(L2H)2][BF4]2 (1[BF4]2). A
solution of Fe[BF4]2ؒ6H2O (0.24 g, 7.2 × 10Ϫ4 mol) and L2H
(0.30 g, 1.4 × 10Ϫ3 mol) in acetone (50 cm3) was stirred at room
temperature for 15 min. The resultant yellow solution was
1
filtered and concentrated to ca. /3 its original volume, where-
upon a yellow precipitate formed. Following overnight storage
at Ϫ30 ЊC, the product was filtered off and washed sequentially
with cold MeOH and Et2O. Small amounts of this product
can be recrystallised from the same solvent mixture, yielding
solvent-free microcrystals. Similar reactions using equivalent
quantities of the appropriate L2R ligand and Fe[ClO4]2ؒ6H2O,
yielded the other complexes in this study. Yields ranged from
59–86%. [CAUTION: while we have experienced no difficulty
Experimental
Unless stated otherwise, all manipulations were performed
in air using commercial grade solvents, except that pre-dried
dmf was purchased from Aldrich and stored under N2. 2,6-
Dipyrazol-1-ylpyrazine (L2H)9 and 3{5}-(2,4,6-trimethyl-
phenyl)pyrazole32 were prepared by the literature procedures.
All other reagents were used as supplied.
Ϫ
in handling the ClO4 salts in this study, metal–organic
perchlorates are potentially explosive and should be handled
with due care in small quantities.]
Synthesis of 2,6-bis(3-methylpyrazol-1-yl)pyrazine (L2Me)
A deoxygenated solution of 3{5}-methylpyrazole (2.0 g, 2.4 ×
10Ϫ2 mol) in dry dmf (250 cm3) was added to solid KH (1.0 g,
2.5 × 10Ϫ2 mol) under N2, and the resultant suspension stirred
at 50 ЊC for 20 minutes. 2,6-Dichloropyrazine (1.8 g, 1.2 × 10Ϫ2
mol) was then added, and the mixture stirred at 90 ЊC for 16 h.
After cooling, a large excess of cold water was added to the
mixture, yielding a white solid which was collected, washed with
water and dried over P2O5. This crude solid contained both
L2Me, and its regioisomer 2-(3-methylpyrazolyl)-6-(5-methyl-
For [Fe(L2H)2][BF4]2 (1[BF4]2): found C, 36.8; H, 2.8; N, 25.5.
Calcd. For C20H16B2F8FeN12: C, 36.7; H, 2.5; N, 25.7%. ES
1
mass spectrum: m/z 499 [56FeF(L2H)2]ϩ, 240 [56Fe(L2H)2]2ϩ. H
NMR spectrum (CD3CN, 295 K): δ 55.2 (8H), 33.2, 11.0 (both
4H) (Py H3/5 ϩ Pz H3–H5) ppm. UV/vis spectrum (MeCN, 295
K): νmax, 103 cmϪ1 (εmax, MϪ1.cmϪ1) 9.2 (sh), 11.4 (9.5), 23.5 (sh),
30.7 (37,300), 37.3 (22,500), 40.9 (49,328), 41.5 (sh), 46.5
(22,300). UV/vis spectrum (MeNO2, 295 K): νmax, 103 cmϪ1
(εmax, MϪ1.cmϪ1) 9.2 (sh), 11.8 (12.0).
1
pyrazolyl)pyrazine in approximately 4 : 1 molar ratio by H
For [Fe(L2H)2][ClO4]2 (1[ClO4]2): found C, 35.4; H, 2.4; N,
24.8. Calcd. For C20H16Cl2FeN12O8: C, 35.4; H, 2.4; N, 24.7%.
ES mass spectrum: m/z 579 [56Fe35ClO4(L2H)2]ϩ, 367
NMR. Recrystallisation of the mixture from acetone allowed
the isolation of pure L2Me as a white solid, albeit with substan-
tial solubility losses. Yield 1.2 g, 40%. Found: C, 59.7; H, 5.0; N,
35.2; Calcd. for C12H12N6: C, 60.0; H, 5.0; N, 35.0%. Mp 164–
166 ЊC. EI mass spectrum: m/z 240 [M]ϩ. NMR spectra (CDCl3,
295 K): 1H; δ 9.08 (s, 2H, Py H3/5), 8.36 (d, 2.0 Hz, 2H, Pz H5),
6.31 (d, 2.0 Hz, 2H, Pz H4), 2.39 (s, 6H, CH3) ppm. 13C; δ 153.1
(Pz C2/6), 144.9 (Pz C3), 130.6, 128.0 (Py C3/5 ϩ Pz C5), 109.1
(Pz C4), 13.8 (CH3) ppm.
[56Fe35ClO4(L2H)]ϩ, 240 [56Fe(L2H)2]2ϩ
.
For [Fe(L2Me)2][BF4]2 (2[BF4]2): found C, 40.6; H, 3.4; N,
23.6. Calcd. For C24H24B2F8FeN12: C, 40.6; H, 3.4; N, 23.7%.
ES mass spectrum: m/z 315 [56FeF(L2Me)]ϩ, 268 [56Fe(L2-
Me)2]2ϩ. 1H NMR spectrum (CD3CN, 295 K): δ 68.6, 68.4, 37.4
(all 4H, Py H3/5 ϩ Pz H4, H5), 7.3 (12H, CH3) ppm. UV/vis
spectrum (MeCN, 295 K): νmax, 103 cmϪ1 (εmax, MϪ1.cmϪ1
)
D a l t o n T r a n s . , 2 0 0 3 , 2 0 5 3 – 2 0 6 0
2058

View Article Online
Table 4 Experimental details for the single crystal structure determinations in this study
1[BF4]2ؒ3CH3NO2
3[BF4]2ؒ0.5{CH3}2COؒ0.1H2O
4[BF4]2ؒ5CH3NO2
Formula
Mr
Crystal class
Space group
C23H25B2F8FeN15O6
837.05
Monoclinic
P21/n
C23H25B2F8FeN15O6
837.05
Monoclinic
P21/n
C29.5H35.2B2F8FeN12O0.6
796.96
Monoclinic
P21/c
C61H71B2F8FeN17O10
1431.82
Hexagonal
P3121
a/Å
b/Å
c/Å
16.5404(2)
8.2833(1)
25.8237(4)
108.2104(6)
3360.89(8)
4
16.9138(2)
8.3045(1)
26.3119(4)
107.4285(4)
3526.13(8)
4
11.3306(1)
15.2392(2)
40.5559(5)
92.8371(6)
6994.17(14)
8
15.1838(1)
–
24.9779(2)
–
4987.09(6)
3
0.320
β/Њ
V/Å3
Z
µ(Mo-Kα)/mmϪ1
T /K
0.558
150(2)
0.531
300(2)
0.518
150(2)
150(2)
Measured reflections
Independent reflections
Rint
23212
7664
0.047
0.064
30653
6902
0.083
0.079
62394
15966
0.128
0.061
89750
7623
0.157
0.057
R(F)a
wR(F 2)b
0.196
0.248
0.162
0.160
Flack parameter
–
–
–
0.004(19)
4
½
a R = Σ[|Fo| Ϫ | Fc|] / Σ|Fo|. b wR = [Σw(Fo2 Ϫ Fc2) / ΣwFo ] .
9.4 (sh), 11.8 (9.9), 23.6 (sh), 30.3 (44,600), 36.6 (27,700), 40.2
(56,400), 45.0 (sh). UV/vis spectrum (MeNO2, 295 K): νmax, 103
cmϪ1 (εmax, MϪ1.cmϪ1) 9.7 (sh), 11.8 (11.3).
See http://www.rsc.org/suppdata/dt/b2/b210368k/ for crystal-
lographic data in CIF or other electronic format.
For [Fe(L2Me)2][ClO4]2 (2[ClO4]2): found C, 39.1; H, 3.2; N,
22.8. Calcd. For C24H24Cl2FeN12O8: C, 39.2; H, 3.3; N, 22.9%.
ES mass spectrum: m/z 635 [56Fe35ClO4(L2Me)2]ϩ, 395 [56Fe-
X-Ray structure determination of [Fe(L2H)2][BF4]2ؒ3CH3NO2
(1[BF4]2ؒ3CH3NO2). Two datasets were collected from the same
crystal of this compound, at 300 and 150 K. At both temper-
35ClO4(L2Me)]ϩ, 269 [56Fe(L2Me)2]2ϩ
.
atures, both BF4 anions and all three CH3NO2 molecules
Ϫ
For [Fe(L2Me2)2][BF4]2 (3[BF4]2): found C, 44.0; H, 4.4; N,
21.8. Calcd. For C28H32B2F8FeN12: C, 43.9; H, 4.2; N, 21.8%.
ES mass spectrum: m/z 611 [56FeF(L2Me2)2]ϩ, 296 [56Fe-
suffered from disorder. Each of these disordered moieties was
modelled over two or three different orientations, although the
precise model used varied slightly between the two refinements.
The following refined restraints were applied to the final
models: at 150 K, B–F = 1.39(2), non-bonded F ؒ ؒ ؒ F = 2.27(2),
C–N = 1.47(2), N–O = 1.22(2) and non-bonded O ؒ ؒ ؒ O =
2.11(2) Å; at 300 K, B–F = 1.39(2), non-bonded F ؒ ؒ ؒ F =
2.27(2), C–N = 1.47(2), N–O = 1.21(2) and non-bonded
O ؒ ؒ ؒ O = 2.10(2) Å. All crystallographically ordered non-H
atoms were refined anisotropically. All H atoms were placed in
calculated positions and refined using a riding model except for
the disordered solvent methyl H groups, which were fixed in
idealised geometries and whose torsions were not refined.
1
(L2Me2)2]2ϩ. H NMR spectrum (CD3CN, 295 K): δ 39.3, 38.3
(both 4H, Py H3/5 ϩ Pz H4), 14.6, 4.3 (both 12H, CH3) ppm.
UV/vis spectrum (MeCN, 295 K): νmax, 103 cmϪ1 (εmax
,
M
Ϫ1.cmϪ1) 12.0 (11.4), 18.5 (sh), 21.2 (3,000), 22.4 (sh), 24.8
(sh), 30.7 (32,500), 37.1 (sh), 40.2 (45,300). UV/vis spectrum
(MeNO2, 295 K): νmax, 103 cmϪ1 (εmax, MϪ1.cmϪ1) 11.8 (11.0),
21.0 (3,000), 22.3 (sh).
For [Fe(L2Me2)2][ClO4]2 (3[ClO4]2): found C, 42.2; H, 4.0; N,
21.4. Calcd. For C28H32Cl2FeN12O8: C, 42.2; H, 4.1; N, 21.2%.
ES mass spectrum: m/z 691 [56Fe35ClO4(L2Me2)2]ϩ, 296
[56Fe(L2Me2)2]2ϩ
.
For [Fe(L2Mes)2][BF4]2ؒH2O (4[BF4]2ؒH2O): found C, 58.9;
H, 5.1; N, 14.8. Calcd. For C56H56B2F8FeN12ؒ.H2O C, 58.8; H,
5.1; N, 14.7%. ES mass spectrum: m/z 476 [56Fe(L2Mes)2]2ϩ. 1H
NMR spectrum (CD3CN, 295 K): δ 9.08 (d, 3.0 Hz, 4H, Pz H5),
9.01 (s, 4H, Py H3/5), 6.73 (s, 8H, Ph H3/5), 6.61 (d, 3.0 Hz, 4H,
Pz H4), 2.36 (s, 12H, 4-CH3), 0.96 (s, 24H, 2,6-CH3) ppm.
X-Ray structure determination of [Fe(L2Me2)2][BF4]2ؒ
0.5{CH3}2COؒ0.1H2O (3[BF4]2ؒ0.5{CH3}2COؒ0.1H2O). The
asymmetric unit of this crystal contains two complex dications,
Ϫ
four BF4 anions, one molecule of acetone, and one weakly
scattering feature not bonded to any other residue that was
modelled as a molecule of water with occupancy 0.2. Three of
the four BF4Ϫ anions was disordered, each being modelled over
three different orientations. All disordered B–F bonds were
restrained to 1.39(2) Å, and non-bonded F ؒ ؒ ؒ F distances
within a given disorder orientation to 2.27(2) Å. All non-H
atoms with occupancy ≥0.5 were refined anisotropically. All H
atoms were placed in calculated positions and refined using a
riding model except for the H atoms attached to the
partial water molecule, which were not included in the final
model. The putative partial water molecule is not positioned to
be a hydrogen-bond donor to any of the pyrazine N atoms.
UV/vis spectrum (MeCN, 295 K): νmax, 103 cmϪ1 (εmax
M
,
Ϫ1.cmϪ1) 11.3 (3.2), 15.2 (sh), 22.4 (3,200), 29.7 (31,500), 35.8
(sh), 38.9 (48,300), 45.0 (sh). UV/vis spectrum (MeNO2, 295 K):
νmax, 103 cmϪ1 (εmax, MϪ1.cmϪ1) 11.3 (3.9), 16.3 (sh), 22.3 (3,400),
23.5 (sh).
For [Fe(L2Mes)2][ClO4]2ؒH2O (4[ClO4]2ؒH2O): found C, 57.5;
H, 4.9; N, 14.6. Calcd. For C56H56Cl2FeN12O8.H2O: C, 57.5; H,
5.0; N, 14.4%. ES mass spectrum: m/z 1051 [56Fe35ClO4-
(L2Mes)2]ϩ, 476 [56Fe(L2Mes)2]2ϩ
.
Single crystal X-ray structure determinations
X-Ray structure determination of [Fe(L2Mes)2][BF4]2ؒ
5CH3NO2 (4[BF4]2ؒ5CH3NO2). The asymmetric unit contains:
half a complex dication, with Fe(1) lying on the crystallo-
graphic C2 axis y, x, Ϫz; one BF4 anion lying on a general
position; two CH3NO2 molecules lying on general positions;
Single crystals of X-ray quality of 1[BF4]2ؒ3CH3NO2, and
3[BF4]2ؒ0.5{CH3}2COؒ0.1H2O and 4[BF4]2ؒ5CH3NO2 were
grown by diffusion of diethyl ether vapour into solutions of the
complexes in the appropriate solvents. Experimental details for
these structure determinations are given in Table 4. All
structures were solved by direct methods (SHELXS 9733) and
refined by full matrix least-squares on F 2 (SHELXL 9734).
CCDC reference numbers 195908–195910 and 199771.
Ϫ
and, half a CH3NO2 molecule lying across the C2 axis Ϫx,
1
Ϫ
–
Ϫxϩy, Ϫz. The BF4 anion is disordered, and was modelled
3
over three equally occupied orientations. All of the CH3NO2
molecules were also disordered. One of these was modelled over
D a l t o n T r a n s . , 2 0 0 3 , 2 0 5 3 – 2 0 6 0
2059

View Article Online
5 J. M. Holland, J. A. McAllister, C. A. Kilner, M. Thornton-Pett,
A. J. Bridgeman and M. A. Halcrow, J. Chem. Soc., Dalton Trans.,
2002, 548.
6 V. A. Money, I. Radosavljevic Evans, M. A. Halcrow, A. E. Goeta
and J. A. K. Howard, Chem. Commun., 2003, 158.
7 J. M. Holland, S. A. Barrett, C. A. Kilner and M. A. Halcrow, Inorg.
Chem. Commun., 2002, 5, 328.
8 J. Elhaïk, C. A. Kilner and M. A. Halcrow, unpublished work.
9 M. Loï, M. W. Hosseini, A. Jouaiti, A. De Cian and J. Fischer, Eur.
J. Inorg. Chem., 1999, 1981.
10 J. C. Rodriguez-Ubis, R. Sedano, G. Barroso, O. Juanes and
E. Brunet, Helv. Chim. Acta, 1997, 80, 86.
11 J. A. Real, M. C. Muñoz, E. Andrés, T. Granier and B. Gallois,
Inorg. Chem., 1994, 33, 3587.
12 B. J. Childs, D. C. Craig, K. A. Russ, M. L. Scudder and
H. A. Goodwin, Aust. J. Chem., 1994, 47, 891.
13 B. J. Childs, J. M. Cadogan, D. C. Craig, M. L. Scudder and
H. A. Goodwin, Aust. J. Chem., 1997, 50, 129.
2 equally occupied orientations, while another was modelled
using three partial CH3NO2 molecules with a 0.50 : 0.25 : 0.25
occupany ratio. The solvent molecule lying across the C2 axis
was modelled using one orientation bisected by the two-fold
axis with occupancy 0.5; and, a second partial molecule on a
general position near the C2 axis, with occupancy 0.25. The
following refined restraints were applied to the final model: B–F
= 1.39(2), non-bonded F ؒ ؒ ؒ F = 2.27(2), C–N = 1.43(2), N–O =
1.24(2) and non-bonded O ؒ ؒ ؒ O = 2.16(2) and C ؒ ؒ ؒ O 2.29(2)
Å. All crystallographically ordered non-H atoms were refined
anisotropically. All H atoms were placed in calculated positions
and refined using a riding model except for the disordered
solvent methyl H groups, which were fixed in idealised
geometries and whose torsions were not refined.
Other measurements
14 V. Niel, J. M. Martinez-Agudo, M. C. Muñoz, A. B. Gaspar and
J. A. Real, Inorg. Chem., 2001, 40, 3838.
Infra-red spectra were obtained as Nujol mulls pressed between
NaCl windows between 600 and 4,000 cmϪ1 using a Nicolet
Avatar 360 spectrophotometer. UV/vis spectra were obtained
with a Perkin-Elmer Lambda 900 spectrophotometer operating
between 3300 and 200 nm, in 1 cm quartz cells. All NMR
spectra were run on a Bruker DPX250 spectrometer, operating
at 250.1 (1H) and 62.9 (13C) MHz. Electron impact mass spectra
were performed on a Kratos MS50 spectrometer, while electro-
spray mass spectra were obtained on a Micromass LCT TOF
spectrometer, employing a MeOH matrix. CHN microanalyses
were performed by the University of Leeds School of
Chemistry microanalytical service. Melting points are
uncorrected. Differential scanning calorimetry data were
obtained using a Perkin-Elmer Pyris calorimeter, with a
15 R. J. Smithson, C. A. Kilner, A. R. Brough and M. A. Halcrow,
Polyhedron, in the press.
16 C. J. O’Connor, Prog. Inorg. Chem., 1982, 29, 203.
17 E. N. Maslen, C. L. Raston and A. H. White, J. Chem. Soc., Dalton
Trans., 1974, 1803.
18 P. Poganiuch, S. Descurtins and P. Gütlich, J. Am. Chem. Soc., 1990,
112, 3270; R. Hinek, H. Spiering, D. Schollmeyer, P. Gütlich and
A. Hauser, Chem. Eur. J., 1996, 2, 1427; Y. Garcia, O. Kahn,
L. Rabardel, B. Chansou, L. Salmon and J.P. Tuchagues, Inorg.
Chem., 1999, 38, 4663.
19 D. Boinnard, A. Bousseskou, A. Dworkin, J. M. Savariault,
F. Varret and J. P. Tuchagues, Inorg. Chem., 1994, 33, 271.
20 D. L. Reger, C. A. Little, A. L. Rheingold, M. Lam, L. M. Liable-
Sands, B. Rhagitan, T. Concolino, A. Mohan, G. J. Long, V. Briois
and F. Grandjean, Inorg. Chem., 2001, 40, 1508; D. L. Reger,
C. A. Little, V. G. Young jr. and M. Pink, Inorg. Chem., 2001, 40,
2870; D. L. Reger, C. A. Little, M. D. Smith and G. J. Long, Inorg.
Chem., 2002, 41, 4453.
21 R. Jakobi, H. Spiering and P. Gütlich, J. Phys. Chem. Solids, 1992,
53, 267; J. A. Real, H. Bolvin, A. Bousseskou, A. Dworkin, O. Kahn,
F. Varret and J. Zarembowitch, J. Am. Chem. Soc., 1992, 114, 4650.
22 L. Pauling, The Nature of the Chemical Bond, Cornell University
Press, Ithaca, New York, 3rd edn., 1960, pp. 257–264.
23 J. M. Holland, C. A. Kilner, M. Thornton-Pett and M. A. Halcrow,
Polyhedron, 2001, 20, 2829.
24 N. K. Solanki, E. J. L. McInnes, F. E. Mabbs, S. Radojevic,
M. McPartlin, N. Feeder, J. E. Davies and M. A. Halcrow, Angew.
Chem., Int. Ed., 1998, 37, 2221; N. K. Solanki, M. A. Leech,
E. J. L. McInnes, J. P. Zhao, F. E. Mabbs, N. Feeder, J. A. K.
Howard, J. E. Davies, J. M. Rawson and M. A. Halcrow, J. Chem.
Soc., Dalton Trans., 2001, 2083.
temperature ramp of 5 K minϪ1
.
Room-temperature magnetic data were measured using a
Sherwood Scientific magnetic susceptibility balance. Variable
temperature magnetic susceptibility measurements were
obtained in the solid state using a Quantum Design SQUID
magnetometer operating at 1000 G. Scans between 5 and 300 K
were run using a continuous temperature ramp, while for
hysteresis measurements, the sample was poised at each
temperature for 1 minute before measurement. Diamagnetic
corrections for the sample and the sample holder were applied
to the data. Magnetic susceptibility measurements in solution
were obtained by Evan’s method35 using a Bruker DRX500
spectrometer operating at 500.13 MHz. A diamagnetic correc-
tion for the sample, and a correction for variation of the density
of the CD3NO2 solvent with temperature,36 were applied to
these data. Diamagnetic corrections were estimated from
Pascal’s constants.16
25 M. A. Halcrow, C. A. Kilner and M. Thornton-Pett, Acta
Crystallogr., Sect. C., 2000, 56, 1425.
26 See e.g. H. L. Chum, J. A. Vanin and M. I. D. Holanda, Inorg.
Chem., 1982, 21, 1146; L. L. Martin, K. S. Hagen, A. Hauser,
R. L. Martin and A. M. Sargeson, J. Chem. Soc., Chem. Commun.,
1988, 1313; D. W. Blakesley, S. C. Payne and K. S. Hagen, Inorg.
Chem., 2000, 39, 1979; S. G. Telfer, B. Bocquet and A. F. Williams,
Inorg. Chem., 2001, 40, 4818.
Acknowledgements
27 A. B. P. Lever, Inorganic Electronic Spectroscopy, Elsevier,
Amsterdam, 2nd edn., 1984, pp. 458–470.
28 K. H. Sugiyarto, D. C. Craig, A. D. Rae and H. A. Goodwin, Aust.
J. Chem., 1993, 46, 1269.
29 A. T. Baker and H. A. Goodwin, Aust. J. Chem., 1986, 39, 209;
A. T. Baker, P. Singh and V. Vignevich, Aust. J. Chem., 1991, 34,
1041.
The authors gratefully thank Dr H. J. Blythe (University of
Sheffield) for the variable temperature susceptibility measure-
ments. The Royal Society (M. A. H.), the EPSRC (J. E.,
V. A. M.) and the University of Leeds are acknowledged for
financial support.
30 A. Albert, R. Goldacre and J. Phillips, J. Chem. Soc., 1948, 2240.
31 J. Elguero, E. Gonzalez and R. Jacquier, Bull. Soc. Chim. Fr., 1968,
5009.
32 A. L. Rheingold, C. B. White and S. Trofimenko, Inorg. Chem., 1993,
32, 3471.
References
1 P. Gütlich, A. Hauser and H. Spiering, Angew. Chem., Int. Ed. Engl.,
1994, 33, 2024.
2 P. Gütlich, Y. Garcia and H. A. Goodwin, Chem. Soc. Rev., 2000, 29,
419.
3 O. Kahn, Science, 1998, 179, 44.
4 J. M. Holland, J. A. McAllister, Z. Lu, C. A. Kilner, M. Thornton-
Pett and M. A. Halcrow, Chem. Commun., 2001, 577.
33 G. M. Sheldrick, Acta Crystallogr., Sect. A., 1990, 46, 467.
34 G. M. Sheldrick, SHELXL 97, Program for the Refinement of
Crystal Structures, University of Göttingen, Germany, 1997.
35 E. M. Schubert, J. Chem. Educ., 1992, 69, 62.
36 P. Oakley, J. Chem. Soc., 1924, 1194.
D a l t o n T r a n s . , 2 0 0 3 , 2 0 5 3 – 2 0 6 0
2060