Photomagnetic properties of iron(II) spin crossover complexes of 2,6-dipyrazolylpyridine and 2,6-dipyrazolylpyrazine ligands

Article abstract of DOI:10.1039/b601366j

The photomagnetic properties of the following iron(ii) complexes have been investigated: [Fe(L¹)₂][BF₄]₂ (1), [Fe(L²)₂][BF₄]₂ (2), [Fe(L ²)₂][ClO

Full text of DOI:10.1039/b601366j

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/dalton | Dalton Transactions
Photomagnetic properties of iron(II) spin crossover complexes of
2,6-dipyrazolylpyridine and 2,6-dipyrazolylpyrazine ligands†
Chiara Carbonera,a Jose´ Sa´nchez Costa,a Victoria A. Money,b Je´roˆme Elha¨ık,c Judith A. K. Howard,*b
Malcolm A. Halcrow*c and Jean-Franc¸ois Le´tard*a
Received 27th January 2006, Accepted 5th May 2006
First published as an Advance Article on the web 17th May 2006
DOI: 10.1039/b601366j
The photomagnetic properties of the following iron(II) complexes have been investigated:
[Fe(L1)2][BF4]2 (1), [Fe(L2)2][BF4]2 (2), [Fe(L2)2][ClO4]2 (3), [Fe(L3)2][BF4]2 (4), [Fe(L3)2][ClO4]2 (5) and
[Fe(L4)2][ClO4]2 (6) (L1 = 2,6-di{pyrazol-1-yl}pyridine; L2 = 2,6-di{pyrazol-1-yl}pyrazine; L3 =
2,6-di{pyrazol-1-yl}-4-{hydroxymethyl}pyridine; and L4 = 2,6-di{4-methylpyrazol-1-yl}pyridine).
Compounds 1–6 display a complete thermal spin transition centred between 200–300 K, and undergo
the light-induced excited spin state trapping (LIESST) effect at low temperatures. The T(LIESST)
relaxation temperature of the photoinduced high-spin state for each compound has been determined.
The presence of sigmoidal kinetics in the HS → LS relaxation process, and the observation of LITH
hysteresis loops under constant irradiation, demonstrate the cooperative nature of the spin transitions
undergone by these materials. All the compounds in this study follow a previously proposed linear
relation between T(LIESST) and their thermal spin-transition temperatures T1/2: T(LIESST) = T0 −
0.3T1/2. T0 for these compounds is identical to that found previously for another family of iron(II)
complexes of a related tridentate ligand, the first time such a comparison has been made.
Crystallographic characterisation of the high- and low-spin forms of 5 and 6, the light-induced
high-spin state of 5, and the low-spin complex [Fe(L4)2][BF4]2 (7), are described.
to exhibit this effect, although in most cases the resultant
Introduction
metastable high-spin state is only long-lived at temperatures
≤50 K.7–9 However, in some rare cases a photoinduced metastable
One of the more fascinating examples of molecular bistability is
the spin crossover (SCO) phenomenon encountered in some 3d4–
3d7 transition metal compounds.1 Materials of this sort can switch
reversibly between high- and low-spin electronic states in response
to an external perturbation, such as a change in temperature or
pressure, upon laser irradiation, or exposure to a strong magnetic
field.2,3 This phenomenon is consequently of particular interest,
given the potential applications of spin-transition materials as
molecular switches in information storage and display devices.4
Spin-crossover effects are also important in some biological and
geological systems.5
state can be observed at 120 K.10 Elucidation of the relation
between the chemical structure of a LIESST compound, and
the lifetime of its photogenerated excited state, will allow pho-
toswitchable compounds operating at even higher temperatures to
be designed.
In this context, we recently observed that three iron(II) com-
plexes, [Fe(L1)2][BF4]2 (1), [Fe(L2)2][BF4]2 (2) and [Fe(L2)2][ClO4]2
(3) (L1 = 2,6-di{pyrazol-1-yl}pyridine; L2 = 2,6-di{pyrazol-1-
yl}pyrazine) also display long lifetimes for their photoinduced
HS states, that are stable below the thermal relaxation temper-
atures [T(LIESST)11] of 80–100 K.12 As a continuation of this
work, we report here a full study of the photomagnetic proper-
ties of these and three other iron(II) complexes [Fe(L3)2][BF4]2
(4), [Fe(L3)2][ClO4]2 (5) and [Fe(L4)2][ClO4]2 (6) (L3 = 2,6-
di{pyrazol-1-yl}-4-{hydroxymethyl}pyridine and L4 = 2,6-di{4-
methylpyrazol-1-yl}pyridine). Our data will be discussed in terms
of the recently reported empirical linear relation between the ther-
mal spin-transition temperature (T1/2), and the thermal stability of
the photochemically generated high-spin state (T(LIESST), eqn
(1)).9
An increasingly important aspect of SCO research concerns the
ability to control the colour or magnetic response of a molecular
material photochemically. This represents a fascinating challenge,
with clear applications in photonic devices. The photogeneration
of long-lived metastable high-spin states in some iron(II) SCO
materials is well-known, as the light-induced excited spin state
trapping (LIESST) effect.6 Several iron(II) compounds are known
aInstitut de Chimie de la Matie`re Condense´e de Bordeaux, UPR CNRS
9048-Universite´ de Bordeaux 1, Groupe des Sciences Mole´culaires, 87
Av. Doc. A. Schweitzer, F-33608, Pessac, France. E-mail: letard@icmcb-
bordeaux.cnrs.fr
T(LIESST) = T0 − 0.3T1/2
(1)
bDepartment of Chemistry, University of Durham, South Road, Durham,
UK DH1 3LE. E-mail: j.a.k.howard@durham.ac.uk
In particular, our results provide support for the observation
that the intercept of this line, T0, varies in an apparently consistent
way between different classes of compound. We also present
detailed X-ray diffraction studies of the structural basis of spin
crossover in the new compounds 5 and 6.
cSchool of Chemistry, University of Leeds, Woodhouse Lane, Leeds, UK.
LS2 9JT. E-mail: m.a.halcrow@leeds.ac.uk
† Electronic supplementary information (ESI) available: Additional details
of the crystal structure of the low-spin complex salt [Fe(L4)2][BF4]2 (7). See
DOI: 10.1039/b601366j
3058 | Dalton Trans., 2006, 3058–3066
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
is ca. 3.5 cm3 K mol−1, showing that these compounds exhibit
a fully occupied quintet, high-spin state at room temperature.
As previously described, these three compounds all show a
complete and abrupt spin-transition upon cooling, with a small
but reproducible hysteresis loop of 3–4 K. The behaviour of 6 is
similar, showing an abrupt spin-transition at 233 K with 3 K
hysteresis. However, its lower room-temperature vMT value of
2.8 cm3 K mol−1 implies that the sample is heterogeneous, with
ca. 20% being low-spin at room temperature. This is addressed
further below. Finally, 4 and 5 show complete but more gradual
spin-transitions, without hysteresis, with mid-points very close to
room temperature. In all cases, vMT below the spin transitions is
close to 0 cm3 K mol−1, as expected for a fully populated low-spin
state ground state.
Thermal spin-crossover can be also monitored by following the
visible spectrum of the sample as a function of the temperature,
measured by diffuse reflectance. Fig. 2 shows the spectral changes
for two of the compounds measured along their thermal spin
transition. The band at 800–850 nm corresponds to a d–d
transition of the high-spin iron(II) centre, while the absorptions
in the 500–650 nm region can be assigned to both d–d and MLCT
transitions of the low-spin material. Consequently, any change
of the signal can be used to directly monitor the thermal SCO
transition at the surface of the sample.18 Hence, the increase of
the reflectivity signal recorded at 650 nm for 4 and at 610 nm
for 5 upon cooling (Fig. 2 inset) corresponds to the increasing
population of the low-spin state following thermal SCO. The
same reflectivity experiment can also monitor any light-induced
phenomena occurring at the surface of the sample. When the
temperature is sufficiently low that relaxation of the photoinduced
high-spin state is slow, the light intensity at the surface of the
sample can be used to tune the spin-state of the complex. In this
sense, the reflectivity signal, which decreases along the thermal
spin transition, reversely increases at lowest temperature (Fig. 2).
The low → high spin photoconversion can be also inves-
tigated using a SQUID magnetometer coupled to an optical
Results and discussion
The synthesis, thermal magnetochemistry and structural charac-
terisation of 1,12–14 2 and 312,15 and 416 have reported previously.
Compound 5 is a new salt of the same complex dication as
in 4, with which it is isostructural. The new ligand L4 was
synthesised by treatment of 2,6-dibromopyridine with 2.3 equiv.
of potassium 4-methylpyrazolide following the usual procedure.17
Complexation of Fe[ClO4]2·6H2O by two equiv. of L4 in acetone
yielded 6 as mustard yellow microcrystals. A similar reaction using
Fe[BF4]2·6H2O afforded the salt [Fe(L4)2][BF4]2 (7), which is low-
spin at room temperature.
Magnetic and photomagnetic measurements
The temperature dependence of the magnetic susceptibility for 1–6
is shown in Fig. 1, and relevant data from these curves are collected
in Table 1. Interestingly, all six complexes exhibit a thermal spin
crossover with a complete conversion. At 290 K, vMT for 1–3
Fig. 1 Temperature dependence of vMT for 1–6. ꢀ = data recorded in cooling and warming modes without irradiation; ᭺ = data recorded with
irradiation at 10 K; ᭜ = T(LIESST) measurement, data recorded in warming mode with the laser turned off after irradiation for one hour. The solid line
through the T(LIESST) measurement shows the fit generated from the deduced experimental thermodynamic parameters. The insets show the derivate
of the dvMT/dT curves, whose minimum corresponds to T(LIESST).
This journal is
The Royal Society of Chemistry 2006
Dalton Trans., 2006, 3058–3066 | 3059
©

View Article Online
Table 1 Magnetic and photomagnetic properties of the compounds in this studya
T1/2↓/K
T1/2↑/K
T(LIESST)/K
T(LITH)↓/K
T(LITH)↑/K
1
2
3
4
5
6
258
218
199
271
284
231
261
221
203
271
284
234
81
91
100
70
65
84
60
88
74
51
48
40
88
99
105
71
65
85
a T1/2↓ and T1/2↑ are the temperatures at which the sample contains 50% of LS and HS spins in the cooling and warming modes. T(LIESST) is the
temperature at which the light-induced HS information was erased in the SQUID cavity. T(LITH)↑ and T(LITH)↓ are the temperatures where there is
apparently 50% of photoconverted HS molecules in cooling and warming modes under continuous irradiation.
The shapes of all the T(LIESST) curves are almost identical.
The vMT product firstly increases upon warming from 10 K due to
zero-field splitting of the high spin iron(II) ion, reaching a plateau
of ca. 3.5 cm3 K mol−1 near 20 K. For all of 1–6, vMT at the
maximum of the T(LIESST) curve is always close to the value
recorded at room temperature. This confirms that for all these
derivatives, light irradiation induces a 100% complete low → high
spin photoconversion, according to the light-induced excited spin
state trapping effect. The T(LIESST) values of 1–6 are collected
in Table 1. As expected, the highest T(LIESST) temperature is
found for 3, which exhibits the lowest thermal spin transition
temperature (T1/2), in agreement with the inverse energy lap law
introduced by Hauser et al.20 Compound 6 consistently displays
an anomaly upon rewarming, where a fraction of the sample
undergoes LIESST relaxation at lower temperature than the bulk
(Fig. 1). This is discussed further below.
The sharpness of the dvMT/dT vs. T minima (Fig. 1 inset)
for 1–6 implies that intermolecular cooperativity may play a role
in thermal relaxation of their photogenerated high-spin excited
states.19 When the relaxation process follows simple stretched
exponential behaviour, as encountered in non-cooperative SCO
systems,21 the minimum of the dvMT/dT vs. T curve is not so well-
defined.19 We have therefore investigated the LIESST relaxation
kinetics of 1–6, at different temperatures between 10 K and the
highest temperatures accessible with our SQUID set up, which
are close to the T(LIESST) values (Fig. 3). The strong deviation
of these relaxation curves from a single exponential is striking,
and they can be modelled using a sigmoidal law, consistent with
the self-accelerated behaviour predicted for strong cooperative
systems. This cooperativity arises from the large difference in
metal–ligand bond lengths between high-spin and low-spin states,
resulting in elastic interactions caused by the change in internal
pressure inside the solids as the spin transition proceeds.18 Thus,
Fig. 2 Changes in the reflectivity spectra of 4 and 5 upon cooling,
and upon laser irradiation. The inset graphs show the intensity of the
the height of the activation barrier to LIESST relaxation changes
as a function of c HS (the fraction of spin centres in the sample
that are high-spin at a given temperature), and the relaxation rate
k*HL(T, c HS) depends exponentially on both c HS and T (eqn (2)
and (3)), where a(T) (=Ea*/kBT) is the acceleration factor at a
given temperature.
absorption at 650 nm under the same conditions.
source.7–9,11,12,18 For all of 1–6, a drastic increase of the magnetic sig-
nal under light irradiation was observed at 10 K. The limit reached
at photosaturation, after typically 1–2 hours of irradiation, was
close to 3.0 cm3 K mol−1. Fig. 1 shows the T(LIESST) curve
recorded for each compound. In this procedure, the irradiation is
maintained until the signal is saturated, then the light is switched
off and the temperature slightly increased at 0.3 K min−1.7–9,11,19
The minimum of the dvMT/dT vs. T curve defines the limiting
temperature T(LIESST), above which the light-induced magnetic
high-spin information is erased in a SQUID cavity.11
(2)
k*HL (T, c HS) = k HL(T)exp[a(T)1 − c HS)]
(3)
Least squares fits of the data for 1–6 were performed using
a sigmoidal model where kHL(T) and a(T) were refined as free
3060 | Dalton Trans., 2006, 3058–3066
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
Fig. 3 Time dependence at various temperatures of the high spin molar fraction generated by light irradiation at 10 K. Each point represents the high-spin
fraction deduced from the magnetic response measured within the SQUID magnetometer around 30 s. The relaxation curves are fitted according to
sigmoidal behaviour (Table 2).
Table 2 Thermodynamic parameters for thermal relaxation from the
LIESST high-spin state, where k0 is the overestimated limit of the
tunnelling rate constant, k∞ and Ea are respectively the pre-exponential
factor and the activation energy of the activated region, and Ea* the
additional activation energy resulting from cooperativity
An elegant way to test the validity of this model is to use the fitted
parameters to reproduce the experimental T(LIESST) curve.19 For
that it is necessary to carefully take into account the time and
temperature dependences. The main difficulty of this simulation
is to satisfactorily estimate the rate constant k0 [i.e. kHL(T → 0)],
for relaxation by quantum mechanical tunnelling. For this, we
consider that the last complete kinetic measurement recorded at
low temperature can be regarded as an upper limit for the k0 value
(Table 2). Based on this, and by using the deduced Ea, k∞ and
a(T) parameters, the agreement between the calculated and the
experimental T(LIESST) curves was very good (Fig. 1).
Another way to measure the effect of cooperativity on the
photoexcitation and subsequent relaxation processes is to monitor
the magnetic properties of a spin crossover material under constant
irradiation. Competition between the constant photoexcitation
and self-accelerated thermal relaxation processes induces a ther-
mal hysteresis loop near T(LIESST), which is termed light-
induced thermal hysteresis (LITH).11,22 Fig. 4 and Table 1 show
k0/s−1
k∞/s−1
Ea/cm−1
Ea*/cm−1
1
2
3
4
5
6
<7 × 10−6
8 × 10−5
8.0 × 106
9.3 × 104
7.9 × 107
2.5 × 109
4.4 × 107
5.1 × 109
1350
1191
1770
1350
1060
1720
340–390
140–180
180–200
120–130
40–70
<4.4 × 10−5
2 × 10−6
<4 × 10−5
1 × 10−5
200–230
parameters. The calculated curves are shown as solid lines in
Fig. 3, and the fitted parameters are listed in Table 2. The apparent
activation energy, Ea, and the apparent pre-exponential factor, k∞,
of the activated region are calculated from the straight line given
by plotting lnkHL(T) vs. 1/T.
Fig. 4 The temperature dependence of vMT for 1–6 measured in warming and cooling modes under constant irradiation, showing their light-induced
thermal hysteresis (LITH) properties. The LITH data from these graphs are listed in Table 1.
This journal is
The Royal Society of Chemistry 2006
Dalton Trans., 2006, 3058–3066 | 3061
©

View Article Online
the LITH loops observed for 1–6 under constant irradiation. From
these data it is also clear that the photo-induced high spin phase
is characterised by a certain degree of cooperativity. There is only
moderate correlation between the activation energy component
arising from cooperativity in the samples (Ea*, Table 2), which
follows the trend:
5 < 4 < 2 < 3 < 6 < 1
and the magnitude of LITH hysteresis (Table 1):
2 < 4 < 5 < 1 ≈ 3 < 6
However, it is notable that 4 and 5, which exhibit the least
cooperative thermal spin-transitions, also show the smallest Ea*
values and, with one exception, the narrowest LITH loops.
Crystallography
A variable temperature crystallographic study of 5 has shown that
−
it is isostructural with the previously described BF4 salt, 4.16
The compound crystallises in the monoclinic space group Cc; the
iron atom occupies a distorted octahedral environment and no
crystallographic phase transition is found over the temperature of
the experiment, selected crystallographic parameters are presented
in Table 3. The hydroxyl groups of the ligands are found to be
thermally disordered at 340 K. However, this disorder is resolved
on cooling to 120 K and appears to play no role in determining the
course of the spin transition. It was possible to refine anisotropic
temperature factors for 5 only for the data collected at 340 K either
due to crystal damage during the spin transition or the intrinsically
poor quality of the crystals. The spin transition in this compound
lies around ambient temperature, the same temperature at which
crystallisation was performed, and this is the probable reason for
the poor crystal quality. Iron–nitrogen bond lengths at 340 K, are
diagnostic for high-spin iron(II), Table 4, on cooling to 30 K there
˚
is a decrease in the mean metal–ligand bond length of 0.180 A,
indicative of a spin transition from the high- to the low-spin
state taking place between these two temperatures. Monitoring
of the unit cell parameters with temperature, between 340 K and
30 K, demonstrates good agreement with the magnetic data with
a sudden decrease in unit cell volume of 1.7% between 290 K
and 270 K, clearly much greater than would be expected due to
thermal contraction alone. These data show that there is little
anisotropy, the a and b unit cell parameters decreasing by 0.5%
between 290 K and 270 K while the c parameter decreases by
1% over the same temperature range. The b angle increases by
0.06%. The reduction in metal–ligand bond lengths on going from
the high- to the low-spin state results in an increase in ligand bite
angle. In common with the majority of spin crossover compounds,
the low-spin state has a closer to ideal geometry than the high-
spin state, as demonstrated by the lower values of the R and t
parameters, as defined by Guionneau et al.23
Irradiation of the crystal in the low spin state at 30 K with
red laser light (k = 632.8 nm, 25 mW) results in a transition to
the metastable high-spin state, initial evidence for the transition
is provided by the colour change of the crystal from brown to
yellow. Full structural analysis reveals that after irradiation 5
is fully in the metastable high-spin state. It was not possible to
refine anisotropic temperature factors for 5, however the results
are sufficiently accurate to allow differentiation of the high- and
3062 | Dalton Trans., 2006, 3058–3066
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
Table 4 Selected geometrical parameters for 5 and 6
Mean Fe–N/A Bite angle/◦
R/◦
t (%)
˚
T/K
5
6
340
30
30
250
30
2.153(8)
1.97(1)
2.150(6)
2.149(8)
1.947(9)
74.0(3)
79.2(4)
73.8(2)
72.7(2)
79.8(4)
147.7
94.0
146.4
158.8
96.3
7.920
3.750
7.895
8.90
3.44
low-spin states from examination of the metal–ligand bond
lengths. The metal–ligand bond lengths after irradiation are diag-
nostic of an iron(II) centre in the high-spin state Table 4 increasing
˚
by 0.168 A. The R parameters are the same for both the HS-1
and HS-2 indicating that complete conversion to the metastable
high-spin state has been achieved. The increase in unit cell volume
on irradiation is equal to 1.6%; very similar to that observed for
the thermal spin transition. Of the previously reported metastable
high-spin state structures two show significantly smaller increases
in unit cell volume for the light induced transition than for
the thermal spin transition.24 A study of 114 and the previously
communicated results for 416 showed that the difference in volume
induced by irradiation with laser light was the same as that seen
on going through the thermal spin transition. That the change
in volume is the same on going through the thermal and light
induced transitions is likely a result of the, comparatively, high
packing density of this family of materials this also accounts for
the comparatively small effect that cooling has on the volume of the
systems. This point is particularly pertinent when it is considered
that the changes observed in both earlier studies, concern the
conformation of the monodentate NCS− or NCSe− ligands.
The addition of a CH2OH group to the 4 position of the central
pyridine ring of the 2,6-dipyrazol-1-yl-pyridine ligand allows the
formation of some weak hydrogen bonds, but in comparison with
1,14 and the BF4 and ClO4 salts of the 2,6-di(3-methylpyrazol-
1-yl)pyrazine complex15a,c the amount of hydrogen bonding is
reduced. Despite this reduced hydrogen bonding, both 4 and
5 show abrupt thermal spin transitions with narrow hysteresis
loops, indicative of high levels of cooperativity. This observation
suggests that it is the extensive network of p interactions in these
structures which is responsible for the cooperative nature of the
transitions. This hypothesis is further supported by the absence
of these interactions in the 2,6-di(3-methylpyrazol-1-yl)pyrazine
complexes in which the thermal spin transition is gradual and
hydrogen bonding plays a critical role.
Fig. 5 Structure of the [Fe(L4)2]2+ cation at 250 K in the HS-1 state.
Hydrogen atoms have been omitted for clarity and thermal ellipsoids are
at 50% probability.
Cooling to 30 K results in a crystallographic phase transition
¯
from P421c to the orthorhombic space group P212121 which
involves a doubling of the crystallographic c axis. The asym-
metric unit is one cation and two anions and there are four
formula units in the unit cell. Neither independent anion is found
to be disordered at this temperature. The mean iron–nitrogen
˚
distance of 1.947(9) A, and the symmetry parameters, indicate that
the iron(II) centre is in the low spin state, in agreement with the
magnetic data. The loss of symmetry is caused by displacement of
the cations such that the iron atoms are no longer aligned parallel
to c but form a zigzag parallel to the c direction being displaced
from each other in both the a and b directions and no longer being
situated at the origin of the unit cell. This distortion of the packing
prevents clashing of the methyl groups on contraction of the unit
cell due to the spin transition. Monitoring the unit cell parameters
with temperature demonstrates that the spin transition takes place
abruptly between 240 K and 230 K. This is in good agreement with
the results of magnetic experiments which estimate T1/2 as being
232 K.
Irradiation of the crystal on the diffractometer at 30 K with red
laser light (k = 632.8 nm, 25 mW) results in a change in colour
of the crystal from the dark yellow-brown associated with the
low-spin state to the yellow of the high-spin state. However, on
attempting to determine the unit cell parameters of the compound
after irradiation, it was found that the crystal had been irreversibly
damaged and the diffraction peaks had become very broad and
diffuse.
In addition to the above behaviour, the photomagnetic studies
on 6 described below showed a weak but consistent anomaly near
50 K, where a small fraction of the photoexcited sample undergoes
LIESST relaxation at a lower temperature than the bulk. In the
light of these results, a powder diffraction study of the same sample
of 6 used for the photomagnetic study was carried out. It was found
that, although the bulk of the sample was isostructural with the
single crystal described above, it also contained ca. 10% of a second
phase. The lower T(LIESST) shown by this contaminating phase
suggests that it should undergo SCO at a substantially higher
temperature than 232 K. While the diffraction peaks from this
¯
6 crystallises in the tetragonal space group P421c with unit cell
˚
˚
parameters of a = 9.587(1) A, c = 17.933(4) A and V = 1648.1(5)
3
˚
A at 250 K. The asymmetric unit consists of a quarter of the
cation and one half of one anion, although at this temperature the
anion is found to be disordered over the two fold axis, selected
crystallographic parameters are presented in Table 3. The iron
atom is situated on a four fold improper axis and, like all the com-
pounds in this series, has a distorted octahedral geometry; bound
equatorially to two ligands through three of their five nitrogen
atoms, Fig. 5. The mean iron–nitrogen distance at this temperature
is consistent with the compound being in the high-spin state. The
iron atom lies on the origin with the iron–pyridine nitrogen (N3)
bonds along the crystallographic c axis. Neighbouring iron atoms,
therefore, are separated by a distance equal to the length of the c
axis and are sited on a straight line through the unit cell.
This journal is
The Royal Society of Chemistry 2006
Dalton Trans., 2006, 3058–3066 | 3063
©

View Article Online
contaminant were too weak to allow a unit cell to be refined, it is
notable that crystals of the BF4− salt of the same complex molecule
are low spin at 300 K. We therefore suggest that the minor phase of
6 is likely to be isostructural with [Fe(L4)2](BF4)2 (7, ESI†). Be that
as it may, the existence of a minor contaminant in bulk samples
of 6 will not perturb the magnetic or photomagnetic behaviour of
the major phase of this material.
high-spin structures.27,28 These values are almost identical to those
shown by 1–6.13–16,25 Hence, the two series undergo very similar
structural changes during SCO, which causes an average change in
˚
Fe–N bond length of 0.21 A. The “stiffness” of the ligand cage, and
the twisting of the ligands around the metal ion during SCO, will
also be very similar in the two series of compounds. So, from our
previous arguments, despite their very different crystal chemistry
it is reasonable that the light-generated metastable high-spin states
of 1–6 and [Fe(bpp)2]X2 should show essentially identical thermal
stability.
T(LIESST) database
We have now investigated the T(LIESST) and T1/2 values of more
than sixty iron(II) spin-crossover materials of nitrogen ligands,
and have found their photomagnetic properties are governed by
a simple linear relation (eqn (1)).7–9 Four parallel T0 lines with
values of 100, 120, 150 and 200 K have been defined so far,
for different series of iron(II) compounds (Fig. 6). To solve the
challenge of achieving room-temperature photo-bistability, we
must now identify the factors controlling T0, so that materials
with higher T0 [and hence T(LIESST)] values can be designed.9
This discussion assumes that the structures of the thermal
and photoinduced high-spin states of a given material are very
similar. This hypothesis is certainly valid for 1,14 416 and 5 (this
work) for which crystallographic data are available from the
photoinduced HS form. The thermal high-spin state at 300 K,
and the photo-induced high-spin state at 30 K, for all three
compounds are isostructural. But this situation is less clear for
the other members of the series, since 2, 315 and 6 (this work) all
undergo crystallographic phase changes upon thermal SCO. In
the absence of a structure determination it is uncertain whether
photoexcitation of 2, 3 or 6 also induces a crystallographic phase
change. This would significantly affect the thermodynamics of
their LIESST relaxation (Tables 1 and 2), although it would have
only a small effect on their T0 parameters.9
Conclusions
We have reported the thermal and photochemical spin-crossover
properties of six iron(II) complexes based on the tridentate ligand.
Although the structural chemistry of their spin transitions differs
significantly between the compounds, they all obey the same linear
T(LIESST) = T0 − 0.3T1/2 law observed previously for a different
family of iron(II) complex salts of a related tridentate ligand. These
results emphasise the importance of the inner coordination sphere
in controlling the lifetime of a photogenerated high-spin excited
state in these materials.
Fig. 6 Variation of T(LIESST) vs. T1/2 for 1–6 (᭺) in comparison with
We have also recorded the kinetics of the LIESST relaxation
processes undergone by these compounds, and detected the
existence of LITH hysteresis loops under constant irradiation.
Both these studies indicate the importance of intermolecular
cooperativity to the rate of relaxation of the LIESST metastable
high-spin state in the solid. There is no obvious correlation
between the parameters in Tables 1 and 2 and T1/2, the width
of the thermal spin transitions, or the structural chemistry of
the spin transitions for 1–6 (e.g. whether or not they undergo
a phase change during thermal SCO). This implies that other
structural factors, that we cannot take account of, also contribute
to the thermodynamics and kinetics of LIESST decay. These might
include: changes in anion positions and/or motion; the onset of
ligand conformational disorder (as observed in 416 and 5 at 300 K,
for example); or, more subtle differences in the geometries and
strength of the interactions between spin centres in the high- and
low-spin forms of the solids. Experiments designed to shed more
light on this question are in progress.
the literature data listed in ref. 9. The region in gray is meaningless as
T(LIESST) must be be ≤T1/2
.
We have previously suggested that ligand denticity and flex-
ibility, and the degree of distortion of the coordination sphere
away from an ideal octahedral geometry, appear to be the most
important factors in determining T0.7–9 In contrast, factors outside
the inner metal coordination sphere (intermolecular cooperativity,
crystal packing, identity of anions, solvent content) result in
only small perturbations to T0. The present work supports these
suggestions, in that 1–6 lie very close together on the T0 = 150 K
line in Fig. 6, that we have previously derived for a series of com-
pounds [Fe(bpp)2]X2·nH2O (bpp = 2,6-di{pyrazol-3-yl}pyridine,
a regioisomer of L1).8 This is the first time that iron(II) complexes
of two different series of ligands that are stereochemically similar,
but otherwise distinct, have been shown to follow the same
photomagnetic correlation. Both 1–6 and [Fe(bpp)2]X2 contain an
iron(II) ion surrounded by six donor nitrogen atoms, contained in
two ligands coordinated in the meridional planes of an (idealised)
octahedron. The structural changes during SCO are very similar in
the two series of compounds. Structure determinations of different
Experimental
members of the [Fe(bpp)2]X2 family25 show that the average Fe–
Compounds [Fe(L1)2][BF4]2 (1),10–12 [Fe(L2)2][BF4]2 (2) and
[Fe(L2)2][ClO4]2 (3),10,13 [Fe(L3)2][BF4]2 (4)14 and L3 14 were prepared
26,27
˚
˚
N distance is 1.95 A in low-spin compounds,
and 2.16 A in
3064 | Dalton Trans., 2006, 3058–3066
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
by the literature procedures. More detail about the crystal structure
of 7 is given in the ESI.† Unless otherwise stated, all manipulations
were carried out in air using reagent grade solvents, except that
bis(2-methoxyethyl) ether was dried over sodium before use.
was adjusted to 5 mW cm−2, and it was verified that this resulted
in no change in magnetic response due to heating of the sample.
Photomagnetic samples consisted of a thin layer of compound
whose weight was obtained by comparison of the thermal spin
crossover curve with that of a more accurately weighed sample of
the same material. Our previously published standardised method
for obtaining LIESST data was followed.11 After cooling slowly
to 10 K the sample, now in the low-spin state was irradiated and
the change in magnetism followed. When the saturation point
had been reached the laser was switched off and the temperature
increased at a rate of 0.3 K min−1. The magnetisation was measured
every 1 K. T(LIESST) was determined from the minimum of a
dvMT/dT vs. T plot for the relaxation process.
Synthesis of 2,6-di{4-methylpyrazol-1-yl}pyridine (L4)
2,6-Dibromopyridine (1.3 g, 5.4 mmol) and potassium 4-methyl-
pyrazolide [prepared in situ by treatment of 4-methylpyrazole
(1.0 g, 12.2 mmol) with KH (0.49 g, 12.2 mmol)] were stirred
in bis(2-methoxyethyl) ether (100 cm3) under N2 at 130 ◦C for
5 days. Quenching the cooled reaction mixture with excess water
yielded L4 as a white solid which was used without further
purification. Yield 0.58 g, 40%. Found C, 65.0; H, 5.5; N, 29.3%.
Calcd for C13H13N5 C, 65.2; H, 5.5; N, 29.3%. Mp 121–123 ◦C. EI
mass spectrum: m/z 240 [M + H]+. 1H NMR (CDCl3): d 2.16 (s,
6H, CH3), 7.56 (s, 2H, Pz H3), 7.74 (d, 2H, 7.5 Hz, Py H3/5), 7.86 (t,
Crystallography
Single crystals of 5–7 were all grown by slow diffusion of Et2O
vapour into MeNO2 solutions of the compounds. Crystallo-
graphic data for 5 and 6 were collected on a Bruker SMART
CCD29 (x-scan, 0.3◦ per frame) diffractometer using graphite-
1
1H, 7.5 Hz, Py H4), 8.29 (s, 2H, Pz H5). 13C{ H} NMR (CDCl3):
d 8.9 (2C, CH3), 108.3 (2C, Py C3/5), 118.5 (2C, Pz C4), 125.4 (2C,
Pz C5), 141.0 (1C, Py C4), 143.2 (2C, Pz C3), 150.0 (2C, Py C2/6).
˚
monochromated Mo-Ka radiation (k = 0.71073 A). Between
Synthesis of [Fe(L3)2][ClO4]2 (5)
360 K and 110 K the crystals were cooled in a flow of chilled
nitrogen gas using an Oxford Cryosystems Cryostream.30 Below
this the crystals were cooled in a flow of chilled helium using an
Oxford Cryosystems HELIX.31 All data processing was carried out
using the SAINT32 and XPREP33 software packages. Absorption
corrections were applied using SADABS.33 Data for 7 were col-
lected using a Nonius KappaCCD diffractometer, with graphite-
monochromated Mo-Ka radiation, and fitted with an Oxford
Cryosystems nitrogen-cooled low-temperature device. All data
processing was carried out using DENZO34 while an absorption
correction was applied using SORTAV.35 The structures were all
solved by direct methods and refined on F2 using full matrix least-
squares methods within the SHELXTL suite. The hydrogen atoms
were placed geometrically and treated with a riding model. For the
LIESST experiments, the sample was irradiated for ten minutes
whilst on the Bruker SMART diffractometer using a He–Ne laser
(k = 632.8 nm, 25 mW).
Treatment of Fe[ClO4]2·6H2O (0.15 g, 0.42 mmol) with L3 (0.20 g,
0.84 mmol) in acetone (30 cm3) at room temperature yielded a
yellow solution which was then filtered. Concentration of the
solution to ca. 10 cm3 yielded mustard yellow microcrystals, which
were stored overnight at −30 ◦C, then isolated, washed with Et2O
and dried in vacuo. Yield 0.18 g, 57%. Found C, 39.1; H, 3.0; N,
19.3%. Calcd for C24H22Cl2FeN10O10 C, 39.1; H, 3.0; N, 19.0%. ES
mass spectrum: m/z 396 [56Fe(L3)(35ClO4)]+, 242 [L3H].
Synthesis of [Fe(L4)2][ClO4]2 (6)
Method as for 5, using L4 (0.20 g, 0.84 mmol). The brown acetone
solution of the complex yielded mustard yellow microcystals upon
standing at −30 ◦C. Yield 0.31 g, 70%. Found C, 42.5; H, 3.6; N,
19.3%. Calcd for C26H26Cl2FeN10O8 C, 42.6; H, 3.6; N, 19.1%. ES
mass spectrum: m/z 394 [56Fe(L4)(35ClO4)]+, 267 [56Fe(L4)2]2+.
CCDC reference numbers 296687–296689 and 296691–296693.
For crystallographic data in CIF or other electronic format see
DOI: 10.1039/b601366j
Synthesis of [Fe(L4)2][BF4]2 (7)
Treatment of Fe[BF4]2·6H2O (0.14 g, 0.42 mmol) with L4 (0.20 g,
0.84 mmol) in acetone (30 cm3) at room temperature yielded a
brown solution which was then filtered. Concentration of the
solution to ca. 10 cm3 yielded brown microcrystals, which were
stored overnight at −30 ◦C, then isolated, washed with Et2O
and dried in vacuo. Yield 0.23 g, 78%. Found C, 44.0; H, 3.8;
N, 19.9%. Calcd for C26H26B2F8FeN10 C, 44.1; H, 3.7; N, 19.8%.
ES mass spectrum: m/z 621 [56Fe(L4)2(11BF4)]+, 553 [56Fe(L4)2F]+,
314 [56Fe(L4)F]+, 267 [56Fe(L4)2]2+. vMT (298 K) 0.090 cm3 mol−1 K.
[CAUTION: While we have experienced no difficulty in han-
dling 5 and 6, metal–organic perchlorates are potentially explosive
and should be handled with due care in small quantities.]
Other measurements
Electron impact and electrospray (MeCN matrix) mass spectra
were respectively performed with VG AutoSpec and Micromass
LCT TOF spectrometers. CHN microanalyses were performed
by the University of Leeds School of Chemistry microanalytical
service. NMR spectra were run on a Bruker ARX250 spectrometer,
operating at 250.1 MHz (1H) or 62.9 MHz (13C).
Acknowledgements
The authors would like to thank Dr H. J. Blythe (Univer-
sity of Sheffield) for help with magnetic measurements, and
C. A. Kilner (University of Leeds) for the crystal structure of
7. Funding by the EPSRC for two studentships (V.A.M., J.E.) and
a Senior Research Fellowship (J.A.K.H.), the Alliance bilateral
action (J.-F.L, M.A.H.), the network of excellence MAGMANet
Magnetism and photomagnetism
The photomagnetic measurements were performed using a Spec-
trum Physics Series 2025 Kr+ laser (k = 532 nm) coupled via
an optical fibre to the cavity of a MPMS-55 Quantum Design
SQUID magnetometer. The optical power at the sample surface
This journal is
The Royal Society of Chemistry 2006
Dalton Trans., 2006, 3058–3066 | 3065
©

View Article Online
[FP6-515767-2] (J.-F.L.) and the Aquitaine Region (J-.F.L.) is
gratefully acknowledged.
I. Radosavljevic Evans, J. Elha¨ık, M. A. Halcrow and J. A. K. Howard,
Acta Crystallogr., Sect. B, 2004, B60, 41; (c) V. A. Money, J. Elha¨ık,
M. A. Halcrow and J. A. K. Howard, Dalton Trans., 2004, 65.
16 V. A. Money, J. Elha¨ık, M. A. Halcrow and J. A. K. Howard, Dalton
Trans., 2004, 1516.
17 D. L. Jameson and K. A. Goldsby, J. Org. Chem., 1990, 55, 4992.
18 F. Varret, M. Nogue`s and A. Goujon, Photomagnetic properties of some
inorganic solids, ed. J. Miller and M. Drillon, Wiley VCH, Weinheim,
Germany, 2002, vol. 2, p. 257.
References
1 See for general reviews: Spin Crossover in Transition Metal Compounds,
Topics in Current Chemistry, ed. P. Gu¨tlich and H. A. Goodwin,
Springer Verlag, Vienna, 2004, vol. 233–235.
2 P. Gu¨tlich, A. Hauser and H. Spiering, Angew. Chem., Int. Ed. Engl.,
19 J.-F. Le´tard, G. Chastanet, O. Nguyen, S. Marce´n, M. Marchivie, P.
Guionneau, D. Chasseau and P. Gu¨tlich, Monatsh. Chem., 2003, 134,
165.
20 A. Hauser, Coord. Chem. Rev., 1991, 111, 275; A. Hauser, J. Jeftic,
H. Romstedt, R. Hinek and H. Spiering, Coord. Chem. Rev., 1999,
190–192, 471.
21 A. Hauser, J. Adler and P. Gu¨tlich, Chem. Phys. Lett., 1988, 152, 468.
22 A. Desaix, O. Roubeau, J. Jeftic, J. G. Haasnoot, K. Boukheddaden,
E. Codjovi, J. Linare`s, M. Nogues and F. Varret, Eur. Phys. J. B, 1998,
6, 183.
23 P. Guionneau, C. Brigonteix, Y. Barrans, A. E. Goeta, J.-F. Le´tard,
J. A. K. Howard, J. Gaultier and D. Chasseau, C. R. Acad. Sci., Ser.
IIc: Chim., 2001, 4, 161; P. Guionneau, M. Marchivie, G. Bravic, J.-F.
Le´tard and D. Chasseau, J. Mater. Chem., 2002, 12, 2546.
24 J. Kusz, H. Spiering and P. Gu¨tlich, J. Appl. Crystallogr., 2001, 34,
229; M. Marchivie, P. Guionneau, J. A. K. Howard, G. Chastanet, J.-F.
Le´tard, A. E. Goeta and D. Chasseau, J. Am. Chem. Soc., 2002, 124,
194; E. J. MacLean, C. M. McGrath, C. J. O’Connor, C. Sangregorio,
J. M. W. Seddon, E. Sinn, F. E. Sowrey, S. J. Teat, A. E. Terry,
G. B. M. Vaughan and N. A. Young, Chem. Eur. J., 2003, 9, 5314;
A. L. Thompson, A. E. Goeta, J. A. Real, A. Galet and M. C. Mun˜oz,
Chem. Commun., 2004, 1390; J. Kusz, D. Schollmeyer, H. Spiering and
P. Gu¨tlich, J. Appl. Crystallogr., 2005, 38, 528.
25 M. A. Halcrow, Coord. Chem. Rev., 2005, 249, 2880.
26 H. Sugiyarto, D. C. Craig, A. D. Rae and H. A. Goodwin,
Aust. J. Chem., 1994, 47, 869; K. H. Sugiyarto, K. Weitzner, D. C.
Craig and H. A. Goodwin, Aust. J. Chem., 1997, 50, 869; E. Coronado,
M. C. Gime´nez-Lo´pez, C. Gimenez-Saiz, J. M. Mart´ınez-Agudo and
F. M. Romero, Polyhedron, 2003, 22, 2375; M. L. Scudder, D. C. Craig
and H. A. Goodwin, CrystEngComm, 2005, 7, 642.
1994, 33, 1024.
3 A. Bousseksou, N. Negre, M. Goiran, L. Salmon, J.-P. Tuchagues,
M.-L. Boillot, K. Boukheddaden and F. Varret, Eur. Phys. J. B, 2000,
13, 451.
4 O. Kahn and C. J. Martinez, Science, 1998, 279, 44; P. Gu¨tlich, Y. Garcia
and T. Woike, Coord. Chem. Rev., 2001, 219–221, 839; J.-F. Le´tard, P.
Guionneau and L. Goux-Capes, Top. Curr. Chem., 2004, 235, 221.
5 D. R. Davydov, G. H. B. Hoa and J. A. Peterson, Biochemistry, 1999, 38,
751; M. F. Perutz, G. Fermi, B. Luisi, B. Shannan and R. C. Liddington,
Acc. Chem. Res., 1987, 20, 309; W. R. Schiedt and C. A. Reed, Chem.
Rev., 1981, 81, 543; J. Badro, G. Fiquet, F. Guyot, J. P. Rueff, V. V.
Struzhin, G. Vanko and G. Monaco, Science, 2003, 300, 789.
6 J. J. McGarvey and I. Lawthers, J. Chem. Soc., Chem. Commun., 1982,
906; S. Decurtins, P. Gu¨tlich, C. P. Ko¨hler, H. Spiering and A. Hauser,
Chem. Phys. Lett., 1984, 105, 1.
7 J.-F. Le´tard, L. Capes, G. Chastanet, N. Moliner, S. Le´tard, J. A. Real
and O. Kahn, Chem. Phys. Lett., 1999, 313, 115.
8 S. Marcen, L. Lecren, L. Capes, H. A. Goodwin and J.-F. Le´tard, Chem.
Phys. Lett., 2002, 358, 87.
9 J.-F. Le´tard, P. Guionneau, O. Nguyen, J. S. Costa, S. Marcen, G.
Chastanet, M. Marchivie and L. Capes, Chem. Eur. J., 2005, 11, 4582.
10 T. Buchen, P. Gu¨tlich and H. A. Goodwin, Inorg. Chem., 1994, 33, 4573;
T. Buchen, P. Gu¨tlich, K. H. Sugiyarto and H. A. Goodwin, Chem.
Eur. J., 1996, 2, 1134; C. H. Wu, J. Jung, P. K. Gantzel, P. Gu¨tlich and
D. N. Hendrickson, Inorg. Chem., 1997, 36, 5339; F. Renz, H. Oshio, V.
Ksenofontov, M. Waldeck, H. Spiering and P. Gu¨tlich, Angew. Chem.,
Int. Ed., 2000, 39, 699; S. Hayami, Z.-Z. Gu, Y. Einaga, Y. Kobayashi,
Y. Ishikawa, Y. Yamada, A. Fujishima and O. Sato, Inorg. Chem., 2001,
40, 3240.
11 J.-F. Le´tard, P. Guionneau, L. Rabardel, J. A. K. Howard, A. E. Goeta,
D. Chasseau and O. Kahn, Inorg. Chem., 1998, 37, 4432.
12 V. A. Money, J. S. Costa, S. Marcen, G. Chastanet, J. Elha¨ık, M. A.
Halcrow, J. A. K. Howard and J.-F. Le´tard, Chem. Phys. Lett., 2004,
391, 273.
13 J. M. Holland, J. A. McAllister, Z. Lu, C. A. Kilner, M. Thornton-Pett
and M. A. Halcrow, Chem. Commun., 2001, 577; 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.
14 V. A. Money, I. Radosavljevic Evans, M. A. Halcrow, A. E. Goeta and
J. A. K. Howard, Chem. Commun., 2003, 158.
15 (a) J. Elha¨ık, V. A. Money, S. A. Barrett, C. A. Kilner, I. Radosavljevic
Evans and M. A. Halcrow, Dalton Trans., 2003, 2053; (b) V. A. Money,
27 K. H. Sugiyarto, W.-A. McHale, D. C. Craig, A. D. Rae, M. L. Scudder
and H. A. Goodwin, Dalton Trans., 2003, 2443.
28 K. H. Sugiyarto, M. L. Scuddler, D. C. Craig and H. A. Goodwin,
Aust. J. Chem., 2000, 53, 755.
29 SMART-1000 CCD, Bruker AXS, Madison, WI.
30 Cryostream Cooler, Oxford Cryosystems Ltd, Oxford, UK.
31 Oxford Helix, Oxford Cryosystems Ltd, Oxford, UK.
32 G. M. Sheldrick, SHELXS-97, SHELXL 97. Program for the Refine-
ment of Crystal Structures, University of Go¨ttingen, Germany, 1997.
33 G. M. Sheldrick, SHELXTL Version 5.1. Bruker AXS., Madison, WI,
1998.
34 Z. Otwinowski and W. Minor, Methods Enzymol., 1997, 276, 307.
35 R. H. Blessing, Acta Crystallogr., Sect. A, 1995, 51, 33.
3066 | Dalton Trans., 2006, 3058–3066
This journal is
The Royal Society of Chemistry 2006
©