Photomagnetism of a sym-cis-dithiocyanato iron(II) complex with a tetradentate N,N'-bis(2-pyridylmethyl)1,2-ethanediamine ligand

Article abstract of DOI:10.1002/chem.201102637

A comprehensive study of the magnetic and photomagnetic behaviors of cis-[Fe(picen)(NCS)₂] (picen=N,N'-bis(2-pyridylmethyl)1,2- ethanediamine) was carried out. The spin-equilibration was extremely slow in the vicinity of the thermal spin-transition. When the cooling speed was slower than 0.1 K min^-1, this complex was characterized by an abrupt thermal spin-transition at about 70 K. Measurement of the kinetics in the range 60-70 K was performed to approach the quasi-static hysteresis loop. At low temperatures, the metastable HS state was quenched by a rapid freezing process and the critical T(TIESST) temperature, which was associated with the thermally induced excited spin-state-trapping (TIESST) effect, was measured. At 10 K, this complex also exhibited the well-known light-induced excited spin-state-trapping (LIESST) effect and the T(LIESST) temperature was determined. The kinetics of the metastable HS states, which were generated from the freezing effect and from the light-induced excitation, was studied. Single-crystal X-ray diffraction as a function of speed-cooling and light conditions at 30 K revealed the mechanism of the spin-crossover in this complex as well as some direct relationships between its structural properties and its spin state. This spin-crossover (SCO) material represents a fascinating example in which the metastability of the HS state is in close vicinity to the thermal spin-transition region. Moreover, it is a beautiful example of a complex in which the metastable HS states can be generated, and then compared, either by the freezing effect or by the LIESST effect. Copyright

Full text of DOI:10.1002/chem.201102637

DOI: 10.1002/chem.201102637
Photomagnetism of a sym-cis-Dithiocyanato Iron(II) Complex with a
Tetradentate N,N’-Bis(2-pyridylmethyl)1,2-ethanediamine Ligand
Jean-FranÅois Lꢀtard,*[a] Saket Asthana,[a, d] Helena J. Shepherd,[c] Philippe Guionneau,[a]
Andrꢀ E. Goeta,[c] Naohiko Suemura,[b] Ryuta Ishikawa,[b] and Sumio Kaizaki[b]
In memory of Andrꢀ E. Goeta
Abstract: A comprehensive study of
the magnetic and photomagnetic be-
perature, which was associated with the
thermally induced excited spin-state-
trapping (TIESST) effect, was mea-
sured. At 10 K, this complex also ex-
hibited the well-known light-induced
excited spin-state-trapping (LIESST)
as a function of speed-cooling and light
conditions at 30 K revealed the mecha-
nism of the spin-crossover in this com-
plex as well as some direct relation-
ships between its structural properties
and its spin state. This spin-crossover
(SCO) material represents a fascinating
example in which the metastability of
the HS state is in close vicinity to the
thermal spin-transition region. More-
over, it is a beautiful example of a com-
plex in which the metastable HS states
can be generated, and then compared,
either by the freezing effect or by the
LIESST effect.
haviors
of
cis-[FeACTHUGNTREN(NUG picen)ACHTGNUTREN(NUGN NCS)2]
(picen=N,N’-bis(2-pyridylmethyl)1,2-
ethanediamine) was carried out. The
spin-equilibration was extremely slow
in the vicinity of the thermal spin-tran-
sition. When the cooling speed was
slower than 0.1 Kminꢀ1, this complex
was characterized by an abrupt thermal
spin-transition at about 70 K. Measure-
ment of the kinetics in the range 60–
70 K was performed to approach the
quasi-static hysteresis loop. At low
temperatures, the metastable HS state
was quenched by a rapid freezing pro-
effect and the TACTHNUTRGNE(NUG LIESST) temperature
was determined. The kinetics of the
metastable HS states, which were gen-
erated from the freezing effect and
from the light-induced excitation, was
studied. Single-crystal X-ray diffraction
Keywords: iron · ligand effects ·
photomagnetism · spin crossover ·
X-ray diffraction
cess and the critical TACTHUNTGRNEUNG(TIESST) tem-
Introduction
phenomenon.[2–5] Materials of this kind combine: 1) low ad-
dressing power (about 5 mWcmꢀ2); 2) short addressing time
(nanosecond scale); 3) perfect reproducibility over succes-
sive cycles, even in a solid matrix; and 4) optical reversibili-
The development of materials that allow us to obtain infor-
mation at the single-molecular level or on assemblies of
molecules is important for the future development of infor-
mation technology.[1] Among all of the reported systems,
one interesting class of molecular switches are those that in-
volve the light-induced excited spin-state-trapping (LIESST)
1
5
ty. The low-spin (LS, A1)!high-spin (HS, T2) transition in
iron(II) spin-crossover (SCO) materials can be induced with
green light and the back-conversion to the LS state can be
induced with red light. In fact, the main limitation comes
from the stability of the photoinduced metastable HS state.
Below 50 K, the lifetime is almost infinite and optical data
storage can be envisaged, whilst at higher temperatures, the
relaxation is governed by the activated regime and the pho-
toinduced HS state decays within seconds or less. To over-
come this limitation, it has been recently reported that, in
some iron(II) molecular materials, the light-induced phe-
nomenon can be generated in the center of a thermal hyste-
resis.[4] However, the integration of such molecular systems
into a “real” device[5] would require an iron(II) spin-cross-
over (SCO) material with a thermal hysteresis loop of
around 100 K in width, a value that has not yet been
reached.
[a] Dr. J.-F. Lꢀtard, Dr. S. Asthana, Dr. P. Guionneau
CNRS, Universitꢀ Bordeaux, ICMCB
Groupe des Sciences Molꢀculaires
87 Av. Doc. A. Schweitzer, 33608 Pessac (France)
Fax : (+33)(0)540002678
E-mail: letard@icmcb-bordeaux.cnrs.fr
[b] N. Suemura, R. Ishikawa, Prof. S. Kaizaki
Department of Chemsitry, Graduate School of Science
Osaka University, Toyonaka
Osaka, 540-0043 (Japan)
[c] H. J. Shepherd, Dr. A. E. Goeta
Crystallography group, Durham University
Durham DH1 3LE (UK)
An alternative strategy is to identify the factors that influ-
ence the stability of the light-induced metastable state at
low temperatures, that is, outside the thermal hysteresis,
through the LIESST phenomenon.[3] Along these lines, we
[d] Dr. S. Asthana
Present address:
Department of Physics, Indian Institute of Technology
Hyderabad (India)
5924
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 5924 – 5934

FULL PAPER
introduced the idea of systematically determining the T-
ACHTUNGTRENNUNG
(LIESST) value,[6a] which represents the limiting tempera-
ture above which the light-induced magnetic HS information
is erased when the temperature in a SQUID magnetometer
is increased at a rate of 0.3 Kminꢀ1. Using this procedure,
we compared the photomagnetic properties of more than
sixty SCO materials[6,7] and we found that factors such as co-
operativity, the nature of the salt, and the degree of hydra-
tion had a relatively negligible effect on the stability of the
metastable state.[8,9] In contrast, the influence of the inner
FeN6 coordination sphere, owing to the nature of the ligand,
appeared to have a significant effect on the stabilization of
the photoinduced HS state.[9,10] Following this idea, we fo-
cused our attention on macrocyclic ligands[11] and we recent-
Figure 1. Magnetic properties as a function of the temperature and cool-
ing rate for both the cooling and warming modes.
ly demonstrated that an [Fe
complex with the L222N5 macrocyclic Schiff-base ligand 2,13-
dimethyl-3,6,9,12,18-pentaazabicyclo[12.3.1]octadeca-
ACHTGNURTNE(NUGN L222N5)(CN)2]·H2O iron(II)
AHCTUNGTRENNUNG
1(18),2,12,14,16-pentaene, which was low-spin at up to about
400 K, exhibited an exceptionally long-lived photoinduced
within the expected region for a HS iron(II) center. When
the temperature was lowered, in agreement with the report
by Toflund et al.,[13] the importance of the HS fraction de-
pended on the cooling rate. When the sample was cooled at
a rate of 7 Kminꢀ1, the cMT product gradually decreased
and no thermal spin-transition was clearly observed. The re-
sidual HS faction (gHS), which was deduced from the ratio
HS state, with a TACHTUNGTRENNUNG
(LIESST) temperature of about 105 K.[11a]
In other words, modifying the inner coordination sphere ap-
peared to be a promising strategy for stabilizing the meta-
stable HS state. Such an alteration around the iron(II) metal
center seemed to affect the kinetics of the structural rear-
rangement that accompanied the photoinduced HS!LS re-
laxation by acting on the vibrational properties[3g,12] and/or
by distorting the inner core of the complex.
cMT/ACHTUNTRGNEU(NG cMT)HT, where (cMT)HT represented the magnetic re-
sponse of the HS state at high temperature, was almost
equal to 0.68 at 40 K, where the influence of the zero-field
splitting was negligible. During a subsequent raising of the
temperature, the magnetic signal exhibited an unusual de-
crease at about 67 K, and then increased at about 70 K to
recover the magnitude of the HS state. When the cooling
rate was 0.5 Kminꢀ1, the thermal spin-transition became
clearer. The HS faction at 40 K was 0.36 and the unusual
bump during warming became less-pronounced. In fact, it
was only when the cooling rate was slower than 0.1 Kminꢀ1
that the thermal spin-transition was almost complete. The
residual HS fraction at 40 K was 0.09. On warming, a classi-
cal spin-transition occurred with an unusual bump at around
65 K. T1/2(fl) and T1/2(›), which define the temperatures at
which there is 50% of both LS and HS molecules during the
cooling and warming modes, respectively, were 65 K and
71 K, which led to a hysteresis loop of about 6 K. These
data demonstrated that the spin-equilibration in this com-
pound was very slow, particularly in the vicinity of the ther-
mal spin-transition. As a consequence, even if the cooling
speed was very slow, that is, 0.1 Kminꢀ1, the measured hys-
teresis would remain an apparent one, as demonstrated by
the unsymmetrical form of the cooling branch versus the
warming branch. In fact, the “real” hysteresis loop would
only be obtained by recording the hysteresis loop over an
infinite time, that is, when a steady state between the popu-
lation and relaxation is achieved. Otherwise, we are dealing
with a dynamic hysteresis.
We investigated the photomagnetic properties of cis-[Fe-
ACHTUNGTRENNUNG(picen)ACHTUNGTRENNUNG(NCS)2], which employed a tetradentate ligand,
picen=N,N’-bis(2-pyridylmethyl)1,2-ethanediamine
(Scheme 1) that was originally synthesized by Toftlund
et al.[13] These authors, and later Gꢂtlich and co-workers[14]
and Kaizaki and co-workers[15], reported that this complex
exhibited a more or less incom-
plete thermal spin-transition at
around 70 K with the possibility
of trapping the HS state at low
temperatures by rapid freezing.
Herein, we report a detailed
study of the photomagnetic
properties and a Raman charac-
terization of this complex. The
potential of this complex for ex-
hibiting metastable HS states
that were either generated by
freezing (named HS2) or by
light irradiation (named HS3) is
Scheme 1. The cis-diisothiocya-
nato-N,N’-bis(2-pyridylmeth-
yl)1,2-ethanediamine iron(II)
complex.
discussed.
Results and Discussion
Magnetic studies: Figure 1 shows the magnetic data of the
complex as function of the cooling rate. The magnetic re-
sponse is expressed as cMT versus T, where cM is the molar
magnetic susceptibility and T is the temperature. At room
temperature, the cMT value (about 3.3 cm3 Kmolꢀ1) was
To approach the final form of the quasi-static (that is,
real) hysteresis loop, we determined some stationary points
on the cooling branch. We used the following procedure:
Chem. Eur. J. 2012, 18, 5924 – 5934
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5925

J.-F. Lꢀtard et al.
The sample was rapidly frozen from 290 K to a given tem-
perature (TQ) over a few seconds and the cMT product was
recorded as a function of time. The experiment was stopped
when the stationary limit, that is, spin-equilibration, was
reached, typically after 5 h. The temperature was finally
raised to 290 K and maintained at this temperature for
5 min to ensure that any LS fraction was erased. Sequential-
ly, the TQ value was fixed at 60, 62, 65, 67, and 70 K.
Figure 2 shows the different kinetics and stationary limits, as
well as a proposed profile of the “real” hysteresis curve. As
expected, this hysteresis was much narrower than the “ap-
parent” one, that is, about 2 K, as compared to 6 K (see
above). Interestingly, the character of the real hysteresis
loop was almost symmetric.
Figure 3. Temperature-dependence of the cMT value of the metastable
HS state that was generated at 10 K by rapid freezing. (^) Data record-
ed in the warming mode (0.3 Kminꢀ1) after the trapping effect; (&) Data
recorded in the cooling (0.1 Kminꢀ1) and warming modes (0.3 Kminꢀ1).
Insets show the derivate of the dcMT/dT curves, whose minimum corre-
sponds to the TACTHNUTRGNEU(GN TIESST) temperature.
over this temperature range.[17] Such a situation was in con-
trast to the behavior at higher temperatures, where at
around 45 K, the cMT product started to decrease dramati-
cally, before finally returning to its initial value at 65 K and
following the warming curve. The relaxation process seemed
to be in the activated regime, which was regarded as a tun-
neling from the thermally populated vibrational levels of the
HS phase.[3d–g] The minimum of the first derivative of the
cMT versus T plot gave a TACTHNUTRGNEUGN(TIESST) value of 61 K.
Figure 2. Evolution of the cMT product versus T (*) at a cooling rate of
0.1 Kminꢀ1 and a warming rate of 0.3 Kminꢀ1. These parameters give the
apparent hysteresis loop. (*) Kinetics of the quenched HS state for dif-
ferent temperatures; each relaxation leads to a stationary limit that de-
scribes the quasi-static hysteresis loop (b).
Following this study of the metastable HS2 state that was
generated by a freezing effect, Figure 4 shows the time de-
pendence of the quenched HS fraction (gHS) at selected tem-
peratures. The kinetics curves showed a clear deviation from
single exponential, whilst the relaxation rate was slow
during a first stage and then accelerated in a second stage.
This behavior was modeled by using a sigmoidal law that
Quenched metastable HS state, HS2: In their original work,
Toflund et al.[13] reported that the metastable HS phase
could be trapped by rapid freezing. Along those lines,
Figure 3 shows the magnetic response after freezing. Rapid-
freezing experiments were performed by cooling the cane
from room temperature to 10 K over a few seconds. The
magnetic response was then recorded and the temperature
was slowly warmed at a rate of 0.3 Kminꢀ1, which corre-
sponded to the measurements on a TACHTUNTRGNEUNG(TIESST) curve. These
data revealed a decrease in the magnetic signal in the range
10–20 K, owing to the zero-field splitting (and thermal pop-
ulation of the higher levels) of a HS iron(II) metal center in
pseudo-octahedral surroundings.[16] At around 20–30 K, the
magnetic signal was more or less constant at
3.19 cm3 molꢀ1 K, which was close to that of the HS phase
(3.28 cm3 molꢀ1 K at 110 K). This result indicated that the
metastable HS2 phase was quantitatively trapped at low
temperatures and, moreover, that the relaxation process was
extremely slow up to 30 K with regards to the time needed
Figure 4. Time-dependence at various temperatures of the high-spin
molar fraction that was generated by rapid freezing at 10 K. Each point
represents the high-spin fraction from the magnetic response that was
measured within the SQUID magnetometer for around 30 s. The relaxa-
tion curves are fitted according to sigmoidal behavior (Table 1).
for recording the T
ACHTUNGTRENNUNG(TIESST) curve. Consequently, the relax-
ation process was clearly governed by the tunneling regime
5926
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 5924 – 5934

Photomagnetism of a sym-cis-Dithiocyanato Iron(II) Complex
FULL PAPER
was consistent with the predicted self-accelerated behavior
for strongly cooperative systems.[3d–g,18] This cooperativity
ꢀ
arises from the large difference in metal ligand bond-
lengths between the HS and LS states, thereby resulting in
elastic interactions that are caused by the change in internal
pressure inside the solids during the spin-transition process-
es. Thus, the height of the activation barrier to relaxation
changes as a function of gHS, and the relaxation rate kꢁ
HLACTHNUTRGENUG(N T,
gHS) depends exponentially on both gHS and T (Equa-
tions (1) and (2), where a(T) (=Ea*/kBT) is the acceleration
factor at a given temperature.
@gHS
@T
¼ kꢁHLgHS
ð1Þ
ð2Þ
Figure 5. Temperature-dependence of the cMT value of the metastable
HS state that was generated at 10 K by the LIESST effect. (&) Data re-
corded in the cooling and warming modes without irradiation; (*) data
kꢁHLðT; gHSÞ ¼ kHLðTÞ exp½aðTÞð1 ꢀ gHSÞꢂ
recorded with irradiation at 10 K; (^) TACTHUNRGTNE(NUG LIESST) values, data recorded
in the warming mode with the laser turned off after irradiation for 1 h.
Insets show the derivate of the dcMT/dT curves, whose minimum corre-
sponds to the TACTHNUTRGNEU(GN LIESST) temperature.
Based on this treatment, the calculated curves of the re-
laxation are shown as solid lines in Figure 4, and the fitted
parameters are listed in Table 1. The apparent activation
energy (Ea) and the apparent pre-exponential factor (k1) of
the activated region were calculated from the straight line
given by plotting lnkHL(T) versus 1/T.
Table 1. Thermodynamic parameters for thermal relaxation from meta-
stable HS states that were generated at 10 K either by rapid freezing or
by optical excitation.[a]
k1 [sꢀ1
]
Ea [cmꢀ1
]
Ea* [cmꢀ1
]
metastable HS2 state
metastable HS3 state
5ꢃ10ꢀ1
1.8ꢃ101
300
350
60–110
60–80
[a] k1 and Ea are the pre-exponential factor and the activation energy of
the activated region, respectively; Ea* is the additional activation energy
that results from cooperativity.
Photoinduced metastable HS state, HS3: The photomagnetic
properties were studied by using a SQUID magnetometer
that was connected to a CW optical source. A dramatic in-
crease of the magnetic signal under light irradiation was ob-
Figure 6. Time-dependence at various temperatures of the high-spin
molar fraction that was generated by rapid freezing at 10 K. Each point
represents the high-spin fraction from the magnetic response that was
measured within the SQUID magnetometer for around 30 s. The relaxa-
tion curves are fitted according to sigmoidal behavior (Table 1).
served at 10 K. Figure 5 shows the TACHTUNTRGNEUNG(LIESST) curve. In this
procedure, the irradiation was maintained until the signal
was saturated, then the light was switched off, and the tem-
perature was raised by 0.3 Kminꢀ1. The minimum of the
dcMT/dT versus T curve defined the limiting temperature, T-
Figure 6 shows the kinetic data that were recorded by
using the SQUID magnetometer over the 40–60 K tempera-
ture range, in which the HS3!LS state relaxation was ther-
mally activated. The relaxation curves strongly deviated
from single exponential and sigmoidal behavior was used.
The calculated curves are shown as solid lines in Figure 6,
and the fitted parameters are listed in Table 1. The photo-
magnetic properties of this complex appeared to be very
close to the metastable HS2 state that was generated by
freezing. The only difference was in the stability of the two
metastable states. Interestingly, the metastable HS3 state
that was generated by optical excitation was slightly more-
stable than the metastable HS2 state. This result was particu-
larly emphasized by comparing the kinetic data at 40 K (cf.
Figure 3 (HS2) and Figure 6 (HS3)). Over more than 12 h,
ACHTUNGTRENNUNG
ACHTUNGTRENNUNG
ACHTUNGTRENNUNG
upon warming from 10 K, owing to the zero-field splitting of
the high-spin iron(II) ion, and reached a plateau at about
30 K. Comparison of the cMT value at the maximum of the
TACHTUNGTRENNUNG(LIESST) curve with the magnetic response that was re-
corded at room temperature indicated that, under bulk con-
ditions, quantitative photoconversion occurred. The mini-
mum of the dcMT/dT versus T curve was used to determine
the limiting TACHTUNGTRENNUNG(LIESST) temperature (61 K).
Chem. Eur. J. 2012, 18, 5924 – 5934
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5927

J.-F. Lꢀtard et al.
the relaxation of the metastable HS3 state at 40 K represent-
ed only few percent, whilst for HS2, almost 90% relaxation
was observed. This difference was also manifested by care-
was characterized by two bands at around 2070 and
2100 cmꢀ1 that were due to the C N stretch of the NCS
ꢀ
ligand (Figure 8). The lower- and higher-frequency bands
corresponded to the high- and low-spin states, respectively.
Nevertheless, the intensities of such Raman bands only
slightly varied with temperature. The high-spin fractions
fully considering the shapes of the T
ACHTUNGTREN(NUNG TIESST) and T-
ACHTUNGTRENNUNG(LIESST) curves. In the latter case, the HS fraction was still
relatively high when the warming branch of the thermal
spin-transition was reached.
were approximated as nHS =IHS/ACHTUNGETRNN(UG IHS+ILS), where IHS (of the
In 1998, it was reported that, under permanent irradia-
high-spin species at around 2070 cmꢀ1) and ILS (the low-spin
species at around 2100 cmꢀ1) are Raman-integrated band-in-
tensities at the observed temperatures. Plots of the high-spin
fraction from the Raman-band intensities versus T did not
behave in a similar manner to those obtained from the mag-
netic susceptibility (cMT; Figure 2 and Figure 3). In contrast
to the deep drop to 100% of the low-spin fraction between
the transition temperatures, the picen complex only exhibit-
ed a shallow dip below T1/2 (Figure 9), which reflected the
slow relaxation from high- to low-spin and confirmed the
shape of the LITH loop (Figure 7). Interestingly, the
LIESST of the picen complex, which was reported to be
about 40% by Mçssbauer spectroscopy using broad-band
(350–650 nm) excitation with a Xe arc lamp at 10 K,[14] was
found to be almost quantitative by photomagnetic study and
Raman spectroscopy, thus illustrating the importance of
light penetration in such deep-colored materials for discus-
sing the occurrence of the photoinduced metastable HS
state.
tion, phase I of the [Fe
BiA=N-2’-pyridylmethylene-4-aminobiphenyl)
a new kind of thermal hysteresis in the vicinity of the T-
ACHUGTTERNN(UG PM-BiA)2ACHTNUGTR(EUNNNG NCS)2] complex (PM-
displayed
ACHTUNGTRENNUNG
(LIESST) temperature.[6a] This phenomenon was termed
light-induced thermal hysteresis (LITH).[6a] More recently,
Varret and co-workers reported a similar effect for the
[FexCo
1ꢀxACHTUNGTRENNUNG(btr)2ACHTUNGTRENNUNG(NCS)2]·H2O system (btr=4,4’-bis-1,2,4-tria-
zole) and a nonlinear macroscopic master equation was pro-
posed that was based on the competition between the con-
stant photoexcitation and the self-accelerated thermal-relax-
ation process.[19] Figure 7 shows the LITH loop for cis-[Fe-
ACHTUNGTRENNUNG(picen)ACHTUNGTRENNUNG(NCS)2]. Under permanent irradiation, the tempera-
ture was slowly cooled to 10 K at a rate of 0.1 Kminꢀ1 and
then warmed to 100 K at a rate of 0.3 Kminꢀ1. On first view,
the shape of the LITH slope was relatively unusual. The HS
fraction remained almost close to unity and the occurrence
of the thermal spin-transition was almost absent. In fact, this
latter property was masked by the population of the photo-
induced HS state, owing to the close vicinity between the
thermal spin-transition and the metastability region, which
induced a very slow spin-equilibration. Thus, this optical
stimulus was enough to almost maintain the HS state over
the whole temperature range.
X-ray diffraction: Variable-temperature single-crystal X-ray
diffraction experiments were performed to determine the
structure–property relationships and to investigate the ther-
mal-spin-crossover mechanism. In addition, the photomag-
netic behavior was explored by single-crystal X-ray diffrac-
tion at very low temperatures coupled with light irradiation
of the sample. First, the crystal structure was solved in the
high spin at a chosen temperature (120 K) that was close
enough to the transition to minimize the thermal effects in
front of the spin-crossover effects on the structural proper-
ties when subsequently comparing the HS and LS crystal
structures. Then, the crystal structure was solved at 30 K, at
which the sample was supposed to be made of a mixture of
HS and LS spin states, depending on the cooling rate. For
this reason, four structure-determinations were performed
at 30 K, which corresponded to different cooling rates: true
flash freezing[10a] was performed first and denoted as the
“flash” crystal structure, then from 120 to 30 K speed cool-
ing of 6, 0.5, and 0.05 Kminꢀ1 were used to give the “fast”,
“medium”, and “slow” crystal structures, respectively. The
light-irradiation experiment was performed starting from
the slow conditions, thereby giving the “slow*” data
(Table 2).
Figure 7. Temperature-dependence of the cMT value of the metastable
HS state that was generated under continuous light irradiation in the
cooling and warming modes (^) and data recorded in the cooling and
warming modes without irradiation (&).
The crystal packing consisted of [FeACHTNUTRGNENUG(picen)ACHUTNGTREN(NUGN NCS)2] sheets
that were parallel to the ac plane (Figure 10). Intermolecu-
lar interactions built the cohesion of the crystal packing
through p stacking and hydrogen bonding within the sheets
and hydrogen-like bonding involving sulfur atoms between
complexes of neighboring sheets. This overall description
was similar in all of these above phases and was not affected
Raman spectroscopy: Next, we attempted to directly ob-
serve the thermal- and light-induced spin-transitions by vari-
able-temperature laser Raman spectroscopy. As previously
indicated,[15] the Raman spectrum of cis-[Fe
ACTHNUTRGNNEG(U picen)ACHTUNTGERN(NUNG NCS)2]
5928
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 5924 – 5934

Photomagnetism of a sym-cis-Dithiocyanato Iron(II) Complex
FULL PAPER
sphere geometry. It is well-es-
tablished that spin-crossover
corresponds
to
important
changes in the Fe–N bond
lengths and a modification of
the
metal
coordination
sphere.[20] For instance, going
from HS to LS for the FeN6 co-
ordination spheres caused the
Fe–N distance to shorten by
about 0.2 ꢄ and the iron envi-
ronment to go from a strongly
distorted octahedron (almost
a trigonal prism) to a slightly
distorted octahedron, with a cor-
responding decrease of 25% in
volume during the transition.
Table 2 lists the Fe–N bond
lengths for [FeACTHNUGTRENNUG(picen)ACHTUNGTRENNUNG(NCS)2]
in all of the phases defined
above. Because the asymmetric
unit of [FeACHTNUTRGENN(UG picen)ACHTUTGNREN(NUNG NCS)2] con-
sisted of one complex, whatever
the phase, there was only one
crystallographically independ-
ent iron site in both the HS and
LS states. As a consequence,
the Fe–N bond lengths reflected
an average of the HS and LS
bond-lengths, with
a relative
weight that corresponded to the
percentage of both species.
Thus, from the Fe–N bond
lengths, given pure HS and
pure LS values, it was possible
to estimate the percentage of
spin-conversion (Table 3). The
data at 120 K clearly showed
a pure HS state whilst, as ex-
pected, other phases revealed
a mixture of the HS and LS
Figure 8. Temperature-dependence of the Raman spectra (excitation: 632.8 nm).
states at 30 K. The percentage
by the spin-conversion. However, the spin-conversion modi-
fied the unit cell parameters (Table 2); comparison of the
HS and 30 K unit cells showed a significant decrease in the
a and c parameters whilst b remained almost constant. This
result indicated that the most-significant modifications in
the crystal packing were due to the effect of spin-crossover
on the intrasheet features. The unit-cell-volume modifica-
tions, which were also represented by the macroscopic-
volume modification of the single crystal owing to spin-
crossover,[20] was 3.7% between the HS 120 K and the LS
slow unit cells, the latter being closer to a pure LS state.
This volume decrease was perfectly in line with the typical
values for mononuclear materials.
of spin-conversion strongly correlated to the speed of cool-
ing and warming between the flash, fast, medium, and slow
modes: the slower the rate of cooling, the larger the conver-
sion. Interestingly, in the Fast experiment, there was a small
amount of HS-to-LS conversion, which meant that the spin-
conversion could not be entirely quenched in a single crys-
tal. The spin-conversion rate increased as the speed of cool-
ing decreased and reached about 25% in the slow mode,
which was slightly higher than expected from the magnetic
measurements. To sum-up, this X-ray-diffraction cooling-
rate study confirmed that the iron HS percentage decreased
continuously when the rate of cooling slowed and that
a direct correlation between the Fe–N bond length and the
rate of spin-conversion was achievable.
The spin-conversion obviously affected the intramolecular
structural features and, in particular, the iron coordination-
Chem. Eur. J. 2012, 18, 5924 – 5934
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5929

J.-F. Lꢀtard et al.
Figure 9. Temperature-dependence of the high-spin fraction (nHS) that
was obtained from the cMT and Raman spectra with excitation at 632.2
and 680 nm.
In addition, this study revealed the mechanism of the spin
phenomenon itself inside this compound. Indeed, because
there was only one crystallographically independent iron
site in the HS/LS phases, the LS sites must have been ran-
domly distributed throughout the crystal structure: there
was no ordering of the low-spin-state distribution. To this
end, the atomic displacements parameters (ADPs) could
provide important information about the system, and as
such were worth examining. Their size and shape were de-
pendent on several factors, two of the most-important being
temperature and disorder.[21] From a crystallographic point
of view, a random distribution of the structurally different
HS and LS sites throughout the lattice manifested itself as
a single crystallographically independent site with a large
variation in the Fe–N bond lengths and angles at the iron
center. This disorder in the spin states resulted in enlarged
ADPs for the whole structure. The ADPs might have been
Figure 10. View along the a axis of the crystal packing of [Fe
(NCS)2] at 120 K. The overall description was the same at lower tempera-
tures.
ACHTUNGTRENNUNG(picen)-
AHCTUNGTRENNUNG
expected to be largest when the ratio of the HS/LS sites was
approximately equal. As mentioned previously, the structur-
al features that underwent the most-pronounced change
during spin-crossover were those that were directly associat-
ed with the FeN6 coordination sphere; hence, it was the size
of the ADPs of the nitrogen atoms that would yield the
most information regarding the HS/LS ratio for structures
that were collected at the same temperature. Figure 11
shows ADPs for all of the determined structures, along with
average Uiso values for the N atoms in each structure.
Examination of the structures after flash, fast, medium,
and slow cooling modes revealed larger ADPs than would
normally be expected at 30 K,
thereby reflecting the observed
random distribution of spin
states throughout the crystal
under these conditions. The
structure that was acquired
after irradiation (slow*) showed
significantly smaller ADPs than
any of the other structures that
were acquired at 30 K, which
was consistent with a full photo-
induced conversion into one
single (HS) state. The ADPs for
this structure were also much
smaller than those in the only
other fully HS structure, which
was collected at 120 K, and
were a direct result of the dif-
Table 2. Selected X-ray diffraction data for [Fe
phases.
ACHUTGTNNERNUG(picen)ACHTUTGNREN(NUNG NCS)2] and the experimental parameters for all of the
HS
120
Flash
Fast
Medium
Slow
Slow*
T [K]
30
instant
30
6
30
0.5
30
0.05
30
0.05
cooling rate [Kminꢀ1
]
space group
a [ꢄ]
b [ꢄ]
P21/n
9.237(1)
13.761(1)
14.653(2)
92.264(2)
1861.1(5)
1.479
9.189(4)
13.712(4)
14.521(5)
92.399(5)
1828.0(9)
1.505
9.154(2)
13.696(2)
14.461(2)
92.279(2)
1811.6(6)
1.519
9.146(2)
13.723(2)
14.379(1)
91.972(2)
1803.4(6)
1.526
9.107(2)
13.742(2)
14.318(1)
91.949(2)
1790.9(6)
1.537
9.167(2)
13.691(3)
14.559(3)
92.621(3)
1825.4(7)
1.508
c [ꢄ]
b [8]
V [ꢄ3]
1calcd [gcmꢀ3
]
unique reflns (Rint
)
3269
3731
4105
4071
3716
3228
A
U
A
E
N
ACHTUNGTRENNUNG
R (all data)
A
U
A
E
N
ACHTUNGTRENNUNG
wR2 (all data)
A
U
A
E
N
ACHTUNGTRENNUNG
5930
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 5924 – 5934

Photomagnetism of a sym-cis-Dithiocyanato Iron(II) Complex
FULL PAPER
Table 3. Fe–N bond lengths [ꢄ], average Fe–N bond length (d [ꢄ]), and the percent-
ACHUTNRGEN(NUG picen)ACHTUNGTREN(NNGU NCS)2], and the relative proportion of HS
and LS states was determined by using structural
methods.
age (%) of the spin-state species at the iron site for all phases.[a]
HS
Flash
Fast
Medium
Slow
Slow*
T [K]
120
30
30
30
30
30
Elsewhere, the photoinduced HS crystal structure
(slow*) appeared to be very close to the high-tem-
perature HS structure. Such a similarity is typically
the case when no structural phase-transition with
a change of symmetry is associated to the spin-
crossover. However, the unit-cell parameters
showed some differences that were attributed to
thermal effects on the crystal packing. It was possi-
ble, starting from the slow* phase at 30 K, to follow
the relaxation to the LS state and then the thermal
spin-crossover from the warming of the single crys-
tal from 30 K to 90 K (Figure 12). The observed
structural behavior was perfectly coherent with the
magnetic properties, which showed the vicinity of
the temperature area of the light-induced and ther-
mal spin-crossover phenomena.
Fe1–N1
Fe1–N2
Fe1–N3
Fe1–N4
Fe1–N5
Fe1–N6
d
2.199
2.214
2.236
2.206
2.087
2.108
2.1750
100
2.164
2.195
2.213
2.171
2.073
2.095
2.1518
88
2.121
2.154
2.145
2.117
2.035
2.051
2.1038
64
2.081
2.119
2.106
2.079
2.010
2.031
2.0708
48
2.026
2.027
2.114
2.036
1.961
2.026
2.0316
28
2.188
2.212
2.231
2.186
2.115
2.099
2.172
99
nHS (%)
nLS (%)
0
12
36
52
72
1
[a] The percentage of spin-state species was calculated from d by using the formula
n
HS =(dꢀdLS)/(dHSꢀdLS) with dHS =2.175 and dLS =1.975, which are the d values for
N
pure HS and LS states, respectively; dHS was known from the high-temperature struc-
ture, dLS could not be strictly known for this compound but it was taken as an average
of typical values in iron materials. Errors in nHS and nLS were estimated to be 5–10%
for all data at 30 K.
Figure 12. Temperature dependence of the unit cell volume from the
30 K photoinduced spin state unit cell (slow*) to the HS unit cell, which
shows the HS* to LS relaxation and thermal spin-crossover.
Conclusion
Thermal spin-crossover of cis-[FeACHTUNGRTENNGU(picen)ACHUTNGTREN(NUGN NCS)2] (picen=
N,N’-bis(2-pyridylmethyl)1,2-ethanediamine) occurred in the
close vicinity of the metastable region of the HS state, that
is, HS2 and HS3, which were either generated by freezing or
by light stimulation. As a consequence, the spin-equilibra-
tion in this compound was extremely slow and the residual
HS fraction at low temperatures strongly depended on the
cooling speed. Only at cooling speeds slower than
0.1 Kminꢀ1 did an almost-complete abrupt thermal spin-
transition occur at about 70 K, with only a residual HS frac-
tion of about 0.09 at 40 K. Nevertheless, even at such a slow
cooling rate, the shape of the thermal hysteresis remained
unsymmetrical. By recording long kinetic relaxations at vari-
ous temperatures in the range 60–70 K, we were able to ex-
trapolate the shape of the quasi-static thermal hysteresis
loop, which appeared to be symmetric in this case.
Figure 11. Molecular structure with atomic displacement parameters
(ADPs) of [Fe(picen)
(NCS)2] and average Uiso values (10ꢀ2 ꢄ2) for the
nitrogen atoms at 120 K and 30 K after various cooling and light-irradia-
tion regimes. ADPs are drawn at 80% probability.
G
ACHTUNGTRENNUNG
ference in temperature between these two measurements.
To conclude, the random-spin-state-distribution mechanism
was clearly revealed by the X-ray-diffraction study of [Fe-
Chem. Eur. J. 2012, 18, 5924 – 5934
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5931

J.-F. Lꢀtard et al.
The high potential of this compound to exhibit metastabil-
ity at low temperatures was used for comparing the charac-
teristics of HS states that were generated by freezing (HS2)
and by light irradiation (HS3). The kinetic parameters that
governed the relaxation process indicated that the two meta-
stable HS states were closely related, although a higher sta-
bility for the photoinduced HS state was observed. Very few
reports have performed such a systematic comparison for
a spin-crossover material.
The close vicinity of the thermal spin-transition and the
relatively slow relaxation process also allowed us to report
an unusual shape of the LITH loop, in which the HS frac-
tion always remained very high. This effect was also con-
firmed by Raman spectroscopy, which found that, whatever
the investigated temperature, the recorded spectrum was
almost always typical of the HS state. In this compound, the
spin-equilibration was so slow that any optical stimulus was
enough for maintaining the HS state.
Single-crystal X-ray diffraction allowed us to elucidate the
phase diagram as a function of speed-cooling and light irra-
diation. The phase diagram showed a direct relationship be-
tween the geometry of the metal coordination sphere and
the degree of spin-conversion. The determination of X-ray
structures in various degrees of spin state at 30 K, from
100% HS to about 30% HS, revealed a spin-conversion
mechanism that consisted of a random distribution of LS
sites throughout the lattice. Without doubt, this mechanism
endowed some flexibility to this material on the macroscop-
ic level, thereby explaining the avoidance of cracking of the
crystalline sample when it was driven by temperature effects
in a mixture of HS and LS spin states. In addition, this crys-
tallography study also confirmed the vicinity of the light-in-
duced and thermal spin-transitions.
involved the systematic determination of the T1/2 and T-
AHCTUNGTRENNUNG
AHCTUNGTRENNUNG
CTHUNGTRENNUNG
AHCTUNGTRENNUNG
T1/2. To check this hypothesis of the close vicinity of the
thermal-spin-crossover phenomenon, we applied hydrostatic
pressure, which is well-known for stabilizing the LS state of
the smaller volume. At 9.9 kbar, the thermal spin-transition
was found, and around 70% of the complex switched into
the LS electronic configuration.[24] All of these results con-
firmed that the relationship between the TACTHNURGTNE(NUG LIESST) and
T1/2 values represents a real guideline that allows us to both
predict the stability of the photoinduced HS state and to un-
derstand the synergetic effects that occur when the two re-
gions are collapsing. The challenge now is to shift the limit-
ing temperature to above room temperature, but for that we
should scrupulously study various SCO materials to identify
the key factor(s) that are acting on the stability of the meta-
stable HS state. In less than 10 years, the TACTHNURTGNEU(GN LIESST) tem-
perature of SCO materials has been improved from 30 K to
almost 130 K,[26] and even 150 K in Prussian blue ana-
logues.[27] The record today is held by a molecular cluster
with a value of about 180 K.[28] The rigidity and the distor-
tion of the inner coordination sphere appear to be the cru-
cial factors.
Experimental Section
N,N’-bis(2-pyridylmethyl)ethylenediamine (picen): Zinc powder (6.7 g)
was stirred in
a
solution of ethylenediamine (1.0 g, 1.67ꢃ10ꢀ2 mol),
EtOH (40 mL), and glacial acetic acid (7 mL) at 60–708C and a solution
of 2-pyridinecarboxaldehyde (3.6 g, 3.33ꢃ10ꢀ2 mol) in EtOH (20 mL)
was gradually added over 2 h. Zinc powder and glacial acetic acid were
added at intervals until a further 23.3 g of each had been added and the
reaction was stirred for a further 24 h. The mixture was allowed to stand
at room temperature overnight and was then filtered. The filtrate was
evaporated to afford a syrupy residue before sodium hydroxide was
added. A brown oil separated and was collected, dried overnight in dieth-
yl ether with anhydrous magnesium sulfate, and concentrated in vacuo to
obtain a yellowish oil. The product was used without further purification.
In summary, this work emphasizes that, when the thermal
spin-transition occurs in the close vicinity of the metastabili-
ty regime, some synergetic effects may happen; that is, the
kinetics of the metastable HS state (that is either generated
by light irradiation or by freezing) overlap with the thermal
transition and, as a consequence, the SCO phenomenon
may appear “hidden”.[22–25] Let us mention three recent con-
tributions that highlight this effect: The first concerned the
(NCS)2]: Because of the air-sensitivity of FeII species, the syn-
[FeACTHNUTRGENNG(U picen)ACHTUGNTRENNUGN
thesis of the complex was carried out under a nitrogen atmosphere using
standard Schlenk techniques.
[FeL2]ACHTUNGTRENNUNG(BF4)2·xH2O complex (L=2,6-bis{3-methylpyrazol-1-
yl}pyridine; x=0–1/3), which exhibited a thermal spin-tran-
sition in two steps that were centered at 147 and 105 K, very
[FeACHTNUTRGNENG(U SCN)2] was prepared in situ by reacting an iron salt with a thiocya-
close to the T
ceptibility experiment revealed only 54% spin-conversion,
but a special thermal treatment between the T(LIESST)
ACHTUNGTRENNUNG
(LIESST) value of 87 K.[22] A magnetic-sus-
nate anion. A solution of KNCS (2 mmol) in MeOH (50 mL) was added
to a mixture of solutions of FeCl2·4H2O (1 mmol) in MeOH and a small
amount of ascorbic acid as a reducing agent in MeOH (50 mL). The solu-
tion was stirred for 20 min and the resulting white precipitate (KCl) fil-
AHCTUNGTRENNUNG
value and the T1/2 temperatures, that is, at 100 K for 2 h, re-
sulted in a slow decrease of the cMT product to zero, thus
showing that the remainder of the spin-crossover can pro-
ceed but is kinetically slow.[22] The second report involved
tered off by using a cannula filter. The colorless solution of [FeACHTUNGTRENNUNG(SCN)2]
was mixed with a solution of picen (1 mmol) in MeOH (100 mL), and the
color of the solution changed to yellow. The resulting mixture was left
undisturbed for 1 day at room temperature. Yellow crystals were formed
and were collected by suction filtration and dried under vacuum (75%
yield).
metal-dilution in the [FexMn
have demonstrated that, in the case of high metal dilution
with x<0.2, T1/2 can reach T(LIESST). In this peculiar situa-
(NCSe)2 series.[23] We
1ꢀxACHUTTGNRENNUG(bpp)2]ACHUTNGTRENNGUN
Magnetic and photomagnetic studies: The measurements were performed
on a Spectrum Physics mixed-gas Ar/Kr laser (l=532 nm) that was cou-
pled via an optical fiber to the cavity of a MPMS-55 Quantum Design
SQUID magnetometer and the power at the sample surface was adjusted
to 5 mWcmꢀ2. The samples consisted of a thin layer of compound, the
AHCTUNGTRENNUNG
tion, where the thermodynamically stable HS region
reached the metastable HS region, the complex remained
HS over the whole temperature range. The third example
5932
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 5924 – 5934

Photomagnetism of a sym-cis-Dithiocyanato Iron(II) Complex
FULL PAPER
weight of which was obtained by comparison of the thermal spin-cross-
over curve with that of a more-accurately weighed sample of the same
compound. Our previously reported standardized method for determin-
ing LIESST properties was followed.[6–9] After cooling slowly to 10 K, the
sample in the LS state was irradiated and a change in magnetism oc-
curred. When the saturation point had been reached, the light was
switched off, the temperature was increased at a rate of 0.3 Kminꢀ1, and
the magnetization was measured every 1 K. TACTHNUTRGNE(UNG LIESST) was determined
from the minimum of the dcMT/dT versus T curve for the relaxation pro-
cess.
[4] a) E. Freysz, S. Montant, S. Lꢀtard, J.-F. Lꢀtard, Chem. Phys. Lett.
2004, 394, 318; b) S. Bonhommeau, G. Molnꢅr, A. Galet, A. Zwick,
J. A. Real, J. J. McGarvey, Angew. Chem. 2005, 117, 4137; Angew.
Chem. Int. Ed. 2005, 44, 4069.
[5] a) O Kahn, C Jay-Martinez, Science 1998, 279, 44; b) J.-F. Lꢀtard, P.
Guionneau, L. Goux-Capes, Top. Curr. Chem. 2004, 235, 221.
[6] a) J.-F. Lꢀtard, P. Guionneau, L. Rabardel, J. A. K. Howard, A. E.
Goeta, D. Chasseau, O. Kahn, Inorg. Chem. 1998, 37, 4432; b) J.-F.
Lꢀtard, L. Capes, G. Chastanet, N. Moliner, S. Lꢀtard, J. A. Real, O.
Kahn, Chem. Phys. Lett. 1999, 313, 115; c) S. Marcꢀn, L. Lecren, L.
Capes, H. A. Goodwin, J.-F. Lꢀtard, Chem. Phys. Lett. 2002, 358, 87.
[7] a) V. A. Money, J. S. Costa, S. Marcꢀn, G. Chastanet, J. Elhaik,
M. A. Halcrow, J. A. K. Howard, J.-F. Lꢀtard, Chem. Phys. Lett.
2004, 391, 273; b) C. Carbonera, J. S. Costa, V. A. Money, J. Elhaik,
J. A. K. Howard, M. A. Halcrow, J.-F. Lꢀtard, Dalton Trans. 2006,
3058.
Raman spectroscopy: Raman spectra were recorded for powdered sam-
ples by using a Raman-excitation HeNe, Ar, or Dye CW unfocused laser
beam (f1mm; about 3 mW) with 32 times accumulation every 20 s by
a
Jasco NR-1800 Raman spectrophotometer. Variable-temperature
Raman spectroscopy was performed by using an Oxford CF1204 cryostat.
X-ray diffraction: Single-crystal structure-determination was carried out
from X-ray data that were collected at 120 K and 30 K by using graphite-
monochromated MoKa radiation (l=0.71073 ꢄ) on a Bruker SMART-
CCD 1 K diffractometer. A series of narrow w-scans (0.38) was per-
formed at several f-settings in such a way as to cover a sphere of data to
a maximum resolution of 0.70 ꢄ. Cell parameters were determined and
refined by using SMART software,[29] and raw frame data were integrated
with the SAINT program.[30] The structures were solved by direct meth-
ods using SHELXS, and refined by full-matrix least-squares on F2 using
SHELXL-97[31] and the graphical user interface Olex2.[32] The tempera-
ture was controlled by a Cryostream N2 open-flow cooling device[33] (for
the 120 K dataset) and an Oxford Cryosystems Helix[34] (an open-flow
He cooling device) for the 30 K datasets. The same crystal was used for
the high-spin, flash, fast, medium, and slow diffraction experiments, with
the slow* experiment being performed on a different crystal of similar
size and quality. The crystal was mounted in an inert oil at room temper-
ature and was cooled at 2 Kminꢀ1 to 120 K where the data were either
collected (HS experiment) or further cooled at 6, 0.5, or 0.05 Kminꢀ1 to
30 K (fast, med, and slow experiments, respectively). The flash experi-
ment was performed by flash freezing the crystal from room temperature
to 30 K.
[8] a) J.-F. Lꢀtard, G. Chastanet, O. Nguyen, S. Marcen, M. Marchivie,
P. Guionneau, D. Chasseau, P. Gꢂtlich, Monatsh. Chem. 2003, 134,
165; b) Molecular Magnets Recent Highlights (Eds.: W. Linert, M.
Verdaguer), Springer, NewYork, 2003, p. 49; c) J.-F. Lꢀtard, J. S.
Costa, S. Marcen, C. Carbonera, C. Desplanches, A. Kobayashi, N.
Daro, P. Guionneau, J. Phys. Conf. Ser. 2005, 21, 23.
[9] a) J.-F. Lꢀtard, P. Guionneau, O. Nguyen, J. S. Costa, S. Marcꢀn, G.
Chastanet, M. Marchivie, L. Capes, Chem. Eur. J. 2005, 11, 4582;
b) J.-F. Lꢀtard, J. Mater. Chem. 2006, 16, 2550.
[10] a) M. Marchivie, P. Guionneau, J. F. Lꢀtard, D. Chasseau, J. A. K.
Howard, J. Phys. Chem. Solids 2004, 65, 17–23; b) P. Guionneau, M.
Marchivie, G. Bravic, J.-F. Lꢀtard, D. Chasseau, Top. Curr. Chem.
2004, 234, 97.
[11] a) J. S. Costa, C. Balde, C. Carbonera, D. Denux, A. Wattiaux, C.
Desplanches, J. P. Ader, P. Gꢂtlich, J.-F. Lꢀtard, Inorg. Chem. 2007,
46, 4114; b) P. Guionneau, J. S. Costa, J.-F. Lꢀtard, Acta. Crystallogr.
Sect.
C 2004, 60, m587m589; c) J. S. Costa, P. Guionneau, J.-F.
Lꢀtard, J. Phys. Conf. Ser. 2005, 21, 67–72; d) S. Bonhommeau, T.
Guillon, L. M. L. Daku, P. Demont, J. S. Costa, J.-F. Lꢀtard, G.
Molnar, A. Bousseksou, Angew. Chem. 2006, 118, 1655–1659;
Angew. Chem. Int. Ed. 2006, 45, 1625–1629.
[12] S. Bonhommeau, N. Brꢀfuel, V. K. Pꢅlfi, G. Molnꢅr, A. Zwick, L.
Salmon, J.-P. Tuchagues, J. S. Costa, J.-F. Lꢀtard, H. Paulsen, A.
Bousseksou, Phys. Chem. Chem. Phys. 2005, 7, 2909–2914.
[13] a) H. Toftlund, E. Pedersen, S. Yde-Andersen, Acta Chem. Scand. A
1984, 38, 693; b) H. Toftlund, Coord. Chem. Rev. 1989, 94, 67–108.
[14] T. Buchen, H. Toflund, P. Gutlich, Chem. Eur. J. 1996, 2, 1129–1133.
[15] N. Suemura, M. Ohama, S. Kaizaki, Chem. Commun. 2001, 1538–
1539.
[16] O. Kahn, Molecular Magnetism, VCH, New York, 1993.
[17] E. Bukhs, G. Navon, M. Bixon, J. Jortner, J. Am. Chem. Soc. 1980,
102, 2918.
[18] a) A. Hauser, Chem. Phys. Lett. 1992, 192, 65; b) A. Hauser, J.
Jeftic, H. Romstedt, R. Hinek, H. Spiering, Coord. Chem. Rev. 1999,
190–192, 471.
Acknowledgements
The authors would like to thank the GIS-Advanced Materials in Aqui-
taine (AMA), the ANR Agency (ULTiMATE and SCOOP), the CEFI-
PRA, and the Aquitaine Region for supporting the development of the
ICPA platform at the International Center of Photomagnetism in Aqui-
taine. We also gratefully thank the PHC ALLIANCE program for fund-
ing the exchange between the Universitꢀ Bordeaux and the University of
Durham.
[19] A. Desaix, O. Roubeau, J. Jeftic, J. G. Haasnoot, K. Boukheddaden,
E. Codjovi, J. Linarꢆs, M. Nogues, F. Varret, Eur. Phys. J. B 1998, 6,
183.
[20] J. F. Lꢀtard, M. Kollmansberger, C. Carbonera, M. Marchivie, P.
Guionneau, C. R. Chim. 2008, 11, 1155–1165.
[21] C. C. Wilson, Crystallogr. Rev. 2009, 15, 3–58.
[22] V. A. Money, C. Carbonera, M. A. Halcrow, J. A. K. Howard, J.-F.
Lꢀtard, Chem. Eur. J. 2007, 13, 5503.
[1] a) G. E. Moore, Electronics 1965, 38, 114; b) G. E. Moore in Digest
of the 1975 International Electron Devices Meeting, IEEE, New
York, 1975, 1113; c) R. P. Feynman, Eng. Sci. 1960, 23, 22; d) F. M.
Raymo, Adv. Mater. 2002, 14, 401.
[2] a) For general reviews, see: P. Gꢂtlich, H. A. Goodwin, Spin Cross-
over in Transition Metal Compounds I-III, Top. Curr. Chem., Spring-
er, Vienna, 2004, Vols. 233–235; b) P. Gꢂtlich, A. Hauser, H. Spier-
ing, Angew. Chem. 1994, 106, 2109; Angew. Chem. Int. Ed. Engl.
1994, 33, 2024; c) P. Gꢂtlich, Y. Garcia, T. Woike, Coord. Chem.
Rev. 2001, 219–221, 839.
[23] C. Baldꢀ, C. Desplanches, P. Gꢂtlich, E. Freysz, J.-F. Lꢀtard, Inorg.
Chim. Acta 2008, 361, 3529.
[24] J.-F. Lꢀtard, C. Carbonera, J. A. Real, S. Kawata, S. Kaizaki, Chem.
Eur. J. 2009, 15, 4146–4155.
[3] a) J. J. McGravey, I. Lawthers, J. Chem. Soc. Chem. Commun. 1982,
906; b) S. Decurtins, P. Gꢂtlich, C. P. Kçhler, H. Spiering, A.
Hauser, Chem. Phys. Lett. 1984, 105, 1; c) A. Hauser, P. Gꢂtlich, H.
Spiering, Inorg. Chem. 1986, 25, 4245–4248; d) A. Hauser, Coord.
Chem. Rev. 1991, 111, 275–290; e) A. Hauser, Comments Inorg.
Chem. 1995, 17, 17–40; f) A. Hauser, Top. Curr. Chem. 2004, 234,
155; g) A. Hauser, C. Enaschescu, M. L. Daku, A. Vargas, N. Am-
stutz, Coord. Chem. Rev. 2006, 250, 1642.
[25] H. Tokoro, S. I. Ohkoshi, Appl. Phys. Lett. 2008, 93, 021906.
[26] S. Hayami, Z.-Z. Gu, Y. Einaga, Y. Kobayashi, Y. Ishikawa, Y.
Yamada, A. Fujishima, O. Sato, Inorg. Chem. 2001, 40, 3240.
[27] a) N. Shimamoto, S.-S. Ohkoshi, O. Sato, K. Hashimoto, Inorg.
Chem. 2002, 41, 678; b) R. Le Bris, C. Mathoniꢆre, J.-F. Lꢀtard,
Chem. Phys. Lett. 2006, 426, 380.
Chem. Eur. J. 2012, 18, 5924 – 5934
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5933

J.-F. Lꢀtard et al.
[28] D. Li, R. Clꢀrac, O. Roubeau, E. Harte, C. Mathoniꢆre, R. Le Bris,
S. M. Holmes, J. Am. Chem. Soc. 2008, 130, 252.
[29] SMART-NT, Data Collection Software, Version 6.1, Bruker Analyti-
cal X-ray Instruments Inc., Madison, WI, USA, 2000.
[33] J. Cosier, A. M. Glazer, J. Appl. Crystallogr. 1986, 19, 105.
[34] A. E. Goeta, L. K. Thompson, C. L. Sheppard, S. S. Tandon, C. W.
Lehmann, J. Cosier, C. Webster, J. A. K. Howard, Acta Crystallogr.
Sect. C 1999, 55, 1243.
[30] SAINT-NT, Data Reduction Software, Version 6.1, Bruker Analyti-
cal X-ray Instruments Inc., Madison, WI, USA, 2000.
[31] G. M. Sheldrick, Acta Crystallogr. Sect. A 2008, 64, 112–122.
[32] O. V. Dolomanov, L. J. Bourhis, R. J. Gildea, J. A. K. Howard, H.
Puschmann, J. Appl. Crystallogr. 2009, 42, 339–341.
Received: August 24, 2011
Revised: December 31, 2011
Published online: March 27, 2012
5934
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 5924 – 5934