Spin crossover in polymeric materials using schiff base functionalized triazole ligands

Article abstract of DOI:10.1002/ejic.201201071

Functionalized 1,2,4-(R)-triazole-based Schiff base (RN= CH-R') ligands have been incorporated into spin crossover 1D materials of the general formula [Fe(Xtrz)₃](A)₂··2H₂O (A = BF₄, ClO₄, Br, NO

Full text of DOI:10.1002/ejic.201201071

FULL PAPER
DOI:10.1002/ejic.201201071
Spin Crossover in Polymeric Materials Using Schiff Base
Functionalized Triazole Ligands
CLUSTER
ISSUE
Hayley S. Scott,[a] Tamsyn M. Ross,[a] Boujemaa Moubaraki,[a]
Keith S. Murray,[a] and Suzanne M. Neville*[a]
Keywords: Spin crossover / Iron / Schiff bases / Magnetic properties / Polymers
Functionalized 1,2,4-(R)-triazole-based Schiff base (R-
N=CH-RЈ) ligands have been incorporated into spin cross-
over 1D materials of the general formula [Fe(Xtrz)3](A)2·
2H2O (A = BF4, ClO4, Br, NO3; Xtrz = triazole liands). The
ligand N-salicylidene 4-amino-1,2,4-triazole (Hsaltrz) and
the novel ligand N-cinimalidene 4-amino-1,2,4-triazole
[A = BF4 (1), ClO4 (2), Br (3)] and [Fe(Hcintrz)3](A)2·1.5(H2O)
[A = BF4 (4), ClO4 (5), Br (6)] revealed spin transitions over
the range T = 100–225 K. The spin crossover properties of
½
the dehydrated materials were also investigated [1–6(dh)]. In
attempts to prepare single crystals of these materials a dinu-
clear {[Fe2(Hsaltrz)5(NCS)4]·2MeOH (7)} and a trinuclear
(Hcintrz) were targeted to assess the implication of incorpo- {[Fe3(Hsaltrz)6(H2O)2(EtOH)4](ClO4)6·2(EtOH) (8)} complex
rating various aromatic functional groups on the spin cross-
over properties of this well-known family of one-dimensional
materials. Magnetic susceptibility measurements on the
polycrystalline powdered materials [Fe(Hsaltrz)3](A)2·2H2O
were obtained; these materials remain high spin over all tem-
peratures and display weak antiferromagnetic exchange
coupling. The structure of the novel ligand Hcintrz (9) is re-
ported.
were made to structurally characterized CuII analogues.[7]
The FeII–NH2trz crystal structure reveals the important
role of interchain interactions and of water in crystal cohe-
sion. Other advances of relevance to future applications, de-
scribe synthetic methods and the probing of physical prop-
erties of nanocrystalline samples of these 1D FeII–triazole
materials,[8] and of liquid crystals of C10–C18-alkyl-substi-
tuted-1,2,4-triazole-FeII mesophases.[9,10]
These developments have inspired and led us, from a
more chemical perspective, to pursue the current study,
which involves the systematic variation of different substitu-
ents on 4-amino-1,2,4-triazole (1,2,4-NH2trz) through ap-
plication of a condensation reaction that generates an imino
(R–N=CH–RЈ) Schiff base. Variations have been investi-
gated in ligand length, aromaticity and hydrogen bonding
capability [Figure 1(a) and (b)]. These ligands have been re-
acted with FeIIX2 in the hopes of forming 1D FeII spin
crossover complexes. In doing so, we hope to further in-
crease our understanding of SCO in relation to the substitu-
ents bound to coordinating triazoles. Notably, past studies
on 4-substituted triazole ligands, 4-amino-1,2,4-triazole, 4-
n-alkyl-1,2,4-triazoles, 4-(3Ј-hydroxypropyl)-1,2,4-triazole
(hyptrz), 4-(2Ј-hydroxyethyl)-1,2,4-triazole (hyetrz) and N-
(4H-1,2,4-triazol-4-yl)benzamide (tba) have produced a
range of SCO materials.[11,12]
Introduction
Using external perturbations to alter the properties of
materials is an attractive pursuit for the development of
memory devices and switches. Utilizing temperature, pres-
sure or light-irradiation as the stimulus is an easily access-
ible means by which to achieve this and has shown promise
in the area of magnetic materials.[1] In particular, spin cross-
over (SCO) materials capable of switching between high
spin (HS) and low spin (LS) states are proving to be of great
interest for such applications and devices, with complexes of
iron(II) being heavily emphasized.[2] Although many ligand
types that yield (N)6 or (N4O2) donor sets have been ex-
plored,[1,3] 1,2,4-triazole-based ligands have received par-
ticular attention and been found, in one-dimensional (1D)
[Fe(1,2,4-triazole)3](A)2 compounds, to give abrupt cross-
overs, sometimes with associated thermal hysteresis loops
being large and spanning room temperature, thus showing
signs of high cooperativity and memory features.[4,5] Recent
advances in development of 1D species include the solving
of the first crystal structure of the family member
{[FeII(1,2,4-NH2trz)3]·(NO3)2·2H2O},[6] prior to which
EXAFS was used to probe FeII structures or comparisons
[a] School of Chemistry, Monash University,
Building 23, Clayton, VIC 3800, Australia
We report here the spin crossover properties of two fami-
lies of such polymeric materials containing the well-known
iron(II)-1,2,4-triazole 1D chain backbone [Figure 1(c)]. We
find that the spin crossover properties of the 1D materials
of the general formula [Fe(Xtrz)3](A)2·n(H2O) (1–6) change
Fax: +61-3-9905-4597
E-mail: suzanne.neville@monash.edu
Homepage: http://monash.edu/science/about/schools/chemistry/
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejic.201201071.
Eur. J. Inorg. Chem. 2013, 803–812
803
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
The thermogravimetric data for complexes 1, 3, 4, and 6
(Figures S3–6) show similar weight loss profiles for each of
the materials. Between room temperature and 100 °C there
is a steady weight loss of 3.1–4.4% which can be attributed
to the loss of 2 water molecules for 1 and 3 and 1.5 water
molecules for 4 and 6. Notably, this loss is much more
abrupt in 4 and 6 which both contain the Hcintrz ligand.
Sharp weight losses between 200 and 300 °C for 1 and 3
and between 200 and 250 °C for 4 and 6 are attributed to
ligand decomposition. We note that thermogravimetric
analysis was not carried out on 2 and 5 as they contain
perchlorate anions which are potentially explosive when
heated.
Room temperature powder X-ray diffraction data were
collected on samples 1–6 in an effort to obtain crystallo-
graphic evidence of the formation of the 1D chains and to
compare the patterns both within one family and between
families (Figure 2 and see Supporting Information). As Fig-
ure 2 highlights, the main feature of all the patterns is a
strong peak at low angle followed by two small peaks and
further varied weak peaks out to about 40°. Although we
were unsuccessful in indexing the patterns (even when high
intensity data was collected using synchrotron radiation)
comparison of these patterns with other similar 1D chain
materials with variously sized triazole ligands reveals dis-
tinct similarities, in particular in the positions and inten-
sities of the low angle peaks.[12] Furthermore, in comparing
the patterns of materials containing the Hsaltrz ligand but
with differing anions (1–3, Supporting Information) the
patterns are very similar with the main peak occurring at
about 5.2–5.3°; the variation in peaks at higher angles is
expected due to the differently sized anions in the lattice.
Figure 1. The functionalized triazole ligands (a) N-salicylidene-4-
amino-1,2,4-triazole (Hsaltrz) and (b) N-cinamalidene 4-amino-
1,2,4-triazole (Hcintrz). (c) The general 1D chain [FeII(Xtrz)3](A)2
structural motif commonly observed with such ligands (X = func-
tionalized triazole, A = anion).
greatly depending on (1) the ligand (Xtrz = Hsaltrz,
–
–
Hcintrz), (2) the anion (A = BF4 , ClO4 , Br–), and (3) the
degree of hydration (n = 0–2). These polycrystalline materi-
als have been characterized using a range of techniques:
powder X-ray diffraction, thermogravimetric analysis, mi-
croanalysis and infrared spectroscopy. In attempting to
form single crystals of these 1D materials dinuclear (7) and
trinuclear (8) complexes were produced; the structures and
magnetism of these complexes are also presented.
Results
Synthesis and Characterization
The Xtrz ligands are insoluble in water; consequently, Similarly, samples 4–6 which all contain the ligand Hcintrz
synthesis of the metal complexes had to be carried out in but having differing anions (Supporting Information) show
an alternate media; the aprotic solvent acetonitrile was similar patterns with the main peaks coinciding at about
found to be most suitable. This resulted in the formation of 5.7–5.8° and subtle differences in the higher angle peaks
off-white polycrystalline powders of 1–6. All attempts to again are due to the different sized anions in the lattice. In
form crystals of these materials suitable for single crystal addition, the upward shift in the high intensity peak posi-
diffraction by controlled diffusion methods (slow diffusion, tion between the families 1–3 and 4–6, is as expected due
ether diffusion, slow cooling) were unsuccessful, resulting to the longer ligand length of Hcintrz compared to Hsaltrz
in dinuclear and trinuclear materials (7 and 8). The forma- likely resulting in different packing efficiencies of adjacent
tion and composition of the intended 1D chain materials chains (see Figure 2).
was thus confirmed using a combination of methods: ele-
mental analysis, thermogravimetric analysis, infrared spec-
troscopy and powder X-ray diffraction.
Acquisition of infrared spectra revealed the presence of
–
–
the non-coordinating anions BF4 or ClO4 in 1, 2, 4, and
5 as evidenced by prominent peaks at 1079, 1082, 1080 and
1082 cm–1, respectively. Furthermore, the 1,2-bridging coor-
dination of the triazole rings in Hsaltrz and Hcintrz in com-
plexes 1–6 was confirmed initially by the presence of strong
peaks at 629, 622, 624, 623, 620 and 620 cm–1, respectively,
relating to the torsional vibration of the triazole ring. Sec-
ondly, an upward shift in the position of the triazole N–N
stretching from the free ligands Hsaltrz and Hcintrz of 1258
and 1171 cm–1 in complexes 1–6: 1271, 1272, 1270, 1189,
1188, and 1188 cm–1 confirmed the bidentate coordina-
tion.[12,13]
Figure 2. Comparison powder X-ray diffraction patterns for (a) 1
and (b) 4, highlighting the shift in peak position with ligand length
variation.
Eur. J. Inorg. Chem. 2013, 803–812
804
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
Magnetism
clear evidence for such coupling in the dinuclear (7) and
trinuclear (8) clusters. For both 1 and 1dh there is a plateau
at 75 K with χMT values of 1.0 and 1.5 cm3 mol–1 K, respec-
tively, followed by a rapid decrease below 10 K due to zero
field splitting and weak antiferromagnetic coupling effects
to reach minimum χMT values of about 0.5 cm3 mol–1 K.
The plateaus are indicative of some HS trapping that occurs
in the major LS iron(II) state.
Magnetic susceptibility measurements with 2 revealed a
gradual one-step, non-hysteretic SCO between 200 and
100 K, with T1/2 ≈ 170 K [Figure 3(b)]. Above 200 K the
χMT values of about 3.2 cm3 mol–1 K are indicative of high
spin iron(II) with the slope of the curve matching that be-
low about 30 K suggestive again of weak antiferromagnetic
HS–HS coupling. Below 100 K there is a plateau at
0.4 cm3 mol–1 K consistent with low spin iron(II). Magnetic
susceptibility measurements on 2dh suggest that a more
gradual one-step SCO than in 2 is occurring below 150 K
[Figure 3(b)] with an inflection at about 70 K. Above 150 K
the χMT values are indicative of high spin iron(II) and be-
low 150 K they gradually decrease to a minimum of
1.0 cm3 mol–1 K, at 10 K, suggesting an incomplete spin
transition. However, the spin transition is not as clear cut
as in 2 and the curve could be largely due to HS–HS anti-
ferromagnetic coupling.
[Fe(Hsaltrz)3](A)2·2(H2O) (1–3)
Magnetic susceptibility measurements were carried out
on this family first in the as-synthesized solvated state and
then with heating to remove any non-coordinated water
molecules (dehydrated state, labeled dh) to examine the in-
fluence of solvent on the SCO behavior. Magnetic suscep-
tibility measurements on both 1 and 1dh reveal a one-step
gradual SCO, for 1 between 150 and 75 K (T1/2 ≈ 130 K,
with a hint of thermal hysteresis) and for 1dh between 125
and 75 K (T1/2 ≈ 100 K) [Figure 3(a)]. Above 150 K the
χMT values of about 3.0–3.5 cm3 mol–1 K are indicative of
HS iron(II) but they do not level off suggesting that weak
antiferromagnetic coupling might be occurring. There is
Magnetic susceptibility measurements on 3 and 3dh both
revealed a high spin state over the range 300–4 K with χMT
values of about 3.0 cm3 mol–1 K at 300 K [Figure 3(c)]. The
two curves for the dehydrated sample 3dh show the warm-
ing curve as the lower plot with a more rapid increase in
χMT at 270 K; the higher plot is the cooling curve that
starts from the same value at 300 K. The warming and cool-
ing curves for 3, with the sample contained in a quartz tube
and dampened with solvent, are identical between 30 and
250 K and indicative of weak antiferromagnetic coupling
with no SCO. The rapid, small change in χMT at 250 K
could be due to a partial spin transition, but solvent loss is
known to occur above room temperature and, so, a spin
transition cannot be assigned with certainty.
[Fe(Hcintrz)3](A)2·2(H2O) (4–6)
Magnetic susceptibility measurements were, as in the
Hsaltrz family, carried out on this family first in the as-
synthesized solvated state and then in the dehydrated state
(dh) to examine the influence of solvent on the SCO behav-
ior (Figure 4).
Magnetic susceptibility measurements on both 4 and 4dh
reveal a one-step, gradual spin transition, for 4 between 200
and 125 K (T1/2 ≈ 170 K) and for 4dh between 225 and
125 K with T1/2 ≈ 200 K [Figure 4(a)]. Above 225 K the
χMT values of about 3.4–3.5 cm3 mol–1 K are indicative of
high spin iron(II) with the curves not reaching a plateau
value, once more suggestive of weak antiferromagnetic HS–
HS coupling. For both there is a plateau at 75 K with χMT
values of 0.8 cm3 mol–1 K indicative of the LS state probably
with some HS trapping. Below about 10 K the more rapid
decrease in χMT is due to HS zero field splitting effects
Figure 3. χMT, per Fe, vs. temperatures spanning 4–300 K for (a)
[Fe(Hsaltrz)3](BF4)2·2(H2O) (1) and [Fe(Hsaltrz)3](BF4)2 (1dh), (b)
[Fe(Hsaltrz)3](ClO4)2·2(H2O) (2) and [Fe(Hsaltrz)3](ClO4)2 (2dh)
and (c) [Fe(Hsaltrz)3](Br)2·2(H2O) (3) and [Fe(Hsaltrz)3](Br)2 (3dh).
Symbols: hydrated form (o) and dehydrated (ᮀ). Note: in (c) the
3dh plots (ᮀ) for the dried powder show the warming curve with
a more rapid increase occurring at 270 K followed by the cooling
curve that shows no step. For 3 the sample was damp with solvent
and measured in a quartz tube (o) first from 300 K with cooling,
then it was warmed and χMT followed the same values. The origin
of the rapid change in χMT at 250 K is not known.
Eur. J. Inorg. Chem. 2013, 803–812
805
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
Magnetic susceptibility measurements on 6 revealed a
multistep spin transition with the low temperature abrupt
step occurring between 225 and 200 K (T1/2 = 213 K) and
the higher temperature step occurring above 250 K and
which is incomplete up to 290 K [Figure 4(c)]. Below 200 K
the large plateau of χMT values of about 0.5 cm3 mol–1
K
are indicative of LS iron(II). In the heating/cooling modes
there is a small hysteresis loop over the range 200–225 K of
3 K. We believe this to be real and not an artifact of solvent
loss that is known to happen, from thermogravimetric stud-
ies, above room temperature. The abruptness and hysteresis
is in line with 6 displaying stronger cooperativity than all
other family members and likely having stronger intermo-
lecular solid state effects. Magnetic susceptibility measure-
ments on 6dh revealed a gradual one-step SCO [Figure 4(c)]
similar to the line shapes exhibited by solvated versions of
other Hcintrz–FeII salts. Above 225 K the χMT values of
about 3.5 cm3 mol–1 K are typical of high-spin iron(II). Be-
low 225 K there is a gradual decrease in χMT values to
150 K (T1/2 = 180 K) then a more gradual plateau to 50 K
reaching a value of 1.4 cm3 mol–1 K, indicative of LS FeII
with remnant HS contribution. Below 50 K there is rapid
decrease in χMT values to reach
a minimum of
0.4 cm3 mol–1 K below 10 K, consistent with zero field split-
ting effects combined with HS–HS antiferromagnetic cou-
pling. There is a small hysteresis loop in the heating and
cooling modes, of 3 K.
Dinuclear and Trinuclear Materials (7 and 8)
Magnetic susceptibility measurements on the dinuclear
and trinuclear materials, 7 and 8, revealed that both remain
in the iron(II) high spin state over the range 300–4 K. Below
about 100 K both materials show a rapid decrease in χMT
values to a minimum of 1.0–0.5 cm3 mol–1 K due to a com-
bination of weak antiferromagnetic coupling and zero field
splitting.[14] Very good fits of the data have been obtained
with our new program PHI[15] by use of an isotropic Hei-
senberg Hamiltonian –2JS1·S2 (using spin-only Si = 2
states; i = 1, 2, 3) that allows for remnant orbital degener-
Figure 4. χMT vs. temperature 4–300 K for (a) [Fe(Hcintrz)3](BF4)2·
2(H2O) (4) and [Fe(Hcintrz)3](BF4)2 (4dh), showing heating and
cooling modes for 4, (b) [Fe(Hcintrz)3](ClO4)2·2(H2O) (5) and
[Fe(Hcintrz)3](ClO4)2 (5dh) and (c) [Fe(Hcintrz)3](Br)2·2(H2O) (6)
and [Fe(Hcintrz)3](Br)2 (6dh) showing heating and cooling modes
for both forms. Symbols: hydrated form (o) and dehydrated (ᮀ).
combined with weak antiferromagnetic coupling. For 4 acy in the HS FeII (5T2g parent states) by allowing g to
there is a small hysteresis loop in the heating and cooling increase from 2.0. The zero-field splitting (axial) parameter,
modes of about 5 K.
D, was not included since we do not have magnetization
Magnetic susceptibility measurements on 5 revealed a data to provide any experimental back-up. The fits are
gradual one-step transition between 225 and 100 K, very shown in Figure 5 and the parameter sets are:
similar to that found in 4 and with T1/2 ≈ 180 K [Fig-
ure 4(b)]. Above 225 K the χMT values of about
Dinuclear (7): g = 2.27; J = –1.2 cm–1
Trinuclear (8): g = 2.16; J(Fe1Fe2 = Fe3Fe2) = –1.34 cm–1;
3.4 cm3 mol–1 K are indicative of high spin iron(II) with the J(Fe1Fe3) = –0.87 cm–1
slope of χMT extrapolating to that below about 30 K, once
Such small J values across FeII(1,2,4-triazole)3FeII brid-
again suggestive of HS–HS weak antiferromagnetic cou- ges are typical of previous reports.[16,17] The χMT plot for
pling. Below 50 K there is a plateau at 0.85 cm3 mol–1
K
the trinuclear complex 8 is different in shape, of course,
consistent with low spin iron(II) and small contributions than those of other trinuclears in which the central FeII un-
from HS trapping. Magnetic susceptibility measurements dergoes a spin transition to LS upon cooling, such that the
on 5dh revealed a high spin nature between 300 and 10 K χMT values below the spin transition are constant (at least
with χMT values of about 3.0 cm3 mol–1 K at 300 K decreas- to about 50 K) and typical of two HS FeII sites.[18,19] The
ing to 1.6 cm3 mol–1 K at about 20 K, indicative of weak very small J(Fe1Fe3) value for trinuclear 8 may be due to
HS–HS exchange coupling between neighboring FeII ions the inter-cluster hydrogen bonding pathways (Figure 8b, be-
[Figure 4(b)]. There is no clear evidence of a spin transition low). However, we believe this small value is the result of a
in this desolvated material.
next-nearest neighbor intra-cluster coupling interaction.
Eur. J. Inorg. Chem. 2013, 803–812
806
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
Figure 5. Plot of χMT (per mol) vs. temperature 4–300 K for (a) 7 and (b) 8 in dc fields of 0.5 T. The red lines are the best fits using the
parameters given in the text.
This trinuclear complex yields an example of spin frustra- are indicative of iron(II) in the high spin state as confirmed
tion with the two competing J values leading to the ground by magnetic susceptibility measurements (Table 1). There
state spin being S = 1 with S = 0 at about 0.13 cm–1, S = 2 are strong intramolecular OH···N=C hydrogen bonding in-
at 1.6 cm–1 and S = 1 at 2.8 cm–1, all other levels being teractions present within each ligand (Table S1). However,
above 4.7 cm–1.
a network of hydrogen bonding interactions between the
lattice water molecules and both thiocyanate anions and the
ligand is evident (see Supporting Information).
Single-Crystal Structural Analysis
Table 1. Selected bond lengths for 7 and 8.
[Fe2(Hsaltrz)5(NCS)4]·4(H2O) (7)
Complex 7
Complex 8
Single crystal diffraction revealed a dinuclear species
where the iron(II) ions are linked centrally by three Hsaltrz
ligands bridging though N-1 and N-2. The octahedral
iron(II) coordination environment is completed by one ni-
trogen from an additional Hsaltrz ligand bound only
Bond
Distance [Å]
Bond
Distance [Å]
Fe···Fe
3.951(3)
2.159(6)
2.085(7)
2.190(6)
2.161(6)
2.196(6)
2.215(6)
2.078(7)
2.145(7)
2.156(7)
2.177(6)
2.191(7)
2.212(6)
Fe1···Fe2
Fe1–N1
Fe1–N5
Fe1–N9
Fe1–O1(w)
Fe1–O2
Fe1–O3
Fe2–N2
Fe2–N6
Fe2–N10
3.865(2)
Fe1–N1
Fe1–N2
Fe1–N5
2.166(10)
2.162(10)
2.186(10)
2.152(8)
through N-1 and two NCS– anions bound through the ter- Fe1–N13
Fe1–N17
Fe1–N21
2.127(9)
minal nitrogen atom. (Figure 6). Crystallizing in the tri-
2.113(10)
2.176(10)
2.176(10)
2.156(11)
¯
clinic space group P1, the entire dinuclear unit is crystallo-
Fe2–N3
Fe2–N4
Fe2–N9
Fe2–N14
Fe2–N18
Fe2–N22
graphically distinct containing two unique iron(II) centers,
Fe(1) and Fe(2), five Hsaltrz ligands and four NCS– ligands.
In the recently reported dinuclear analogue [Fe2(Hsaltrz)5-
(NCS)4]·4(MeOH) which crystallized in the monoclinic
space group C2/c, there is only one unique iron(II) center
with the full dinuclear unit being generated by symmetry.[19]
The Fe–N bond lengths in 7 for Fe(1) and Fe(2) at 123(2) K
[Fe3(Hsaltrz)6(H2O)2(EtOH)4](ClO4)6·2(EtOH) (8)
Single crystal diffraction of a crystal of 8 revealed a tri-
nuclear species where the central iron(II) site is bridged to
the exterior iron(II) atoms by three Hsaltrz ligands bridging
through both N-1 and N-2 of each ligand [Figure 7(a)]. The
octahedral coordination environment of the external
iron(II) ions is completed by one water molecule and two
ethanol molecules. The Fe–N bond lengths of all iron(II)
atoms at 123 K is indicative of iron(II) in the HS state
(Table 1). This agrees with magnetic susceptibility measure-
ments where a HS nature is observed over the range 4–
300 K. There are strong intramolecular OH···N=C hydro-
gen bonding interactions present within each ligand
[Table 1, Figure 8(b)].
Figure 6. Structural representation of dinuclear 7. Iron(II) atoms
are shown as spheres, all other atoms as sticks, hydrogen atoms,
solvent molecules and anions have been omitted for clarity.
Eur. J. Inorg. Chem. 2013, 803–812
807
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
cessful they are similar to patterns of the previously re-
ported material [Cu(tba)3](CF3SO3)2·3(H2O) [tba = N-(4H-
1,2,4-Triazol-4-yl)benzamide] which contains a function-
alized triazole ligand similar in length to Hsaltrz.[12]
The magnetic properties of the two families of new SCO
materials of the form [Fe(L)3](A)2·n(H2O) show comparable
trends to other similarly 4-substituted analogues, in that
they are influenced by the degree of hydration, the anion
and the ligand length. Interestingly, we find that the spin
crossover of all these materials occurs at a significantly
lower temperature than most examples of this type from
previous reports. This observation may, in part, be due to
the insolubility of the ligands in aqueous media resulting in
formation of a finer precipitate upon addition of the
iron(II) salt water solution. Furthermore, the shapes of the
χMT vs. T plots in Figure 3 and Figure 4 are reminiscent of
those recently described for iron(II) polymers, in their as-
synthesized and desolvated forms, made using triazole-
based dendritic ligands.[13] The plots, supplemented by
Mössbauer spectra, were generally interpreted in terms of
HS to LS spin transitions, with remnant HS forms responsi-
ble for broad plateaus observed at intermediate tempera-
tures and with more rapid decreases in χMT below about
50 K assigned to zero field splitting effects. Weak antiferro-
magnetic coupling between nearest neighbor FeII sites was
possible but not considered. Additionally, a 1D [FeL3](A)2
family incorporating 1,2,4-propionic acid substituted tri-
azole ligands, with H-bonding possibilities and of a similar
length to Hsaltrz, in the solvated and desolvated forms and
with a variety of anions A, also showed similar magnetic
plots to those given here.[21] On the other hand, the family
of 1D chain materials containing the ligand N-(4H-1,2,4-
triazol-4-yl)benzamide (tba) which is of similar length to
Hsaltrz show abrupt spin transitions, some with hysteresis
and near room temperature.[12]
Figure 7. Structural representation of (a) the trinuclear 8 and
(b) hydrogen bonding network present between bound water and
perchlorate anions. Iron(II) atoms are shown as spheres, all other
atoms as sticks, hydrogen atoms, solvent molecules and anions have
been omitted for clarity.
Figure 8. Structural representation of (a) the ligand Hcintrz (b) the
packing in one unit cell viewed along the a-axis.
Hcintrz (9)
The ligand Hcintrz crystallizes in an orthorhombic sym-
metry where there is a slight torsion (40°) between the tri-
azole and benzene rings [Figure 8(a)]. The ligands pack in
the unit cell in a herringbone arrangement [Figure 8(b)]
with a large array of hydrogen bonding and edge-to-face π-
stacking interactions clearly visible.
After solvent removal in all complexes shown here, we
see a dramatic change in magnetic properties. Generally, we
see a more gradual spin transition and a decrease in T or
½
HS nature (i.e. a destabilization of the LS state). There is a
vast range of consequences seen in other 1D chain triazole
materials with solvent removal, but overall the trend seems
to parallel observations reported here.[5,12,21]
Assessing the magnetic behavior of each ligand, overall
the Hcintrz materials show higher spin transition tempera-
tures than the Hsaltrz analogues. Interestingly, longer li-
gand lengths in the past have been attributed to less cooper-
Discussion
1D chain materials
The 1D chain structure and composition of 1–6 was con- ative spin transitions as the interchain communication is
firmed using a combination of elemental analysis, IR spec- presumably weakened. This is seen clearly when investiga-
troscopy, thermogravimetric analysis and powder X-ray dif- ting dendrimer triazole ligands.[9,13,21,22] Here, Hcintrz, be-
fraction. Attempts to synthesize single crystals of the cop- ing significantly longer than Hsaltrz, we find the opposite
per(II) analogues of 1–6, as is common practice with the trend. This suggests that intermolecular interactions be-
iron(II) species being difficult to crystallize – except in one tween the chains, ligands, anions and solvent molecules play
case,[6] were unsuccessful. The fact that the powder X-ray the largest role in defining magnetic properties. Of course,
patterns of the two families, 1–3 and 4–6 are similar but in the absence of structural data for these materials, there
shifted suggests a similar structure for these complexes as is no way to assess these features further.
different packing and lattice arrangements would yield dif-
From all these measurements of the Schiff base function-
ferent crystallographic patterns even for similar linear chain alized triazole complexes it is very clear that the magnetic
structures. Although indexing of these patterns was unsuc- behavior is highly dependent on factors such as the nature
Eur. J. Inorg. Chem. 2013, 803–812
808
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
of the counteranions, the length and functionality of the 4- trast to [Fe3(p-MeOptrz)6(H2O)6](Tos)6 which undergoes a
substitutued triazole ligands, and the degree of hydration.
SCO and also has a coplanar ligand. Thus, it is likely a
combination of ligand and anion dependencies that ac-
counts for the noted spin transition.
Dinuclear Complex
Dinuclear complex 7 is well characterized, structurally
and magnetically, without showing a spin transition, but it
displays weak antiferromagnetic coupling across the triple
Conclusions
The synthesis of 4-substituted triazole ligands using sim-
triazole bridges. A pseudo-isomorph of 7 was recently re- ple Schiff base chemistry has enabled production of func-
ported,[20] [Fe2(Hsaltrz)5(NCS)4]·4MeOH, differing in the tional ligands for use in SCO materials synthesis. We pres-
non-coordinating solvent, which shows a spin transition. ent here two ligands, Hsaltrz and the novel ligand Hcintrz,
Indeed, at 123 K for 7 the inter-dinuclear Fe···Fe distance which contain differing substituent lengths, hydrogen bond-
[3.951(3) Å] is comparable to that reported for [Fe2(Hsaltrz)5- ing capabilities and aromaticity. In this way we have shown
(NCS)4]·4(MeOH) in the HS state. Furthermore, in that SCO properties of 1D chain materials of the form
[Fe2(Hsaltrz)5(NCS)4]·4(MeOH) a large array of hydrogen [Fe(L)3](A)2·n(H2O) (L = Hsaltrz, Hcintrz; A = BF4, ClO4,
bonds and π-stacking interactions exist. However, in 7, Br, n = 2, 0) vary vastly based on ligand identity, anion and
while the dinuclear units are quite closely packed and degree of hydration. We are continuing these studies with a
aligned, there are no significant interactions.[20] It is well wider range of anions and substituted ligands as well as
established that intermolecular interactions make a large longer ligands, more aromatic ligands, and ligands with
contribution to the existence or absence of spin transitions. greater opportunities for hydrogen bonding.
This is especially evident in polymorphic materials where
We have also shown that the symmetry, crystal packing,
the ligand field surrounding the iron(II) center is iden- lattice interactions and solvation play an important role in
tical.[23]
the presence or absence of SCO in dinuclear and trinuclear
materials. These similar species have previously been re-
ported to undergo a spin transition, yet here, for a variety
of reasons, they remain high spin.
Trinuclear Complex
We have fully characterized trinuclear complex 8 struc-
turally and magnetically and found no spin transition and
only weak antiferromagnetic coupling across the triple–tri-
azole bridges. There have been several other iron(II) trinu-
clear triazole species reported in the past[17–19] that undergo
a spin transition. Notably, they all comprise coordinated
water molecules, whereas 8 contains coordinated ethanol
molecules. Some of these examples have been structurally
characterized in both the HS and LS states and we find that
in 8 the Fe···Fe distances and Fe–N and Fe–O bond lengths
are in each case comparable to HS state structures
(Table 1). With respect to the previously published
{[Fe3(hyetrz)6(H2O)6](CF3SO3)6 [hyetrz = 4-(2Ј-hydroxy-
ethyl)-1,2,4-triazole]} significance was placed on intermo-
lecular interactions and their contribution to cooperativity
of the spin transition. This suggests that more interactions
result in a more rigid structure which stabilizes the LS state
resulting in a less abrupt spin transition.[19] There is a large
network of intermolecular interactions present in 8 between
the coordinated water molecules and the perchlorate anions
Experimental Section
General Remarks: All reagents and solvents were purchased from
Sigma–Aldrich Pty. Ltd. and used as received. Infrared spectra
were measured with a Bruker Equinox 55 FT-IR fitted with a
71Judson MCT detector and Specac “Golden Gate” diamond
ATR. ESI-MS were recorded using a Micromass (now Waters)
Platform II QMS connected to a Waters Alliance 2690 LC system
for automatic flow injections. Microanalyses were performed by
Campbell Microanalytical Laboratory, Department of Chemistry,
University of Otago, Dunedin, New Zealand. Magnetic suscep-
tibility data were collected using a Quantum Design MPMS 5
SQUID magnetometer under an applied field of 1 T over the range
300–4 K. Samples were measured both in the solvated and desol-
vated (dh) states. Desolvation was achieved with heating (100 °C) in
vacuo. Thermogravimetric analysis measurements were performed
with a Mettler Toledo TGA/SDTA851e thermogravimetric ana-
lyzer. Operating software: STARe Software v. 8.10 (Mettler Toledo
GmbH, 1993–2004).
Crystallographic Data Collection and Refinement
[Figure 7(b), Table 1]. In an early report of the trinuclear Crystallographic data and parameters for 7–9 are summarized in
Table 2. Single crystal diffraction data for 7 and 8 were collected
with a Bruker APEX X8 diffractometer using Mo-Kα (λ =
0.71073 Å) radiation and equipped with an Oxford Instruments ni-
trogen gas cryostream. Crystals were mounted on a MiTeGen
MicroMounts fibre in a small amount of oil. Diffraction data
analysis was performed using SAINT+ within the APEX2 software
package.[24] Empirical absorption corrections were applied to all
data using SADABS.[25] Single crystal data for 9 was collected at
the Australian Synchrotron using the MX2 beamline operating at
16 keV (λ = 0.77506 Å). Data were collected using the Blue Ice
materials [Fe3(p-MeOptrz)8(H2O)4](BF4)6 and [Fe3(p-MeO-
ptrz)6(H2O)6](Tos)6(4-p-methoxypheny1)-1,2,4-triazole, it
was found that the Tos analogue undergoes a spin transi-
–
tion; however, the BF4 analogue remains HS as does 8.[17]
From structural analysis it was determined the main differ-
ence is the dihedral angles between the triazole and phenyl
rings of the bridging p-MeOptrz ligands. The latter are al-
most coplanar thus contributing to the overall electron do-
nating effect. In 8, the triazole and phenyl rings are nearly
coplanar (7.7–17.9°) and the complex remains HS in con- software.[26] Initial data processing was carried out using the XDS
Eur. J. Inorg. Chem. 2013, 803–812
809
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
Table 2. Crystal data and refinement details.
7
8
9
Formula
C49H40Fe2N24O9S4
C68H68Cl6Fe3N24O41
C11H10N4
376.30
FW [gmol–1]
1348.99
triclinic
2217.69
triclinic
Crystal system
Space group
a [Å]
b [Å]
c [Å]
α [°]
β [°]
orthorhombic
P212121
5.5890(11)
9.1730(18)
19.453(4)
–
¯
¯
P1
P1
11.402(2)
13.654(3)
21.364(4)
96.029(9)
104.798(11)
107.114(9)
3014.3(10)
1.486
12.5727(19)
12.974(3)
16.800(3)
67.592(4)
70.867(5)
84.783(5)
2391.7(8)
1.540
–
–
γ [°]
V [Å3]
997.3(3)
1.253
0.082
ρcalcd. [Mgm–3]
μ [mm–1]
0.694
0.715
data/restraints/parameters
R(F) {IϾ2σ(I), all} [%]
Rw(F2) {IϾ2σ(I), all} [%]
GoF
11500/0/784
0.1163 {0.1753}
0.2997 {0.3402}
1.092
7584/0/622
0.1480 {0.2230}
0.3716 {0.4332}
1.336
3055/0/136
0.0612 {0.0810}
0.1451 {0.1659}
1.061
package.[27] The structures were solved using SHELXS and refined
using SHELXL-97 within X-SEED.[28] All non-hydrogen atoms in
the structures were refined anisotropically and hydrogen atoms
were generated using the riding model.
ized water (1 mL, where A = Br–). This solution was filtered into
a warm acetonitrile (15 mL) solution of Xtrz (2.12 mmol), giving
an instant white precipitate. The mixture was stirred for 15 min
while cooling to room temperature, and then left undisturbed over-
night for complete precipitation. The solid was filtered, washed
with acetonitrile and dried in air. The composition of each material
was confirmed by microanalysis and the degree of hydration assess
by thermogravimetric analysis.
CCDC-901403 (for 7), -901404 (for 8), and -901405 (for 9) contain
the supplementary crystallographic data for this paper. These data
can be obtained free of charge from The Cambridge Crystallo-
graphic Data Centre via www.ccdc.cam.ac.uk/datarequest/cif.
[FeII(Hsaltrz)3](BF4)2·2(H2O) (1): Hsaltrz (200 mg, 1.0 mmol) was
dissolved in MeCN (10 mL). Separately Fe(BF4)2·6(H2O) (0.11 g,
0.34 mmol) was dissolved in MeCN (1 mL). The two solutions were
then combined to produce a white precipitate. Mw 794.01; yield
Powder X-ray diffraction analyses were performed with a Bruker
D8 Focus diffractometer, equipped with a Cu-Kα source and a
graphite monochromator. Patterns were collected from 2° to 80°
2θ, using a step size of 0.2° at a scan rate of 0.1°min–1.
189 mg (70.00%); IR: ν = 3266 (br), 3125 (s), 2250 (w), 2084 (w),
˜
Ligand Synthesis
1619 (s), 1607 (s), 1572 (m), 1533 (s), 1483 (m), 1458 (m), 1392 (w),
1334 (m), 1271 (s), 1203 (m), 1155 (w), 1079 (s), 1022 (w), 977 (w),
913 (w), 885 (w), 806 (w), 758 (s), 629 (s) cm–1.
C27H24N12Fe1O3B2F8: calcd. C 40.84, H 3.05, N 21.17; found C
40.73, H 3.21, N 20.94
Hsaltrz: The ligand Hsaltrz was synthesized according to reported
procedures.[29]
Hcintrz: The synthesis of Hcintrz was adapted from previously de-
scribed methods.[30] 4-amino-1,2,4-triazole (3 g, 35 mmol), cinnam-
aldehyde (5.5 g, 42 mmol) and 7 drops of concentrated H2SO4 were
dissolved in EtOH (75 mL) and the solution refluxed and stirred
for 2 h. After this time the solution was removed from heat/stirring
and left to reach ambient temperature. After this time, H2O
(70 mL) was added to the solution and a white precipitate formed
following 10 min. of stirring. The precipitate was subsequently fil-
tered and washed with 20 mL H2O, followed by 10 mL hexane.
Crystals of Hcintrz suitable for single-crystal X-ray diffraction were
grown by recrystallization from acetonitrile and water. Mw 198.22;
yield 4.8 g (69.2%); 1H NMR (400 mHz, [D6]DMSO): δ = 9.060
(s, 2 H), 8.829 (m, 1 H), 7.707 (m, 2 H), 7.428 (m, 3 H), 7.341 (m,
1 H), 7.158 (m, 1 H) ppm. 13C NMR (400 mHz, [D6]DMSO): δ
= 159.591, 145.413, 138.781, 134.868, 130.085, 129.291, 128.030,
123.489 ppm. ESI-MS found [M + H]+ peak at m/z 199.1; found
[FeII(Hsaltrz)3](ClO4)2·2(H2O) (2): Hsaltrz (200 mg, 1.0 mmol) was
dissolved in MeCN (10 mL). Separately Fe(ClO4)2·6(H2O) (90 mg,
0.34 mmol) was dissolved in MeCN (1 mL). The two solutions were
then combined to produce a white precipitate. The precipitate was
filtered off and washed with MeCN. Mw 819.30; yield 176 mg
(63.18%); IR: ν = 3110 (s), 1620 (s), 1606 (s), 1572 (m), 1535 (s),
˜
1483 (m), 1458 (m), 1391 (w), 1334 (m), 1272 (s), 1204 (m), 1082
(s), 1052 (m), 991 (m), 914 (m), 807 (m), 759 (s), 622 (s) cm–1.
C27H24N12Fe1O11Cl2 (+H2O not in formula): calcd. C 38.73, H
3.13, N 20.07; found C 38.01, H 3.04, N 20.26.
[FeII(Hsaltrz)3](Br)2·2(H2O) (3): Hsaltrz (200 mg, 1.0 mmol) was
dissolved in MeCN (10 mL). Separately FeBr2 (72 mg, 0.34 mmol)
was dissolved in H2O (1 mL). The two solutions were then com-
bined to produce a white precipitate, which was filtered and washed
[M + Na]+ peak at m/z 221.1. IR: ν = 3124 (s), 3088 (s), 1629 (s),
˜
with MeCN (10 mL). Mw 780.21; yield 146 mg (55.03%); IR: ν =
˜
1587 (s), 1517 (s), 1496 (s), 1449 (s), 1291 (m), 1259 (m), 1171 (s),
1053 (s), 979 (s), 877 (s), 840 (s), 752 (s), 649 (s), 620 (s) cm–1.
C11H10N4: calcd. C 66.65, H 5.08, N 28.26; found C 66.80, H 5.09,
N 28.51,%
3260 (br), 3042 (s), 2971 (s), 1083 (w), 1619 (s), 1606 (s), 1572 (m),
1533 (s), 1482 (m), 1459 (m), 1391 (w), 1333 (m), 1270 (s), 1219
(m), 1201 (m), 1153 (s), 1075 (s), 1020 (m), 991 (m), 912 (m), 806
(m), 756 (s), 624 (s) cm–1. C27H24N12Fe1Br2O3 + 4H2O (not in-
cluded in formula): calcd. C 38.05, H 3.78, N 19.72; found C 38.21,
H 3.36, N 20.03.
Synthesis of 1D Chain Iron(II) Complexes
[Fe(Xtrz)3](A)2·2H2O (Xtrz = Hsaltrz, Hcintrz, A = BF4, ClO4, Br):
Fe(A)4·n(H2O) (0.85 mmol) and ascorbic acid (0.57 mmol) were
[FeII(Hcintrz)3](BF4)2·1.5(H2O) (4): Hcintrz ligand (300 mg,
1.5 mmol) was dissolved in MeOH (20 mL). Separately Fe(BF4)2·
–
dissolved in acetonitrile (3 mL, where A = BF4– or ClO4 ) or deion-
Eur. J. Inorg. Chem. 2013, 803–812
810
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
6(H2O) (168 mg, 0.50 mmol) and ascorbic acid (20 mg) were dis-
Acknowledgments
solved in MeOH (3 mL). The two solutions were combined drop-
wise and a white precipitate immediately formed. The precipitate
was collected by filtration and washed with MeOH. Mw 794.01;
We thank the Australian Research Council for providing an Austra-
lian Research Fellowship to S. M. N. and Discovery Grants to
S. M. N and K. S. M. to support this work. The help provided by
Nicholas Chilton in fitting the magnetic data for 7 and 8 by use of
his program PHI is gratefully acknowledged.
yield 311 mg (78.33%); IR: ν = 3130 (s), 2081 (w), 1627 (s), 1586
˜
(s), 1532 (s), 1449 (s), 1396 (m), 1337 (m), 1301 (m), 1189 (s), 1080
(s), 988 (s), 889 (m), 751 (s), 689 (m), 623 (s) cm–1.
C33H30N12Fe1B2F8: calcd. C 48.09, H 3.67, N 20.39; found C 49.09,
H 4.04, N 20.76.
[FeII(Hcintrz)3](ClO4)2·1.5(H2O) (5): Hcintrz (300 mg, 1.5 mmol)
was dissolved in MeOH (20 mL). Separately Fe(ClO4)2·6(H2O)
(128 mg, 0.50 mmol) and ascorbic acid (20 mg) were dissolved in
MeOH (3 mL). The two solutions were combined dropwise and
an off-white precipitate immediately formed. The precipitate was
collected by filtration and washed with MeOH. Mw 849.42; yield
[1] P. Gütlich, H. A. Goodwin, Top. Curr. Chem. 2004, 233, 1.
[2] O. Kahn, C. J. Martinez, Science 1998, 279, 44.
[3] Y. Garcia, V. Niel, M. Carmen Munoz, J. A. Real, Top. Curr.
Chem. 2004, 233, 229.
[4] J. G. Haasnoot, Coord. Chem. Rev. 2000, 200–202, 131.
[5] P. J. v. Koningsbruggen, Top. Curr. Chem. 2004, 233, 123.
[6] A. Grosjean, N. Daro, B. Kauffmann, A. Kaiba, J.-F. Létard,
P. Guionneau, Chem. Commun. 2011, 47, 12382.
424 mg (51.3%); IR: ν = 3112 (s), 1627 (s), 1586 (s), 1531 (s), 1449
˜
[7] M. M. Dirtu, C. Neuhausen, A. D. Naik, A. Rotaru, L. Spinu,
Y. Garcia, Inorg. Chem. 2010, 49, 5723.
(m), 1395 (m), 1337 (m), 1188 (s), 1082 (s), 1054 (s), 989 (s), 750
(s), 688 (s), 620 (s) cm–1. C33H30Fe1N12O8Cl2: calcd. C 46.66, H
3.56, N 19.79; found C 46.45, H 3.70, N 19.45.
[8] A. Bousseksou, G. Molnar, L. Salmon, W. Nicolazzi, Chem.
Soc. Rev. 2011, 40, 3313; I. Boldog, A. B. Gaspar, V. Martínez,
P. Pardo-Ibañez, V. Ksenofontov, A. Bhattacharjee, P. Gütlich,
J. A. Real, Angew. Chem. 2008, 120, 6533; Angew. Chem. Int.
Ed. 2008, 47, 6433; L. Catala, D. Brinzei, Y. Prado, A. Gloter,
O. Stéphan, G. Rogez, T. Mallah, Angew. Chem. 2008, 120, 1;
Angew. Chem. Int. Ed. 2008, 47, 1; E. Coronado, J. R. Galán-
Mascarós, M. Monrabal-Capilla, J. García-Martinez, P. Pardo-
Ibáñez, Adv. Mater. 2007, 19, 1359; T. Forestier, S. Mornet, N.
Daro, T. Nishihara, S.-I. Mouri, K. Tanaka, O. Fouché, E. Fre-
ysz, J.-F. Létard, Chem. Commun. 2008, 4327; S. M. Neville, C.
Etrillard, S. Asthana, J.-F. Létard, Eur. J. Inorg. Chem. 2010,
282.
[FeII(Hcintrz)3](Br)2·1.5(H2O) (6): Hcintrz ligand (300 mg,
1.5 mmol) was dissolved in MeOH (20 mL). Separately FeBr2
(100 mg, 0.50 mmol) and ascorbic acid (20 mg) were dissolved in
MeOH (3 mL). The two solutions were combined dropwise and a
white precipitate immediately formed. The precipitate was collected
by filtration and washed with MeOH. Mw 810.32; yield 333 mg
(82.1%); IR: ν = 2959 (s), 1649 (sh), 1626 (s), 1584 (s), 1529 (s),
˜
1448 (m), 1317 (w), 1188 (s), 1074 (s), 990 (s), 750 (s), 688 (s),
620 (s) cm–1. C33H30N12 Fe1Br2 + 5H2O (solvent not included in
formula): calcd. C 44.02, H 4.48, N 18.67; found C 43.94, H 3.88,
N 17.91
[9] T. Fujigaya, D.-L. Jiang, T. Aida, J. Am. Chem. Soc. 2005, 127,
5484.
[10] P. Grondin, D. Siretanu, O. Roubeau, M.-F. Achard, R. Clerac,
Inorg. Chem. 2012, 51, 5417.
Attempts to form crystals of the 1D chain materials [Fe(Xtrz)3](A)2·
nH2O by slow evaporation, slow diffusion and solvothermal meth-
ods were unsuccessful; in each case polycrystalline samples were
produced. However, when using A = NCS– or NCSe/ClO4– crystals
of the formula [Fe2(Hsaltrz)5(NCS)4]·4(H2O) (7) and [Fe3(Hsaltrz)6-
(H2O)2(EtOH)4](ClO4)6 (8) were obtained by slow diffusion. An
isostructural cobalt(II) analogue of 7, [Co2(Hsaltrz)5(NCS)4]·H2O
has been previously reported but differs in its degree of hy-
dration.[31] Additionally, an isomorph of 7 was recently reported,[19]
[Fe2(Hsaltrz)5(NCS)4]·4 MeOH, differing in the solvent.
[11] L. G. Lavrenova, V. N. Ikorski, V. A. Varnek, I. M. Oglezneva,
S. V. Larionov, Koord. Khim. 1986, 12, 207; O. Roubeau, J. M.
Alcazar-Gómez, E. Balskus, J. J. A. Kolnaar, J. G. Haasnoot, J.
Reedijk, New J. Chem. 2001, 25, 144; Y. Garcia, P. J. van Kon-
ingsbruggen, P. Lapouyade, L. Rabardel, O. Kahn, M. Wier-
czorek, R. Bronisz, Z. Ciunik, M. F. Rudolf, C. R. Acad. Sci.
Paris IIc 1998, 1, 523; Y. Garcia, P. J. van Koningsbruggen, R.
Lapouyade, L. Fournès, L. L. Rabardel, O. Kahn, V. Kseno-
fontov, G. Levchenko, P. Gütlich, Chem. Mater. 1998, 10, 2426.
[12] M. Seredyuk, A. B. Gaspar, M. C. Muñoz, M. Verdaguer, F.
Villain, P. Gütlich, Eur. J. Inorg. Chem. 2007, 4481.
[13] Y. Wei, P. Sonar, M. Grunert, J. Kusz, A. D. Schlüter, P.
Gütlich, Eur. J. Inorg. Chem. 2010, 3930.
[14] Y. Garcia, P. Guionneau, G. Bravic, D. Chasseau, J. A. K.
Howard, O. Kahn, V. Ksenofontov, S. Reiman, P. Gütlich, Eur.
J. Inorg. Chem. 2000, 1531.
[15] N. F. Chilton, PHI Computer Program, unpublished 2012.
[16] D. Savard, C. Cook, G. D. Enright, I. Korobkov, T. J. Burch-
ella, M. Murugesu, CrystEngComm 2011, 13, 5190.
[17] M. Thomann, O. Kahn, J. Guilhem, F. Varret, Inorg. Chem.
1994, 33, 6029.
[18] G. Vos, R. A. G. de Graaff, J. G. Haasnoot, A. M.
van der Kraan, P. de Vaal, J. Reedijk, Inorg. Chem. 1984, 23,
2905; G. Vos, R. A. Le Febre, R. A. G. de Graaff, J. G. Haas-
noot, J. Reedijk, J. Am. Chem. Soc. 1983, 105, 1683.
[19] Y. Garcia, P. Guionneau, G. Bravic, D. Chasseau, J. A. How-
ard, O. Kahn, V. Ksenofontov, S. Reiman, P. Gütlich, Eur. J.
Inorg. Chem. 2000, 1531.
[20] Y. Garcia, F. Robert, A. D. Naik, G. Zhou, B. Tinant, K. Rob-
eyns, S. Michotte, L. Piraux, J. Am. Chem. Soc. 2011, 133,
15850.
[Fe2(Hsaltrz)5(NCS)4]·2MeOH (7): Single crystals were prepared in
a H-shaped tube by slow diffusion. A solution of Fe(ClO4)2·6(H2O)
(128 mg, 0.50 mmol), NaNCS (81 mg, 1.0 mmol) and ascorbic acid
(5 mg) in MeOH (3 mL) was placed in one arm of the tube and a
solution of Hsaltrz (282 mg, 1.5 mmol) in MeOH (3 mL) was
placed in the other arm. The remainder of the tube was carefully
filled with neat MeOH and over a period of 2 weeks yellow crystals
of 7 formed.
[Fe3(Hsaltrz)6(H2O)2(EtOH)4](ClO4)6·2(EtOH) (8): Single crystals
were prepared in a H-shaped tube by slow diffusion. A solution of
Fe(ClO4)2·6(H2O) (128 mg, 0.50 mmol) and ascorbic acid (5 mg) in
of EtOH (3 mL) was placed in one arm of the tube and a solution
of Hsaltrz (94 mg, 0.5 mmol) in EtOH (3 mL) was placed in the
other arm. The remainder of the tube was carefully filled with neat
methanol and over a period of 2 weeks colorless crystals of 8
formed.
Supporting Information (see footnote on the first page of this arti-
cle): Powder X-ray diffraction, crystallographic information, and
thermogravimetric analyses.
[21] A. P. Railliet, A. D. Naik, A. Rotaru, J. Marchand-Brynaert,
Y. Garcia, Hyperfine Interact. 2012, 205, 51.
Eur. J. Inorg. Chem. 2013, 803–812
811
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
[22] Y. Wei, P. Sonar, M. Grunert, J. Kusz, A. D. Schlüter, P.
Gütlich, Eur. J. Inorg. Chem. 2010, 3930.
[23] M. Marchivie, P. Guionneau, J.-F. Letard, D. Chasseau, Acta
Crystallogr., Sect. B 2003, 59, 479; S. Matouzenko, A. Boussek-
sou, S. Lecoca, P. J. van Koningsbruggen, M. Perrin, O. Kahn,
A. Kollet, Inorg. Chem. 1997, 36, 5869.
[24] SMART, SAINT and XPREP. Area Detector and Data Integra-
tion and Reduction Software, Bruker Analytical Instruments
Inc., Madison, Wisconsin, USA, 1995.
[25] G. M. Sheldrick, SADABS. Empirical Adsorption Correction
Program for Area Detector Data, University of Göttingen, Ger-
many, 1996.
[27]
[28]
W. J. Kabsch, J. Appl. Crystallogr. 1993, 26, 795.
SHELXL97. Program for Crystal Structural Solution and Re-
finement, Bruker Analytical Instruments Inc., Madison, Wis-
consin, USA, 1997; L. J. Barbour, University of Stellenbosch,
South Africa, 1999.
F. Robert, A. D. Naik, F. Hidara, B. Tinant, R. Robiette, J.
Wouters, Y. Garcia, Eur. J. Org. Chem. 2010, 621.
[29]
[30] A. Colautti, R. J. Ferlauto, V. Maurich, M. De Nardo, C. Nisi,
F. Rubessa, C. Runti, Chim. Ther. 1971, 6, 367; M. Mazza, L.
Montanari, F. Pavanetto, Farmaco Ed. Sci. 1976, 31, 334.
[31] Y. Zhang, S. Chen, G. Fan, Z. Zhao, S. Gao, Chin. J. Chem.
2009, 27, 1697.
[26] T. M. McPhillips, S. E. McPhillips, H. J. Chiu, A. E. Cohen,
A. M. Deacon, P. J. Garman, A. Gonzalez, N. K. Sauter, R. P.
Phizackerley, S. M. Soltis, P. Kuhn, J. Synchrotron Radiat.
2002, 9, 401.
Received: September 14, 2012
Published Online: January 16, 2013
Eur. J. Inorg. Chem. 2013, 803–812
812
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim