Figure 11. LIESST effect of (a) 2, (b) 3, and (c) 4. Initial thermal spin transition in the cooling mode from 300 K to 10 K at a sweep rate of 2 K
−
1
min , 532 nm light irradiation at 10 K, and thermal relaxation after switching the light off are indicated as blue, green, and red triangles,
−
1
respectively. The thermal relaxation process was recorded in the warming mode from 10 K at a sweep rate of 0.3 K min until the completion of
−
1
thermal relaxation (150, 130, and 130 K for 2, 3, and 4, respectively) and then the temperature was increased at 2 K min to 300 K.
are retained at lowered temperature, and interestingly, at 275 K
in the heating mode (see Figures S22e, S22f, S23e, and S23f).
Consequently, this conformational bistability induces hyste-
natures, the temperature of the middle point (HS:LS =
50:50) of the overall SCO between complete HS and LS states
for 2 and 3 or the middle point of the complete HS−HS and
retic small χ T changes between 273 K and 285 K (see also
HS−LS states for 4 was used as the T value. The estimated
M
1/2
magnetic properties and DSC study sections). Except for the
T
values are 165, 171, and 189 K for 2, 3, and 4,
1
/2
above-mentioned conformational change, NTf anions show
respectively, resulting the estimated T value of 160 K for the
2
0
no further conformational change or disordering between 296
K and 105 K. In this way, in contrast to methyl-substituted 1,
the lattice of methoxy-substituted 4 can accommodate the
present triazole-containing tripodal hexadentate ligand system.
It is noteworthy that the T value of 160 K for 2−4 is much
0
higher than that of the related complexes bearing similar
55,56
conformational change of the NTf anion and the terminal
imidazole-containing N tripodal ligand (T = 100 K).
2
6 0
6
2
MeO group of the ligand, allowing the partial volume change
of the complex cations associated with SCO.
DFT Studies. DFT calculations for 1−4 were performed.
Based on the results of calculations with TPSSh, M06L, and
B3LYP functionals (see the Supporting Information), and
DFT studies reported by Cirera and co-workers, we chose
LIESST Experiments. Photomagnetic studies were also
performed for 2−4, since they showed a variety of SCO
behaviors. As shown in Figure 11, as well as Figure S24 in the
Supporting Information, after initial cooling from 300 K to 10
K at a sweep rate of 2 K min− in darkness, green laser light
63
TPSSh functional in the present discussion.
First, we performed single-point DFT energy calculations on
experimental crystal structures (Table S14 in the Supporting
Information). The HS−LS energy differences (ΔEHS−LS) for
1−4 at 296 K (−9700, −9715, −9542, −9716, and −9460
1
−
2
irradiation (532 nm, the power of the light is ca. 5 mW cm )
to the polycrystalline samples at 10 K for 15−20 min affords an
−
1
increase of the χ T values with the saturated values of ca. 3.0,
cm for 1, 2, 3, Fe1 site of 4, and Fe2 site of 4, respectively)
M
3
−1
2
.9, and 2.9 cm K mol for 2, 3, and 4, respectively,
reveal that the HS states are more stable in all complexes at
indicating efficient and almost quantitative conversion of the
Fe site from the LS (for 2 and 3) or 1/2(HS + LS) (for 4) to
RT. On the other hand, at 100−105 K, ΔE
values are
HS−LS
II
−1
12 783, 11 471, and 10 644 cm for 2, 3, and Fe2 site of 4,
the photoinduced metastable HS state. After the light was
switched off, the thermal relaxation was studied. Upon
respectively, suggesting that the LS states are more stable,
−
1
except for the value of −10 058 cm for Fe1 site of 4 (the HS
elevating the temperature, the χ T values increase to reach
state is more stable). At 185 K for 2, ΔE
HS−LS
values are
M
3
−1
−1
the maximum values of ca. 3.5, 3.3, and 3.4 cm K mol for 2,
−9135, −9123, and 11 769 cm for Fe1, Fe2, and Fe3 site,
3, and 4, respectively, and then slightly decrease. The increase
respectively, indicating that the HS state is more stable in Fe1
of the χ T values from 10 K to ca. 25 K can be attributed to
and Fe2 sites, whereas the LS state is more stable in Fe3 site.
M
II
−1
the zero-field splitting of the trapped Fe HS molecules (S =
Finally, at 160 K for 3, ΔE
value of 10 821 cm for Fe2
site reveals that the LS state is more stable, while the value of
HS−LS
6
0
2
). Finally, the χ T values decrease abruptly in a single-step
M
−
1
manner at ∼95−120 K (for 2 and 3), and at ∼85−105 K (for
−4282 cm for Fe1 site, reflecting the existence of both HS
and LS species. These results are consistent with the magnetic
properties and single-crystal X-ray structure analyses.
4
) to reach the thermally stable LS state (for 2 and 3), and 1/
61
2
(HS + LS) state (for 4). The T(LIESST) values determined
from the dχ T/dT vs T curve in the warming mode at a sweep
rate of 0.3 K min are 113, 111, and 97 K for 2, 3, and 4,
respectively.
Next, we performed DFT geometry optimization calculation
in the gas phase on 1−4 at both LS and HS states using
experimental crystal structures as initial geometry. The
geometrical parameters are summarized in Table S15 in the
Supporting Information, and the optimized structures are
shown in Figure S25 in the Supporting Information. Optimized
structural parameters for 1−4 are similar to each other,
irrespective of the remote substituents Me, Cl, Br, and MeO,
and are in agreement with experimental crystal structures in
M
−
1
Letard et al. have reported the linear correlation between
́
T(LIESST) and T1 values with the general equation
/2
T(LIESST) = T − 0.3T1/2, which was constructed by the
0
comprehensive examination of a large number of SCO
6
1
compounds. To reveal the T value of the present ligand
0
system, first, the T1/2 values of 2−4 were estimated. Since 2−4
have a variety of thermal SCO profiles including the
combination of stepwise, gradual, abrupt, and hysteretic
each spin states. In addition, the ΔE
values of these DFT
HS−LS
optimized structures are also similar to each other (Table S19
K
Inorg. Chem. XXXX, XXX, XXX−XXX
Hagiwara, Hiroaki
