Kinetic study of thermal Z to e isomerization reactions of azobenzene and 4-dimethvlaniino-4′-nitroazobenzene in Ionic Liquids [1-R-3- methyliniidazolium bis(trifluoromethylsulfonyl)imide with R = Butyl, Pentyl, and Hexyl]

Article abstract of DOI:10.1002/chem.200600081

Thermal Z to E isomerization reactions of azobenzene and 4-dimethylamino-4′-nitroazobenzene were examined in three ionic liquids of general formula 1-R-3-methylimidazolium bis(trifluoromethylsulfonyl)imide (R = butyl, pentyl. and hexyl). The first-order rate constants and activation energies for the reactions of azobenzene measured in these ionic liquids were consistent with those measured in ordinary organic solvents, which indicated that the slow isomerization through the inversion mechanism with a nonpolar transition state was little influenced by the solvent properties, such as the vis cosity and dielectric constant, of ionic liquids. On the other hand, the rate constants and the corresponding frequency factors of the Arrhenius plot were significantly reduced for the isomerization of 4-dimethylamino-4′- nitroazobenzene in ionic liquids compared with those for the isomerization in ordinary organic molecular solvents with similar dielectric properties. Although these ionic liquids are viscous, the apparent viscosity dependence of the rate constant could not be explained either by the Kramers-Grote-Hynes model or by the Agmon-Hopfield model for solution reactions. It is proposed that the positive and the negative charge centers of a highly polar rotational transition state are stabilized by the surrounding anions and cations, respectively, and that the ions must be rearranged so as to form highly ordered solvation shells around the charge centers of the reactant in the transition state. This requirement for the orderly solvation in the transition state results in unusually small frequency factors of 10⁴-10⁷ s ^-1.

Full text of DOI:10.1002/chem.200600081

DOI: 10.1002/chem.200600081
Kinetic Study of Thermal Z to E Isomerization Reactions of Azobenzene and
4
-Dimethylamino-4’-nitroazobenzene in Ionic Liquids [1-R-3-
Methylimidazolium Bis(trifluoromethylsulfonyl)imide with R=Butyl, Pentyl,
and Hexyl]
[
a]
[a]
[a]
[a]
[a]
Keita Baba, Hajime Ono, Eri Itoh, Sumitaka Itoh, Kyoko Noda,
[a]
[
a]
[b]
[c]
Toshinori Usui, Koji Ishihara, Masahiko Inamo, Hideo D. Takagi,* and
Tsutomu Asano*
[
d]
Abstract: Thermal Z to E isomeriza-
tion reactions of azobenzene and 4-
cosity and di
A
H
R
U
G
constant could not be explained either
by the Kramers–Grote–Hynes model
or by the Agmon–Hopfield model for
solution reactions. It is proposed that
the positive and the negative charge
centers of a highly polar rotational
transition state are stabilized by the
surrounding anions and cations, respec-
tively, and that the ions must be rear-
ranged so as to form highly ordered
solvation shells around the charge cen-
ters of the reactant in the transition
state. This requirement for the orderly
solvation in the transition state results
in unusually small frequency factors of
liquids. On the other hand, the rate
constants and the corresponding fre-
quency factors of the Arrhenius plot
were significantly reduced for the iso-
dimethylamino-4’-nitroazobenzene were
A
C
H
T
R
E
U
N
G
examined in three ionic liquids of gen-
eral formula 1-R-3-methylimidazolium
bis(trifluoromethylsulfonyl)imide (R=
butyl, pentyl, and hexyl). The first-
order rate constants and activation en-
ergies for the reactions of azobenzene
measured in these ionic liquids were
consistent with those measured in ordi-
nary organic solvents, which indicated
that the slow isomerization through the
inversion mechanism with a nonpolar
transition state was little influenced by
the solvent properties, such as the vis-
merization of 4-dimethylamino-4’-nitro-
azobenzene in ionic liquids compared
with those for the isomerization in or-
dinary organic molecular solvents with
similar dielectric properties. Although
these ionic liquids are viscous, the ap-
parent viscosity dependence of the rate
Keywords: azobenzenes
liquids · isomerization · kinetics
4
7
À1
10 –10 s .
Introduction
[
a] K. Baba, H. Ono, E. Itoh, Dr. S. Itoh, K. Noda, T. Usui,
Prof. Dr. H. D. Takagi
Research Center for Materials Science
Nagoya University
Nagoya 464-8602 (Japan)
Fax: ( +81)52-789-5473
E-mail: h.d.takagi@nagoya-u.jp
Ionic liquids have attracted much attention from chemists in
various fields because of their unique physical and chemical
properties, such as 1) fluidity over wide temperature ranges,
2) nonvolatility even in a vacuum, and 3) relatively large di-
electric constants. In addition to these properties, some
series of ionic liquids have been synthesized to exhibit a
wide range of viscosity by varying the substituents on the
[
b] Prof. Dr. K. Ishihara
Department of Chemistry
Waseda University
[1–5]
cationic and/or anionic part of the molecule.
Although
Shinjuku, Tokyo 169-8555 (Japan)
various types of organic/inorganic reactions in ionic liquids
have been reported to date, systematic studies concerning
the effect of the physical properties of ionic liquids on the
[
c] Prof. Dr. M. Inamo
Department of Chemistry
Aichi University of Education
Kariya 448-8542 (Japan)
[6]
reaction process are scarce.
[
d] Prof. Dr. T. Asano
Thermal Z to E isomerization of substituted azobenzenes
Department of Applied Chemistry
Oita University
Oita 870-1192 (Japan)
[7–13]
has been intensively studied in organic solvents,
and the
mechanism has been examined in various media including
[14]
E-mail: asano-at@oct-net.ne.jp
supercritical CO2.
It is known that the isomerization of
5328
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 5328 – 5333

FULL PAPER
unsubstituted azobenzene (AB)
proceeds through the inversion
transition state 1, in which one
of the nitrogen atoms is sp hy-
bridized. On the other hand,
pressure. The ionic liquids, 1-R-3-methylimidazolium bis(trifluoromethyl-
sulfonyl)imide (R=butyl, pentyl, and hexyl) are abbreviated as B-IL,
P-IL, and H-IL, respectively, throughout this article.
It was found that essentially colorless ionic liquids were obtained when
the reactions were carried out with excess amounts of active charcoal.
Treatment of the final products with charcoal did not significantly im-
prove the purity of the ionic liquids: the purity was examined by using
microanalysis as well as by the faint color exhibited by the impure prod-
uct.
the reaction of push–pull-substituted azobenzenes, such as
-dimethylamino-4’-nitroazobenzene (DNAB), takes place
4
through two independent mechanisms: in nonpolar solvents,
by inversion of one of the nitrogen atoms as in AB, and in
polar solvents, by rotation around the NÀN bond through
Because analyses using the Karl Fisher method revealed that the synthe-
sized ionic liquids contained large amounts of water (typically
À1
1
20 mmolkg for 1-butyl-3-methylimidazolium bis(trifluoromethylsulfo-
the polarized transition state 2 (Scheme 1). In solvents such
nyl)imide), each ionic liquid was dried in a vacuum oven at 608C for
8 h; it was shown from the preliminary experiments to examine the opti-
mum conditions for drying ionic liquids at 608C that synthesized B-IL
4
À1
À1
contained 120 mmolkg of water, which was reduced to 70 mmolkg
À1
À1
after 2 h, to ꢀ10 mmolkg after 24 h, and to 5.0 mmolkg after 48 h.
The temperature of the oven was kept at 608C, because the drying proce-
dure at temperatures higher than 808C caused partial decomposition, as
observed by the coloring of B-IL. The amount of water in the treated
À1
ionic liquids was less than 5 mmolkg for all of the ionic liquids used in
this study. Further drying under the same conditions did not make a sig-
nificant improvement.
Scheme 1. The isomerization reaction of DNAB in polar solvents by rota-
tion around the NÀN bond through polarized transition state 2.
1
B-IL: H NMR ([D
1
1
]chloroform): d=8.75 (s, 1H; NCHN), 7.30 (m, J=
), 3.94 (s, 3H;
), 1.36 (sext, J=7.5 Hz, 2H; N-
), 0.96 ppm (t, J=7.5 Hz, 3H; CH CH ); elemental analysis
: C 28.64, N 10.02, H 3.61; found: C 28.22,
.5 Hz, 2H; NCHCHN), 4.17 (t, J=7.4 Hz, 2H; NCH
CH N), 1.85 (q, 2H; NCH CH
(CH CH
2
3
2
2
A
H
R
U
G
2
)
2
2
3
2
calcd (%) for C10
15 3 4 2 6
H N O S F
as benzene, the two mechanisms compete with each other
and a shift of the mechanism was identified by monitoring
N 10.21, H 3.61.
1
P-IL: H NMR ([D
1
]chloroform): d=8.76 (s, 1H; NCHN), 7.30 (m, 2H;
), 3.94 (s, 3H; CH N), 1.87 (q, J=6.9, 2H;
), 1.33 (m, J=7.9 Hz, 4H; CH (CH ), 0.90 ppm (t, J=5.1 Hz,
3H; CH CH ); elemental analysis calcd (%) for C H N O S F : C
[12]
the temperature dependence of the activation energy.
These mechanisms were later supported by quantum-chemi-
NCHCHN), 4.16 (t, 2H; NCH
NCH CH
2
3
2
2
3
A
H
R
U
G
2 2
)
[15]
cal calculations.
3
2
11 17
3
4
2
6
In this study, we focus on the interaction between the
solute and the ionic liquids for the Z to E isomerization re-
actions of AB and DNAB at different temperatures. The
alkyl substituent on the imidazolium ring was varied to con-
trol the viscosity of the ionic liquids, while the dielectric
properties were not largely altered.
30.49, N 9.70, H 3.95; found: C 30.29, N 9.92, H 3.83.
1
H-IL: H NMR ([D
1
]chloroform): d=8.79 (s, 1H; NCHN), 7.29 (m, 2H;
), 3.95 (s, 3H; CH N), 1.86 (m, 2H;
(CH ), 0.88 ppm (t, J=6.9 Hz, 3H;
); elemental analysis calcd (%) for C12 : C 32.21, N
.39, H 4.28; found: C 32.69, N 9.49, H 4.65.
NCHCHN), 4.17 (t, 2H; NCH
NCH CH ), 1.32 (m, 6H; CH
CH CH
2
3
2
2
3
A
T
E
N
2 3
)
3
2
19 3 4 2 6
H N O S F
9
Measurements: Sample solutions were prepared by using a vacuum line
to avoid contamination with water, and sealed after deaeration. Concen-
trations of the azobenzenes in each sample solution were determined
spectrophotometrically. Amounts of water in each sample solution were
determined by using a Mitsubishi Kasei KA-07 Karl Fisher apparatus
after each measurement by breaking the sample cells: it was found that
the amount of water in each solution was typically less than
Experimental Section
Materials: All of the common organic solvents were purified as recom-
[
16]
À1
À1
À1
mended. AB (Wako Pure Chemical Ind.) was purified by sublimation,
whereas DNAB (TCI) was recrystallized from toluene. 1-Chlorobutane,
2.74 mmolkg
5
for R=butyl, 4.12 mmolkg
for R=pentyl, and
.00 mmolkg for R=hexyl, thus indicating that contamination with
1
-chloropentane, and 1-chlorohexane (Tokyo Kasei Ltd.; GR grade)
water was successfully avoided by using this sampling method.
were dried over anhydrous MgSO for several days before distillation. 1-
4
Z isomers of AB and DNAB were prepared by irradiating the samples
with light (l>350 by using a UV-cutoff filter) with sufficient agitation by
using a Ushio high-pressure Hg lamp (1000 W) for 30 min. During the ir-
radiation, the temperature of the sample solutions was controlled by
using thermostatted water, so as to make the measurements of the ther-
mal Z to E isomerization possible immediately after irradiation. Kinetic
measurements of the isomerization of the azobenzenes were carried out
by measuring the absorption change of the n–p* transition-band maxima
at lꢀ440 and ꢀ480 nm for AB and DNAB, respectively, by utilizing a
JASCO V-570 spectrophotometer with a thermostatted cell compartment.
The temperature of the sample solutions was controlled to within
Æ0.28C by means of the circulation of the thermostatted water. All the
reactions were found to clearly be of first order, and such an example of
the absorbance change corresponding to the Z to E thermal isomeriza-
Methylimidazole (Aldrich) was dried over sodium and distilled under re-
duced pressure. The ionic liquids were synthesized by modifying the re-
[
2,4]
ported methods.
-R-3-Methylimidazolium chloride (R=butyl, pentyl, and hexyl) was ob-
tained by refluxing each solution of 1-methylimidazole in toluene (ca.
1
0
0
.5 mol) and 1-chlorobutane, 1-chloropentane, or 1-chlorohexane (ca.
.8 mol), respectively, at 112–1138C under an argon atmosphere for 3 h.
Each oily product was separated from toluene under reduced pressure.
The yield was nearly 100%. The bis(trifluoromethylsulfonyl)imide salts
(
ionic liquids) of each imidazolium cation were prepared by using the fol-
lowing method: Lithium bis(trifluoromethylsulfonyl)imide (0.09 mol) was
added to a solution of 1-R-3-methylimidazolium chloride (0.08 mol) dis-
solved in distilled water (100 mL), and the mixture was left under reflux
at 1008C for 1 h. After cooling the mixture to room temperature, the or-
ganic layer was extracted by using dichloromethane (200 mL). The final
products were obtained by removing dichloromethane under reduced
1
tion of AB in B-IL is shown in Figure 1. H NMR signals were measured
by using a Bruker AMX-400WB spectrometer at 400 MHz.
Chem. Eur. J. 2006, 12, 5328 – 5333
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5329

H. D. Takagi, T. Asano et al.
Figure 1. Absorbance change at l=438 nm for the reaction of azoben-
À4
À1
zene in B-IL. [AB]=9.12”10 mmolkg , T=318.0 K.
Figure 2. Relationship between the position of the n–p* band maximum
of azobenzene and the dielectric constant of organic solvents.
Results and Discussion
Table 1. Absorption maximum of the n–p* band of azobenzene in ionic
liquids at various temperatures.
Shift of the n–p* transition band in various solvents includ-
ing ILs: Reports on the dielectric properties of ionic liquids
are limited, because of 1) the uncertainty in the measure-
ments that originates from the relatively high dielectric con-
stants of this type of fluid, and 2) the nonexistence of estab-
lished formulae that describe the dielectric properties of
molten salts other than the qualitative theory postulated by
T [8C]
l
max
(n–p*) [nm]
A
T
E
N
B-IL
P-IL
H-IL
6
5
0
5
443
444
444
442
442
445
445
443
445
443
445
443
442
443
50
4
2
5
3–40
[17]
Doucet. Dielectric constants have been reported only for
[
18]
molten ethylammonium nitrate, a molten Ca(NO ) /KNO
A
H
R
U
G
3 2
[20]
3
Table 2. Absorption maxima of the n–p* and p–p* bands of azobenzene
in organic solvents at 258C.
[19]
mixture,
molten trioctylammonium chloride,
and
[21]
molten nitrate salts of alkali metals. Weingartner reported
that the relative dielectric constant of molten ethylammoni-
um nitrate was small (27.1 at 158C) and decreased monoton-
ically to 22.2 at 808C, although measurements of dielectric
properties were difficult at a low frequency. It is also known
that the Walden product, the product of viscosity and con-
[a]
e
rel
l
max
A
H
R
U
G
l
max
(p–p*)
A
C
H
T
R
E
U
N
G
[
[nm]
1
,4-dioxane
2.209
2.381
20.56
24.55
35.94
64.92
448.5
447
444
443
441
438
318
320
toluene
acetone
ethanol
acetonitrile
propylene carbonate
316.5
316
319
[18]
ductivity of a fluid, is almost constant for this ionic liquid.
In our study, we utilized a rather classical method for the
rough estimation of the relative dielectric constants of ionic
liquids: the shift of the n–p* transition band of AB was
monitored in various organic solvents and in the ionic liq-
uids. It has been known that the n–p* band shifts toward
the shorter wavelength region in polar solvents, because the
electronic configuration in the ground sate is more stabilized
compared with the p* excited state. The results are shown
in Figure 2 and Tables 1 and 2: the n–p* band maximum
was monotonically blueshifted on increasing the relative di-
[
a] Relative dielectric constant.
Z to E isomerization of AB in B-IL, P-IL, and H-IL: The
results for the thermal Z to E isomerization of AB exam-
ined in ionic liquids are summarized in Tables 3 and 4. The
[
a]
Table 3. Rate constants and kinetic parameters observed for Z to E iso-
merization of azobenzene in three ionic liquids at various temperatures.
À4
À1
B-IL, [AB]=9.12”10 molkg
T [K] 310.2
k [s
A
C
H
T
R
E
U
N
G
electric constant of the solvent. In addition, the position of
the band maximum was not sensitive to temperature (22 to
08C). From these results, the relative dielectric constants
315.0
6.63”10
319.5
1.11”10
324.2
1.69”10
328.9
2.97”10
À1
À6
À6
À5
À5
À5
À5
À5
À5
À5
À5
À5
]
3.82”10
6
À3
À1
for B-IL, P-IL, and H-IL were determined to lie in the
range of 15 to 20. This range of dielectric constant is consis-
tent with the value for ethylammonium nitrate at room tem-
P-IL, [AB]=1.00”10 molkg
T [K] 310.9
k [s
315.5
6.69”10
319.8
1.24”10
323.9
1.91”10
327.6
3.49”10
À1
À6
À6
]
3.30”10
[18]
perature. The reason for such small dielectric constants of
these ionic liquids may be attributed to the formation of ion
pairs that have a longer lifetime than the dielectric relaxa-
À4
À1
H-IL, [AB]=9.54”10 molkg
T [K] 310.6
k [s
315.6
7.73”10
319.8
1.29”10
324.0
2.15”10
329.4
3.51”10
À1
À6
À6
]
4.77”10
À10
[18]
tion (ca. 10 s at room temperature ), or to the low mobi-
lity of the large cations with asymmetric structures.
[
a] Typical errors for the rate constants at each temperature were within
2% for two runs.
5330
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 5328 – 5333

Kinetic Study of Thermal Isomerization
FULL PAPER
[
a]
Table 4. Activation energies and frequency factors for the isomerization
reaction of azobenzene in ionic liquids.
Table 5. Rate constants observed for Z to E isomerization of DNAB in
[
b]
three ionic liquids at various temperatures.
À1
À1
À5
À1
Ionic liquid
E
a
[kJmol
]
A [s
]
B-IL, [DNAB]=3.02”10 molkg
T [K] 280.9 284.2
3.86Æ0.02 4.99Æ0.01 6.55Æ0.03 7.82Æ0.03 8.78Æ0.02
[
a]
9
10 [a]
14 [a]
288.7
290.4
293.3
B-IL
P-IL
H-IL
89.5Æ2.1 (91.6Æ2.0)
4.8”10 (1.1”10 )
1.3”10 (1.4”10 )
3.5”10
3
À1
[a]
12
k”10 [s ]
104.5Æ3.1 (117Æ3.0)
94.4Æ4.7
1
0
À5
À1
P-IL, [DNAB]=2.01”10 molkg
T [K] 280.9 284.2
0.94Æ0.04 1.18Æ0.03 1.89Æ0.02 2.15Æ0.03 2.51Æ0.03
[
a] Data obtained in ionic liquids that contain ꢀ100 mmolkgÀ1 of water
288.7
290.4
293.3
are shown in the parentheses.
3
À1
k”10 [s
]
À5
À1
H-IL, [DNAB]=2.01”10 molkg
T [K]
284.1
1.06Æ0.02 1.40Æ0.03 1.76Æ0.03 2.15Æ0.04 3.93Æ0.03
316.7 320.9 324.8
6.35Æ0.07 7.58Æ0.05 9.01Æ0.15
288.7
293.4
297.9
307.1
3
À1
rate constants (k) were comparable to the reported values
k”10 [s
T [K]
]
]
for common organic solvents and the activation energies,
3
À1
À1
k”10 [s
E =90 (B-IL), 104 (P-IL), and 95 kJmol (H-IL), were
a
[
a] Standard deviations given are for up to four runs). [b] Activation pa-
also in the range of reported values for ordinary organic sol-
À1 [8,22,23]
rameters obtained from these data are shown in Table 6.
vents, that is, 84–104 kJmol .
Neither the rate con-
stants nor the activation parameters in Tables 3 and 4 show
a clear correlation with the viscosity of the ionic liquids, 52
(
B-IL), 57 (P-IL), and 60 cP (H-IL), at 258C. The reaction
stants for these ionic liquids are as large as 15 to 20: the
rate constants measured for molecular solvents that have
similar dielectric constants to those of the ionic liquids are
unequivocally larger. Such large rate constants observed in
dipolar solvents have been explained by the dominance of
the rotation mechanism in which the nitrogen–nitrogen p
bond is heterolytically cleaved and the reaction is effected
became faster on increasing the chain length on the imida-
zolium ring, probably reflecting a small decrease of the po-
[24]
larity of the medium. Therefore, it is reasonable to con-
clude that the thermal Z to E isomerization of AB was not
influenced by the viscosity of the medium and that the reac-
tion can be safely described by the ordinary transition-state
theory (TST) even in these ionic liquids.
[7–12]
by rotation around the remaining NÀN bond.
It is evi-
dent from Table 6 that the frequency factors (A) are very
Z to E isomerization of DNAB in B-IL, P-IL, and H-IL:
The results for the thermal Z to E isomerization of 4-dime-
thylamino-4’-nitroazobenzene
small for the reactions in ionic liquids, although the activa-
tion energies are in the range of those (42–52 kJmol ) re-
À1
(
DNAB) examined in B-IL, P-
Table 6. Kinetic parameters reported for the Z to E isomerization of DNAB in various solvents at 258C.
IL, and H-IL are summarized
in Table 5. The Arrhenius plots
were linear and therefore no
significant change in the reac-
tion pathway was indicated in
the temperature range exam-
ined in this study. Summarized tetrahydrofuran
in Table 6 are the kinetic pa-
rameters reported for the Z to
E isomerization of DNAB in
various solvents, together with N,N-dimethylformamide
the relative dielectric constants dimethylsulfoxide
(
solvent. In Figure 3, the plots of
the activation energy and the
rate constant of DNAB against tetrachloromethane
the relative dielectric constant
of the solvents at 258C are
Solvent
e
rel
h
E
a
A
k
[s ]
À1 [a]
À1 [b]
À1 [a]
ACHTREUNG
[
cP]
A
H
R
U
G
]
A
H
R
U
G
]
[
c]
14
cyclohexane
hexane
benzene
2.023
1.880
2.275
2.209
7.58
0.898
0.2985
0.6028
88.3
83.7
102.0
108.4
48.1
52.3
46
49.0
48.1
42
1.2”10
1.3”10
3.5”10
3.4”10
1.6”10
1.2”10
6.8”10
8.8”10
1.6”10
4.1”10
9.1”10
1.4”10
1.2”10
1.2”10
0.04
1
1
2
5
0.002
0.0108
0.0142
0.513
0.692
2.5
17.39
47.03
50
[d]
16
8
1,4-dioxane
1.439
0.460
1.324
0.3040
1.078
0.5445
0.802
1.996
9
o-dichlorobenzene
acetone
ethanol
methanol
9.93
8
20.70
24.55
32.70
36.71
46.68
9
10
8
8
38
120
[
e]
6
[f]
[f]
[f]
B-IL
P-IL
H-IL
pentane
ꢀ17.5
ꢀ17.5
ꢀ17.5
52
57
60
45.6Æ2.5
56.9Æ3.3
41.0Æ0.8
0.00922
0.00350
0.00225
0.00271
0.0027
0.0111
0.162
0.0152
0.180
0.90
3.0
1.48
8.22
8.35
1640
3.34
erel) and viscosities (h) of each
[
e]
7
[e]
4
[c]
1.844
0.225
1.038
0.5516
0.757
[
c]
[d]
2.24
toluene
chlorobenzene
diethyl ether
chloroform
2.379
5.621
4.335
4.806
8.93
12.40
10.36
19.92
[
c]
0.237
shown. It is seen that the rate
constants measured in the ionic dichloromethane
liquids at 258C are close to pyridine
those observed in solvents with
relatively small dielectric con-
stants, such as hexane, benzene,
[c]
0.5413
0.4117
0.884
0.7823
2.1297
2.0197
3.302
[
c]
1
2
,2-dichloroethane
-propanol
[c]
cyclohexanone
formamide
18.3
[c]
111
6.0
[
d]
[c]
8
1
,4-dioxane, pentane, and tetra- glycerol triacetate
chloromethane, although the es-
timated relative dielectric con- reference [1]. [f] This work.
16.9
49.8
6.6”10
[a] From references [7] and [8]. [b] Frequency factors for Arrhenius plot. [c] At 208C. [d] At 158C. [e] From
Chem. Eur. J. 2006, 12, 5328 – 5333
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5331

H. D. Takagi, T. Asano et al.
Figure 4. Plots of the logarithmic value of the isomerization rate constant
against the logarithmic value of the viscosity of the solvents. Only data
for molecular solvents (&) with similar dielectric constants to those of
the ionic liquids (
&
) were used.
[37,38]
(
=10 Pas).
It also has to be pointed out that in this “dif-
fusion-controlled” region, the reaction was characterized by
nonlinear Arrhenius plots with large downward deviations
[37,38]
at low temperatures,
P-IL, and H-IL.
while the plots were linear in B-IL,
A change of the mechanism from the rotation in polar
molecular solvents to the nitrogen inversion in the ionic liq-
uids can also be rejected on the basis of the activation pa-
rameters shown in Table 6. The slow isomerization in the
ionic liquids was caused by smaller frequency factors where-
as similarly slow isomerization in nonpolar molecular sol-
vents was a result of large activation energies. There is little
question that the reaction proceeded via the highly polar ro-
tational activated complex 2 (Scheme 1) in the ionic liquids.
All of these experimental results indicate that the very
small rate constants observed in the ionic liquids were not
caused by the large macroscopic viscosity (52–60 cP at 258C
decreasing to ca. 20 cP at 408C). In other words, the reac-
tion is still in the TST-valid region in the present ionic liq-
uids and the results must be interpreted on the basis of “sol-
vation equilibria” in the initial and the transition states. In
the initial state, specific interactions between the reactant Z
isomer and the surrounding ions is not as strong as those in
the transition state, because the reactant does not carry any
formal charges. In such a case, the main role of the solvent
is to provide a cavity that hosts the reactant. In the transi-
tion state, on the other hand, newly formed positive and
negative charge centers must be stabilized by contact with
the anion(s) and the cation(s), respectively. Though we do
not know exactly how the ions are oriented around each
charge center, random arrangements should not result in the
maximum stabilization. The cationic and anionic parts of
Figure 3. Activation energy (top) and lnk (bottom) for the isomerization
reaction of DNAB in various solvents plotted against the relative dielec-
tric constant of the solvents.
ported for the reactions in solvents with relative dielectric
constants of 10 to 30, such as acetone, ethanol, methanol,
and so forth. It should be noted that no significant decrease
of the frequency factor was observed for the slow isomeriza-
tion reaction of AB through the inversion mechanism in
these ionic liquids.
Log–log plots of the rate constant against the viscosity for
relatively polar molecular solvents (10<e ) and for ionic
rel
liquids are illustrated in Figure 4. Clearly, the rate constant
shows little correlation with the viscosity for common organ-
ic solvents. In the ionic liquids, however, the reaction was
markedly slower and logk seems to be linearly correlated
v
a
with logh. The activation energy of viscous flow (E ) was
roughly estimated by using the reported viscosities at vari-
v
a
À1
ous temperatures for B-IL to give E =41Æ7 kJmol ,
which may be compared with the activation energy of the
isomerization reaction of DNAB in this fluid, that is, E =
a
À1
4
6Æ6 kJmol . Therefore, it is tempting to attribute this re-
tardation of the isomerization to slow thermal fluctuations
of the cations and the anions surrounding the reactant.
[
25–27]
However, neither the Kramers–Grote–Hynes model
the more comprehensive Agmon–Hopfield-type models
nor
each ionic liquid must be placed at fairly strict positions
[
28–36]
+
with suitable orientations, around the Me N = and the =
ACHTREUNG
2
À
can rationalize these retardations because the slope of the
plot, À10, was much larger than the value predicted by
these models, 0<Àslope<1. In addition, the isomerization
of DNAB had been found to enter the Kramersꢁ “high
NO2 atomic groups. We therefore conclude that each cati-
onic and anionic part of the ionic liquid independently sol-
vates the polarized structure of DNAB in the transition
state and the counterions have to be positioned, again in a
relatively strict manner, next to the nearest neighbors of
4
viscosity region” only when the viscosity exceeded 10 cP
5332
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 5328 – 5333

Kinetic Study of Thermal Isomerization
FULL PAPER
each cationic/anionic part. A longer carbon chain on the
imidazolium ring would have imposed further restrictions
for the arrangement of the ions and influenced the rate of
the isomerization through the frequency factor. The failure
to observe any strong retardations of the isomerization of
AB in the ionic liquids implies the importance of the elec-
trostatic interactions in the highly polar rotational transition
state 2. In the inversion transition state of AB (1), the inter-
actions between the surrounding ions and the reactant are
not as strong as those in 2 and looser and less strict arrange-
ments of the ions (ionic liquids) would suffice to produce
optimum solvation: considerably larger frequency factors
were observed for the reactions of AB as a result.
[8] T. Asano, T. Okada, S. Shinkai, K. Shigematsu, Y. Kusano, O.
Manabe, J. Am. Chem. Soc. 1981, 103, 5161–5165.
[
9] T. Asano, T. Yano, T. Okada, J. Am. Chem. Soc. 1982, 104, 4900–
904.
10] T. Asano, T. Okada, J. Phys. Chem. 1984, 88, 238–243.
[11] T. Asano, T. Okada, J. Org. Chem. 1984, 49, 4387–4391.
4
[
[
[
[
12] T. Asano, T. Okada, J. Org. Chem. 1986, 51, 4454–4458.
13] D.-M. Shin, D. G. Whitten, J. Am. Chem. Soc. 1988, 110, 5206–5208.
14] T. Usui, S. Funahashi, H. D. Takagi, T. Asano, Chem. Lett. 1998, 27,
1
159–1160.
[15] R. Cimiraglia, T. Asano, H.-J. Hofmann, Gazz. Chim. Ital. 1996,
126, 679–684.
16] J. A. Riddick, W. B. Bunger, Organic Solvents, 3rd ed., Wiley-Inter-
science, New York, 1970.
17] Y. Doucet, G. Gimenez, R. Santini, A. TÞte, J. Chim. Phys.-Phys.
Chim. Biol. 1972, 69, 1760–1769.
18] H. Weingärtner, A. Knocks, W. Schrader, U. Kaatze, J. Phys. Chem.
A 2001, 105, 8646–8650.
19] K. Funke, J. Hermeling, J. Kꢂmpers, Z. Naturforsch. A 1988, 43,
[
[
[
[
[
Conclusion
1
094–1102.
20] R. Elsebrock, M. Stockhausen, G. Czechowski, J. Jadzyn, Pol. J.
Chem. 1998, 72, 2463–2470.
[21] A. Morabin, A. TÞte, Rev. Gen. Electr. 1968, 77, 961–971.
The present results unequivocally demonstrate the impor-
tance of electrostatic solute–solvent interactions in ionic liq-
uids. In a reaction where the polarity of the transition state
does not differ greatly from that of the reactant, ionic liq-
uids behave more or less in the same manner as the ordina-
ry molecular liquids. On the other hand, if a reactant in-
creases its polarity during the activation process, an exhaus-
tive rearrangement of the solvation environment, which in-
cludes fairly strict positioning and orientations of the cations
and anions around the charge centers of the activated com-
plex, has to take place so as to allow the polar transition
state to be stabilized, and this, in turn, results in the retarda-
tion of the reaction and a smaller frequency factor. In other
[
22] J. Halpern, G. W. Brady, C. A. Winkler, Can. J. Res. Sect. B 1950, 28,
40–155.
23] J. M. Nerbonne, R. G. Weiss, J. Am. Chem. Soc. 1978, 100, 5953–
954.
1
[
5
[24] The rate constants in various solvents show a small but clear tenden-
cy to decrease with increasing polarity of the solvent. This solvent
dependence was attributed to a polarity decrease in the rehybridiza-
tion of one of the nitrogen atoms during the activation step. See, P.
Haberfield, P. M. Block, M. S. Lux, J. Am. Chem. Soc. 1975, 97,
5804–5806.
[25] H. A. Kramers, Physica 1940, 7, 284–304.
26] R. F. Grote, J. T. Hynes, J. Chem. Phys. 1980, 73, 2715–2732.
27] J. T. Hynes, J. Stat. Phys. 1986, 42, 149–168.
28] N. Agmon, J. J. Hopfield, J. Chem. Phys. 1983, 78, 6947–6959.
[29] N. Agmon, J. J. Hopfield, J. Chem. Phys. 1983, 79, 2042–2053.
[30] H. Sumi, R. A. Marcus, J. Chem. Phys. 1986, 84, 4894–4914.
31] H. Sumi, J. Phys. Chem. 1991, 95, 3334–3350.
32] A. M. Berezhkovskii, V. Yu. Zitserman, Chem. Phys. Lett. 1989, 158,
[
[
[
words, a large charge density on ÀNO in the transition state
2
induces the cationic parts of the ionic liquid to restructure
the hydrogen-bonding network that had been partly de-
stroyed by the dissolution of the azobenzenes. The impor-
tance of the hydrogen-bonding network for understanding
the physical properties of ionic liquids was previously point-
[
[
3
69–374.
[
33] A. M. Berezhkovskii, Yu. A. Dꢁyakov, V. Yu, Zitserman, J. Chem.
Phys. 1998, 109, 4182–4189.
[34] A. M. Berezhkovskii, V. Yu. Zitserman, D.-Y. Yang, J. Kuo, S. H.
[39]
ed out by Dupont in his recent review.
Lin, Phys. A 1998, 251, 399–429.
[
[
[
[
[
35] M. V. Basilevsky, V. M. Ryaboy, N. N. Weinberg, J. Phys. Chem.
990, 94, 8734–8740.
36] M. V. Basilevsky, V. M. Ryaboy, N. N. Weinberg, J. Phys. Chem.
991, 95, 5533–5540.
37] T. Asano, H. Furuta, H. Sumi, J. Am. Chem. Soc. 1994, 116, 5545–
550.
38] T. Asano, K. Cosstick, H. Furuta, K. Matsuo, H. Sumi, Bull. Chem.
Soc. Jpn. 1996, 69, 551–560.
39] J. Dupont, J. Braz. Chem. Soc. 2004, 15, 341–350.
1
[
[
1] S. V. Dzyuba, R. A. Bartsch, ChemPhysChem 2002, 3, 161–166.
2] P. Bonhôte, A.-P. Dias, N. Papageorgiou, K. Kalyanasundaram, M.
Grätzel, Inorg. Chem. 1996, 35, 1168–1178.
1
[
[
3] T. Welton, Chem. Rev. 1999, 99, 2071–2084.
4] P. J. Dyson, M. C. Grossel, N. Srinivasan, T. Vine, T. Welton, D. J.
Williams, A. J. P. White, T. Zigras, J. Chem. Soc. Dalton Trans. 1997,
5
3
465–3469.
[
[
[
5] A. S. Larsen, J. D. Holbrey, F. S. Tham, C. A. Reed, J. Am. Chem.
Soc. 2000, 122, 7264–7272.
6] P. Wasserscheid, W. Keim, Angew. Chem. 2000, 112, 3926–3945;
Angew. Chem. Int. Ed. 2000, 39, 3772–3789.
7] P. D. Wildes, J. G. Pacifici, G. Irick, Jr., D. G. Whitten, J. Am. Chem.
Soc. 1971, 93, 2004–2008.
Received: January 19, 2006
Published online: April 19, 2006
Chem. Eur. J. 2006, 12, 5328 – 5333
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5333