Complexation characteristics of permethylated cycloinulohexaose, cycloinuloheptaose, and cycloinulooctaose with metal cations

Article abstract of DOI:10.1039/b102710g

The host-guest complexation behavior of cyclic oligosaccharides, permethylated cycloinulohexaose 1, permethylated cycloinuloheptaose 2, and permethylated cycloinulooctaose 3 with metal cations has been characterized by means of UV-visible, NMR, and electrospray ionization (ESI) mass spectrometry. In the crystal state, the structures of 1·K⁺, 1·Rb⁺, and 1·Cs⁺ were same as that of the 1·Ba²⁺ complex which has a u-u-d-u-u-d (u = up, d = down) furanose ring arrangement for the plane of the crown ring moiety. The association constants (K_s) in THF and in [²H₆]acetone at 298 K were evaluated. The binding ability of host 1 with metal cations was of the same degree as that of calix[6]arene derivative 4 and much higher than those of hosts 2 and 3. The thermodynamic parameters of the complexation of host 1 with metal cation in THF were determined, and it was suggested that the entropy change for the solvation of the metal cations was one of the important factors in the complexation equilibrium. It was clarified that the structure of the host 1·K⁺ complex in solution at low temperature (furanose ring arrangement: u-d-u-d-u-d) was different from that in the crystal state (u-u-d-u-u-d arrangement) by the coalescence behavior in ¹H-NMR. The relative peak intensity of the complex ions of host 1 or 2 with two alkali metal ions in ESI mass spectrometry (in acetone) showed a correlation in the first order approximation with the ratio of the corresponding complex ion concentrations estimated from the K_s values.

Full text of DOI:10.1039/b102710g

View Article Online / Journal Homepage / Table of Contents for this issue
Complexation characteristics of permethylated cycloinulohexaose,
cycloinuloheptaose, and cycloinulooctaose with metal cations†
Motohiro Shizuma,*a Yoshio Takai,b Mishio Kawamura,c Tokuji Takedaa and
Masami Sawada*b
2
a The Technochemistry Department, Osaka Municipal Technical Research Institute, Joto-ku,
Osaka 536-8553, Japan
b Materials Analysis Center, The Institute of Scientific and Industrial Research,
Osaka University, Ibaraki, Osaka 567-0047, Japan
c Department of Biology, Osaka Kyoiku University, Kashiwara, Osaka 583-8582, Japan
Received (in Cambridge, UK) 23rd March 2001, Accepted 4th June 2001
First published as an Advance Article on the web 29th June 2001
The host–guest complexation behavior of cyclic oligosaccharides, permethylated cycloinulohexaose 1, permethylated
cycloinuloheptaose 2, and permethylated cycloinulooctaose 3 with metal cations has been characterized by means of
UV-visible, NMR, and electrospray ionization (ESI) mass spectrometry. In the crystal state, the structures of 1ؒKϩ,
1ؒRbϩ, and 1ؒCsϩ were same as that of the 1ؒBa2ϩ complex which has a u-u-d-u-u-d (u = up, d = down) furanose ring
arrangement for the plane of the crown ring moiety. The association constants (KS) in THF and in [2H6]acetone at
298 K were evaluated. The binding ability of host 1 with metal cations was of the same degree as that of calix[6]arene
derivative 4 and much higher than those of hosts 2 and 3. The thermodynamic parameters of the complexation of
host 1 with metal cation in THF were determined, and it was suggested that the entropy change for the solvation of
the metal cations was one of the important factors in the complexation equilibrium. It was clarified that the structure
of the host 1ؒKϩ complex in solution at low temperature (furanose ring arrangement: u-d-u-d-u-d) was different
from that in the crystal state (u-u-d-u-u-d arrangement) by the coalescence behavior in 1H-NMR. The relative peak
intensity of the complex ions of host 1 or 2 with two alkali metal ions in ESI mass spectrometry (in acetone) showed
a correlation in the first order approximation with the ratio of the corresponding complex ion concentrations
estimated from the KS values.
ation behavior of permethylated cycloinulohexaose 1, per-
methylated cycloinuloheptaose 2, and permethylated cyclo-
inulooctaose 3 with metal cations.
Introduction
It is well-known that some kinds of peptides and natural iono-
phores such as monensin, valinomycin, etc., play important
roles as ion carriers and as ion channels in the typical metal ion-
transport system which is one of the fundamental mechanisms
for accumulation of energy, and the medium by which muscle is
controlled, and information passed on in living systems.1–4 On
the other hand, various hosts such as crown ethers,5 cryptands,6
spherands,7 cavitands,8 calixarenes,9 etc. have been synthesized
and widely studied as artificial ion carriers. In particular,
calix[4]arene derivatives have been reported to have high selec-
tivity toward alkali metal ions (Naϩ–Kϩ = ca. 400 at 298 K in
MeOH).10
Cyclofructans are one of the cyclic oligosaccharides and are
structural isomers of cyclodextrins.11 Cyclodextrins are cyclic
oligomers composed of α(1→4) linked -glucopyranose units,
while cyclofructans consist of β(2→1) linked -fructofuranose
units. Therefore, their structure and properties are quite dif-
ferent. Cyclofructans are expected to be a new carbohydrate
type of ionophore because of the existence of the crown ether
ring skeleton in the center of the molecules.12 However, there
have been few reports on the complexation of cyclofructans
and their derivatives with metal cations as yet,13–15 and further
development of this field has been expected.
Results
Solid 13C-NMR
The 13C-NMR spectra of host 1 in [2H]chloroform, host 1ؒKϩ
complex (counter anion: picrate) and host 1ؒBa2ϩ complex
(counter anion: SCNϪ) in the crystal state, at room temperature
are shown in Fig. 1. In the solution state, 6 peaks of the fructo-
furanose moiety (C1–C6) and 3 peaks of the methoxy groups
(3-OMe, 4-OMe and 6-OMe) were observed, respectively. In the
solid state, each 13C-NMR peak in both of host 1ؒKϩ complex
and host 1ؒBa2ϩ complex was observed as three peaks. The
signals in solid 13C-NMR spectra of host 1ؒRbϩ complex and
host 1ؒCsϩ complex were also observed as three split peaks,
respectively.
UV-visible spectrometry
Bathochromic shifts of absorption maxima (∆λ of λmax) of the
picrate anion induced by adding a host to a metal picrate
in THF at 298 K were observed (Fig. 2 and Table 1).16 The
shifts induced by adding calix[6]arene derivatives 4 are also
shown for comparison with permethylated cyclofructans.17 The
association constants (KS) of complexation between the host
and a metal cation guest were then determined from the
absorption changes at certain wavelengths at 298 K (Table 2).
The thermodynamic parameters were estimated from KS at
various temperatures.18 In the case of host 1 with Kϩ, Rbϩ, and
In this paper, we report a further insight into the complex-
† Electronic supplementary information (ESI) available: a plot of log
KS for 1/T in the complexation of host 1 with Kϩ (Fig. S1), a typical
ESI mass spectrum of the complex ions (Fig. S2), and a chromato-
graphic chart of cyclofructans (n = 6–8) (Fig. S3). See http://
www.rsc.org/suppdata/p2/b1/b102710g/
1306
J. Chem. Soc., Perkin Trans. 2, 2001, 1306–1314
This journal is © The Royal Society of Chemistry 2001
DOI: 10.1039/b102710g

View Article Online
Fig. 1 A liquid 13C-NMR spectrum of host 1 and solid 13C-NMR
spectra of host
1 complex with metal cations. (a) Host 1 in
[2H]chloroform, (b) crystalline host 1ؒKϩ complex (counter anion:
picrate), (c) host 1ؒBa2ϩ complex (counter anion: SCNϪ).
Fig. 2 UV-visible spectra of a 5.0 × 10Ϫ5 M THF solution of
potassium picrate in the presence of different amounts of 1 at 298 K.
Molar ratio of the host to metal picrate guest ([H]/[G]): (a) 0, (b) 0.3, (c)
0.7, (d) 1.4, (e) 2.7, (f ) 7.0. Addition of variable amounts of 1 caused a
bathochromic shift of λmax from 358 to 382 nm (∆λ = 24 nm).
for the other metal cations, only small shifts were observed. In
the case of host 3, no shifts were observed. The KS ordering of
host 1 with the metal picrates was Kϩ > Rbϩ > Csϩ > Naϩ. The
KS value of host 2 with Rbϩ was about 102 times smaller than
that of host 1. For the thermodynamic parameters, the ordering
of the enthalpy (∆H) and the entropy (∆S) of host 1 was
Kϩ > Rbϩ > Csϩ (Table 2).
Table 1 Bathochromic shifts (∆λ of λmax, nm) of the picrate anion
probe by adding permethylated cyclofructans 1, 2 and 3 and calixarene
derivative 4 in THF at 298 K
Guest cation
Host
1
2
3
4
Naϩ
Kϩ
∼5a
24
22
1
0
0
0
0
6 (6b)
∼1
11
2
2
0
24 (25b)
21
Rbϩ
Csϩ
18
20 (20b)
ϩ
NH4
∼7a
Ba2ϩ
∼3a
a Isosbestic correlation disappeared, so we could not detect the exact
∆λ. b Ref. 17 (at 303 K).
Csϩ, large bathochromic shifts were observed. In the case of
host 1 with Naϩ, a larger shift was observed by adding excess
host 1, but the isosbestic point disappeared. In the case of host
2 with Rbϩ, a clear bathochromic shift was observed. However,
J. Chem. Soc., Perkin Trans. 2, 2001, 1306–1314
1307

View Article Online
Table 2 LogKS and thermodynamic parameters for the complexation of host 1, 2, 4 and 18-crown-6 (18C6) with various metal cations (counter
anion: picrate) in THF at 298 K.
Host
Cation
LogKS
∆GЊ/kcal molϪ1
∆HЊ/kcal molϪ1
T ∆SЊ/kcal molϪ1
1
Naϩ
Kϩ
3.25 0.03
4.34 0.04
4.23 0.03
4.03 0.04
2.71 0.01
4.20 0.01a
5.12 0.04b
1
Ϫ6.0
Ϫ5.8
Ϫ5.6
Ϫ4.6
Ϫ1.8
Ϫ3.4
Ϫ9.1
Ϫ4.2
4.2
2.4
Ϫ3.3
Ϫ0.3
1
Rbϩ
Csϩ
Rbϩ
Kϩ
1
2
4
18C6
Rbϩ
a LogKS = 4.13 at 303 K (ref. 17). b At 303 K (∆λmax = 8 nm).
Table 3 LogKS and 1H-NMR induced shifts of permethylated cyclofructans (1–3) with metal cations (counter anion: SCNϪ) in [2H6]acetone
Low-field induced shift (ppm)
Host
Guest
LogKS
H-1
H-1Ј
H-3
H-4
H-6
H-6Ј
3-OMe
4-OMe
6-OMe
1
1
1
1
1
1
1
2
2
2
2
2
3
3
3
3
3
Liϩ
1.4 0.1
2.2 0.1
3.8 0.0
3.6 0.0
2.9 0.0
0.36 0.02b
4.3
(6)a
(3)
(2)
(3)
(4)
(3)
(1)
(4)
(9)
(6)
(8)
(3)
(4)
(3)
(3)
(2)
(3)
Ϫ0.21
Ϫ0.09
Ϫ0.23
0.02
0.06
Ϫ0.01
0.04
0.02
0.05
0.17
0.18
0.02
0.12
0.10
0.16
0.14
0.08
0.08
0.35
0.07
0.10
0.09
0.03
0.07
0.16
0.19
0.05
0.01
0.28
0.09
0.08
0.10
0.10
0.06
0.07
0.16
0.04
0.06
0.05
0.02
0.05
0.12
0.14
0.06
0.02
0.18
0.05
0.04
0.05
0.06
0.03
0.04
0.09
0.03
0.04
0.03
0.01
0.06
0.14
0.19
0.11
0.01
0.18
Naϩ
Kϩ
Rbϩ
Csϩ
Ca2ϩ
Ba2ϩ
Naϩ
Kϩ
Ϫ0.09
Ϫ0.06
Ϫ0.05
0.23
Ϫ0.01
0.07
0.21
0.36
1.1 0.1c
1.5 0.1
2.0 0.1
1.9 0.1
0.71 0.06
0.94 0.06
0.65 0.07
1.0 0.1
1.4 0.1
1.1 0.0
Ϫ0.03
Ϫ0.04
0.11
0.07
0.05
0.07
0.07
0.03
0.07
0.03
Rbϩ
Csϩ
Ba2ϩ
Naϩ
Kϩ
0.04
Ϫ0.05
Rbϩ
Csϩ
Ba2ϩ
a Values in ( ) are numbers of monitored peaks. b We can not remove completely crystal water by heating under vacuum. c The association constant of
the 1 : 2 complex was estimated (KS 1:2 = 0.1 0.03 MϪ1).
including the values reported before.14b The ordering of KS of
host 1 with metal cations was Ba2ϩ > Kϩ > Rbϩ > Csϩ > Naϩ >
Liϩ > Ca2ϩ. In the case of host 2, the ordering was Rbϩ >
Kϩ > Csϩ > Naϩ > Ba2ϩ.Inhost3,theorderingwasCsϩ > Ba2ϩ
>
Rbϩ > Kϩ > Naϩ. In the case of hosts 1 and 2, the downfield
shifts of the three singlet signals of the OMe groups were
observed, and the ordering of the downfield shifts was 3-
OMe > 4-OMe > 6-OMe. For the ring protons, one of the two
H-1 protons showed large upfield shifts. The H-4 protons
showed large downfield shifts. For host 3, no specific cation-
induced shifts were observed.
Temperature-dependent 1H-NMR
1H-NMR spectra of the free host 1 or the complexes of the host
1 with metal cations (counter anion: SCNϪ) were measured at
various low temperatures in order to observe the coalescence of
the proton peaks (Table 4). The mixture of solvents [2H]chloro-
form and [2H2]dichloromethane was used because of the ease
of treatment at low temperatures. The coalescence behavior of
the proton peaks of host 1 was observed in the temperature
range of 185 to 205 K, and the peaks split into two (Fig. 4).
The number of splitting peaks suggests the symmetry of the
molecule or the peaks, etc., as reported before.14b In the case of
the complexes of host 1 with Naϩ, Kϩ, and Csϩ, the signals split
into two. In the complex of the host 1 with Ba2ϩ, the peak was
separated into three peaks. The activation free energy (∆GC≠) of
the conformational interconversion was evaluated from the
splitting width (∆ν) and the coalescence temperature (T C).19
Fig. 3 1H-NMR (360 MHz) spectral changes of 2 with added Rbϩ
(SCNϪ) in [2H6]acetone (298 K) ([H]0 = 0.726 mM). Ratio of
concentration [H]/[G]: (1) 0, (2) 0.36, (3) 1.09, (4) 2.54, (5) 5.10, (6) 8.75,
(7) 14.2, (8) 21.5, (9) 28.8, (10) 36.0.
NMR titration
The induced shifts of the 1H-NMR peaks of the hosts in
[2H6]acetone at 298 K on adding metal cations (counter anion:
SCNϪ) were observed, and the KS values were determined
from the shifts (Fig. 3). The values are summarized in Table 3
≠
The ordering of the ∆GC was free host 1 < host 1ؒNaϩ < host
1ؒKϩ < host 1ؒBa2ϩ
.
1308
J. Chem. Soc., Perkin Trans. 2, 2001, 1306–1314

View Article Online
Table 4 Coalescence behavior and expected structure of host 1 and the complexes with metal cations (counter anion: SCNϪ) in temperature-
dependent 1H-NMR (360 MHz)a
Guest
Number of separated peaks
Molecular symmetry
Furanose ring arrangement
T C/K
∆GC≠/kcal molϪ1
—
2
2
2
2
3
C3
C3
C3
C3
C2
u-d-u-d-u-d
u-d-u-d-u-d
u-d-u-d-u-d
u-d-u-d-u-d
u-u-d-u-u-d
185–205
205–220
230–240
9.2
10.4
12.0
Naϩ
Kϩ
Csϩ
Ba2ϩ
285–295
13.2
a Solvent: [2H]chloroform–[2H2]dichloromethane (v/v = 1 : 3). The concentration of complexes was estimated at about 1.63 mM. Activation free
energy (∆GC≠) of the conformational interconversion was estimated by the modified Erying equation depending on coalescence temperature (T C) and
separating width (∆ν).
Fig. 5 Conformation of 18-crown-6 ring moiety in crystalline host
1ؒBa2ϩ complex.
Table 5 Relative peak intensity of the complex ions of permethylated
cyclofructans (1, 2) with metal cations (counter anion: SCNϪ) in elec-
trospray ionization (ESI) mass spectrometry
Host
Metal cation 1
Metal cation 2
I(hostؒM)ϩ/I(hostؒK)ϩa
1
1
1
2
2
2
Naϩ
Rbϩ
Csϩ
Naϩ
Rbϩ
Csϩ
Kϩ
Kϩ
Kϩ
Kϩ
Kϩ
Kϩ
14
66
26
81
122
52
Fig. 4 Temperature-dependent 1H-NMR (360 MHz) spectra of host 1
with Kϩ (SCNϪ) in [2H]chloroform–[2H2]dichloromethane (v/v = 1 : 3).
The concentration of complex was estimated at about 1.6 mM.
ESI mass spectrometry
ESI mass spectra of a three-component mixture in an acetone
solution including the host and two kinds of metal cations
(counter anion: SCNϪ) were measured. The relative peak inten-
sity [I(hostؒM)ϩ/I(hostؒK)ϩ] of the complex ions of the host
with alkali metal cation (Mϩ) for the complex ion with Kϩ is
shown in Table 5. For comparison between the binding ability
of host 1 and that of host 2, ESI mass spectra of a three-
component mixture containing two hosts and Kϩ were also
measured. In the case of host 1, ordering of the relative peak
intensity was Kϩ > Rbϩ > Csϩ > Naϩ. For host 2, the ordering
was Rbϩ > Kϩ > Csϩ > Naϩ. Compared with complex ions
with Kϩ, the peak intensity of the complex ions of host 1 was
larger than that of host 2.
a All values were normalized as I(hostؒK)ϩ = 100 except b. b The value
was normalized as I(host 1ؒK)ϩ = 100.
Guest
Kϩ
Host 1
Host 2
I(host 2ؒK)ϩ/I(host 1ؒK)ϩ
1
2
11b
furanose moieties in six furanose units. This fact accounts for
the results of solid 13C-NMR. The solid 13C-NMR spectra of
host 1ؒKϩ, host 1ؒRbϩ, host 1ؒCsϩ complexes are similar to that
of host 1ؒBa2ϩ complex. Thus, the solid NMR spectra suggest
that these complexes have a C2 symmetry axis in the crystal
state. Although in the case of the crystalline complexes of host
1 with Kϩ, Rbϩ and Csϩ (counter anion: picrate), a few parts of
the methoxy groups on the edge and others have not been
solved yet, it was found that the conformation of the 18-crown-
6 ring skeleton was the twist-boat type (gϩgϩgϩgϩgϩgϩ) and
the four fructofuranose rings were located on the ring plane
(u-u-d-u-u-d arrangement).21 In the crystalline complexes of
host 1 with a series of metal cations (Kϩ, Rbϩ, Csϩ and Ba2ϩ),
host 1 has a regular conformation (see later).
Discussion
Structure of crystalline complexes
The crystalline host 1ؒBa2ϩ complex (counter anion: SCNϪ)
had been already analyzed by means of X-ray crystal-
lography.14b In the complex, the barium ion is not located in
the center of the 18-crown-6 ring skeleton but in the binding
site consisting of the six oxygens of the crown ether ring
skeleton and the four oxygens of the 3-OMe groups. Four
fructofuranose rings which bind to the cation were above the
crown ring plane, and two furanose rings were below the ring
plane. The furanose ring arrangement is u-u-d-u-u-d (u = up,
d = down) for the crown ring plane. The conformation of the
crown ring skeleton was the twist boat type, and the arrange-
ment of the –O–C–C–O– torsion angles was gϩgϩgϩgϩgϩgϩ.20
The complex has a C2 symmetry axis (Fig. 5). Therefore, the
crystalline complex has three stereochemically different fructo-
Distance between metal cation and picrate
With the change in distance between the oxygen and potassium
ions in the potassium picrate model, whose structure was opti-
mized by the PM3 semiempirical molecular orbital method,22
the UV-visible spectra of the modified potassium picrate were
simulated by means of the ZINDO calculation method.23 The
difference (∆λ) between the wavelength (λmax0) of the absorption
maxima in the initial optimized structure and the wavelength
J. Chem. Soc., Perkin Trans. 2, 2001, 1306–1314
1309

View Article Online
Fig. 6 Bathochromic shifts (BS, ∆λ) of λmax of picrate anion vs. K–O
distance (r). The line is the fitting curve of BS = Ϫk(1/r2) (where k is
a constant). The bathochromic shifts were calculated by ZINDO
simulation.
λmax was plotted against the Kϩ–O distance (Fig. 6). The longer
the Kϩ–O distance, the larger the bathochromic shift. It is
deduced from the plot curve that the Kϩ–O distance is ca. 7 Å
because the experimental bathochromic shift is 24 nm. There-
fore, it is suggested that the potassium picrate becomes a
completely separated ion pair by the complexation with host 1
in THF.
Fig. 7 LogKS vs. cation radius plots for the complexation of (a)
permethylated cyclofructans (in [2H6]acetone) or (b) benzo crown ethers
(in chloroform) with alkali metal cations. B18C6 = benzo-18-crown-6,
B21C7 = benzo-21-crown-7, B24C8 = benzo-24-crown-8.
Binding ability with metal cations
The binding ability of host 1 was very high in comparison with
hosts 2 and 3. The binding ability of host 1 was approximately
as high as that of host 4 in THF. The logKS value in THF at
298 K is linearly correlated with logKS in [2H6]acetone at 298 K.
It is then suggested that difference between the counter anions
(SCNϪ or picrate) is negligible. The alkali metal cation-
selectivity of host 1 (Kϩ > Rbϩ > Csϩ > Naϩ) is different from
the selectivity of host 4 (Csϩ > Kϩ > Naϩ), and the difference is
assumed to be the difference in flexibility of the binding cavities
in the hosts. The binding cavity of host 1 consists of the 18-
crown-6 ring skeleton and the 3-OMe groups of furanose rings
spiro-anellated on the crown ring skeleton. On the other hand,
the cavity of host 4 consists of phenol oxygens and carbonyl
oxygens of the ethoxycarbonylmethyl moiety. The former
would be more rigid than the latter. Host 1 then showed a clear
selectivity for metal cation size. On the other hand, host 4 can
change the cavity size flexibly according to the metal cation size,
so that it showed higher binding ability to all alkali metal
cations.
the solvent around the metal cations would be completely torn
off in the complexation with the cryptands. Therefore, in this
case, the enhancement of entropy due to desolvation becomes
large. This increase in the entropy negates the reduction in the
entropy due to the association. The X-ray crystal structure of
the complexes of host 1 with metal cations clearly shows that
the cations are completely encapsulated within the cavity of the
host. Therefore, it appears reasonable that the complexation of
host 1 shows the entropy change (∆S) as that of the cryptands.
The ordering of the solvation affinity was Kϩ > Rbϩ > Csϩ.26
The ordering of the obtained ϪT ∆S was in good agreement
with the solvent affinity. In the complexation with Kϩ and Csϩ,
from a large negative ∆H, the distortion of host 1 seems to be
larger in the former case probably because Kϩ is smaller than
Csϩ. Host 1 seems to form a complex with Csϩ with less stress.
Relationship between bathochromic shift and logKS
For host 1 and 18-crown-6, logKS is plotted against the
bathochromic shifts (Fig. 8). Host 1 with Naϩ, Kϩ, Rbϩ, and
Csϩ shows a linear relation. Linearity is observed for 18-crown-
6 with Kϩ, Rbϩ, and Csϩ, but not with Naϩ.27 The magnitude of
the bathochromic shifts depends on the average distance
between the metal cation and the picrate anion. In the com-
plexes of 18-crown-6 with Kϩ, Rbϩ, and Csϩ, the arrangements
of –O–C–C–O– torsion angles in the crown rings are repre-
sented by gϩgϪgϩgϪgϩgϪ.28 The arrangement is gϩgϪgϩgϩgϩgϪ
in the complex with Naϩ.29 The complexes of flexible host 4
are assumed to be different structures depending on the metal
cation species. Shinkai et al. reported that there is no linearity
between the bathochromic shifts and logKS in the case of host 4
with alkali metal ions.17
Correlation between metal cation selectivity and cavity size of
hosts
From the correlation between the alkali metal ion radius and
the KS values, the cations having the highest KS values of host 1,
host 2, and host 3 among the alkali metal series (in [2H6]-
acetone) have the same selectivity as the corresponding benzo
crown ether series (in chloroform) (Fig. 7).24 Characteristically,
the larger the ring size of the permethylated cyclofructan series,
the lower the binding ability. It is suggested that the binding
sites of the hosts do not consist of only the crown ether ring
skeleton but also the oxygens of the furanose rings in solution.
Thermodynamics
The contribution of the entropy factor (∆S) to the free energy
(∆G) of the complexation equilibrium of host 1 with metal
cations was large. In the case of host 1 with Kϩ and Rbϩ,
∆H < 0, ∆S > 0, as in the case of cryptand[2.2.1] with alkali or
alkaline earth ions such as Naϩ, Kϩ and Mg2ϩ in water.25 In
general, the entropy is reduced in an intermolecular association
system. However, because cryptands are cage-like structures,
Binding points between hosts and guests in solution
1
Induced shifts in the H-NMR peaks of host 1 by adding the
present metal cations showed the following features. In the two
types of H1 and H1Ј in the crown ring skeleton, high-field shift
and low-field shift were observed. Among the three types of the
OMe protons, the largest shifts were observed on the 3-OMe
1310
J. Chem. Soc., Perkin Trans. 2, 2001, 1306–1314

View Article Online
Fig. 8 LogKS vs. bathochromic shift plots for the complexion of host
1 (᭺) and 18-crown-6 (18C6) (᭹). Data for the host 1 series are
determined in tetrahydrofuran. Data for 18C6ؒNaϩ, 18C6ؒKϩ, and
18C6ؒCsϩ are from ref. 24.
Fig. 9 Comparison of complex ion peak intensities in ESIMS and
concentration ratio of complex ions calculated from KS values in
[2H6]acetone at 298 K.
protons, a larger low-field shift on H-4, which is located cis to 3-
OMe, and the oxygens of the 3-OMe groups are then believed
to be coordinated to metal cations. These features were also
observed in host 2. In solution, host 1 and host 2 would form
complexes with metal cations at binding points including the
oxygens in the crown ring skeleton and the oxygens of the
3-OMe groups in the fructofuranose rings.
from the u-furanose rings and the d-furanose rings, respectively.
Therefore, a possible complex structure in solution at low tem-
perature may be expected to have C3 symmetry and u-d-u-d-u-d
furanose rings arrangement as in the crystalline structure of
cycloinulohexaose (Table 4).12 The structure of the host 1ؒBa2ϩ
complex, which shows the highest binding ability in this series,
may be expected to have C2 symmetry (furanose ring arrange-
ment: u-u-d-u-u-d) in good agreement with the result of the
X-ray crystal structure analysis.
Fructofuranose ring arrangement of complexes in solution at low
temperature
The activation free energy change (∆GC≠) estimated from
coalescence behavior of the signals in temperature-dependent
1H-NMR is assumed to involve an intramolecular conform-
ation change in host 1 or its complexes as shown in Scheme 1. If
In our previous report, the charge induced shifts of host 1
were estimated on the basis of induced-shift differences in
1
the H-NMR spectra of the Kϩ and Ba2ϩ complexes at room
temperature.14b As indicated by Immel and Lichtenthaler,30 the
estimation of the charge-induced shift is valid only if both
complexes exhibit identical or similar conformations in a strict
sense. From the obtained linear relationship between batho-
chromic shift and logKS, this seems to be true at room tem-
perature for a qualitative level. Considering the symmetry of
the host–metal complexes in solution at low temperatures, the
evaluation might not be quantitatively suitable for this system.
Relative peak intensity in ESI mass spectrometry
The relative peak intensity of the complex ions of host 1 and
host 2 with a given alkali metal ion observed in ESI mass
spectra is plotted against the ratio of the complex concen-
tration estimated from KS in [2H6]acetone at 25 ЊC. The relative
intensity is in agreement with the ratio of the concentration,
especially in host 1 (Fig. 9). So far, it has been reported that the
relative peak intensity in ESI mass spectrometry reflects the
ratio of the complex concentration, in the case of crown ethers
and cryptands with metal cations,31 and metal cation selectivity
of various artificial ionophores studied in detail using ESI
mass spectrometry under controlled conditions.32 The peak
intensity in ESI mass spectra depends on various factors such
as charges of ions, solvents, counter ions, and ionization con-
ditions (accelerating cone voltage, capillary temperature, etc.).
Although it is also reported that, in the case of modified crown
ethers with alkylammonium ions, the relative intensity differs
from the concentration ratio in solution.33 Generally, the trans-
ferability of the metal ions to the gas phase from the liquid
phase largely depends on the kind of cations because of the
differences in the solvation effect.34 Herein, acetone and thio-
cyanate were also used in order to estimated KS values in solu-
tion as solvent and counter anion, respectively. However, under
our conditions, we found good agreement between the relative
intensity and the concentration ratio in solution, suggesting
Scheme 1
the binding ability between a host and a guest is high, it is
expected that ∆GC is larger because of controlled conform-
ation inversion. Indeed, the ordering of ∆GC agrees well with
≠
≠
that of KS. The two peaks (integration 1 : 1) of complexes of
host 1 with Naϩ, Kϩ and Csϩ at low temperature would result
J. Chem. Soc., Perkin Trans. 2, 2001, 1306–1314
1311

View Article Online
that there are few transferability differences of the present
complex ion. For complex ions consisting of metal cations with
a host, which is similar to solvent molecules in solvation, the
transferability factors may be leveled off to the same extent as
the free metal cations.
p-tert-Butylcalix[6]arene ester 4 was prepared by adding ethyl
bromoacetate to a slurry of a commercial p-tert-calix[6]arene
(Tokyo Kasei. Co.) and anhydrous potassium carbonate in
acetone at room temperature38 and identified by mp, 1H-NMR
and FT-IR. Commercial 18-crown-6 was used without purifica-
tion (Aldrich).
Metal picrates were prepared by a neutralization reaction of
picric acid and metal hydroxide in methanol and recrystallized
in water.39 The products were dried in vacuo for 5–10 h at 100 ЊC
and used in UV-visible titration experiments. The metal picrates
were identified by elemental analysis.
Metal thiocyanates were used without purification of com-
mercial reagents which were dried under vacuum at 100 ЊC
overnight (Na and Ca salts, Wako Co.; Li, K, Cs, and Ba salts,
Kishida Chemical Co.). RbSCN was prepared by anion
exchange through a liquid chromatographic column (anion
exchange resin, Dowex AG1-X2 SCN form; water) from metal
chloride (Wako Co.). Their formation was identified by atomic
absorption analysis (Rb and Na) and ion chromatography (ClϪ
and SCNϪ). The products were dried in vacuo at 80 ЊC before
the experiment.
Summary
The binding ability of host 1 with metal cations is relatively
large compared to that of host 2 and host 3 in [2H6]acetone.
Unlike host 4, which shows approximately the same binding
ability in THF, host 1 was suggested to complex rather rigidly
with a series of alkali metal cations in solution. In the case of
host 1ؒKϩ, host 1ؒRbϩ, and host 1ؒCsϩ complexes, the liquid
NMR spectrometrically deduced structure (furanose ring
arrangement: u-d-u-d-u-d) at low temperature is clearly dif-
ferent from the crystalline structure (u-u-d-u-u-d) by the solid
NMR.35
Properties and structures of cyclofructan complexations
have been studied and deduced by means of various spectral
experiments and molecular simulations. Recently, the chiral
discrimination ability of the permethylated cyclofructans
toward amino acid derivatives has appeared.36 Further study of
these cyclic oligosaccharides will be continued to develop them
as a new, interesting class of ionophore.
General
Solid 13C-NMR (70 MHz) spectra were taken with a JEOL EX-
270 spectrometer. UV-visible spectra were taken with a Hitachi
U-3410 UV spectrometer. 1H-NMR (360 MHz) and 13C-NMR
(90 MHz) spectra were taken with a Bruker AM400 spec-
trometer. TMS was used as the internal standard. ESI mass
spectra were measured by using a Finnigan TSQ7000.
Experimental
Materials
Liquid column chromatography for the isolation and the
purification of cyclofructans was carried out on a Yamazen LC
apparatus with an RI detector under appropriate medium
pressure. Gel-permeation liquid chromatography was per-
formed using a LC-09 (Nihon Bunseki Kougyou Co.) with an
RI detector. The compounds used were identified with the
following methods. Melting points were measured with a
Permethylated cyclofructans (1, 2, and 3). Free cyclofructans
(hydroxy type) were isolated from a cyclic oligomer mixture
(n = 6–8) obtained by enzyme reaction (Mitsubishi Kagaku
Co.) using chromatography with an eluent of water through a
column (size, ∅ 25 × 350 mm (Ca) and ∅ 25 × 550 mm (K))
filled with cation exchange resin (Dowex AG-50W) calcium
and potassium forms under medium pressure (ca. 3.5 kgf cmϪ2
,
1
flow 1.0 mL minϪ1). The ordering of the retention time was
octamer, heptamer and hexamer and they were obtained in the
amounts of 7.8, 18.9 and 65.5 mg, respectively, from 100 mg of
the mixture. The isolated cyclofructans were permethylated by
the Hakomori method.37 The products were identified by mp,
FABMS, NMR, FT-IR, and elemental analyses.
Yanaco micro melting point apparatus. H-NMR (270 MHz)
spectra were taken with a JEOL EX-270 spectrometer. FT-IR
spectra were taken with an Analect RFX-65 or a Shimadzu
FT-IR 8100 spectrometer. FAB mass spectra were taken with
a JEOL JMS-600 (NBA matrix). Elemental analyses were
performed using a Perkin-Elmer 2400. Ion chromatographic
analyses were performed using a Dionex QIC analyzer. Atomic
absorption analysis data were taken with a Dainiseikosha SAS/
727.
Permethylated cycloinulooctaose (3). Under nitrogen gas flow,
a dry DMSO solution of dry crude cycloinulooctaose 270 mg
(0.21 mmol) was added to a stirred DMSO solution of dimethyl
sulfinyl carbanion (ca. 12 mmol), which was prepared from
NaH and dry DMSO, and was stirred for 4 h at room tem-
perature. The reaction mixture was cooled in an ice bath,
methyl iodide 2.7 mL (43.3 mmol) was added dropwise at ca.
10 ЊC, and stirred at room temperature overnight. To the reac-
tion mixture was added dichloromethane, the organic layer
was washed with aq. Na2S2O3 and water, and the organic layer
was dried over anhydrous MgSO4 overnight. After filtration,
the solution was evaporated in vacuo to give a residue that was
purified by liquid column chromatography [filler, silica gel
60, Merck; eluted solvent, methanol–benzene = 1 : 8 (v/v)],
and recrystallized from dichloromethane–petroleum ether to
give 168 mg. The obtained white powder 110 mg was purified by
preparative gel-permeation liquid chromatography (eluted
solvent, chloroform) and recrystallized from dichloromethane–
petroleum ether to yield 3 (50 mg, 15 %): colorless powder, mp
128–133 ЊC; 1H-NMR (270 MHz, CDCl3) δ (ppm) 4.08 (d, 8H,
3J3,4 = 7.1 Hz, H-3), 3.83 (m, 8H, H-5), 3.71 (d, 8H, 2J1,1Ј = 10.0
Hz, H-1), 3.71 (dd, 8H, 3J3,4 = 6.9 Hz, H-4), 3.56 (m, 24H, H-1Ј,
H-6, H-6Ј), 3.48 (s, 24H, -CH3), 3.43 (s, 24H, -CH3), 3.39 (s,
24H, -CH3) from TMS; FT-IR (KBr-disk) 1121, 1101, 1073
(ether CO) cmϪ1; FABMS (m/z) 1671 (M ϩ K)ϩ; Anal. Calcd.
for C72H108O40: C, 52.93; H, 7.90. Found: C, 52.96; H, 7.91.
Solid 13C-NMR
The crystalline structures of 1ؒKϩ (picrate) complex, 1ؒRbϩ
(picrate) complex, 1ؒCsϩ (picrate) complex, and the 1ؒBa2ϩ
(SCNϪ)2 complex were analyzed using the solid 13C-NMR as
follows. An aqueous solution of 1 with an excess amount of
metal cation was allowed to evaporate slowly. The obtained
single crystals were powdered by a mortar and about 100 mg
of the powder was used for the solid 13C-NMR measurement.
The single crystals were not suitable for X-ray crystallography
because of their instability in air. A cross-polarization magic
angle spinning (CPMAS) pulse sequence was applied. The spin
ratio of the samples was about 6000 kHz. Adamantane was
used as an indirect reference (chemical shift of the methyne
carbon peaks: δ = 37.8 ppm).
MO calculation
The relationship between the bathochromic shift in UV-visible
spectra and the cation–anion distance of potassium picrate was
estimated with ZINDO calculation software in the CAChe
work system (CAChe Scientific, Inc.). The initial structure of
potassium picrate was obtained from the PM3 optimization in
MOPAC (ver. 94. 10). Only the K–O distance in the fixed model
1312
J. Chem. Soc., Perkin Trans. 2, 2001, 1306–1314

View Article Online
was changed ranging from 3.27 to 15 Å and the simulation of
the configuration interaction in a ZINDO (parameter, INDO/1;
SCF type, RHF; C. I. level, 8) at each K–O distance gave
an electronic spectral pattern, i.e., λmax in UV-visible. A plot of
the K–O distance (r) vs. bathochromic shifts (BS) demonstrated
the relationship of BS = Ϫk(1/r2) (where k is a constant) that
suggested a principal contribution of the electrostatic effect to
the bathochromic shifts.
of a 1 : 2 (2–metal) complex. The KS1 : 1 and KS1 : 2 values were
estimated by a non-linear simulation method for the 1 : 2 com-
plexation.41 In the other case, the KS values were estimated by
the simulation for the 1 : 1 complexation.
Low temperature NMR measurements
The solution of host 1 complexes with metal cations was pre-
pared by the treatment in a supersonic homogenizer of host
1 [2H]chloroform–[2H2]dichloromethane (v/v = 1 : 3) solution
containing an excess amount of metal salt. The decanted solu-
tions were used for the NMR experiment. The initial concen-
tration of the host was 1.63 mM. The concentration of the
complex was estimated at about 1.6 mM because of the large
association ability in the low-polar solvent. The probe of the
NMR instrument was slowly refrigerated by liquid nitrogen
until the temperature was low enough (ca. 180 K) to observe
coalescence behavior. 1H-NMR spectra were measured at each
temperature with increases of 10 K. The activation free energy
(∆GC≠) values for the apparent conformational changes were
estimated by the Eyring equation from the coalescence
temperature (T C) and the splitting width (∆ν) of their proton
peaks (3–5 peaks).
UV-visible experiments
The bathochromic shifts (∆λ) of λmax of a given metal picrate
caused by adding hosts (1–4) in THF were determined as
follows. Permethylated cyclofructans were gradually added to
a ca. 5 × 10Ϫ5 M THF solution of metal picrate stirred mag-
netically in a UV cell until no shift was observed or until the
isosbestic relationship disappeared. A new λmax was then deter-
mined from the UV-visible spectra. Commercial THF (spectro-
photometric grade, Dojin Co.) was used without purification.
The KS values of 1 and 2 with metal picrate were determined
at constant temperature in THF. To maintain the temperature
of the UV cell, water regulated at that temperature in a
NESLAB endocal refrigerated circulating bath was allowed to
flow into a circulating cell holder.
For a typical example, the case of host 1 with Kϩ (picrate) at
18 ЊC: to 3 mL of a 4.43 × 10Ϫ5 M potassium picrate solution
in THF was added Apart of a 3.17 × 10Ϫ2 M host 1 solution in
THF (Apart = 1, 2, 4, 8, 15, 30 µL, thus Atotal = 60 µL), and the
UV-visible spectra of these solutions were measured. These
concentration ratios [H]/[G] of the host and the guest after
the addition amounted to 0.239, 0.717, 1.67, 3.59, 8.26, and
14.3, respectively. The KS values were determined using typical
non-linear methods (1 : 1 simulation). The calculated KS
values were 2.69 × 104 MϪ1 (382 nm), 2.68 × 104 MϪ1 (394 nm),
and 2.69 × 104 MϪ1 (400 nm). The averaged value was
ESIMS experiments
ESIMS sample solutions were prepared by the following pro-
cedure. 10 µL of a 0.5 mM acetone solution of the host and 10
µL each of 5 mM acetone solutions of two different metal thio-
cyanates were mixed and diluted to 50 µL by adding acetone.
The concentration of the host and two metal cations were 0.10,
1.0 and 1.0 mM. In all samples, potassium thiocyanate was used
as the second metal salt for comparison. To maintain con-
gruence between the peak intensity of the host 1ؒmetal complex
ion and host 2ؒmetal complex ion, the relative peak intensity of
the host 2ؒmetal complex ion was corrected by the relative peak
intensity [I(host 2ؒK)ϩ/I(host 1ؒK)ϩ = 0.11] estimated from an
ESI mass spectrum of a three component sample including
hosts 1, 2 and potassium thiocyanate.
ESI mass spectra were measured under the following con-
ditions: syringe pump rate, 5 mL minϪ1; mass range, m/z
0–2000; capillary temperature, 150 ЊC; spray interface voltage,
4.50 kV; skimmer pump, 740 mTorr; scan time, 10 s; accumula-
tion times, 10 scans. The sample solution was directly infused
into an ion source.
(2.69 0.01) × 104 MϪ1
.
Thermodynamic parameters were estimated from the linear
relationship between logKS measured at various temperatures
and the reciprocal of the temperature (1/T ). The slope and the
intercept of the linear relationship fitted by least squares
methods were estimated as ∆H and ∆S, respectively.
The UV and visible sources were D2 and W lamps. The
scan range was 280–500 nm and the scan speed was 120 nm
minϪ1. All temperatures in the UV cell were measured using
a thermometer (thermocouple type, TAKARA thermistor
instruments Co., D226).
Acknowledgements
1H-NMR experiments
We are very grateful to Mitsubishi Kagaku Co. for the kind
donation of the cyclofructan mixture (n = 6–8), to Professor
Emeritus Takao Uchiyama (Osaka Kyouiku University) for his
kind advice, to Dr Zenzaburo Totsuka, Mr Koji Matsuda, and
Mr Yasuhiro Yamato (Fujisawa Pharm. Co.) for measurements
of the ESI mass spectra, to Mr Hitoshi Yamada (ISIR, Osaka
University) for the FABMS measurements, to Mrs Fusako
Fukuda (ISIR) for elemental analyses, to Professors Takahiro
Kaneda and Yoshiteru Sakata (ISIR) for permission to use the
preparative gel-permeation liquid chromatography instrument
in the early stage of our works, to Dr Shingo Yamamura
(OMTRI) for the ion chromatographic analyses and atomic
absorption analyses and to Dr Masaya Takahashi (OMTRI) for
the solid NMR measurements.
1
1H-NMR spectra were measured using the H-NMR titration
technique reported before.14 Commercial [2H6]acetone (Aldrich,
99.5 atom% D) was used as a solvent without purification. The
KS values of complexation equilibrium between the permethyl-
ated cyclofructans and metal thiocyanate at 25 ЊC (298 K) were
determined by the following procedure. To an NMR tube con-
taining a 0.5 mL [2H6]acetone solution of permethylated cyclo-
fructan of which the exact concentration was known was added
dropwise a [2H6]acetone solution of metal thiocyanate, whose
concentration was exactly prepared, and 1H-NMR spectra
were measured in order. To maintain the sample solution at a
constant temperature, the NMR tube was kept on the probe,
which was maintained at 25 ЊC (298 K), for over 15 min before
measurement of the NMR spectra. The averages of the KS
values estimated from the concentration ratio [G]/[H] and the
corresponding shifts of the resulting proton peaks (1–9 peaks)
were adopted. The KS values were calculated using a general
non linear method.14
References
1 B. C. Pressman, E. J. Harris, W. S. Jagger and J. H. Johnson, Proc.
Natl. Acad. Sci. USA, 1967, 58, 1949.
2 A. Agtarap, J. W. Chamberlin, M. Pinkerton and L. K. Steinrauf,
J. Am. Chem. Soc., 1967, 89, 5737.
3 (a) B. C. Pressman, Fed. Proc., 1968, 27, 1283; (b) B. C. Pressman,
Annu. Rev. Biochem., 1976, 45, 501.
The stoichiometry of host 1 with all alkali metal ions and
host 2 with rubidium ion were confirmed to be 1 : 1 complexes
by the Job’s plot.40 In the case of host 2 with NaSCN, the
1
behavior of H-NMR induced-shifts suggested the formation
J. Chem. Soc., Perkin Trans. 2, 2001, 1306–1314
1313

View Article Online
4 (a) M. Noda, S. Shimizu, T. Tanabe, T. Takai, T. Kayano, T. Ikeda,
H. Takahashi, H. Nakayama, Y. Kanaoka, N. Minamino,
K. Kangawa, H. Matsuo, M. A. Raftery, T. Hirose, M. Notake,
S. Inayama, H. Hayashida, T. Miyata and S. Numa, Nature, 1984,
312, 121; (b) T. Tanabe, M. Takeshima, A. Mikami, V. Flockerzi,
H. Takahashi, K. Kangawa, M. Kojima, H. Matsuo, T. Hirose and
S. Numa, Nature, 1987, 328, 313.
5 (a) C. J. Pedersen, J. Am. Chem. Soc., 1967, 89, 2495; (b) D. J. Cram,
Angew. Chem., Int. Ed. Engl., 1988, 27, 1009.
6 (a) B. Dietrich, J.-M. Lehn and J. P. Sauvage, Tetrahedron Lett.,
1969, 2885; (b) J.-M. Lehn, Angew. Chem., Int. Ed. Engl., 1988, 27,
90.
20 The conformation of the crown ether is mainly represented by the
arrangement of the –O–C–C–O– torsion angles: gϩ = ∼0 to ϩ120,
gϪ = 0 to Ϫ120, and t = Ϫ120 to 180Њ.
21 Space group: 1ؒKϩ (PicϪ) complex, P43212; 1ؒRbϩ (PicϪ) complex,
P41212; 1ؒCsϩ (PicϪ) complex, P41212.
22 J. J. P. Stewart, J. Comput. Chem., 1989, 10, 209.
23 J. A. Pople and D. L. Beveridge, in Approximate Molecular Orbital
Theory, McGraw-Hill Book Company, New York, 1970.
24 R. M. Izatt, K. Pawlak and J. S. Bradshaw, Chem. Rev., 1991, 91,
1721.
25 J. J. Christensen, D. J. Eatough and R. M. Izatt, Chem. Rev., 1974,
74, 351.
7 D. J. Cram, T. Kaneda, R. C. Helgeson and G. M. Lein, J. Am.
Chem. Soc., 1979, 101, 6752.
26 Y. Marcus, in Ion Solvation, John Wiley, New York, 1985.
27 R. D. Boss and A. I. Popov, Inorg. Chem., 1986, 25, 1747.
28 J. D. Dunitz, M. Dobler, P. Seiler and R. P. Phizackerley, Acta
Crystallogr., Sect. B, 1974, 30, 2733.
29 M. Dobler, J. D. Dunitz and P. Seiler, Acta Crystallogr., Sect. B,
1974, 30, 2741.
30 S. Immel and F. W. Lichtenthaler, Liebigs. Ann. Chem., 1996,
39.
31 (a) E. Leize, A. Jaffrezic and A. V. Dorsselaer, J. Mass Spectrom.,
1996, 31, 537; (b) S. M. Blair, E. C. Kempen and J. S. Brodbelt,
J. Am. Soc. Mass Spectrom., 1998, 9, 1049; (c) K. Wang and
G. W. Gokel, J. Org. Chem., 1996, 61, 4693.
8 D. J. Cram, Science, 1983, 219, 1177.
9 (a) C. D. Gutche and R. Muthukrishnan, J. Org. Chem., 1978,
43, 4905; (b) C. D. Gutche, B. Dhawan, K. H. No and R.
Muthukrishnan, J. Am. Chem. Soc., 1981, 103, 3782; (c) S. Shinkai,
K. Araki and O. Manabe, J. Am. Chem. Soc., 1988, 110, 7214.
10 A. F. Danil de Namor, R. M. Cleverley and M. L.
Zapata-Ormachea, Chem. Rev., 1998, 98, 2495.
11 (a) M. Kawamura, T. Uchiyama, T. Kuramoto, Y. Tamura and
K. Mizutani, Carbohydr. Res., 1989, 192, 83; (b) M. Kawamura
and T. Uchiyama, Carbohydr. Res., 1994, 260, 297.
12 M. Sawada, T. Tanaka, Y. Takai, T. Hanafusa, T. Taniguchi,
M. Kawamura and T. Uchiyama, Carbohydr. Res., 1991, 217, 7.
13 (a) T. Uchiyama, M. Kawamura, T. Uragami and H. Okuno,
Carbohydr. Res., 1993, 241, 245; (b) N. Yoshie, H. Hamada,
S. Takada and Y. Inoue, Chem. Lett., 1993, 353.
14 (a) Y. Takai, Y. Okumura, S. Takahashi, M. Sawada, M. Kawamura
and T. Uchiyama, J. Chem. Soc., Chem. Commun., 1993, 53;
(b) Y. Takai, Y. Okumura, M. Sawada, S. Takahashi, M. Siro,
M. Kawamura and T. Uchiyama, J. Org. Chem., 1994, 59, 2967.
15 T. Kida, K. Isogawa, W. Zhang, Y. Nakatsuji and I. Ikeda, Bull.
Chem. Soc. Jpn., 1998, 71, 1201.
16 (a) M. Bourgoin, K. H. Wong, J. Y. Hui and J. Smid, J. Am.
Chem. Soc., 1975, 97, 3462; (b) A. Arduini, A. Pochini, S. Reverberi,
R. Ungaro, G. D. Andreeti and F. Ugozzoli, Tetrahedron, 1986,
42, 2089.
17 T. Arimura, M. Kubota, T. Matsuda, O. Manabe and S. Shinkai,
Bull. Chem. Soc. Jpn., 1989, 62, 1674.
32 M. Goodall, P. M. Kelly, D. Parker, K. Gloe and H. Stephan,
J. Chem. Soc., Perkin Trans. 2, 1997, 59.
33 M. Sawada, Y. Takai, H. Yamada, J. Nishida, T. Kaneda,
R. Arakawa, M. Okamoto, K. Hirose, T. Tanaka and K. Naemura,
J. Chem. Soc., Perkin Trans. 2, 1998, 701.
34 (a) M. Yamashita and J. B. Fenn, J. Phys. Chem., 1984, 88, 4511;
(b) M. Yamashita and J. B. Fenn, J. Phys. Chem., 1984, 88, 4671;
(c) P. Kebarle and L. Tang, Anal. Chem., 1993, 65, 972A.
35 For example, D. Live and S. I. Chan, J. Am. Chem. Soc., 1976, 98,
3769.
36 M. Sawada, M. Shizuma, H. Adachi, Y. Takai, T. Takeda and
T. Uchiyama, Chem. Commun., 1998, 1453.
37 S. Hakomori, J. Biochem. (Tokyo), 1964, 55, 205.
38 M. A. McKervey, E. M. Seward, G. Ferguson, B. Ruhl and
S. J. Harris, J. Chem. Soc., Chem. Commun., 1985, 388.
39 M. A. Coplan and R. M. Fuoss., J. Phys. Chem., 1964, 68,
1177.
18 Y. Takeda and O. Arima, Bull. Chem. Soc. Jpn., 1985, 58, 3403.
19 M. Ohki, in Applications of Dynamic NMR Spectroscopy to Organic
Chemistry, VCH Publishers, Inc., Weinheim, 1985,chapter 3, p. 11.
40 H. Tsukube and H. Sohmiya, J. Org. Chem., 1991, 56, 875.
41 R. P. Bonar-Law and J. K. M. Sanders, J. Am. Chem. Soc., 1995, 117,
259.
1314
J. Chem. Soc., Perkin Trans. 2, 2001, 1306–1314