Temporary crown ether compounds induced by UV irradiation

Article abstract of DOI:10.1021/jp013164q

A new reversible photoresponsive system using a photophysical process is presented and proved in detail. 1,n-bis(3-(1-pyrenyl)propylcarboxy)oxaalkanes (DP3n: n = 3 to 6), linear oligooxyethylenes with pyrenyl groups at both ends, were shown to form a relatively stable cyclic structure analogous to a crown ether in fluid solution during their photoexcited states, occurring due to the formation of an intramolecular excimer between two terminal pyrenyl groups. Moreover, DP34 molecules were found to capture and transport cadmium ions more effectively when they were irradiated. Accordingly, these compounds are referred to as photoinduced crown ether compounds (PICs) because they work as a crown ether when irradiated and they can select metal ions to capture according to their diameters. Our study is the first to attempt to use excimers to fix the conformation of a molecule for relatively long periods and to put this photophysical process to practical use.

Full text of DOI:10.1021/jp013164q

3316
J. Phys. Chem. B 2002, 106, 3316-3322
Temporary Crown Ether Compounds Induced by UV Irradiation
Hideyuki Itagaki,* Wakako Masuda, Yukiko Hirayanagi, and Kazuya Sugimoto
Department of Chemistry, School of Education and Graduate School of Electronic Science and Technology,
Shizuoka UniVersity, 836 Ohya, Shizuoka 422-8529, Japan
ReceiVed: August 20, 2001; In Final Form: NoVember 29, 2001
A new reversible photoresponsive system using a photophysical process is presented and proved in detail.
1,n-bis(3-(1-pyrenyl)propylcarboxy)oxaalkanes (DP3n: n ) 3 to 6), linear oligooxyethylenes with pyrenyl
groups at both ends, were shown to form a relatively stable cyclic structure analogous to a crown ether in
fluid solution during their photoexcited states, occurring due to the formation of an intramolecular excimer
between two terminal pyrenyl groups. Moreover, DP34 molecules were found to capture and transport cadmium
ions more effectively when they were irradiated. Accordingly, these compounds are referred to as photoinduced
crown ether compounds (PICs) because they work as a crown ether when irradiated and they can select metal
ions to capture according to their diameters. Our study is the first to attempt to use excimers to fix the
conformation of a molecule for relatively long periods and to put this photophysical process to practical use.
Introduction
In a short communication, we have described a novel
reversible photoresponsive system that employs a photophysical
process.1 We introduced pyrenyl groups onto both ends of linear
oligooxyethylenes (Figure 1). When irradiated with UV light,
the pyrenyl groups link together to form an excimer with a
lifetime in the sub-microsecond range.2 This structure is
analogous to that of crown ethers, which are cyclic ethylene
glycols capable of transporting ionic compounds into an organic
phase. We showed that one of these compounds indeed had the
properties of a crown ether: when irradiated, it could capture
calcium ions and transport them through a CCl4 phase. Accord-
ingly, we have referred to them as photoinduced crown ether
compounds (PICs).
Here we report that a PIC is a general phenomenon. To prove
that a PIC can select a metal ion whose size fits the cavity of
its closed form, we examine a case in which cadmium ions are
added into a PIC system: water is easily polluted by calcium
because it is present everywhere, but water has no chance of
being spontaneously contaminated by cadmium.
Figure 1. A schematic drawing of a photoinduced crown ether
compound. The probability that a compound will be in the closed form,
which is similar to a crown ether, is usually very small in the absence
of light. However, if one of the terminal pyrenyl groups of oligooxy-
ethylene compounds is irradiated, it links to the other pyrenyl group to
form an excimer that stabilizes this conformation. Here shows an
extended conformation as an example of open form (I), but a longer
oxyethylene chain prefers helical conformations.
On the other hand, several studies have investigated oli-
gooxyethylenes having aromatic compounds at both ends. These
studies can be classified into three types: (i) studies on the
expansion or cyclization of an oligomer or polymer in fluid
solution,3,4 (ii) attempts to develop a fluorescent reagent to
monitor metal ions captured by oligooxyethylenes,5-8 and (iii)
attempts to develop a compound to capture metal ions by
irradiation.9,10 In the first two types of studies, excimers formed
between aromatic end groups were used as probes to obtain
the information regarding the physical properties or the amounts
of metal ions present. The third type of study used photodimer-
ization, which is a photochemical reaction, to permanently link
the aromatic end groups. Accordingly, our work was the first
attempt to use excimers to fix the conformation of a molecule
for relatively long periods. Needless to say, it is reversible
because a compound capturing metal ions dissociates its excimer
in the dark and would release them in water. Since the
concentration of metal ions in a PIC solution can be changed
by irradiation, such solutions might have applications as
photoswitches.
Experimental Section
Materials. The structure of the 1,n-bis(3-(1-pyrenyl)propyl-
carboxy)oxaalkanes (DP3n; n ) 3, 4, 5, 6) studied in the present
work is shown in Scheme 1. To prepare the DP3n’s, 1-pyreneb-
utyric acid (PB) was first converted to the acid-chloride, 3-(1-
pyrenyl)propylcarbonyl chloride (PPCC), and then reacted with
triethylene glycol (OE3; n ) 3), tetraethylene glycol (OE4; n
) 4), pentaethylene glycol (OE5; n ) 5), and hexaethylene
glycol (OE6; n ) 6) to produce DP33, DP34, DP35, and DP36,
respectively.
* Author to whom correspondence should be addressed.
10.1021/jp013164q CCC: $22.00 © 2002 American Chemical Society
Published on Web 03/05/2002

Temporary Crown Ether Compounds
J. Phys. Chem. B, Vol. 106, No. 12, 2002 3317
1
SCHEME 1
DP35. H NMR (CD3CN)(δ in ppm, J in Hz): 2.06 (qui,
4H, J ) 7.5), 2.41 (t, 4H, J ) 7.3), 3.32 (t, 4H, J ) 7.7), 3.36-
3.47 (m, 12H), 3.55 (t, 4H, J ) 4.8), 4.12 (t, 4H, J ) 4.8),
7.84-8.32 (m, 18H). 13C NMR (CD3CN)(δ in ppm): 26.3, 31.8,
33.3, 62.9, 69.0, 69.3, 70.0, 70.4, 123.3, 124.5, 124.6, 125.8,
126.3, 127.0, 127.2, 127.3. IR (neat): 3040, 2880, 1730, 1600,
1590, 1460, 1420, 1350, 1250, 1180, 1130, 960, 850, 760, 720.
HRFAB-MS: (M + H)+ calcd. 778.3506 for C50H51O8.
Found: 778.3491.
1
DP36. H NMR (CD3CN)(δ in ppm, J in Hz): 2.07 (qui,
4H, J ) 7.6), 2.42 (t, 4H, J ) 7.2), 3.33 (t, 4H, J ) 7.7), 3.35-
3.49 (m, 16H), 3.57 (t, 4H, J ) 4.8), 4.13 (t, 4H, J ) 4.8),
7.85-8.33 (m, 18H). 13C NMR (CD3CN)(δ in ppm): 26.5, 31.9,
33.2, 63.1, 68.5, 69.8, 69.9, 69.9, 123.2, 124.5, 124.7, 125.8,
126.3, 127.0, 127.2, 127.3. IR (neat): 3040, 2880, 1730, 1600,
1460, 1420, 1350, 1250, 1180, 1110, 960, 850, 760, 720.
HRFAB-MS: (M + H)+ calcd. 823.3846 for C52H54O9.
Found: 823.3828.
The solvents used for photophysical measurements were THF,
which was freshly distilled, and acetonitrile (Wako, luminasol
grade) without purification. Cadmium perchlorate hexahydrate
and calcium thiocyanate tetrahydrate were purchased from
Kishida Chemical Ltd. and Wako Ltd., respectively.
To a solution of 2.0 g of PB in 190 mL of benzene, 4.3 g of
thionyl chloride was added dropwise with stirring. The solution
was refluxed at 80 °C for 4 h. A 2.2 g quantity of PPCC was
obtained by evaporating the solvent and dried under vacuum.
PPCC (1.0 g) was then reacted in 100 mL of benzene including
1 mL of pyridine (PPCC/pyridine ) 1 g/1 mL) under N2 flux
with OE3, OE4, OE5, and OE6, respectively. The mole ratio
of PPCC and OEn (n ) 3-6) was 4. The solution was refluxed
at 80 °C for 7 h. After evaporating the solvent, 150 mL of 1%
aqueous HCl was added and the organic products were extracted
into diethyl ether. The combined extracts were washed with 1%
aqueous HCl (three times), 1% aqueous NaOH (three times),
and distilled water (two times), and saturated NaCl solution,
and then were evaporated and dried under vacuum. The products
were separated with a Sephadex L-20 (Pharmacia) chromato-
graphic column using THF as the eluent. Since impurities having
a pyrenyl group at only one end do not show excimer
fluorescence, the fractions whose fluorescence spectra consisting
of monomer and excimer fluorescence were identical with one
another were collected all together. Because an impurity having
only one terminal pyrenyl group shows strong fluorescence of
monomeric pyrenyl group about 100 times as much as DP3n
does, note that the fluorescence measurements are quite effective
to show the purity: for example, we can detect the existence
of the impurity with only one terminal pyrenyl group even if
its fraction is 0.1%.
Measurements. Fluorescence spectra and fluorescence ex-
citation spectra were measured on a Hitachi F-4500 spectro-
fluorometer. The sample temperature was controlled by an
Oxford DN1704 cryostat with an ITC-4 digital temperature
controller, which can regulate to better than (0.1 K. Independent
temperature measurement was carried out by means of a second
thermocouple and a potentiometer. All samples were kept at
each set temperature and spectra were run repeatedly for more
than 30 min even after perfect duplication was obtained to make
sure that the equilibrium state was attained. Fluorescence decay
curves were obtained by using a Horiba NAES550 single photon
counting machine with a nanosecond flash lamp. The excitation
wavelength was separated to be 340 nm through an interference
filter. The emission decay was measured at 25 °C through an
interference filter of 383- (monomeric singlet), or 479-nm
(excimer). We analyzed fluorescence decay curves by the
deconvolution method after O’Connor and Phillips11 using the
Durbin-Watson factor (DW)12,13 to assess the validity of the
trial fitting function. DW is calculated from
N
N
DW )
(R - R )2/( R2)
∑
∑
i
i-1
i
i)2
i)1
where the weighted residual Ri ) (Yi - Fi)/Yi1/2, Yi and Fi are
the values of the experimental data and trial calculation value
corresponding to the time channel i, and N is the number of
experimental points. For the best fit, the value of DW approaches
2.0.
1
DP33. H NMR (CD3CN)(δ in ppm, J in Hz): 2.03 (qui,
4H, J ) 7.5), 2.36 (t, 4H, J ) 7.2), 3.28 (t, 4H, J ) 8.0), 3.50
(s, 4H), 3.54 (t, 4H, J ) 4.8), 4.09 (t, 4H, J ) 4.8), 7.82-8.30
(m, 18H). 13C NMR (CD3CN)(δ in ppm): 26.5, 31.9, 33.1, 63.0,
68.5, 69.9, 123.2, 124.5, 124.6, 125.8, 126.3, 126.9, 127.2,
127.3. IR (neat): 3040, 2950, 1730, 1600, 1510, 1460, 1250,
1180, 1130, 960, 850, 760, 720. HRFAB-MS: (M + H)+ calcd.
691.3060 for C46H43O6. Found: 691.2965.
All NMR experiments of DP3n in CD3CN were performed
using a JEOL GX400 FT-NMR spectrometer with a static field
1
strength of 399.65 MHz for H and 100.4 MHz for 13C. The
concentration of cadmium ion transferred was determined with
a Hitachi Z-8270 polarized Zeeman atomic absorption spectro-
photometer.
1
DP34. H NMR (CD3CN)(δ in ppm, J in Hz): 2.06 (qui,
4H, J ) 7.5), 2.40 (t, 4H, J ) 7.1), 3.31 (t, 4H, J ) 7.7), 3.40-
3.45 (m, 8H), 3.52 (t, 4H, J ) 4.8), 4.10 (t, 4H, J ) 4.8), 7.84-
8.31 (m, 18H). 13C NMR (CD3CN)(δ in ppm): 26.6, 31.9, 33.2,
63.1, 68.5, 69.9, 69.9, 123.3, 124.5, 124.7, 125.8, 126.3, 127.0,
127.2, 127.3. IR (neat): 3040, 2880, 1730, 1600, 1460, 1250,
1110, 960, 850, 760, 720. HRFAB-MS: (M + H)+ calcd.
735.3322 for C48H47O7. Found: 735.3265.
Results and Discussion
Photophysical Processes of DP3n. Figure 2 shows the
fluorescence spectra of DP3n in aerated acetonitrile at 25 °C.
The peaks near 378 and 480 nm were assigned to the pyrenyl
monomer singlet and excimer emission, respectively.14 In fact,
the excitation spectra for the fluorescence between 380 and 550

3318 J. Phys. Chem. B, Vol. 106, No. 12, 2002
Itagaki et al.
TABLE 1: Fluorescence Decay Profiles of DP3n in Degassed
Solutions of THF and Acetonitrile
1/λ1
(ns)
1/λ2 1/λe
kEM
τEM
(ns)
samples
A1
A2 (ns) (ns) (10-8 s-1
)
In THF
1-pyrenebutyric acid 1.0 210
DP33
DP34
DP35
DP36
0.93
0.93
0.92
5.2 0.069 200 52
5.6 0.069 200 51
6.8 0.078 200 50
1.9
1.7
1.4
1.2
5.3
5.8
7.0
8.2
7.9
52
In Acetonitrile
1-pyreneacetic acid 1.0 267
DP33
DP34
0.91
0.89
0.96
0.89
0.94
1.8 0.090 200 54
2.1 0.11 200 53
2.0 0.039 200 51
2.2 0.11 200 49
2.9 0.057 200 53
5.5
4.7
5.0
4.5
3.4
1.8
2.1
2.0
2.2
2.9
DP34 + Cd2+ b
DP35
Figure 2. Fluorescence spectra of DP3n in aerated THF (concentration
of pyrenyl groups is less than 1 µM). All the spectra are normalized to
the peak of excimer fluorescence. The excitation wavelength is 342
nm.
DP36
a [DP3n] ∼ 1 µM. b [Cd2+] ∼ 30 µM.
SCHEME 2a
nm of all samples were almost identical. They were also
identical to their absorption spectra. No concentration depen-
dence was observed for the fluorescence spectra of the DP3n’s
below 10 µM, meaning that all the excimers shown in Figure 2
are formed intramolecularly. The fluorescence spectra of DP3n
in THF, water, chloroform, and dichloromethane all showed
strong excimer fluorescence and weak monomeric fluorescence,
indicating that DP3n readily takes a closed form similar to a
crown ether as shown in Figure 1.
The ratio of excimer fluorescence peak intensity, IE, to
monomer fluorescence peak intensity, IM, was inversely related
to the number of oxyethylene units: the order of IE/IM was DP33
> DP34 > DP35 > DP36. The intramolecular excimers are
formed after excitation of one of the pyrenyl moieties in a
diffusional encounter with the other pyrenyl moiety. The IE/IM
of two terminal aromatic groups attached to a chain is already
known to have such a similar dependence on a chain length as
the dependence of the yields of organic ring closure reactions
on ring size.15,16 Thus, the order of IE/IM shown above is quite
reasonable because the average distance between two terminal
pyrenyl groups must be in the order of DP33 < DP34 < DP35
< DP36.
a M* and E* are monomer singlet and excimer, respectively. kFM
and kFE are the rate constants for the radiative deactivation from excited
monomer and excimer state, respectively. kIM and kIE are those for the
nonradiative deactivation, and kEM is the rate constant for excimer
formation, respectively.
However, the problem is that there is a smaller decay constant,
λ2, which is the same as the monomer singlet lifetime (∼200
ns) of pyrenyl moiety, although Scheme 2 never predicts its
existence. The decay component with a time constant of 200
ns means that some DP3n’s cannot form excimers and that this
fraction is more than 6%.24
Two explanations for this are possible: one is because there
are some conformations that can never change to the conforma-
tion possible to form excimers within the excited lifetime of
the pyrenyl group, and another is because some pyrenyl groups
are photodegraded during the measurements of fluorescence
decay curves, and eventually a DP3n with a degraded pyrenyl
group at one end cannot form an excimer intramolecularly.
The photodegradation of pyrenyl groups was assured because
the excimer fluorescence of DP3n decreased after the irradiation
of UV light. Thus, when we repeatedly measured the fluores-
cence decay curves of one sample, the fraction of component
with 200 ns (1/λ2) was found to increase. However, the decay
curves of each DP3n were ascertained to agree with one another
in the case where we measured each new degassed solution with
irradiating them for a short time period. Thus, we think that
some conformations of DP3n could be impossible to form an
intramolecular excimer within the excited lifetime of the pyrenyl
group (200 ns), but the information on their fraction was
contaminated by the photodegradation of a terminal pyrenyl
group. In the present paper, we would like to consider only the
larger decay constant, λ1.
To examine the dynamic process of the excimer formation
in the DP3n system, we measured the time profile of monomer
and excimer fluorescence of DP3n (n ) 3-6) in degassed THF
and acetonitrile by means of the nanosecond single photon
counting technique. The monomer fluorescence of DP3n in both
solvents was determined to decay dual-exponentially (eq 1) with
the reciprocal of decay constants of 2 to 8 ns and 200 ns.17
IM(t) ) A1 exp(-λ1t) + A2 exp(-λ2t), A1 + A2 ) 1 (1)
The decay curve of excimer fluorescence was fitted well to
IE(t) ) A3 exp(-λet) - A4 exp(-λ3t)
(2)
Roughly speaking, λ1 and λ3 almost agree with each other and
A3 is a little larger than A4. The calculated decay parameters
are summarized in Table 1. In the case in which λ3 is short, the
error is large, so we present only λe values in Table 1.
The first conclusion deduced from the transient measurements
is that the excimer dissociation of DP3n can be neglected even
at room temperature, because the decay component with τe is
not detectable in the monomer fluorescence. Therefore the
kinetics of excimer formation in the DP3n system is considered
to conform to the conventional model shown in Scheme 2.23
Here the correspondence between λ1 and λ3 is explicable.
Under the assumption that Scheme 2 where excimer dis-
sociation can be neglected as described above is valid, eqs 3
and 4 hold.
λ1 ) kFM + kIM + kEM
(3)
(4)
IE/IM ) kFEkEM/(kFM(kFE + kIE))
Because kFM + kIM can be substituted by the experimental value

Temporary Crown Ether Compounds
J. Phys. Chem. B, Vol. 106, No. 12, 2002 3319
Figure 3. Temperature dependence of IE/IM for 1.0 µM DP35 in
degassed THF. All the IE/IM values are equilibrium values.
Figure 4. Cd2+ concentration dependence of IE/IM of DP3n in aerated
acetonitrile: DP33 (O), DP34 (2), DP35 (3), and DP36 (9). (IE/
2+
2+
IM)[Cd
and (IE/IM)[Cd are the values of IE/IM in the absence and
TABLE 2: Values of Activation Energies of DP3n in THF
(kJ/mol)
])0
]
presence of Cd2+ ions. All the concentrations of DP3n’s are 0.50 µM.
DP33
20
DP34
19
DP35
19
DP36
13
extended form but could have a helical conformation where an
excimer conformation is formed differently from the case of
DP3n with n e 5.
∆EEM
of PB, each kEM of DP3n can be calculated as shown in Table
1. As a reference, the reciprocal of kEM is defined as τEM, which
indicates the time required for two terminal pyrenyl groups to
encounter each other, i.e., the average time required for taking
a closed conformation as II in Figure 1. The transient results
are found to agree well with the photostationary results. As
shown in eq 4, IE/IM is proportional to the rate of excimer
formation, i.e, to the rate of the cyclization process. The order
of IE/IM among the DP3n’s was identical to the order of the
Thus far, we have clarified the information regarding the
process of intramolecular excimer formation in the DP3n system.
The fluorescence results show that (i) DP3n normally (i.e., in
the dark) exists in the open form (for example, I in Figure 1),
(ii) it requires a few nanoseconds to form a closed form (II in
Figure 1), (iii) τEM becomes longer with longer chains, and (iv)
roughly 40% (1/e) of DP3n would remain in the closed
conformation at even 50 ns after irradiation.
Photophysical Processes of DP3n when Cadmium Ions are
Added. We examined whether the addition of cadmium ions
influences the excimer formation of DP3n. We measured the
dependence of DP3n fluorescence on Cd2+ concentration using
3 µM DP3n (absorbance ∼ 0.05) and [Cd(ClO4)2] ranging from
10-5 to 10-2 M. Since DP3n is more stable in acetonitrile than
in water, the measurements were made in acetonitrile. The
fluorescence of the pyrenyl group of DP3n and its monomeric
model, PB, was found to be not quenched at all by the addition
τEM values, and indicates that the rate of the cyclization process
is dependent on the average distance between the two pyrenyl
end groups. In conclusion, the DP3n with a shorter oxyethylene
chain readily takes a closed conformation, i.e., DP33 > DP34
> DP35 > DP36.
Next we measured the temperature dependence of DP3n
fluorescence in THF. The peak wavelength of the excimer
emission is not changed at low temperatures, although a slight
blue-shift is observed above 260 K. Figure 3 demonstrates the
temperature dependence of IE/IM of DP35 in THF. A plot of
ln(IE/IM) versus 1/T was found to give a straight line below 240
K, where the isoemissive point was observed, and to yield a
value of 19 kJ/mol as an activation energy of excimer formation,
∆EEM: kEM ) A exp(-∆EEM/RT). Table 2 summarizes the values
of ∆EEM of DP3n obtained by using the same method. Because
the formation process of these intramolecular excimers is
diffusion-controlled and the kEM should be dependent on the
average distance between the two pyrenyl end groups, it is quite
reasonable that the values of ∆EEM are identical with one
another.
It is needless to say that the excimer formation between two
terminal pyrenyl groups of DP3n is influenced by the viscosity
of solvent, which is dependent on temperature. Thus, the values
of ∆EEM in Table 2 are not the net values for the molecular
motion of only DP3n’s own but rather include the temperature
dependence of solvent viscosity. Here, viscosity, η, follows a
relationship of η ) A exp(∆Evis/RT) where ∆Evis is the activation
energy for solvent molecules to flow. By using the reference
values between 248 and 323 K,26 ∆Evis of THF is calculated to
be 4.4 kJ/mol. Because ∆EEM is five times as large as ∆Evis,
the results show that the activation energy is not dependent on
the unit number of the oxyethylene chain (n) among n ) 3 to
5. It means that ∆EEM is determined only by the local motion
of the pyrenyl group.
of Cd2+. Figure 4 shows the Cd2+ concentration dependence of
2+
2+])0
IE/IM of DP3n: if there is no change, (IE/IM)[Cd ]/(IE/IM)[Cd
should be 1. Addition of Cd2+ up to 3 mM did not significantly
change the IE/IM values of DP33, 35, and 36. The change of
IE/IM observed when Cd2+ was added into DP3n is by far smaller
than that when Ca2+ was added into DP3n. Thus, the interaction
between DP3n and Cd2+ ions is assumed to be weaker than
that between DP3n and Ca2+ ions. Accordingly, we have to
comment that the fluorescence of any DP3n is useless for
monitoring Cd2+. However it should be noted that the excimer
formation efficiency of DP34 increased by 10% when Cd2+ was
added. The increase in IE/IM is not so noticeably large, but it is
reproducible. Then, we measured the fluorescence decay curves
of degassed acetonitrile solutions of DP34 when Cd2+ was added
into them. However, the decay patterns of DP34 with Cd2+ were
found to be almost identical with those of DP34 without Cd2+
:
the values of three decay constants were perfectly the same.
The difference is only the fraction of two components observed
in monomer fluorescence. The results might suggest that the
addition of Cd2+ decreases the fraction of the conformations
unable to form an intramolecular excimer within the excited
lifetime of the pyrenyl group (200 ns).
We do not think that this is a strong evidence for the
association of DP34 with Cd2+. However, we may safely say
that the unique change was observed for DP34 with Cd2+
.
Interaction between DP3n and Cadmium Ions Monitored
1
∆EEM of only DP36 was found to be lower than those of
other DP3n’s. It could be due to the fact that polyoxyethylene
chains prefer helical conformations instead of extended ones
shown in Figure 1. Six units of oxyethylene cannot have a linear
by NMR Measurements. We used H NMR and 13C NMR
measurements employing CD3CN as a solvent to examine the
ability of the DP3n’s to form a complex with Cd2+. First, we
1
measured the H NMR and 13C NMR spectra of oligooxyeth-

3320 J. Phys. Chem. B, Vol. 106, No. 12, 2002
Itagaki et al.
Figure 5. The averaged difference values, ∆δ, of the 1H NMR
chemical shifts of OEn (n ) 3-6) in CD3CN when the Cd2+ ions
([Cd2+]/[OEn] ∼ 50) were added (δ[Cd ]) and absent (δ[Cd ])0): ∆δ
Figure 7. The averaged increased values, ∆δ, of some 1H NMR
chemical shifts of DP3n (n ) 3-6) in CD3CN when the Ca2+ ions
2+
2+
2+
2+
) δ[Cd - δ[Cd
.
])0
]
([Ca2+]/[DP3n] ∼ 50) were added (δ[Ca ]) and absent (δ[Ca ])0): ∆δ
2+
2+
2+
2+
) δ[Ca - δ[Ca
.
])0
]
Figure 6. The averaged increased values, ∆δ, of some 1H NMR
chemical shifts of DP3n (n ) 3-6) in CD3CN when the Cd2+ ions
([Cd2+]/[DP3n] ∼ 50) were added (δ[Cd ]) and absent (δ[Cd ])0): ∆δ
2+
2+
2+
2+
) δ[Cd - δ[Cd
.
])0
]
ylenes (OEn; n ) 3 to 6) with and without Cd2+. In the case of
13C NMR spectra of OEn, the addition of Cd2+ did not induce
any meaningful shifts of any signals: the largest change was
72.1 ppm to 68.7 ppm observed for OE3, and the other shifts
were found to be around 1 ppm. However, the addition of Cd2+
Figure 8. A quartz cell for determining whether irradiated PICs of
different sizes are capable of transporting cadmium ions.
1
increased the H NMR chemical shifts of methylene protons
by 0.25-0.27 ppm (for example, in the case of OE5, 3.57 to
3.84 ppm (0.27), 3.50 to 3.77 ppm (0.27), and 3.49 to 3.76 ppm
Cd2+ transported by a DP3n in the presence and absence of
UV light, after a method for measuring the transportation yield
of metal ions.28 A solution of 1 mM DP3n in CCl4 was placed
in part (C). An aqueous solution of 10 mM Cd(ClO4)2 and 10
µM DP3n, and pure water were simultaneously layered on top
of (C) in (A) and (B), respectively. The transportation yield of
Cd2+ in (B) was measured using atomic absorption spectroscopy.
In the case of Cd2+, the limit of detection is 2.2 nM. Cadmium
ions are normally insoluble in CCl4. Thus, when we did not
dissolve any DP3n in (A) and (C), no Cd2+ was observed in
(B) even after 10 min either with and without UV irradiation.
To measure transportation yields of Cd2+ by DP3n, part (A)
was either kept in the dark for 3 min, or irradiated at 6.4 ×
1018 photons/s for 3 min.29 Figure 9 shows that these experi-
ments have a relatively large error. We did not stir the solution
because we were afraid that the splash drop might directly go
into (B) from (A). Nevertheless, Figure 9 clearly shows that a
greater amount of Cd2+ was transported to (B) by way of (C)
when (A) was irradiated. These experiments provided that DP34
1
(0.27)). Figure 5 shows the average shifts of H NMR signals
of OEn when Cd2+ was added. Although there exists an
interaction between OEn and Cd2+, its strength was not
significantly different among the different OEn’s.
1
The H NMR and 13C NMR chemical shifts of DP3n’s did
not change significantly even by the addition of Cd2+ (see Figure
6), while those of OEn’s changed very much: in particular, no
remarkable changes were observed for 13C NMR spectra. These
results are completely different from those for DP3n and Ca2+
1
(Figure 7). For example, the H NMR signals assigned to the
hydrogen atoms of the oxyethylene part of DP3n with Ca2+
shifted as follows: (unit, ppm) 0.015 (DP33), 0.11 (DP34), 0.33
(DP35), and 0.19 (DP36) in midchain (-O-CH-), 0.019
(DP33), 0.12 (DP34), 0.18 (DP35), and 0.079 (DP36) in end-
chain (-COO-CH). Thus, the NMR results show that DP35
has the strongest interaction with Ca2+ among the DP3n’s (n )
3-6). Compared with these changes, the changes shown in
Figure 6 can be neglected. If we dare to find any small
differences in Figure 6, the H NMR signals of only DP34
worked effectively as a PIC and that it transported 4 × 10-7
M
1
shifted to a direction different from those of the other DP3n’s,
although the shifts were not so remarkable.
after irradiation for 3 min.
Using the same method as described above, we measured
the amount of Cd2+ transported by DP3n when part (A) was
irradiated and not irradiated. More than five experiments were
performed to obtain the values of Cd2+ transported. Figure 10
shows the average values of Cd2+ transported by DP3n. In the
case of DP33 and DP36, the values were almost the same in
Transportation of Cadmium Ions by DP3n with and
without Irradiation of Light. There is no effectively direct
way to measure the ability of DP3n to capture metal ions when
it is irradiated. Therefore, we employed a quartz cell with a
partition as shown in Figure 827 to determine the amount of

Temporary Crown Ether Compounds
J. Phys. Chem. B, Vol. 106, No. 12, 2002 3321
Figure 9. Transport of cadmium ions by DP34 in the presence and
absence of UV light.
Figure 11. Schematic drawing of the behavior of DP3n during
measurements using a cell shown in Figure 8 as a photoinduced crown
ether compound. Only part (A) is irradiated.
In the case of DP35 and Ca2+, some strong evidence indicates
that they form a sort of complex even in the dark: by the
addition of Ca2+, (i) 1H NMR chemical shifts of DP35 changed
to a great degree (nearly 0.4 ppm), and (ii) the efficiency of the
excimer formation decreased very much.1 The greater transport
of Ca2+ by DP35, which has five units of oxyethylene, is
consistent with its size: the diameter of the hole in 5-unit-crown
ether, 15-crown-5 (15-C-5), is 170 to 220 pm31 while that of
Ca2+ is 200 pm. Thus, we assume that DP35 molecules do not
start with capturing Ca2+ after forming an excimer, but rather
they have some interaction with Ca2+ before being excited. The
intramolecular excimer formation between the pyrenyl end
groups is considered to make DP35 bind with Ca2+ very tightly.
Eventually, excimer formation plays an important role in
creating the strong interaction between DP3n and Ca2+. Because
it is somewhat difficult for the crown ether conformation bearing
Ca2+ to be solvated by water molecules, the Ca2+-bearing DP35
molecules are assumed to be more stable in CCl4.
Figure 10. Transport of cadmium ions by DP3n where n is the number
of oxyethylene units in the presence (O) and absence (b) of UV light
irradiation.
the presence and absence of UV light, meaning that they do
not work as a PIC at all. On the contrary, DP34 and DP35
transported Cd2+ when they were irradiated, and in particular
DP34 showed the greatest amount of Cd2+ transported.
As we have shown in a short communication,1 DP35
transported the greatest amount of Ca2+, whereas it is DP34
that transported the greatest amount of Cd2+. This difference
suggests that the size of oligooxyethylene is strongly related to
the capture of metal ions.
What Occurs in This PIC system? Here we describe a
probable mechanism governing the process taking place during
the experiments using a cell shown in Figure 8, and also describe
some possible weaknesses regarding our experiments.
The Cd2+ is assumed to be transported to (B) by the following
mechanism, which is shown as an illustration in Figure 11.
When irradiated, PICs in (A) capture Cd2+. Some of those near
(C) would diffuse into (C). Although (C) is not irradiated, the
Cd2+ ion cannot be released from the PIC because it is not
soluble in CCl4. Some of the Cd2+-bearing PIC molecules would
diffuse into (B), where they could then release the Cd2+ ion.
Thus, some of the Cd2+ would end up in (B).
One problem here is that only diffusion for 3 min at room
temperature would result in not so many Cd2+-bearing PICs
being transported between the two aqueous solutions (A) and
(B). For example, if the diffusion coefficient, Λ, of a PIC
molecule is 10-5 cm2/s, the rms distance, L, of a molecule
diffused for 3 min is calculated to be 1 mm since L ) (6Λt)1/2
for t seconds. Thus, we have to consider that the convection
plays a role in the migration of Cd2+ between compartments
under our experimental conditions because infrared photons were
not filtered out during UV irradiation. However, no Cd2+ was
observed in (B) without any DP3n’s after the UV irradiation
for even 10 min. Thus, it is obvious that the increased degree
of Cd2+ transportation is not due to the heating effect when the
solution is irradiated.
In the case of Cd2+ ions, we obtained no strong evidence
that they are captured by DP3n’s in the dark, compared with
the change of DP35 with Ca2+. If DP34 has no interaction with
Cd2+ in the dark and starts with binding Cd2+ after the UV
irradiation, this is the ideal system of the photoinduced
temporary crown ethers that we hope to develop. However, we
do not have clear evidence so far about whether DP34 starts
with capturing Cd2+ just after the irradiation or it only
strengthens the binding of Cd2+ when irradiated. In the latter
case, the following results could support the existence of some
interaction between DP34 and Cd2+, although they are not strong
evidences: (i) the NMR results of DP34 with Cd2+ are a bit
different from those of DP3n with Cd2+, and (ii) the excimer
formation efficiency of DP34 increased by the addition of Cd2+
.
In any rate, it is clear that intramolecular excimer formation
between two terminal pyrenyl groups plays an important role
in making the strong interaction of DP3n with Cd2+ and in the
transportation of Cd2+ shown in Figures 9 and 10.
As a matter of fact, the diameter of Cd2+, which is 190 pm,
is slightly smaller than that of Ca2+ but larger than the cavity
size of 12-crown-4 (12-C-4), which is around 144 pm.32 This
means that Cd2+ cannot be captured as tightly by a 12-C-4
molecule. However, some reports have shown that Cd2+ is
stabilized by the formation of a sandwich dication such as [(12-

3322 J. Phys. Chem. B, Vol. 106, No. 12, 2002
Itagaki et al.
C-4)2Cd]2+. Colton et al. used electrospray mass spectrometry
to investigate the interactions in solutions between some crown
ethers and 17 metal ions.33 Upon mixing with 12-C-4, only
cadmium, zinc, and lead were found to give ions containing
the metal and 12-C-4. As expected from their radii, they all
gave species with two crown ether ligands coordinated to the
metal ion. Bobrowski et al. examined the complexation of
essentially nonsolvated cations including Cd2+ with some crown
ethers.34 They found that a strong complex of Cd2+ was formed
in dichloromethane with 12-C-4 and that the ratio of the ligand
to cadmium ion in the complex is larger than 1:1. Zhang et al.
showed that the [(12-C-4)2Cd]2+ dications in [(12-C-4)2Cd][Cd2-
(SCN)6] and [(12-C-4)2Cd] [Cd2(SCN)8] serve not only as the
spacer of the crystal packing, but also as a template for the
formation of the two-dimensional layers.35,36 This also consti-
tutes evidence that [(12-C-4)2Cd]2+ dication has a sort of stable
structure.
(10) Desvergne, J.-P.; Fages, F.; Bouas-Laurent, H.; Marsau, P. Pure
Appl. Chem. 1992, 64, 1231, and references therein.
(11) O’Connor, D. V.; Phillips, D. Time-correlated Single-photon
Counting; Academic Press: London, 1985.
(12) Durbin, J.; Watson, G. S. Biometrika 1950, 37, 409.
(13) Durbin, J.; Watson, G. S. Biometrika 1951, 38, 159.
(14) Birks, J. B. Photophysics of Aromatic Molecules; Wiley-Inter-
science: New York, 1970; p 302.
(15) Hirayama, F. J. Chem. Phys. 1965, 42, 3163.
(16) Zachariasse, K. A.; Ku¨hnle, W. Z. Phys. Chem. (Wiesbaden) 1976,
101, 267.
(17) Some compounds were reported to show fluorescence from an
intramolecular excimer having two pyrenyl groups overlapped partially with
each other.18,19 The kinetic scheme including the formation of two excimers
was discussed by some authors.19-22 However, our DP3n systems can be
best fitted to the conventional kinetic scheme shown in Scheme 2 with
DW factors being near 2.0.
(18) Collart, P.; Demeyer, K.; Toppet, S.; De Schryver, F. C. Macro-
molecules 1983, 16, 1390.
(19) Zachariasse, K. A.; Duveneck, G.; Busse, R. J. Am. Chem. Soc.
1984, 106, 1045.
The present study is the first to use an excimer to fix the
conformation of a molecule, and shows that these photoinduced-
crown-ether compounds effectively capture and transport Cd2+
only when they are irradiated by light. In this paper, we have
presented evidence proving our concept of a photoinduced
crown ether.
(20) Itagaki, H.; Obukata, N.; Okamoto, A.; Horie, K.; Mita, I. J. Am.
Chem. Soc. 1982, 104, 4469.
(21) Itagaki, H.; Horie, K.; Mita, I. Macromolecules 1983, 16, 1395.
(22) De Schryver, P.; Collart, J.; Vandendriessche, R.; Goedeweeck,
A. M.; Swinnen, M.; Van der Auweraer, Acc. Chem. Res. 1987, 20, 159.
(23) Klo¨pffer, W. In Organic Molecular Photophysics; Birks, J. B., Ed.;
Wiley: New York, 1974; p 371.
(24) The existence of a fluorescence component having the same lifetime
as the isolated monomeric fluorescence was observed for oxyethylene
compounds with pyrenyl groups at both ends.25
(25) Nakamura, H. (Hokkaido University), Private communication.
(26) CRC Handbook of Chemistry and Physics, 76th ed.; Lide, D. R.,
Ed.; CRC Press: Boca Raton, 1995.
(27) The cell is different from that employed in ref 1.
(28) Rosano, H. L.; Schulman, J. H.; Weisbuch, J. B. Ann. N.Y. Acad.
Sci. 1961, 92, 457.
Acknowledgment. This work was supported by a Grant-
in-aid for Scientific Research (B) (09450348) from the Ministry
of Education, Science, Sports and Culture of Japan, and by the
Iwatani Naoji Foundation. The authors are grateful to A. Yagi
for operating FAB-MS to determine molecular weights of
DP3n’s.
(29) The intensity of the irradiation light for pyrenyl moieties was
References and Notes
determined by using the potassium ferrioxalate actiometry method.30
(1) Itagaki, H.; Masuda, W.; Hirayanagi, Y. Chem. Phys. Lett. 1999,
30, 402.
(2) Itagaki, H. In Experimental Methods in Polymer Science: Modern
Methods in Polymer Research and Technology; Tanaka, T., Ed.; Academic
Press: New York, 2000; Chapter 3.
(30) Murov, S. L. Handbook of Photochemistry; Marcel Dekker: New
York, 1973; p 119.
(31) Vo¨gtle, F. Supramolecular Chemistry: An Introduction; Wiley:
Chichester, 1989; Chapter 2.
(32) Christensen, J. J.; Eatough, D. J.; Izatt, R. M. Chem. ReV. 1974,
74, 351.
(33) Colton, R.; Mitchell, S.; Traeger, J. C. Inorg. Chim. Acta 1995,
231, 87.
(34) Bobrowski, A.; Bond, A. M.; Ellis, S. R. Inorg. Chim. Acta 1999,
293, 223.
(35) Zhang, H.; Wang, X.; Zhu, H.; Teo, B. K. J. Am. Chem. Soc. 1997,
119, 5463.
(3) Horie, K. Kobunshi 1983, 32, 196.
(4) Mita, I.; Horie, K. J. Macromol. Sci. ReV. Macromol. Chem. Phys.
1987, C27, 91.
(5) Tung, C.-H.; Wang, Y.-M. J. Am. Chem. Soc. 1990, 112, 6322.
(6) Kakizawa, Y.; Akita, T.; Nakamura, H. Chem. Lett. 1993, 1671.
(7) Suzuki, Y.; Morozumi, T.; Kakizawa, Y.; Bartsch, R. A.; Hayashita,
T.; Nakamura, H. Chem. Lett. 1996, 547.
(8) Kawakami, J.; Komai, Y.; Ito, S. Chem. Lett. 1996, 617.
(9) Desvergne, J.-P.; Bouas-Laurent, H. J. Chem. Soc. Chem. Commun.
1978, 403.
(36) Zhang, H.; Wang, X.; Zhu, H.; Xiao, W.; Zhang, K.; Teo, B. K.
Inorg. Chem. 1999, 38, 886.