Single-molecule fluorescence photoswitching of a diarylethene- perylenebisimide dyad: Non-destructive fluorescence readout

Article abstract of DOI:10.1021/ja110686t

Single-molecule fluorescence photoswitching plays an essential role in ultrahigh-density (Tbits/inch²) optical memories and super-high-resolution fluorescence imaging. Although several fluorescent photochromic molecules and fluorescent proteins

Full text of DOI:10.1021/ja110686t

ARTICLE
pubs.acs.org/JACS
Single-Molecule Fluorescence Photoswitching of a Diarylethene-
Perylenebisimide Dyad: Non-destructive Fluorescence Readout
Tuyoshi Fukaminato,*,†,‡ Takao Doi,§ Nobuyuki Tamaoki,† Katsuki Okuno,z Yukihide Ishibashi,z
Hiroshi Miyasaka,z and Masahiro Irie*,||
†Research Institute for Electronic Science, Hokkaido University, N20, W10, Kita-ku, Sapporo 001-0020, Japan
‡PREST, Japan Science and Technology Agency (JST)
§Department of Chemistry and Biochemistry, Graduate School of Engineering, Kyushu University, Motooka 744, Nishi-ku, Fukuoka
819-0395, Japan
zDivision of Frontier Materials Science, Graduate School of Engineering Science, and Center for Quantum Science and Technology
under Extreme Conditions, Osaka University, Toyonaka, Osaka 560-8531, Japan
Department of Chemistry and Research Center for Smart Molecules, Rikkyo University, Nishi-Ikebukuro 3-34-1, Toshima-ku, Tokyo
171-8501, Japan
S Supporting Information
b
ABSTRACT: Single-molecule fluorescence photoswitching plays
an essential role in ultrahigh-density (Tbits/inch2) optical mem-
ories and super-high-resolution fluorescence imaging. Although
several fluorescent photochromic molecules and fluorescent pro-
teins have been applied, so far, to optical memories and super-high-
resolution imaging, their performance is unsatisfactory because of
the absence of “non-destructive fluorescence readout capability”.
Here we report on a new molecular design principle of a molecule
having non-destructive readout capability. The molecule is composed of acceptor photochromic diarylethene and donor fluorescent
perylenebisimide units. The fluorescence is reversibly quenched when the diarylethene unit converts between the open- and the
closed-ring isomers upon irradiation with visible and UV light. The fluorescence quenching is based on an electron transfer from the
donor to the acceptor units. The fluorescence photoswitching and non-destructive readout capability were demonstrated in solution
(an ensemble state) and at the single-molecule level. Femtosecond time-resolved transient and fluorescent lifetime measurements
confirmed that the fluorescence quenching is attributed to the intramolecular electron transfer.
’ INTRODUCTION
prepared fluorescent photochromic diarylethenes having a bis-
(phenylethynyl)anthracene or a perylenebisimide (PBI) unit.3-5
Similar molecules have also been prepared by Hell et al.12,13 In
these molecules, the fluorescence spectrum overlaps well with
the absorption spectrum of the closed-ring isomer of the DE unit,
while the absorption spectrum of the open-ring isomer locates
shorter than the fluorescence spectrum. Therefore, fluorescence
quenching does not take place in the open-ring isomer, while the
fluorescence is efficiently quenched by intramolecular energy
transfer when the DE unit converts from the open- to the closed-ring
isomers. Although the molecules are successfully used to demonstrate
photoswitching of fluorescence at the single-molecule level,3-5 un-
fortunately the fluorescence quenching inherently induces the cyclor-
eversion reaction of the DE unit. The excited energy of the fluorescent
unit is intramolecularly transferred to the closed-ring isomer and
induces the ring-opening reaction. This reaction leads to the destruc-
tion of information in optical memories. For application to write-many-
read-many (WMRM) devices, the fluorescence switching and readout
processes should be decoupled and independently controllable.
Fluorescent photochromic molecules, which have both photo-
chromic and fluorescent properties in a single molecule, have
attracted increasing interest because of their potential applications
in optical memories and as molecular switches and fluorescent
biological markers.1-18 A typical example of such molecules is a
dyad composed of photochromic and fluorescent dye units, which
modulates the fluorescence emission by the photochromic reaction.
Not only small organic molecules but also photoactivatable fluor-
escent proteins (PAFPs),19 such as “Dronpa”,20 are suggested as
fluorescent photochromic chromophores. Fluorescence photo-
switching of such molecules and proteins at the single-molecule
level has opened possibilities to develop ultra-high-density (Tbits/
inch2) optical memories3-5,20-22,27-29 and super-high-resolution
fluorescence imaging.13,14,25,26,30-35 PAFPs contribute signifi-
cantly to the development of high-resolution imaging.25,26,31-33
So far, we have developed various types of fluorescent
diarylethene (DE) derivatives for single-molecule optical
memories.3-5,9-11 In a series of our studies, two types of
fluorescence quenching mechanisms are employed, intramole-
cular energy transfer3-5 and electron transfer mechanisms.9-11
To photoswitch the fluorescence by the former mechanism, we
Received: November 29, 2010
Published: March 10, 2011
r
2011 American Chemical Society
4984
dx.doi.org/10.1021/ja110686t J. Am. Chem. Soc. 2011, 133, 4984–4990
|

Journal of the American Chemical Society
ARTICLE
Scheme 1. Molecular Structures of Dyad 1, Model DE 2, and Model PBI 3
To overcome the problem of destructible readout, several
methods or ideas have been proposed, such as readout using the
changes of the refractive index or IR absorption at the wavelengths
out of the switching unit,36-38 stabilized signals by means of acid
or base reactions,39 switching of the electron-transfer effi-
ciencies,40 the molecular polarity,18 or the supramolecular confor-
mational changes,41 and so on.42 Although some of them have the
potential to achieve non-destructive readout43 in solid ensemble
systems, no suitable molecular systems that can be applied at the
single-molecule level have been developed due to inherent limita-
tions of sensitivity, stability, and efficiency in their molecular
properties and/or difficulty of operation in the solid state.
To avoid such problems and achieve reversible and efficient
fluorescence photoswitching with non-destructive readout capabil-
ity, we employed photoinduced electron transfer as the photo-
switching mechanism.8,40 The mechanism requires changing the
oxidation/reduction potentials of the photoswitching DE unit.
When the oxidation or reduction potential differences are large
enough to induce electron transfer between the fluorescent unit and
one of the isomers of the DE unit, isomerization of the DE unit
induces the photoswitching of the fluorescence. The electron-
transfer mechanism allows molecular design in such a manner that
the absorption bands of both DE isomers are shorter than the
fluorescence spectrum of the fluorescent unit. The shorter absorp-
tion bands decouple the photochromic reaction from the fluores-
cence detection.
According to this concept, DE derivatives linked with fluorescent
PBI have been synthesized.9-11 Although the cycloreversion reac-
tion along with the energy-transfer quenching was successfully
avoided and efficient reversible fluorescence quenching via an
electron transfer was observed, we failed to demonstrate non-
destructive readout performance because of an unexpected photo-
cyclization through the triplet state.11 To prohibit the triplet route,
we shifted the absorption spectrum of the PBI unit to longer
wavelengths. When the S1 state of the PBI unit is lower than the T1
state of the DE unit, the triplet route is anticipated to be lost. After
several trials, we succeeded in designing a diarylethene-perylene-
bisimide (DE-PBI) dyad (1, Scheme 1), which has non-destructive
readout capability. Here we report on the fluorescence photoswitch-
ing and non-destructive fluorescence readout of 1 in solution as well
as at a single-molecule level.
’ RESULTS AND DISCUSSION
Molecular Design. Figure 1 shows the absorption and fluor-
escence spectra of component molecules [open- (2a) and
4985
dx.doi.org/10.1021/ja110686t |J. Am. Chem. Soc. 2011, 133, 4984–4990

Journal of the American Chemical Society
ARTICLE
Figure 3. Relative fluorescence quantum yields as a function of di-
electric constant for 1a (O) and 1b (b).
Figure 1. Absorption and fluorescence spectra of each component in
1,4-dioxane. Absorption spectra of the open-ring isomer of diarylethene
unit (black), the closed-ring isomer of diarylethene unit (blue), and the
PBI unit (red), and fluorescence spectrum of the PBI unit (green).
Dyad 1 was synthesized by procedures similar to those
reported in a previous paper.10,44-46 Detailed procedures of
the synthesis and compound data are described in the SI.
Fluorescence Photoswitching in Solution. Figure 2a,b
shows the absorption and fluorescence spectral changes upon
irradiation with visible (λ = 445 nm) and UV (365 nm) light in a
binary solution (1,4-dioxane-methanol 50/50) (ε = 17.4) of 1,
respectively. Upon visible (λ = 445 nm) light irradiation, the
absorption band around 360 nm gradually increases. From the
absorption spectrum of 1b isolated by HPLC, the conversion
from 1a to 1b in the photostationary state upon irradiation with
445 nm light was estimated to be 85%. Upon irradiation with UV
(365 nm) light, the absorption spectrum turns back to the initial
one (92%). Reversible fluorescence intensity change was also
observed along with the photochromic reaction. When the
solution reaches the photostationary state under irradiation with
445 nm light, the fluorescence intensity decreases to 14% of the
initial value. The fluorescence intensity recovers upon UV
(365 nm) light irradiation. On the other hand, in pure 1,4-
dioxane (ε = 2.21), no fluorescence photoswitching was ob-
served along with the photocyclization and photocycloreversion
reactions of the DE unit (see SI Figure S2). This solvent
dependence strongly suggests that the fluorescence quenching
is caused by intramolecular electron transfer.
In order to closely elucidate the quenching mechanism, the
relative fluorescence quantum yields of both isomers 1a and 1b
were measured as a function of the dielectric constants of the
solvents. Figure 3 shows the plots of the relative fluorescence
quantum yields (ΦF/ΦFM) of the isomer against the dielectric
constants (ε) at room temperature (SI Table S1 summarizes the
results). Binary mixtures of 1,4-dioxane and methanol were used
to change the solvent dielectric constant. As can be seen from
Figure 3, ΦF/ΦFM of 1a decreases slightly with increasing
dielectric constant. In contrast, ΦF/ΦFM of 1b decreases dra-
matically when the dielectric constant increases above 5. In a
binary polar solution (1,4-dioxane-methanol 40/60, ε = 20.46),
the fluorescence is strongly quenched to ΦF/ΦFM < 0.047, while
1b in pure 1,4-dioxane shows fluorescence intensity comparable
to that of 1a. No remarkable change in the absorption and
fluorescence spectra depending on the solvent polarity was
observed (see SI Table S2).
Table 1. Reduction and Oxidation Potentials of Model DE 2
and PBI 3, Spectroscopic Excited-State Energy E00, Center-to-
Center Distances Rcc between DE Moiety and PBI Moiety,
and Estimated Gibbs Free Energies ΔG°
E00
Ered
(2-/2) [V]a
Eox
(3þ/3) [V]b
Rcc
[Å]c
ΔG°
[kcal/mol]
compd
[eV]
1a
1b
2.25
2.25
-1.39
1.04
1.04
12.1
12.8
1.23
-0.97
-8.55
a See ref 10. b See SI. Center-to-center distances Rcc of both isomers
c
were estimated from the AM1 calculation in MOPAC.
Figure 2. Absorption and fluorescence spectral changes of 1 in a binary
solution (1,4-dioxane-methanol 50/50) upon irradiation with 445 and
365 nm light: open-ring isomer (1a) (solid line), the closed-ring isomer
(1b) (dashed line), and the photostationary state under irradiation with
445 nm light (dotted line).
closed-ring isomers (2b) of the photochromic DE unit and
fluorescent N,N0-bis(ethylpropyl)-1,7-bis[4-(1,1,3,3-tetramethyl-
butyl)phenoxy]perylene-3,4:9,10-tetracarboxylbisimide (3),
Scheme 1] in 1,4-dioxane. The fluorescence spectrum of 3 does
not overlap with the absorption spectra of 2a and 2b. Therefore,
singlet-singlet intramolecular energy transfer from 3 to 2a and
2b is avoided. The absorption maximum of the PBI unit was
observed at 537 nm, which is 15 nm longer than the maximum of
PBI without 4-(tetramethylbutyl)phenoxy substituents.11 The
small red-shift prohibits the triplet route.
In order to explore the feasibility of an intramolecular electron-
transfer photoswitching, energy gaps (ΔG°) for charge separa-
tion in the excited states of 1a and 1b in a given solvent were
estimated by applying the reduction and oxidation potentials of
model compounds 2 and 3 to the Rehm-Weller equation
(detailed parameters are shown in Table 1) (see Supporting
Information (SI)).28 The energy gaps (ΔG°) were calculated to
be 1.23 kcal/mol for the open-ring isomer (1a) and -8.55 kcal/mol
for the closed-ring isomer (1b) in dichloromethane (ε = 8.93).
From the energy values it is probable that electron transfer takes
place only for the closed-ring isomer (1b) in dichloromethane.
Fluorescence Quenching Dynamics. To directly elucidate
the fluorescence quenching mechanism, time-resolved fluores-
cence and transient absorption spectral measurements were
carried out. Figure 4a,b shows fluorescence decays of 1a in pure
1,4-dioxane and in a 1,4-dioxane-methanol (50/50) binary solu-
tion, respectively. Both decays are well reproduced by single-
exponential decay functions with time constants of 4.9 ( 0.1 ns
in 1,4-dioxane and 3.9 ( 0.1 ns in the binary solution. These time
constants are similar to those for 3 in both solutions (see SI),
4986
dx.doi.org/10.1021/ja110686t |J. Am. Chem. Soc. 2011, 133, 4984–4990

Journal of the American Chemical Society
ARTICLE
Figure 5. Spectra corresponding to 180 (O) and 30 ps (b) time
constants, obtained by analysis of the time profiles of transient absorp-
tion of 1b in a 1,4-dioxane-methanol (50/50) binary solution,
excited with a femtosecond 530 nm laser pulse. The bold solid line is
the SnrS1 absorption spectrum of PBI unit, measured for 1b in pure
1,4-dioxane.
Figure 4. Fluorescence decays of 1a and 1b in pure 1,4-dioxane (a,c)
and in a 1,4-dioxane-methanol (50/50) binary solution (b,d). Excita-
tion and monitoring wavelengths were 532 and 570 nm, respectively.
Instrumental response function (IRF) was 32 ps. Red lines are results
calculated for the analysis (see text).
indicating that no effective quenching takes place in the excited
PBI unit by the open-ring isomer in either solution. Figure 4c,d
shows the fluorescence decays of 1b in pure 1,4-dioxane and in
the binary solution, respectively. The decay of 1b in pure 1,4-
dioxane is reproduced by single-exponential decay with a time
constant of 4.8 ( 0.1 ns. This time constant is similar to the
fluorescence lifetime of 3 and 1a in 1,4-dioxane, showing that no
effective quenching of the PBI unit by the closed-ring isomer
occurs in pure 1,4-dioxane. On the other hand, the decay in the
binary solution shows rapid decay in the sub-nanosecond time
region, followed by a further decay with a much longer time
constant. The red line in Figure 4d is the curve calculated with a
double-exponential decay function with the faster time constant
of 200 ps and the slower one of 3.9 ns. Pre-exponential factors for
the faster and the slower time constants are respectively 0.93 and
0.07. The fast decay of 200 ( 10 ps indicates that the PBI unit in
the fluorescent state is effectively quenched by the closed-ring
isomer of the DE unit in the binary solution. On the other hand,
the slower one might be due to the residual trace of 1a in the
solution.
To clarify the mechanism of the quenching directly, we
employed transient absorption spectroscopy with a femtosecond
laser pulse at 530 nm as an excitation source. Time evolution of
transient absorption spectra of 1b in 1,4-dioxane was ascribed to
the deactivation of the S1 state of the PBI unit with a time
constant of ca. 5 ns, which is in agreement with the fluorescence
lifetime in 1,4-dioxane (see SI). This result indicates again that
the S1 state of the PBI unit is not affected by the closed-ring
isomer of the DE unit in pure 1,4-dioxane. On the other hand,
time evolution of transient absorption spectra of 1b in the binary
solution (1,4-dioxane-methanol 50/50) showed rapid decay in
a few hundreds of picoseconds time region (SI). The main time
constants of the time profiles were 30 and 180 ps, the latter of
which is in agreement with the fluorescence decay time constant
of 1b in the binary solution, as shown in Figure 4. The spectra
corresponding to these two time constants, which were obtained
by the analysis of time profiles with different monitoring
wavelengths (SI), are exhibited in Figure 5. The component
corresponding to the decay with 180 ps time constant is safely
Figure 6. Non-destructive fluorescence readout capability of both
open- and closed-ring isomers. Fluorescence intensities at 568 nm upon
excitation at 532 nm of 1a (O) and 1b (b) were plotted against
irradiation time. Both isomers had the same absorbance at 532 nm
(∼0.1). λex = 532 nm laser light (2.5 mW/cm2).
attributed to the S1 state of the PBI unit on the basis of the
spectral band shape and the absorption maxima.10 On the other
hand, the spectrum appearing with 180 ps and decaying with
30 ps time constants corresponds to the radical cation of PBI.10
These results clearly lead to the conclusion that the charge
separation reaction between the excited PBI unit and the closed-
ring isomer of DE takes place with a time constant of 180 ps,
followed by charge recombination with a 30 ps time constant.
Non-destructive Fluorescence Readout. Non-destructive
fluorescence readout capability is an indispensable property for
optical memories.8,18,36-42 As mentioned in the Introduction,
previous DE derivatives connected with the fluorescence unit lack
the non-destructive fluorescence readout capability owing to the
quenching by energy transfer3-5 and contribution of the triplet
channel in the photocyclization reaction.10,11 On the other hand, the
present dyad 1 gives the non-destructive performance. The small
red-shift of the absorption spectrum of the PBI unit avoids the triplet
channel. The fluorescence intensities at 568 nm of 1a and 1b stay
constant even when the binary solution (1,4-dioxane-methanol
50/50) of 1a and 1b is irradiated with 532 nm laser light (2.5 mW/
cm2) for 2 h, as shown in Figure 6. Absorption spectral changes of
both isomers were not observed even after 2 h. This result indicates
that the absorbed energy of the PBI unit does not induce the
isomerization of the DE unit. Non-destructive readout capability is
achieved in this dyad system.
Fluorescence Photoswitching at the Single-Molecule Le-
vel. Fluorescence photoswitching and non-destructive readout
were also observed at the single-molecule level. Samples were
4987
dx.doi.org/10.1021/ja110686t |J. Am. Chem. Soc. 2011, 133, 4984–4990

Journal of the American Chemical Society
ARTICLE
Figure 7. Single-molecule fluorescence photoswitching in a PMA thin film. (a-g) Wide-field fluorescence images of 1 embedded in a PMA thin film.
Conditions: excitation wavelength, 532 nm from a solid laser; UV (330-380 nm) light; visible (440-490 nm) light; N2 atmosphere; exposure time, 300
ms. (h) Time trace of fluorescence intensity of a single molecule of 1, represented as a white circle in panels a-g. Green, violet, and blue bars indicate
irradiation period with 532 nm laser light, UV light, and visible light, respectively.
prepared by spin-casting a toluene solution of 2 wt % poly-
(methyl acrylate) (PMA, Tg = 9 °C) or 1.2 wt % Zeonex (Tg ≈
130 °C) containing 1b (1.5 ꢀ 10-11 M) to prepare all molecules
in the photoreactive anti-parallel conformation. When closed-ring
isomers are converted to the open-ring isomers, all molecules are in
the anti-parallel conformation.1,4 PMA, a polar and soft polymer,
was selected as the host matrix in this study to observe the reversible
photoswitching, referring to the recent result reported by W€urthner
and Brouwer’s group.47 Zeonex, an apolar and rigid polymer, was
used to determine the environmental polarity dependence of the
fluorescence photoswitching behavior.
Figure 7a-g shows the wide-field fluorescence microscopy
images taken on the same area under irradiation with 532 nm as
the fluorescence excitation light source (SI Movie S1). Visible
(440-490 nm) light from a Hg lamp was used as the light source
to induce the photocyclization reaction, and UV (330-380 nm)
light from a Hg lamp was used as the light source to induce the
photocycloreversion reaction. Initially, the PMA film containing
1b shows a dark image under irradiation with only 532 nm laser
light (Figure 7a). At 16.8 s, when UV light is irradiated to the
sample for ca. 5 s, fluorescence bright spots corresponding to
individual 1a molecules appear (Figure 7b). After UV light
irradiation is stopped, the fluorescent state persists under irradia-
tion with 532 nm laser light. At 29.1 s, when visible light is
irradiated onto the sample, the fluorescence bright spots dis-
appear abruptly (Figure 7c). After stopping the visible light
irradiation, the dark state remains until the next irradiation with
UV light. These ON-OFF fluorescence switching behaviors can
be observed several times (Figure 7d-g).
Only one-step irreversible photobleaching was observed, as shown
in SI Figure S7 (see also SI Movie S2). Intramolecular electron-
transfer quenching of fluorescence requires polar environments. As
evidenced in solution experiments, the fluorescence switching takes
place only in polar environments.
Figure 7h shows the fluorescence intensity trajectory of one
molecule, which is indicated by the white circle in Figure 7a-g.
As can be seen in Figure 7h, one-step ON-OFF fluorescence
switching is clearly observed upon visible and UV light irradia-
tion. The fluorescence “OFF” state is initially observed under
irradiation with only 532 nm laser light. The fluorescence “OFF”
state changes to the fluorescence “ON” state in one step upon
UV light irradiation. After stopping UV light irradiation, the
fluorescence “ON” state remains stable during irradiation with
only 532 nm laser light. The fluorescence “ON” state turns back
to the fluorescence “OFF” state when visible light is irradiated
onto the sample. After stopping visible light irradiation, the dark
state remains until the next irradiation with UV light. The
fluorescence switching is safely ascribed to the photochromic
reaction of individual dyad 1. It is worth noting that both
fluorescence “ON” and “OFF” states remain stable during
excitation with 532 nm laser light. The result clearly indicates
that non-destructive fluorescence readout can be achieved even
at the single-molecule level.
’ CONCLUSION
A photoswitchable fluorescent DE-PBI dyad 1 was synthe-
sized, and its fluorescence photoswitching and non-destructive
fluorescence readout were studied in solution and at the single-
molecule level. The fluorescence quantum yield of 1b is strongly
dependent on solvent polarity and decreases with increasing
solvent dielectric constant. It is concluded from the time-
resolved fluorescent and transient absorption spectroscopies that
On the other hand, in the rigid apolar Zeonex polymer film
containing 1, no clear fluorescence switching was observed upon
irradiation with visible and UV light (SI Figure S7). The fluores-
cence intensity of an individual molecule embedded in the Zeonex
polymer film is constant upon visible and/or UV light irradiation.
4988
dx.doi.org/10.1021/ja110686t |J. Am. Chem. Soc. 2011, 133, 4984–4990

Journal of the American Chemical Society
ARTICLE
the fluorescence is quenched by the intramolecular electron-
transfer process. Non-destructive fluorescence readout upon
excitation with 532 nm laser light is confirmed for both 1a and
1b under excitation with 532 nm laser light in solution as well as
at the single-molecule level. The results indicate that the dyad
system is useful for ultra-high-density optical memories and
super-high-resolution fluorescence imaging.
8 MHz repetition rate was employed. A photomultiplier tube
(Hamamatsu Photonics, R3809U-50) with an amplifier (Hamamatsu
Photonics C5594) and a counting board (PicoQuanta, PicoHarp 300)
were used for signal detection. A monochromator was placed in front of
the photomultiplier tube. The instrumental response function was
estimated by the fwhm of the scattered light from a colloidal solution
for the excitation light pulse. In the present measurements, it was 32 ps.
For the evaluation of the analyzed result, we used a reduced χ2 value.50,51
Single-Molecule Fluorescence Detection. For single-mole-
cule measurements, molecules were sparsely distributed around 100 nm
thin films of Zeonex and poly(methyl acrylate) (PMA) on clean quartz
glass coverslips by spin-coating one drop of mixed solution of a ∼10-11
M the closed-ring isomer (1b) and 1.2 wt % Zeonex or 2 wt % PMA in
toluene at 4000 rpm. Single-molecule fluorescence imaging and fluor-
escence intensity trajectory collection were performed under a nitrogen
atmosphere, using an inverted fluorescence optical microscope (Nikon,
TEi) equipped with a 100ꢀ 1.45 NA oil-immersion objective (Nikon),
appropriate filters (Chroma and Semrock) to remove the excitation
light, and a back-illuminated electron-cooled CCD camera (iXon,
Andowr). Circularly polarized 532 nm light from a solid laser (Spectra
Physics, Excelsior 532) was used as the excitation light source.
’ EXPERIMENTAL SECTION
General. General chemicals were purchased from Tokyo Chemical
Industries, Wako Pure Chemicals, or Aldrich Chemical Co. and used
without further purification. Solvents used in photochemical measure-
ments were of spectroscopic grade and were purified by distillation
before use. 1H NMR spectra were recorded on an NMR spectrometer
(Bruker AVANCE 400, 400 MHz). Samples were dissolved in CDCl3
with tetramethylsilane as an internal standard. Mass spectra were
measured with a mass spectrometer (Shimadzu GCMS-QP5050A and
Applied Biosystems, Voyger). Cyclic voltammograms was measured by
using an ALS-600 electrochemical analyzer. Absorption and fluores-
cence spectra were measured with a Hitachi U-3500 absorption spectro-
photometer and a Hitachi F-2500 fluorescence spectrophotometer,
respectively. Fluorescence quantum yields were determined by using
monochromatic 500 nm as the excitation light source, a very dilute
solution, the absorbance of which is less than 0.02 at the excitation
wavelength for all solvents, and N,N0-bis(1-hexylheptyl)perylene-
3,4:9,10-tetracarboxylbisimide (Φf = 0.99, in dichloromethane) as the
reference. Photoirradiation was carried out using an USHIO 500 W
xenon lamp as the light sources. Monochromic light was obtained by
passing the light through a monochromator (Jobin-Yvon) or a band-pass
filter (Δλ1/2 = 15 nm).
’ ASSOCIATED CONTENT
S
Supporting Information. Synthetic procedures, charac-
b
terization data, electrochemical data, dielectric constant depen-
dence of photophysical properties, photochromic property in a
1,4-dioxane solution, time-resolved spectroscopic data, control
experiment of the single-molecule fluorescence switching, and
two movie files concerning the single-molecule fluorescence
detection in a PMA film and in a Zeonex film. This material is
available free of charge via the Internet at http://pubs.acs.org.
Transient Absorption Measurements. For the detection of
dynamic behaviors in the femtosecond to nanosecond time region, a dual
NOPA/OPA laser system was used for transient absorption measurements.
The details of the system were described elsewhere.48,49 Briefly, the output
of a femtosecond Ti:sapphire laser (Tsunami, Spectra-Physics) pumped by
the second harmonic generation of a continuous-wave Nd3þ:YVO4 laser
(Millennia Pro, Spectra-Physics) was regeneratively amplified with 1
kHz repetition rate (Spitfire, Spectra-Physics). The amplified pulse
(802 nm, 0.9 mJ/pulse energy, 85 fs fwhm, 1 kHz) was divided into
two pulses with the same energy (50%). One of these pulses was guided
into a non-collinear optical parametric amplifier (NOPA) system
(TOPAS-white, Light-Conversion). The NOPA output can cover the
wavelength region between 500 and 1000 nm, with 1-40 mW output
energy and ca. 20 fs fwhm. In the present work, the excitation wavelength
was set to 532 nm. The white light continuum, which was generated by
focusing the fundamental light at 802 nm into a 1 mm CaF2 window,
covers the wavelength region from 350 to 720 nm. Polarization of two
pulses was set at the magic angle for the entire measurement. The signal
and the reference pulses were respectively detected with multichannel
photodiode array systems (PMA-10, Hamamatsu), and the detected
signals were sent to a personal computer for further analysis. In order to
correct the transient absorption spectrum for the dispersion of the probe
light, we measured the optical Kerr effect of CCl4 and determined the
wavelength-dependent arrival times of the femtosecond white light at
the sample position. From the cross-correlation trace between the
NOPA output and the super-continuum at the sample position, the
response pulse duration was ca. 80 fs. A rotating sample cell with an
optical length of 2 mm was used, and the absorbance of the sample at the
excitation wavelength was set to ∼1.0.
’ AUTHOR INFORMATION
Corresponding Author
tuyoshi@es.hokudai.ac.jp; iriem@rikkyo.ac.jp
’ ACKNOWLEDGMENT
This work was partly supported by Grant-in-Aids for Scientific
Research on Priority Areas “New Frontiers in Photochromism
(471)” (No. 19050008) from the Ministry of Education, Culture,
Sports, Science, and Technology (MEXT) of Japanese Govern-
ment. T.F. acknowledges support from PREST, JST, and Grant-
in-Aids for Young Scientist (B) (No. 21750149).
’ REFERENCES
(1) Irie, M. Chem. Rev. 2000, 100, 1685.
(2) Giordano, L.; Jovin, T. M.; Irie, M.; Jares-Erijman, E. A. J. Am.
Chem. Soc. 2002, 124, 7481–7489.
(3) Irie, M.; Fukaminato, T.; Sasaki, T.; Tamai, N.; Kawai, T. Nature
2002, 420, 759.
(4) Fukaminato, T.; Sasaki, T.; Tamai, N.; Kawai, T.; Irie, M. J. Am.
Chem. Soc. 2004, 126, 14843.
(5) Fukaminato, T.; Umemoto, T.; Iwata, Y.; Yokojima, S.; Yoneya-
ma, M.; Nakamura, S.; Irie, M. J. Am. Chem. Soc. 2007, 129, 5932.
(6) Tian, H.; Wang, S. Chem. Commun. 2007, 781.
(7) Yun, C.; You, J.; Kim, J.; Huh, J.; Kim, E. J. Photochem. Photobiol.
C: Photochem. Rev. 2009, 10, 111.
(8) Berberich, M.; Krause, A. M.; Orlandi, M.; Scandola, F.;
W€urthner, F. Angew. Chem., Int. Ed. 2008, 47, 6616.
(9) Odo, Y.; Fukaminato, T.; Irie, M. Chem. Lett. 2007, 36, 240.
Fluorescence Lifetime Measurements. For the measurement
of the fluorescence time profiles, the time-correlated single-photon
counting (TCSPC) method using a picosecond Nd3þ:YAG laser
(DPM-1000&SBR-5080-FAP, Coherent, 532 nm, fwhm ca. 30 ps) with
4989
dx.doi.org/10.1021/ja110686t |J. Am. Chem. Soc. 2011, 133, 4984–4990

Journal of the American Chemical Society
ARTICLE
(10) Fukaminato, T.; Tanaka, M.; Doi, T.; Tamaoki, N.; Katayama,
T.; Mallick, A.; Ishibashi, Y.; Miyasaka, H.; Irie, M. Photochem. Photobiol.
Sci. 2010, 9, 181.
(45) Jeong, Y.-C.; Yang, S. I.; Kim, E.; Ahn, K.-H. Tetrahedoron 2006,
62, 5855.
(46) Fukaminato, T.; Tanaka, M.; Kuroki, L.; Irie, M. Chem. Com-
mun. 2008, 3924.
(11) Fukaminato, T.; Doi, T.; Tanaka, M.; Irie, M. J. Phys. Chem. C
2009, 113, 11623.
(12) Bossi, M.; Belov, V.; Polyakova, S.; Hell, S. W. Angew. Chem.,
Int. Ed. 2006, 45, 7462.
(13) F€olling, J.; Polyakova, S.; Belov, V.; van Blaaderen, A.; Bossi,
M. L.; Hell, S. W. Small 2008, 4, 134.
(14) Hu, D.; Tian, Z.; Wu, W.; Wan, W.; Li, A. D. Q. J. Am. Chem.
Soc. 2008, 130, 15279.
(15) Soh, N.; Yoshida, K.; Nakajima, H.; Nakano, K.; Imato, T.;
Fukaminato, T.; Irie, M. Chem. Commun. 2007, 5206.
(16) Zhu, L.; Wu, W.; Zhu, M.-Q.; Han, J. J.; Hurst, J. K.; Li, A. D. Q.
J. Am. Chem. Soc. 2007, 129, 3524.
(47) Siekierzycka, J. R.; Hippius, C.; W€urthner, F.; Williams, R. M.;
Brouwer, A. M. J. Am. Chem. Soc. 2010, 132, 1240.
(48) Miyasaka, H.; Murakami, M.; Okada, T.; Nagata, Y.; Itaya, A.;
Kobatake, S.; Irie, M. Chem. Phys. Lett. 2003, 371, 40.
(49) Ishibashi, Y.; Murakami, M.; Miyasaka, H.; Kobatake, S.; Irie,
M.; Yokoyama, Y. J. Phys. Chem. C 2007, 111, 2730.
(50) O’Connor, D. V.; Phillips, D. Time-Correlated Single-Photon
Counting; Academic Press: London, 1984.
(51) Boens, N.; et al. Anal. Chem. 2007, 79, 2137.
(17) Zou, Y.; Yi, T.; Xiao, S.; Li, F.; Li, C.; Gao, X.; Wu, J.; Yu, M.;
Huang, C. J. Am. Chem. Soc. 2008, 130, 15750.
(18) Liang, Y. C.; Dvornikov, A. S.; Rentzepis, P. M. Proc. Natl. Acad.
Sci. U.S.A. 2003, 100, 8109.
(19) Tsien, R. Y. Annu. Rev. Biochem. 1998, 67, 509.
(20) Ando, R.; Mizuno, H.; Miyawaki, A. Science 2004, 306, 1370.
(21) Dickson, R. M.; Cubitt, A. B.; Tsien, R. Y.; Moerner, W. E.
Nature 1997, 388, 355.
(22) Moerner, W. E. J. Phys. Chem. B 2002, 106, 910.
(23) Habuchi, S.; Ando, R.; Dedecker, P.; Verheijen, W.; Mizuno, H.;
Miyawaki, A.; Hofkens, J. Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 9511.
(24) Andresen, M.; Wahl, M. C.; Steal, A. C.; Grater, F.; Schafer,
L. V.; Trowitzsch, S.; Weber, G.; Eggeling, C.; Grubmuller, H.; Hell,
S. W.; Jakobs, S. Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 13070.
(25) Flors, C.; Hotta, J.-I.; Uji-i, H.; Dedecker, P.; Ando, R.; Mizuno,
H.; Miyawaki, A.; Hofkens, J. J. Am. Chem. Soc. 2007, 129, 13970.
(26) Dedecker, P.; Hotta, J.-I.; Flors, C.; Sliwa, M.; Uji-i, H.;
Roeffaers, M. B. J.; Ando, R.; Mizuno, H.; Miyawaki, A.; Hofkens, J. J.
Am. Chem. Soc. 2007, 129, 16132.
(27) An overview of optical data storages based on photoswitchable
fluorescent proteins: Adam, V.; Mizuno, H.; Grichine, A.; Hotta, J.-I.;
Yamagata, Y.; Moeyaert, B.; Nienhaus, G. U.; Miyawaki, A.; Bourgeois,
D.; Hofkens, J. J. Biotechnol. 2010, 149, 289.
(28) Holman, M. W.; Liu, R.; Zang, L.; Yan, P.; DiBenedetto, S. A.;
Bowers, R. D.; Adams, D. M. J. Am. Chem. Soc. 2004, 126, 16126.
(29) Heilemann, M.; Margeat, E.; Kasper, R.; Sauer, M.; Tinnefeld,
P. J. Am. Chem. Soc. 2005, 127, 3801.
(30) Hell, S. W.; Wichmann, J. Opt. Lett. 1994, 19, 780.
(31) Hell, S. W. Nat. Biotechnol. 2003, 21, 1347.
(32) Betzig, E.; Patterson, G. H.; Sougrat, R.; Lindwasser, O. W.;
Olenych, S.; Bonifacino, J. S.; Davidson, M. W.; Lippincott-Schwartz, J.;
Hess, H. F. Science 2006, 313, 1642.
(33) Hell, S. W. Science 2007, 316, 1153.
(34) Heilmann, M.; van de Linde, S.; Sch€uttpelz, M.; Kasper, R.;
Seefeldt, B.; Mukherjee, A.; Tinnefeld, P.; Sauer, M. Angew. Chem., Int.
Ed. 2008, 47, 6172.
(35) Heilemann, M.; Dedecker, P.; Hofkens, J.; Sauer, M. Laser
Photon. Rev. 2009, 3, 180.
(36) Kardinahl, T.; Franke, H. Appl. Phys. A: Mater. Sci. Process. 1995,
61, 23.
(37) Seibold, M.; Port, H. Chem. Phys. Lett. 1996, 252, 135.
(38) Uchida, K.; Saito, M.; Murakami, A.; Kobayashi, T.; Nakamura,
S.; Irie, M. Chem.—Eur. J. 2005, 11, 534.
(39) Yokoyama, Y.; Yamane, T.; Kurita, Y. J. J. Chem. Soc., Chem.
Commun. 1991, 1722.
(40) Myles, A. J.; Branda, N. R. J. Am. Chem. Soc. 2001, 123, 177.
(41) K€arnbratt, J.; Hammarson, M.; Li, S.; Anderson, H. L.; Albins-
son, B.; Andrꢀeasson, J. Angew. Chem., Int. Ed. 2010, 49, 1854.
(42) Lim, S.-J.; Seo, J.; Park, S. Y. J. Am. Chem. Soc. 2006, 128, 14542.
(43) “Non-destructive readout” means that readout light does not
influence the written information.
(44) Jeong, Y.-C.; Yang, S. I.; Ahn, K.-H.; Kim, E. Chem. Commun.
2005, 2503.
4990
dx.doi.org/10.1021/ja110686t |J. Am. Chem. Soc. 2011, 133, 4984–4990