An organic optical transistor operated under ambient conditions

Article abstract of DOI:10.1002/anie.201104193

It takes three: The key functionalities of an optical transistor, gating and amplification, are demonstrated exploiting the photophysical properties of a molecular triad (see picture). Two building blocks of the triad are highly efficient fluorophores, whereas the third building block is a photochromic molecule that can be reversibly interconverted between two bistable forms by light.

Full text of DOI:10.1002/anie.201104193

DOI: 10.1002/anie.201104193
Organic Optical Transistors
An Organic Optical Transistor Operated under Ambient Conditions**
Martti Pꢀrs, Christiane C. Hofmann, Katja Willinger, Peter Bauer, Mukundan Thelakkat, and
Jꢁrgen Kçhler*
Transistors are one of the most important inventions of the
last century and have paved the road for the triumphal
procession of electronic technologies into everyday life.[1]
However, driven by the continuous pressure for miniaturiza-
tion, transistor technology has arrived at a point at which
further increasing the speed of data processing has become
increasingly difficult to accomplish. Consequently, the
demand for faster control and trigger elements has stimulated
the search for devices that operate with photons instead of
electrons as signal carriers. The first steps towards this
direction have been taken already by the demonstration of
transistor functionalities that exploit nonlinear light–matter
interactions.[2,3] However, these systems featured either a low
gain, and/or had to be operated at cryogenic temperatures.
Herein we demonstrate that a molecular triad that consists of
fluorophores covalently linked to a photochromic molecule
can be operated as an optical transistor under ambient
conditions.
A promising class of organic materials for data processing
are photochromic systems, that is, molecules that can be
interconverted between two bistable forms by light. In the
past photochromic systems based on cis–trans isomerization
or photocyclization reactions have been studied[4–7] and apart
from fascinating results also severe problems that concern
insufficient discrimination between read-out and switching,
low fatigue resistance, and/or weak contrast between the
signals registered from the two states have been reported.[8–14]
Herein we present a photochromic system that consists of two
Figure 1. a) Schematic sketch of the molecular triad consisting of two
perylene bisimide (PBI) units that are covalently linked to a
dithienylcyclopentene (DCP; for details of the chemical
synthesis see the Supporting Information). The DCP unit
undergoes a photoinduced cyclization reaction from the open
to the closed form upon irradiation with light in the UV
spectral range (280–310 nm) and vice versa a photoinduced
ring-opening reaction upon irradiation with light in the visible
spectral range (500–650 nm; Figure 1a). The design strategy
of this triad facilitates the combination of the superior fatigue
perylene bisimide (PBI) units that are covalently linked to a dithienyl-
cyclopentene (DCP). Top: Closed form, bottom: open form. b) Absorp-
tion spectra of the PBI–DCP–PBI triad dissolved in toluene at a
concentration of 1.5ꢀ10ꢀ6 molLꢀ1 for the open (red) and the closed
(blue) form of the DCP unit. For comparison the dashed line shows
the absorption spectrum of pure PBI in the same solvent at a
concentration of 1.5ꢀ10ꢀ6 molLꢀ1
.
resistance of the diarylethene derivatives[15] with the superior
fluorescence properties of the perylene derivatives.[16–19] The
absorption spectrum of the triad is shown in Figure 1b. For
the open state of the DCP unit (Figure 1b, red line), the triad
features a strong absorption band around 300 nm and three
bands at 460 nm, 492 nm, and 529 nm, respectively. By
comparison with the absorption spectrum from a solution of
pure PBI, (Figure 1b, dashed line), the three bands can be
[*] Dr. M. Pꢁrs, Dr. C. C. Hofmann, Prof. Dr. J. Kçhler
Experimental Physics IV, University of Bayreuth
95440 Bayreuth (Germany)
E-mail: juergen.koehler@uni-bayreuth.de
K. Willinger, Dr. P. Bauer, Prof. Dr. M. Thelakkat
Applied Functional Polymers, University of Bayreuth
95440 Bayreuth (Germany)
!
assigned to characteristic vibronic bands of the S1 S0
transition of the PBI units. For the closed form of the DCP
unit, (Figure 1b, blue line), the absorption spectrum of the
triad features, in addition to the identical bands from the
perylene units, broad absorption bands between 300–400 nm
and between 500–700 nm.
[**] We thank S. Frank and Prof. Dr. S. Kꢂmmel for fruitful discussions
and gratefully acknowledge financial support from the German
Science Foundation (DFG; GRK-1640).
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201104193.
Angew. Chem. Int. Ed. 2011, 50, 11405 –11408
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
11405

Communications
For the exploitation of PBI–DCP–PBI as an optical gate, Figure 2. The large contrast can be understood as a modu-
the triads are immobilized in a low-molecular-weight poly- lation of the energy transfer efficiency between the PBI and
styrene matrix and are studied by fluorescence microscopy. In the DCP subunits as a function of the conformation of the
order to convert the DCP to the closed state, the sample was DCP (Figure 2, inset bottom right). In the closed conforma-
illuminated at 300 nm with an intensity of 30 mWcmꢀ2 tion, the lowest excited singlet state of the DCP is lower in
(exposure time 500 ms) and the fluorescence was probed by energy than the lowest excited singlet state of PBI allowing
illumination at 514 nm with an intensity of 23 mWcmꢀ2 for efficient PBI!DCP energy transfer and subsequent
(300 ms). This conversion/probe sequence was repeated m radiationless decay.[12] In the open conformation the ordering
times (here m = 50). Subsequently, the sample was exposed to of the DCP and PBI singlet energies is reversed and energy
an analogous illumination sequence (300 nm: 130 mWcmꢀ2, transfer from PBI to DCP becomes impossible. Hence the
250 ms; 514 nm: 96 mWcm2, 250 ms) to convert the DCP DCP acts as externally controlled trap for the PBI excitation
back to the open state. In the following such a full cycle of 2m energy.
conversion/probe sequences will be referred to as one switch-
Subsequently we investigated the achievable contrast as a
ing cycle. This illumination scheme allows us to modulate the function of the intensities of the conversion beams. In order to
fluorescence intensity from the PBI units. This process is do so, the parameters of the 635 nm illumination were kept
illustrated in the inset of Figure 2 top left that shows the fixed as given above and the intensity of the 300 nm radiation
relative change of the observed fluorescence intensity for was varied between 0.4 mWcmꢀ2 and 30 mWcmꢀ2. The result
converting the DCP unit from closed!open!closed con- of this experiment reveals that the ON–OFF contrast of the
formations. For brevity, in the following we will refer to the modulation of the fluorescence intensity increases for
intensity levels as ON and OFF (corresponding to the open increasing UV intensity and is shown by the blue data
and closed conformation of DCP), being aware that OFF points in Figure 2. Next the experiment was reversed and the
means low intensity rather than no intensity. For further parameters for the 300 nm illumination were fixed to the
quantification we determine the relative change of the initial values, while the intensity of the 635 nm beam was
fluorescence intensity before and after switching, that is, varied between 0 and 16.7 Wcmꢀ2 (Figure 2, red data points).
(IfinalꢀIinitial)/ION = DI/ION where Ifinal/initial refer to ION OFF
/
Also the OFF–ON contrast for the change of the fluorescence
depending on the chronological order of the switching intensity increases for an increase of the intensity of the
process. Obviously the magnitude of this ratio corresponds 635 nm beam. Yet this experiment reveals as well that an ON–
to the contrast of the modulation of the fluorescence intensity, OFF contrast of about 0.38 is achieved already without any
which amounts to 0.73 for the example shown in the inset of radiation at 635 nm. This observation reflects that also the
probing beam at 514 nm induces the ring-opening reaction in
the DCP units and converts the sample partly to the ON state.
Further we investigated the fatigue resistance of the
triads. For that experiment, the sample was illuminated at
635 nm with an intensity of 43 Wcmꢀ2 (exposure time 250 ms)
and the fluorescence was probed by illumination at 514 nm
with an intensity of 96 mWcmꢀ2 (250 ms). This conversion/
probe sequence was repeated 5 times for one switching cycle.
Subsequently, the sample was exposed to an analogous
illumination sequence (300 nm: 130 mWcmꢀ2
; 250 ms/
514 nm: 96 mWcmꢀ2; 250 ms) to convert the DCP back to
the closed state. The result of an experiment, in which the
triads have been irradiated for more than 3000 cycles is shown
in Figure 3. The main figure shows the contrast of the
fluorescence as a function of the number of switching cycles.
From its initial value of 0.83 it drops to about 0.38 after 3000
cycles. The two insets at the bottom of Figure 3 show the
modulation of the fluorescence intensity at the beginning and
at the end of the experiment reflecting a telegraph-like
alternation between a high- and a low-intensity level.
For these traces the abscissa has been converted to real
time, thus demonstrating the reversible conversion of the
triad for more than five hours. Analysis of the intensity of the
fluorescence in the ON and OFF state revealed a decrease of
the ON intensity as a function of time, indicating a slight
photobleaching of the PBI chromophores. Nevertheless, we
note that the end of the experiment was determined by the
experimentalist and not by the photo(in)stability of the
sample. Identifying the S1 and S0 states of the PBI units as
“source” and “drain” for the flow of fluorescence photons and
Figure 2. Fluorescence response of the sample as a function of the
conversion of the DCP unit. Main panel: change of the fluorescence
with respect to the ON state signal as a function of the intensities of
the conversion beams; left hand side (blue data points): variation of
the intensity of the 300 nm beam; right hand side (red data points):
variation of the intensity of the 635 nm beam. For more details see
text. Inset top: Example for the change of the normalized fluorescence
intensity for a closed!open!closed switching cycle. The bars on top
indicate the timing of the illumination periods in the various spectral
ranges. The changes from OFF!ON and ON!OFF, respectively, as
indicated by the arrows, correspond to the outermost data points of
the main panel. Inset bottom: Schematic representation of the
modulation of the fluorescence from the PBI units caused by energy
transfer from PBI to DCP unit in the closed conformation.
11406 www.angewandte.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 11405 –11408

is a basic requirement for encoding information. In the ON
state, the flow of fluorescence photons can be significantly
reduced by illumination with UV light. Reading out the
system, that is, detecting the fluorescence intensity under
continuous illumination at 514 nm, restores (in part) the ON
state because the ring-opening reaction of the DCP units can
also be induced at this wavelength which offers an interesting
option for nondestructive read-out. For possible applications
it will be challenging to develop triads with reduced crosstalk
between the read-out and switching channels. Another field
that might benefit from these findings is superresolution
imaging. At low temperatures the classical diffraction limit of
optical microscopy can be circumvented by taking advantage
of the slight spectral variations in the absorption spectra of
single molecules.[20–23] Under ambient conditions, newly
developed techniques take advantage of chromophores that
can be converted either stochastically (STORM)[24] or by
photoactivation (PALM)[25] to nonfluorescing states, which
has led to a growing demand for (photo)switchable chromo-
phores.[26,27]
We have demonstrated the key functionalities of an
optical transistor, gating and amplification, exploiting the
photophysical properties of a molecular triad. This is a crucial
step towards the realization of future devices that can be
operated with photons instead of electrons. It is shown that
the triad can be operated under ambient conditions for
several hours. Our next challenges are to extend these
experiments to individual triads thereby reducing the
amount of material that is required for storing one bit of
information to a single molecule, and to speed up the gating
process, which is limited in principle only by the fluorescence
lifetime of the PBI units (5 ns).
Figure 3. Contrast of the modulation of the fluorescence intensity as a
function of the number of switching cycles that each consist of 2ꢀ5
conversion/probe sequences. Each data point corresponds to the
average over 50 switching cycles. The illumination intensities (expo-
sure times) were 96 mWcmꢀ2 (250 ms) at 514 nm, 130 mWcmꢀ2
(250 ms) at 300 nm, and 43 Wcmꢀ2 (250 ms) at 635 nm. The kinks
after about 1000 and 2500 switching cycles reflect slight readjustments
of the fourth-harmonic generation due to drifts during the long-term
experiment. The line serves as a guide for the eye. Insets bottom:
Modulation of the fluorescence intensity at the beginning and the end
of the experiment. Inset top: Transistor analogy identifying the S1 and
S0 state of PBI as source and drain, the conversion beams as gate
voltage of different polarity, and the optical pumping as external
circuit, respectively.
the conversion beams as “gate”, the triads can be associated
with an optical transistor or an optical gate (see inset Figure 3,
top right). In this analogy the type of conversion beam,
300 nm or 635 nm, corresponds to the polarity of the gate
voltage of a conventional transistor, which commonly con- Experimental Section
Two 100 mL solutions of toluene, one containing 3 ꢀ 10ꢀ6 molLꢀ1 of
trols the flow of electrons between source and drain. Here it
controls the flow of photons emitted from the PBI units and
thereby the contrast ratio of the fluorescence. In order to
obtain the efficiency of the photoconversion processes in
terms of the number of photons per molecule required to
induce a conversion process it would be desirable to measure
the change of the absorption of the sample in real time.
Unfortunately this is impossible in our setup and therefore we
have to rely on the change of the fluorescence signal allowing
us at least to obtain an order of magnitude estimate for these
numbers. This approach is based on several approximations
(details are given in the Supporting Information). From these
calculations we find that the photocyclization reaction
(open!close) requires only about 10 photons per triad at
300 nm, whereas the ring-opening reaction (close!open)
needs about 100 photons at 514 nm and about 5000 photons at
635 nm per triad, respectively. These numbers should be read
as rough estimates, because of the approximations made in
the calculations. Nevertheless, the accuracy is sufficient to
testify that the flow of fluorescence photons from the PBI
units, which are emitted at a rate of 107–108 photonssꢀ1 per
molecule,[18] can be controlled by relatively few photons that
trigger the conversion processes. In terms of a logical gate we
can distinguish unambiguously between ON and OFF, which
the triads, and one containing polystyrene/toluene in a mass ratio of
1:3 were prepared. Equal volume fractions of these two solutions
were mixed. From the mixture we dropped 30 mL on a microscope
cover slip and let the solvent evaporate for about 5 min. Subsequently,
a second cover slip was put on top of the sample resulting in
polystyrene films with a thickness of about 10 mm and a triad
concentration of about 1.4 ꢀ 10ꢀ5 mollꢀ1. Then the samples were
mounted in a home-built inverted microscope. The optical setup
provides light beams with wavelengths of 300 nm and 635 nm for the
photoconversion of the DCP between its bistable forms, and a probe
beam of 514 nm to excite the fluorescence of the PBI. The Supporting
Information gives more details about the synthesis, sample prepara-
tion and the optical setup.
Received: June 17, 2011
Published online: October 5, 2011
Keywords: chromophores · fluorescence · optical transistors ·
.
photoswitches
[1] I. M. Ross, Proc. IEEE 1998, 86, 7 – 28.
[2] J. Hwang, M. Pototschnig, R. Lettow, G. Zumofen, A. Renn, S.
Gçtzinger, V. Sandoghdar, Nature 2009, 460, 76 – 80.
Angew. Chem. Int. Ed. 2011, 50, 11405 –11408
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 11407

Communications
[3] D. E. Chang, A. S. Sørensen, E. A. Demler, M. D. Lukin, Nat.
[16] T. Weil, T. Vosch, J. Hofkens, K. Peneva, K. Mꢁllen, Angew.
Chem. 2010, 122, 9252 – 9278; Angew. Chem. Int. Ed. 2010, 49,
9068 – 9093.
[17] E. Lang, F. Wꢁrthner, J. Kçhler, ChemPhysChem 2005, 6, 935 –
941.
[18] E. Lang et al., ChemPhysChem 2007, 8, 1487 – 1496.
[19] F. Wꢁrthner, Chem. Commun. 2004, 1564 – 1579.
[20] A. M. van Oijen, J. Kçhler, J. Schmidt, M. Mꢁller, G. J. Braken-
hoff, Chem. Phys. Lett. 1998, 292, 183 – 187.
[21] A. M. van Oijen, J. Kçhler, J. Schmidt, M. Mꢁller, G. J. Braken-
hoff, J. Opt. Soc. Am. A 1999, 16, 909 – 915.
[22] A. V. Naumov, A. A. Gorhelev, Y. G. Vainer, L. Kador, J.
Kçhler, Angew. Chem. 2009, 121, 9930 – 9934; Angew. Chem. Int.
Ed. 2009, 48, 9747 – 9750.
[23] A. V. Naumov, A. A. Gorshelev, Y. G. Vainer, L. Kador, J.
Kçhler, Phys. Chem. Chem. Phys. 2011, 13, 1734 – 1742.
[24] B. Huang, W. Wang, M. Bates, X. Zhuang, Science 2008, 319,
810 – 813.
[25] E. Betzig et al., Science 2006, 313, 1642 – 1645.
[26] M. Heilemann, P. Dedecker, J. Hofkens, M. Sauer, Laser
Photonics Rev. 2009, 3, 180 – 202.
[27] J. Vogelsang, T. Cordes, C. Forthmann, C. Steinauer, P.
Tinnefeld, Proc. Natl. Acad. Sci. USA 2009, 106, 8107 – 8112.
Phys. 2007, 3, 807 – 812.
[4] H. Dꢁrr, Angew. Chem. 1989, 101, 427 – 445; Angew. Chem. Int.
Ed. Engl. 1989, 28, 413 – 431.
[5] H. Bouas-Laurent, H. Dꢁrr, Pure Appl. Chem. 2001, 73, 639 –
665.
[6] Molecular Switches (Ed.: B. L. Feringa), Wiley-VCH, Weinheim,
2001.
[7] Y. Yokoyama, Chem. Rev. 2000, 100, 1717 – 1739.
[8] N. P. M. Huck, W. F. Jager, B. de Lange, B. L. Feringa, Science
1996, 273, 1686 – 1688.
[9] B. L. Feringa, W. F. Jager, B. de Lange, Tetrahedron 1993, 49,
8267 – 8310.
[10] T. Fukaminato et al., Photochem. Photobiol. Sci. 2010, 9, 181 –
187.
[11] M. Irie, T. Fukaminato, T. Sasaki, N. Tamai, T. Kawai, Nature
2002, 420, 759 – 760.
[12] T. Fukaminato et al., J. Am. Chem. Soc. 2007, 129, 5932 – 5938.
[13] A. Fernaꢂndez-Acebes, J.-M. Lehn, Chem. Eur. J. 1999, 5, 3285 –
3292.
[14] H. Hu et al., J. Photochem. Photobiol. A 2007, 189, 307 – 313.
[15] M. Irie, Chem. Rev. 2000, 100, 1685 – 1716.
11408 www.angewandte.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 11405 –11408