Multicomponent Molecularly Controlled Langmuir ... - ACS Publications

DOI: 10.1021/jp1009862@proofing. Copyright © American Chemical Society. * To whom correspondence should be addressed. E-mail: [email protected]...
0 downloads 0 Views 834KB Size
Download PDF
J. Phys. Chem. C 2010, 114, 8559–8567

8559

Multicomponent Molecularly Controlled Langmuir-Blodgett Systems for Organic Photovoltaic Applications Paola Vivo,*,† Tommi Vuorinen,‡ Vladimir Chukharev,† Antti Tolkki,† Kimmo Kaunisto,† Petri Ihalainen,§ Jouko Peltonen,§ and Helge Lemmetyinen† Department of Chemistry and Bioengineering, Tampere UniVersity of Technology, P.O. Box 541, FI-33101 Tampere, Finland, VTT Technical Research Centre of Finland, P.O. Box 1300, FI-33101 Tampere, Finland, and Center of Excellence for Functional Materials, Laboratory of Paper Coating and ConVerting, Åbo Akademi UniVersity, Porthaninkatu 3-5, FI-20500 Turku, Finland ReceiVed: February 2, 2010; ReVised Manuscript ReceiVed: March 23, 2010

The capability of Violanthrone-79 (V-79) and N,N′-bis(2,5-di-tert-butylphenyl)-3,4:9,l0-perylenebis(dicarboximide) (PDI) to act as electron acceptors, with respect to poly(3-hexylthiophene) (PHT) and to the photoinduced fullerene anion of porphyrin-fullerene (P-F) dyad, was demonstrated in Langmuir-Blodgett (LB) films by the time-resolved Maxwell displacement charge method. The introduction of V-79 and PDI in oriented multilayered films led to improved light harvesting and increased lifetime of the charge separation, enhancing the photocurrent generation measured using a three-electrode photoelectrochemical cell. The best solar cell performance was achieved for the multifunctional film structure where efficient PHT-phthalocyanine heterojunction (PHT|ZnPH4) was combined with the P-F|V-79 system. 1. Introduction The use of organic materials for solar energy conversion is gaining more and more attention, due to the substantial future prospects. Several promising results have been recently reported, as regards both the efficiency1–5 and the stability of organic photovoltaic devices.6–8 Vectorial photoinduced electron transfer (VPET) through multicomponent films is a key mechanism in the light-to-electricity conversion. The understanding of photophysical processes taking place in organic thin films is crucial for designing molecular devices with specific characteristics. Oriented donor-acceptor (D-A) pairs have been intensively used to perform primary VPET with high efficiency.9–12 In such dyads, donor and acceptor moieties are covalently linked together, ensuring an efficient intramolecular charge separation. After the charge-separated (CS) state is formed, an additional secondary acceptor layer next to the acceptor side of the dyad can increase the charge separation distance, making the recombination of charges less probable. Recently, the orientation of a D-A pair has been utilized for the first time as initiation for electron transfer (ET) through multicomponent films, and the interlayer interactions have been thoroughly studied by photochemical and photophysical methods.13–16 Triads, including an electron-donor layer added to the dyad system, were investigated and a multistep VPET was detected. This consists of a primary process taking place between the donor and the acceptor of the dyad, followed by a secondary ET, in which electrons are delivered by an additional donor layer (secondary donor) to the dyad. Electron transfer in thin Langmuir-Blodgett (LB) systems including a secondary acceptor (PPQ) with the dyad has been reported,14 even though that specific compound showed poor electron conducting properties for the photovoltaic applications. * To whom correspondence should be addressed. E-mail: paola.vivo@ tut.fi. Phone: +358 40 1981 128. Fax: +358 3 3115 2108. † Tampere University of Technology. ‡ VTT Technical Research Centre of Finland. § Åbo Akademi University.

The main goal of this work is to test Violanthrone-79 (V79) and N,N′-bis(2,5-di-tert-butylphenyl)-3,4:9,l0-perylenebis(dicarboximide) (PDI) as secondary electron acceptors. To the best of our knowledge, V-79 is studied in the LB films for the first time. Perylene derivatives are widely used in organic photovoltaics, and properties of PDI in LB films have been reported.17–19 Dutta et al. studied mixed LB films of PDI in a stearic acid matrix,17 whereas a detailed investigation on the dependence of the photophysical characteristics of the mixed LB films of PDI in different matrices on the various LB parameters was conducted by Hussain and co-workers.18 In addition, this paper presents the development of multicomponent systems, consisting of D-A pair, secondary donor, and secondary acceptor/antenna including the evaluation of generated photocurrent under illumination in a three-electrode photoelectrochemical cell. The results aim at presenting a novel strategy, which can lead to efficient photovoltaic devices. When an acceptable internal efficiency is achieved, the total absorbance will increase by stacking or folding the layers. In layered cells, separate active layers of donor(s) and acceptor(s) are deposited one on top of the other between the electrodes. Essential core of the cell is the D-A pair, consisting of a porphyrin-fullerene dyad.20 The approach is based on the use of the D-A molecules in oriented molecular films, in order to achieve a vectorial electron transfer zone as the primary step in the consecutive electron transfer processes. The primary excited photoactive molecules (electron donating moieties) and the electron accepting moieties are oriented anisotropically, e.g., by applying the Langmuir-Blodgett technique. Thus the photoinduced electron transfer takes place from the donor to the acceptor in one defined direction throughout the whole film structure. 2. Experimental Section 2.1. Materials. Molecular structures of the studied compounds are presented in Figure 1. Commercially available (Sigma Aldrich) V-79 (16,17-bis(octyloxy)anthra[9,1,2-cde-

10.1021/jp1009862  2010 American Chemical Society Published on Web 04/08/2010

8560

J. Phys. Chem. C, Vol. 114, No. 18, 2010

Vivo et al.

Figure 1. Molecular structures and acronyms of the compounds used.

]benzo[rst]pentaphene-5,10-dione, known as Violanthrone-79) and PDI were used without further purification. The porphyrinfullerene dyads (P-F and F-P) were synthesized as described elsewhere.20 Difference between the P-F and F-P compounds is the position of the hydrophilic end, located on the porphyrin or fullerene moiety, respectively. Highly regioregular (90-93%) poly(3-hexylthiophene) (PHT) was purchased from Rieke Metals. The phthalocyanine derivative (ZnPH4) was synthesized according to the literature21 and was used in PC measurements. Chloroform of analytical grade (Merck) was used as received, for preparing the solutions and as a spreading solvent. Octadecylamine (ODA) of 99% grade (Sigma Aldrich) was used as a matrix molecule for the LB film preparation. Stock solutions for all the compounds used, were prepared in chloroform with concentrations of ∼1 mg/mL. Spreading solutions were obtained by dilution from the stock solutions to concentration equal or less than 1 mM prior to the film preparation. 2.2. Sample Preparation and Characterization. All the studied compounds were deposited as LB films, if not otherwise specified. LB deposition systems (LB 5000, LB Minitrough, and LB Alternate) from KSV Instrument Ltd. (Finland) were utilized for recording the isotherms and for the film deposition. The subphase was phosphate buffer (0.5 mM Na2HPO4 and 0.1 mM NaH2PO4 in Milli-Q water) with pH ∼7. During the whole process, the temperature of the subphase was kept at 18 ( 0.5 °C by a thermostat. LB films of V-79 were prepared by spreading a 100% V-79 solution in chloroform (c ) 0.87 mM), while in the case of

PDI dye the mixing with ODA in ratio 40:60, respectively in a chloroform solution (c ) 1.26 mM) was required. PHT was deposited as 60 mol % film (per monomer unit) in ODA matrix, and P-F and F-P dyads as 10 mol % film in ODA.13 The orientation of the dyad in the film structures was determined by selecting P-F or F-P molecule, and by depositing the dyad layer in either upward (from water to air) or downward (from air to water) direction.28 Deposition pressures and rates for PHT, P-F, and F-P where chosen according to previous studies.15 For V-79 and PDI, the details about the film preparation are described in the Supporting Information (SI). After every upward deposition, the films were let to dry for 10 min, and their quality was checked through absorption measurements. Some samples were prepared for photocurrent measurements by spin-coating PHT and ZnPH4 (PHT from a 2 g/L chloroform solution, ZnPH4 from a 3.3 g/L methanol solution). For spectroscopic studies and profilometric characterization, films were deposited onto quartz substrates, cleaned by the standard procedure,22 and plasma etched for 15 min in a lowpressure nitrogen atmosphere with a plasma cleaner PDC-23G (Harrick). Samples for the photoelectrical measurements were fabricated on glass substrates coated with a transparent conducting indium-tin-oxide (ITO) anode, whose sheet resistance was ∼10 Ω/square. The ITO substrates were cleaned in an ultrasonic bath first in acetone and then in chloroform. Subsequently, a N2 plasma cleaning process was applied for 10 min prior to use. In samples for photovoltage measurements, in order to

LB Systems for Organic Photovoltaic Applications insulate the active layers from the two electrodes, 11 ODA layers were deposited onto ITO slides, and 20 ODA layers were added onto the active layers to finish the sample. In the following sections the term “single layer” is used to mean two LB monolayers of the same molecule, if not otherwise specified. Samples for photocurrent experiments were similar to those for PV measurements, but did not contain ODA layers between the electrodes and the active layers. Film thicknesses were determined by measuring the step height of LB layers with a Wyko NT1100 optical profilometer (Veeco) in the phase-shifting interferometry mode with 20× objective (horizontal resolution 0.75 µm). Surface was imaged with 50× objective (horizontal resolution 0.55 µm). Vertical resolution is 1 Å in thickness measurements according to manufacturer’s specifications. For V-79 and PDI, multilayers were prepared (6 and 5 LB, respectively) because of the unreliability of imaging one monolayer. Steps were produced by removing part of the film by cotton wetted with chloroform. Atomic force microscopy (AFM) measurements were performed with Nanoscope IIIa scanning probe microscope equipped with a JVT-scanner in MultiMode (Digital Instruments, Inc., USA). The steady-state absorption spectra for all the studied compounds in solutions and films were recorded with a Shimadzu UV-3600 spectrophotometer. 2.3. Differential Pulse Voltammetry. The differential pulse voltammetry (DPV) technique was used to estimate the oxidation and reduction potentials of V-79, PDI, and P-F using a Ag/AgCl wire as a pseudoreference electrode. 0.1 M tetra-nbutylammonium tetrafluoroborate (TBABF4) in benzonitrile (PhCN) was used as the supporting electrolyte. After measuring the background, a PhCN solution of the sample was added to the electrochemical cell. For each sample, the measurements were repeated after adding ferrocene/PhCN solution in order to fix the reference potential. The measurements were carried out under a nitrogen flow in two directions: toward the positive and the negative potential. The final values of oxidation and reduction potentials were calculated as an average of the two scans relative to a ferrocene standard as reference. 2.4. Photoelectrical Measurements. All of the measurements were carried out at room temperature, without protecting the active layers against air exposure. PhotoWoltage Measurements. The vectorial photoinduced electron transfer (VPET) in films was studied by the timeresolved Maxwell displacement charge (TRMDC) method, which provides a useful way to verify the vertical orientation of the donor and acceptor part in thin films.13–15,24–30 The measuring system of the TRMDC method is described elsewhere.28 Samples were excited by a 10 ns laser pulse at 430 nm from the second harmonic of a titanium-sapphire laser (adjustable in the wavelength range 340-450 nm) pumped by the second harmonic of a Q-switched Nd:YAG laser (532 nm). Samples were excited also at 532 and 690 nm. A typical sample structure for the photovoltage (PV) measurements was ITO|11-12 ODA|active layers|20 ODA|InGa liquid-metal drop electrode (InGa). Since the photoactive layers are isolated from the two electrodes (ITO and InGa), the measured PV signals are caused only by the electron movement inside the active layers, in a direction perpendicular to the plane of the films. The TRMDC signal amplitudes are proportional to the charge displacement and to the number of CS states. The signal direction is connected with the direction of the displacement: negative signals are obtained when electrons move from ITO toward the InGa

J. Phys. Chem. C, Vol. 114, No. 18, 2010 8561

Figure 2. Absorption spectra of the monolayers of the studied compounds.

electrode and vice versa. Moreover, the signal decay describes the recombination of the CS states in the layers. The built-in electric field31 due to the difference in the work functions of the electrodes (4.7 eV for ITO, 4.2 eV for InGa) can be canceled by applying an external bias voltage of -0.5 V. Photocurrent Measurements. The steady-state photocurrent (PC) measurements were performed in the electrochemical cell, where the top metal electrode was replaced by an aqueous electrolyte solution containing an electron acceptor. Thus the photogenerated electrons were transported from the topmost active layer of the film to the counter electrode by the acceptor in the electrolyte. The measurements were performed in a threeelectrode cell, where ITO was the working electrode (area )0.28 cm2), a Pt wire the counter electrode, and Ag/AgCl (sat. KCl) the reference electrode. 1,1′-dimethyl-4,4′-bipyridinium dichloride (methyl viologen, MV2+) was used as electron acceptor (concentration 5 mM) in an aqueous electrolytic solution. Supporting electrolyte was 100 mM KCl. The current-voltage (I-V) characteristics, the steady-state photocurrent measurements with stepwise excitation, and the incident photon-to-current conversion efficiency (IPCE) spectra were measured by using E5272A source/monitor unit (Agilent), by exciting at the absorption maxima (430, 540, 690, and/or 705 nm, depending on the sample). Excitation light source was a Xe-lamp coupled with a monochromator (bandwidth approximately 8 nm), used to select the excitation wavelengths. 3. Results and Discussion Details about the LB film preparation of 100 mol % V-79 and 40 mol % PDI in ODA are discussed in the SI (Figure S1). The formation of PHT (60 mol % in ODA), P-F (10 mol % in ODA) and F-P (10 mol % in ODA) LB monolayers have been previously described in detail elsewhere.28,32 3.1. Film Characterization. The average thicknesses for V-79 and PDI monolayers were estimated to be approximately 1 nm and for 2.5 nm, respectively, by profilometry measurements. In general, the film surfaces appeared smooth, although some micrometer sized aggregates can be seen at the surface. Step profiles and surface images are shown in the SI (Figures S2 and S3). AFM studies were carried out for V-79 and PDI monolayers, deposited on silicon substrates by the LB technique. A short description of the AFM results and figures are presented in the SI (Figures S4 and S5). Spectroscopic Measurements. Absorption spectra of the monolayers under investigation are presented in Figure 2. All the films were made by upward LB deposition. P-F and F-P have a typical narrow absorption band centered at 430 nm and PHT has a broad band with the maximum at 540 nm. The PDI absorption maximum overlaps with the absorption of PHT,

8562

J. Phys. Chem. C, Vol. 114, No. 18, 2010

CHART 1: Energy Levels Diagram of the Studied Compoundsa

a HOMO and LUMO values are estimated from DPV measurements (P-F, V-79, and PDI) or from literature (ITO,35 PHT,36 and methyl viologene37).

while V-79 has a characteristic band at longer wavelengths (maximum at 690 nm). The maximum fraction of light absorbed of any LB monolayer is roughly 2%. In order to obtain high absorption in a broad wavelength range, a multilayered structure is required. The multilayer usually consists of different components (donors, acceptors) having absorption maxima at different wavelengths. Thus, the spectrum of the resulting multilayered structure covers a wider region of the solar spectrum. In the studied film structures the absorption spectrum for a multilayer reveals that the overall absorption is a linear combination of the absorptions of each molecular layer, thus indicating a low spectroscopic interaction between the molecules of different layers in their ground state. Absorption at the Soret band of 100 mol % V-79 and 40 mol % PDI films are reported as a function of the number of deposited layers (SI, Figures S6a and S7a). This represents a straightforward test of the reproducibility of the transfer of a chromophore containing monolayer onto a solid substrate (quartz in this case). The absorption increases linearly with increasing film thickness, in accordance with Beer’s law, for both V-79 and PDI (Figures S6b and S7b). Differential Pulse Voltammetry. According the DPV experiments vs Ag/AgCl, the oxidation and reduction potentials were found for V-79 to be +0.47 and -1.21 eV and for PDI +1.23 and -0.99 eV, respectively. These values are referred to the ferrocene standard. For the P-F dyad the oxidation and reduction potentials were 0.59 and -1.06 eV, respectively. From the data obtained by DPV, it is possible to derive the HOMO and LUMO energies for the studied compounds, by taking into account the energy of the ferrocene reference (4.8 eV below the vacuum level34). The energy level diagram of the studied compounds is presented in Chart 1. For ITO,35 PHT,36 and methyl viologene (MV2+),37 the HOMO/LUMO energies are taken from the literature. The energy levels of V-79 and PDI reveal that both compounds could be used as electron acceptors for the fullerene anion of the dyad. PDI seems to be slightly more suitable than V-79, due to a better matching of its LUMO energy with that of the fullerene moiety of P-F. These results will be confirmed by the photovoltage measurements, as presented in the next section.

Vivo et al. 3.2. TRMDC Photovoltage Measurements. The acceptor characteristics of V-79 and PDI LB layers were verified with the time-resolved Maxwell displacement charge (TRMDC) technique, since the polarity of the TRMDC signals gives valuable information about the VPET direction. In this work, double and multilayer systems were constructed by combining PHT, P-F, F-P, V-79, and/or PDI monolayers. PHT is a well-known conducting polymer with electron donating properties. P-F and F-P are oriented D-A systems, for which the VPET process has been studied in detail earlier.28 Let us refer first to samples containing V-79. PV responses of PHT|V-79 and V-79|PHT were measured at different bias voltages (-0.5 V; 0 V; +0.5 V) at the excitation wavelengths of λexc ) 532 nm (where mainly PHT is excited) and λexc ) 690 nm (where V-79 has its absorption maximum). A response with negative polarity was observed for PHT|V-79, while a nearly mirror image of positive polarity signal was observed for the V-79|PHT structure (Figure 3). This result, obtained at the excitation wavelengths of both PHT and V-79, indicates electron transfer from PHT to V-79. The change of the external bias voltage had an effect on the PV amplitude, but polarity of the signals remained the same, indicating that the direction of the electron transfer is determined by the sample structure, but not by the built-in field caused by the difference in electrodes work function. The obtained signal polarities are supported by the results provided by DPV in solution for V-79 and the published HOMO/LUMO values for PHT36 (Chart 1). Similar experiments were carried out with P-F and V-79. According to the orientation of V-79 and P-F (or F-P) with respect to ITO (i.e., whether P-F or V-79 are deposited from the ITO side), two different polarities of the signals were obtained. When V-79 was on top of fullerene (P-F|V-79) the signal was negative. On the contrary, when it was deposited under the fullerene layer (V-79|F-P) the signal was positive. From these results one can conclude that V-79 is able to accept electrons from P-F dyad, when either P-F or V-79 is excited. At the excitation wavelength of 430 nm (where porphyrin absorbs), electrons transfer from the fullerene anion of dyad to V-79 because of the very fast and efficient ET taking place in P-F.28,38 When V-79 is photoexcited (at 690 nm), the source of electrons which are then transferred to V-79 cannot be identified without carrying out additional photochemical studies. The amplitudes of the PV signals are shown in Table 1 as sensitivity, voltage divided by the excitation energy density (V cm2/mJ). The advantage of this is the possibility to compare the amplitudes of signals recorded at different excitation densities. The photovoltage responses for single layers (PHT, V-79, or PDI) were also measured, for comparison. At the zero bias voltage, all unoriented single layers produced negative signals,31 but their amplitudes were very low, as shown in Table 1. The comparison of the amplitudes of single layers and bilayers gives interesting insights. Let us consider the bilayer structures producing positive PV signals, since negative signals can be also obtained from unoriented films, due to the built-in electric field.31 At the excitation wavelength of 532 nm, for the bilayer structure V-79|PHT, the PV amplitude is about 4 times higher with opposite sign than the sum of the amplitudes of individual PHT and V-79 layers. Similarly, when exciting the structure V-79|F-P at 430 nm, roughly 2 times higher amplitude with opposite sign was obtained, compared to the sum of F-P and V-79 layers. Positive signals indicate that the studied structures can generate PV against the internal field. If the negative signals are also considered (presented in Table 1 as well), even higher

LB Systems for Organic Photovoltaic Applications

J. Phys. Chem. C, Vol. 114, No. 18, 2010 8563

Figure 3. Photovoltage responses at three different external applied bias voltages (-0.5 V; 0 V; +0.5 V) for film structures PHT|V-79 (a) and V-79|PHT (b) at λexc ) 690 nm.

TABLE 1: Photovoltage Response Amplitudes (Reported As Sensitivities) for the Described Sample Structures, at Specific Excitation Wavelengthsa

active layers V-79 PDI PHT P-F V-79|PHT PDI|PHT V-79|F-P PDI|F-P PHT|V-79 PHT|PDI P-F|V-79 P-F|PDI V-79|PDI

PV amp. at 430 nm, V cm2/mJ

PV amp. at 532 nm, V cm2/mJ

PV amp. at 690 nm, V cm2/mJ

-0.01 -0.02

-0.03 -1.22 -2.15 -0.04 8.49 110

-0.07

-17.50 -133

-10

-0.17

0.28 11.34 -1.48 -44.60

Figure 4. Comparison between PHT|V-79 and PHT|PDI photovoltage responses in µs time scale. The PV amplitudes are reported as sensitivities (V m2/J), to ensure a fair comparison between curves recorded at different excitation energy densities.

-0.25 -21 -31.30

-43.80

a The full sample structure is ITO|ODA|active layers|ODA|InGa. The listed values include the amplifier gain factor. Sensitivity of P-F correlates very well with previously published data.28 F-P has a sensitivity value close to that of P-F (but opposite sign), as it can be estimated from ref 28.

amplitudes of the bilayer signals compared to the single layers were achieved. Since the PV amplitude is proportional to the charge separation distance, the increase in the signal amplitudes of the bilayers, compared to those of the single layers, can be attributed to the increased charge separation distance,26 because the total absorption is the same, and thus the number of CS states can be assumed unchanged. In the bilayer P-F|V-79, a two-step electron transfer process takes place. First, an intramolecular electron transfer occurs in the dyad from porphyrin to fullerene. This is followed by a secondary electron transfer from the fullerene radical anion in P-F to V-79. Such a multistep vectorial electron transfer process is supported by several other studies, where the secondary ET takes place mainly from a donor (e.g., PHT) to the porphyrin cation of P-F.13,15,29 Similar structures as those with the V-79 acceptor, were made with the PDI acceptor, by using the same donors (PHT, P-F). Similarly, the PV results (Table 1) support a vectorial electron transfer process from PHT or P-F to PDI. The PV amplitude increases roughly 40 times for PHT|PDI bilayer at 532 nm excitation and roughly 262 times for P-F|PDI bilayer at the 430 nm excitation, when compared to the sum of the amplitudes of individual layers. Hence, the most efficient electron transfer takes place between P-F and PDI. By comparing the increases of the amplitudes for PDI samples to those for V-79 samples, it can be concluded that PDI is a stronger electron acceptor than V-79. The addition of a V-79 layer on top of P-F increased the lifetime of the CS state as well, as shown in Figure S8a. After

2 µs, the photovoltage amplitude of P-F film was reduced by 3 times, while that of P-F|V-79 remains roughly the same. Similar behavior can be observed in samples containing P-F or F-P with PDI, with even longer lifetimes. The PV signals recorded in ms time scale for F-P and PDI|F-P systems are shown in Figure S8b. After 2 ms, the PV amplitude of the bilayer PDI|F-P was approximately 2 times lower than the initial value, while that of F-P was reduced by roughly 4 times. The increase in the lifetime observed in bilayer systems implies that the recombination of the separated charges in the triad system (P-F|V-79 or PDI|F-P) is retarded compared to the dyad alone. This is an additional proof of how the primary charge separation (taking place in P-F or F-P) is prolonged in time by the secondary electron transfer from the fullerene anion to the acceptor. Another difference between the V-79 and PDI samples relates to the shape of the photovoltage responses, which is in turn related to the lifetime of the CS state. The PV signals of all the presented structures were quite long-lived, in time scale from µs to ms. However, PDI structures had longer lifetimes than V-79 ones. In Figure 4, it can be noticed that for PHT|PDI the signal amplitude was reduced by 14% after 2 µs, while for PHT|V-79 by 34%. Analogous behavior was observed in different layer configurations containing V-79 and PDI. Interaction between V-79 and PDI. Possible photoinduced interactions between V-79 and PDI were analyzed as well. For this bilayered structures containing single layers of V-79 and PDI were prepared. In the sample configuration V-79|PDI, the PV response had negative sign, when either V-79 or PDI are excited (at 690 and 532 nm, respectively), Figure S9 in SI. The amplitudes of PV signals are reported in Table 1. The reverse structure PDI|V-79 generated a positive signal. Thus electrons transfer from V-79 to PDI in both bilayer configurations and PDI acts as electron-acceptor toward V-79. The result is as it was expected since the reduction energy of

8564

J. Phys. Chem. C, Vol. 114, No. 18, 2010

TABLE 2: Photovoltage Response Amplitudes (Reported As Sensitivities) for the Described Sample Structures, at Specific Excitation Wavelengthsa sample structure

λexc, nm

PV amp., V cm2/mJ

P-F| 2 V-79 P-F| 6 V-79 P-F| 12 V-79 P-F| 2 PDI P-F| 6 PDI P-F| 12 PDI PHT| 2 V-79 PHT| 6 V-79 PHT| 12 V-79 PHT|PDI PHT| 6 PDI PHT| 12 PDI

430 430 430 430 430 430 532 532 532 532 532 532

-0.001 -0.003 -0.02 -0.006 -0.02 -0.05 -0.05 -0.15 -0.25 -0.16 -0.72 -1.44

a The full sample structure is ITO|ODA|active layers|ODA|InGa. PV measurements were carried out without amplifier, thus the PV amplitudes are roughly 100 times smaller than previous series (see Table 1).

PDI is lower than that of V-79, and in the previously discussed PV results PDI behaves as a stronger acceptor than V-79. Influence of the Number of Layers on the PhotoWoltage. After demonstrating the electron transfer from P-F and PHT to V-79 and PDI, the influence of number of layers on the PV responses could be analyzed in order to understand if the charge separation in a multilayered sample (which contains several acceptor layers) is limited not only by the interface between donor and acceptor but also by the acceptor layers. The effect of V-79 and PDI multilayers on the TRMDC signal was studied by measuring the photovoltage with a slightly different device configuration than that described previously. The time resolution of the TRMDC system is limited by the bandwidth of the amplifier, so that processes faster than 20 ns cannot be detected. The strong intensities of the PV signals enable direct PV measurements without the amplifier and thus higher time resolution can be achieved. The samples were excited with a 10 ns laser pulse, and the responses recorded with a fast sampling oscilloscope (Tektronix TDS5032B). Without the amplifier, the PV amplitudes were roughly 100 times lower than those presented in the previous parts of this Section. The measured samples were constituted by a donor monolayer (P-F or PHT) and by 2, 6, and 12 V-79 or PDI layers. Sample structures and their PV signal intensities are presented in Table 2. When the samples containing P-F and PDI were excited at 430 nm, the PV amplitudes always increased with increasing the number of PDI layers. Intensity for the sample P-F|2 PDI was -0.006 V cm2/mJ. When raising the number of PDI layers from 2 to 6, intensity increased by 3-fold (being -0.02 V cm2/mJ for P-F|6 PDI). Finally, switching from 6 to 12 PDI layers, the intensity further increased by 2.5 times (being -0.05 V cm2/mJ for P-F|12 PDI). When considering the raise in PV amplitudes, it is worth mentioning that at 430 nm only the porphyrin moiety absorbs, and the PDI absorption can be neglected. Such kind of increase in the PV amplitudes with the number of PDI layers can be explained by considering the ET through PDI layers, which originates from the diffusion of charges inside PDI layers. The charge mobility of PDI is high enough at its ground state, so that the lack of photo excitation of the PDI moiety does not influence the diffusion. Similar reasoning can be applied to the P-F|V-79 structures, with the only difference that V-79 films are 100 mol %, and thus the ET through the layers is facilitated by the lack of ODA.

Vivo et al. The results related to the structures containing P-F|PDI or P-F|V-79 show that photovoltage signals are higher for P-F|PDI compared to P-F|V-79 samples. However, the electron transfer through the layers seems more affected by the number of layers in V-79 than in PDI samples. The signal of the sample containing 12 V-79 layers was 20 times higher than that of the sample with 2 V-79 layers, while for PDI the increase in sample thickness from 2 to 12 PDI layers led to an increase in signal intensity by only 8 times. This could be explained by the higher electron conductivity of V-79 films compared to PDI ones, which can arise also from the fact that V-79 films are 100 mol %, while PDI layers contain 60 mol % ODA matrix. In fact, ODA is an insulator which decreases the conductivity of the films. Moreover, ODA is an inert matrix, and thus it reduces the coverage of active molecules in the films. Hence, the probability of an interaction between two active sites will be lower with high ODA ratios. Samples consisting of different thicknesses of V-79 or PDI together with PHT as a donor were also measured. For V-79 samples 2, 6, and 12 V-79 layers and for PDI samples 1, 6, and 12 PDI layers were used. In PHT|PDI samples, when excited at 532 nm, both PHT and PDI absorbed light. This explains the high amplitudes of the PV signals. The absorbance increased roughly 4-fold when switching from PHT|PDI to PHT|6PDI, and 8-fold when switching from PHT|PDI to PHT|12 PDI. The increase in the PV amplitudes was not very significant, when the rise in the absorbances at the excitation wavelength is taken into account. This can be understood considering that only photons absorbed in PHT layer affect to the PV signals. In other words, PHT, but not PDI, is the key-component to be photoexcited in order to generate the ET. Different behavior was observed for PHT|V-79 samples, where the absorbance did not increase very much by increasing the number of V-79 layers. The PV amplitudes, however, were several times higher than those of the PHT|2V-79 sample. This behavior can be explained by supposing the ET to proceed through the V-79 layers. 3.3. Photocurrent Measurements. Additional aim of this work was to evaluate the photocurrent (PC) generated from structures containing the studied molecular films by means of a three-electrode electrochemical cell. Samples for PC measurements were similar as those used in the PV measurements, with the obvious differences due to the different characteristics of the two methods. For PC, the active layers were deposited directly on the ITO anode and no insulating ODA layers were deposited on top of the active layers. In Table 3, first series of samples is presented, having the following general structure: ITO|PHT|X|electrolyte, where X ) V-79, or PDI. The steady-state PC measurements were carried out with the sample excitation of 532 nm, which corresponds to the wavelength where PHT and PDI have their maximum absorptions, and at 690 nm where mainly V-79 absorbs (Please refer to SI Figures and to Table 3 for absorption). A reference sample containing only PHT was also made for comparison, but it showed very low PC. In PHT|V-79 systems, the generated photocurrent at the excitation wavelength of 532 nm increased from 3 nA/cm2, for PHT alone, to roughly 45 nA/cm2 in the case of 2 V-79 layers, and to 300 nA/cm2 in the case of 6 V-79 layers. Hence, when the number of V-79 increased, the cell performance enhanced by approximately 7 times. Samples containing PDI on the top of PHT showed a different behavior. The highest PC, and thus the highest efficiency, was obtained for 2 PDI layers, but switching from

LB Systems for Organic Photovoltaic Applications

J. Phys. Chem. C, Vol. 114, No. 18, 2010 8565

TABLE 3: Short-Circuit Current (Isc) from Steady-State PC Measurements, Absorption (A), External (ΦE) and Internal (ΦI) Quantum Yields, Relative Photovoltage (PV), and Photocurrent (PC) Amplitudes at Different Excitation Wavelengths (λexc)a active layers

λexc, nm

Isc, nA/cm2

A

Φ E, %

Φ I, %

PHT|2 V-79

532 690 532 690 532 532 532

44.6 17.0 299.2 107.5 189.2 148.9 3

0.028 0.016 0.031 0.040 0.033 0.054 0.008

0.02 0.01 0.14 0.06 0.09 0.07

0.34 0.28 2.00 0.72 1.20 0.59

PHT|6 V-79 PHT|2 PDI PHT|6 PDI PHT a

PV rel. ampl.

PC rel. ampl.

1

1

3

6.7

3b 14.4 0.12

4.2 3.3 0.07

The full sample structure is ITO|active layers|electrolyte. b PV relative amplitude of PHT|PDI film.

2 to 6 PDI layers, the PC was reduced. Moreover, a double PDI layer was more effective than a double V-79 layer. These results are coherent with the presented PV results, as shown in Table 3. In PV the most efficient CT was obtained for PDI samples, and the effect of multilayered V-79 on PHT was greater than that of the multilayered PDI. The external quantum yield (ΦE), i.e., the fraction of incident photons converted to electrical current, and the internal quantum yield (ΦI), which represents the fraction of absorbed photons converted to electrical current, are presented in Table 3 as derived from steady-state PC measurements. The highest internal quantum yield for this series was 2%, reached with the PHT|6 V-79 cell at the excitation wavelength of 532 nm. The bilayer system consisting of PHT and P-F is known to undergo multistep VPET with yield close to unity.13 Starting from such a bilayer structure (PHT|P-F), another set of samples was measured consisting of PHT, P-F and multilayered V-79 or PDI, as shown in Table 4. The steady-state PC experiments with stepwise excitation revealed that the bilayer PHT|P-F showed a better performance than the sample containing only a PHT single layer (see Tables 3 and 4). Furthermore, the addition of V-79 or PDI to the PHT|P-F system was beneficial to the PC generation. By increasing the number of V-79 layers (2, 6, and 12, prepared by LB technique), the current increased at each excitation wavelength, and consequently the quantum yields increased as well. The best results were obtained with 12 V-79 layers, which generated (at the dyad excitation wavelength) an internal quantum yield of roughly 7%. The fact that the best performance is achieved at 430 nm emphasizes the crucial role of excited P-F in these cells, which is responsible for starting the efficient multistep VPET through the layers. As for data reported in Table 3, also in this PDI-sample series (PHT|P-F|PDI), the increase in the number of acceptor layers leads to lower PC amplitudes. Efficiencies were dramatically reduced when the number of PDI layers was increased from 2 to 12 layers, as shown in Table 4. However, these results demonstrate again a better interaction between the dyad at the CS state (P+-F-) and PDI compared to V-79 (according to Chart 1, LUMOs of PDI and V-79 are very closed to each other and thus one can assume a similar interaction of such molecules with MV2+), as already pointed out for PV measurements. In fact, the internal quantum yield for the structure PHT|P-F|2PDI is 4 times higher (being 6.5%) than that of the corresponding V-79 sample (1.6%), as shown in Table 4. In Figure 5, the steady-state photocurrent responses of three samples are presented. When the sample was illuminated at 430 nm, the sample containing PHT|V-79 generated PC amplitude of roughly 100 nA/mW. By replacing the V-79 layers with P-F, the amplitude increased to 150 nA/mW, and when a trilayer system PHT|P-F|V-79 was developed, the PC raised up to 300 nA/mW. The observed increase in the PC amplitude with the

TABLE 4: Short-Circuit Current (Isc) from Steady-State PC Measurements, Absorption (A), and External (ΦE) and Internal (ΦI) Quantum Yields at Different Excitation Wavelengths (λexc)a active layers PHT|P-F PHT|P-F|2 V-79 PHT|P-F|6 V-79 PHT|P-F|12V-79 PHT|P-F|2 PDI PHT|P-F|6 PDI PHT|P-F|12 PDI a

λexc, nm

Isc, nA/cm2

A

Φ E, %

Φ I, %

430 540 690 430 540 690 430 540 690 430 540 690 430 540 430 540 430 540

60.9 50.0 0.9 109.8 177.1 51.8 357.7 466.2 188.9 434.4 570.4 365.0 442.5 452.8 299.0 326.0 63.0 137.0

0.017 0.025 0.001 0.023 0.029 0.016 0.036 0.024 0.039 0.02 0.03 0.04 0.020 0.027 0.029 0.046 0.035 0.077

0.04 0.02 0.001 0.08 0.09 0.03 0.25 0.21 0.11 0.31 0.25 0.21 0.29 0.21 0.18 0.12 0.04 0.06

1.18 0.42 0.19 1.55 1.34 0.85 3.14 3.99 1.30 6.85 4.30 2.41 6.54 3.54 2.77 1.15 0.45 0.37

The full sample structure is ITO|active layers|electrolyte.

Figure 5. PC responses from steady-state measurements with stepwise excitation for the structures of indicated configurations, when excited at 430 nm.

trilayer structure PHT|P-F|V-79 can be explained by the multistep VPET from porphyrin to the fullerene moiety of the dyad, followed by a secondary electron-transfer from PHT to the porphyrin cation and from the fullerene anion to V-79. The action spectra normalized to the excitation intensity (per Watt), measured for the same film structures, clarify even better this behavior, as depicted in Figure 6. It is also worth pointing out once again the important role played by the excitation of PHT. From the action spectrum of PHT|P-F|V-79 sample, it is evident that the biggest contribution to the PC generation takes place at PHT excitation wavelength. For the bilayer sample PHT|P-F, the clear contribution of the dyad at 430 nm was more important than V-79 contribution at higher wavelengths (690 nm) observed in PHT|V-79 sample.

8566

J. Phys. Chem. C, Vol. 114, No. 18, 2010

Figure 6. Action spectra for the indicated structures, reported as PC amplitude (in arbitrary units).

From this on, we show how it is possible to achieve more efficient photovoltaic devices, by using the essential information obtainable from the VPET studies. Multifunctional layered systems were developed, as presented by the structures in Table S1, SI. The idea for these configurations came from recently published PV and PC studies, where the VPET from PHT to ZnPH4 (see section 2.2) was demonstrated, and it was shown that the photovoltaic devices based on the heterojunction “PHT|ZnPH4” can be prepared.15 In the samples of Table S1, the next step was to add the system “P-F|V-79”, for which the CT was studied in section 3.2, on top of PHT|ZnPH4. Interaction between ZnPH4 and P-F was also studied in ref 15, where a very efficient electron donation from ZnPH4 to the porphyrin cation of P-F was demonstrated by photovoltage method. Finally, the ET from P-F to V-79 was demonstrated in the previous Section by TRMDC method. The addition of “P-F|V-79” onto “PHT|ZnPH4” was beneficial to the device performance. The sample with highest internal quantum yield (slightly over 7% at 430 nm) was the one having 6 V-79 layers on top of P-F. For such a configuration, the current-voltage (I-V) characteristics are shown at different excitation wavelengths in Figure 7a, and the photovoltaic parameters calculated from the I-V curves are presented in Table S1 of the SI. For this structure, the power conversion efficiency (η) was 0.05% at dyad excitation wavelength. The enhancement in the PC (and quantum yield) results from the vectorial multistep ET from PHT to V-79. In such a complex film structure, the lifetime of ET is prolonged as a consequence of the increased charge separation distance, and of the lateral diffusion of electrons in the acceptor layer. This retards the undesired back electron transfer process. Moreover, the better performance can be attributed to the enhanced coverage of the solar spectrum in the multilayered structure, to the polarization of neighboring layers that support the ET, and/or to additional energy transfer process which results in better light harvesting to obtain the ET state. Finally, in Figure 7b the action spectrum of the same sample is presented, together with the absorption spectrum. As it can

Vivo et al. be noticed, the two spectra showed similar peaks at corresponding wavelengths thus indicating that the photocurrent was induced by the active layers. The main peaks correspond to the absorption of dyad at 430 nm, PHT around 530 nm, ZnPH4, and V-79 at roughly 700 nm. As already mentioned, even though PDI showed a better interaction than V-79 with P-F in PV measurements, V-79 samples were more efficient than PDI samples in PC measurements. A possible reason for this behavior is the low charge mobility in PDI. In fact, from data in Tables 3 and 4, it is evident that a sample with 2 PDI layers is almost as efficient as a sample with 6 V-79 layers. This indicates a more efficient PC generation in PDI samples, which is limited in multilayered structures because of PDI resistivity. The aim of the present work was to discuss the advantage of combining several layers with different functions to enhance the PC generation in the cell, as can be predicted based on the results from the multistep VPET through the films. It is demonstrated that, if the absorbance and the lifetime of the charge separation increase (switching from simpler to more complex structures), the PC generation will be enhanced as expected. It is however not always possible to find a straightforward correlation between PV and PC measurements. In PV measurements, the active layers are insulated from the electrodes (ITO and InGa). In a solar cell additional interfaces need to be considered, like the two junctions between each electrode and the active layers. Moreover, to enhance the PC generation, the thickness of the active layers must be carefully optimized, in order to find the best compromise between the sample absorption and serial resistance. 4. Conclusions In this paper, V-79 and PDI were studied as molecularly controlled thin films deposited by the LB technique. After the film characterization by optical profilometer, AFM, absorption spectroscopy, and DPV, the capability of V-79 and PDI to act as electron acceptors toward a set of known compounds (PHT, P-F) was demonstrated by the TRMDC photovoltage method. In particular, it was interesting to show that both V-79 and PDI are able to accept electrons from the photoinduced fullerene anion of P-F dyad. Interaction between V-79 and PDI was also studied, revealing that PDI is able to accept electrons from V-79, thus being a stronger electron-acceptor. By developing multilayered film structures with increased number of acceptor layers, higher sensitivities of PV signals were obtained, even when considering the changes in absorption. This indicated that both PDI and V-79 are quite conductive and we can suppose that the ET, initiated by the P-F, proceeds through the acceptor layers.

Figure 7. I-V curves at different excitation wavelengths (a) and action spectrum with corresponding absorption spectrum (b) for structure PHT|ZnPH4|P-F|6 V-79.

LB Systems for Organic Photovoltaic Applications Additional aim of this work was to prove that V-79 and PDI improve the PC generation, when introduced in multilayered structures. The results from PC measurements in a threeelectrode photoelectrochemical cell are coherent with those from PV experiments. PDI seemed to interact more efficiently with P-F, even though the thickness of PDI films in the cell must be accurately optimized, due to its resistivity. More complex structures, with the introduction of an efficient heterojunction presented in literature15 (PHT|ZnPH4), were finally considered. The most efficient cell (ITO|PHT|ZnPH4|PF|6V-79|electrolyte) showed an internal quantum yield of 7.1%. The efficiencies of the prepared devices are in general quite low, compared to the most recent state-of-the-art devices. One main reason for low efficiencies could be the film degradation, due to the preparation and characterization of the cells in air. Moreover, the total absorption of the measured structures was very low, and no optimization of the active layers thickness was at all considered. Finally, P-F and PDI were deposited as 10 and 40 mol % films in ODA matrix, thus an insulating matrix was introduced in part of the active layers. In this study, the perylene derivative PDI showed more promising electron-acceptor characteristics than V-79. Thus, an additional work focusing on the improvement of the perylene derivative films quality and on the omission of any insulating matrix (ODA) from the active layers of the cells will be reported in the future. Acknowledgment. This work was supported by the Finnish Funding Agency for Technology and Innovation (TEKES), project “Organic Solar Cells”. Supporting Information Available: Details about V-79 and PDI LB deposition; profilometer surface images and thickness evaluation for V-79 and 40 mol % PDI; AFM images for 100 mol % V-79, 40 mol % PDI; absorption spectra of multilayers of V-79 (and PDI), and dependence of absorption on the number of layers; PV responses at different excitation energies and at different applied bias voltages; photovoltaic parameters as calculated from I-V curves. This material is available free of charge via the Internet at http://pubs.acs.org. References and Notes (1) Gra¨tzel, M. Inorg. Chem. 2005, 44, 6841. (2) Kim, J. Y.; Lee, K.; Coates, N. E.; Moses, D.; Nguyen, T.; Dante, M.; Heeger, A. J. Science 2007, 317, 222. (3) Bruder, I.; Karlsson, M.; Eickemeyer, F.; Hwang, J.; Erk, P.; Hagfeldt, A.; Weis, J.; Pschirer, N. Sol. Energy Mater. Sol. Cells 2009, 93, 1896. (4) Bai, Y.; Cao, Y.; Zhang, J.; Wang, M.; Li, R.; Wang, P.; Zakeeruddin, S. M.; Gra¨tzel, M. Nat. Mater. 2008, 7, 626. (5) Rand, B. P.; Genoe, J.; Heremans, P.; Poortmans, J. Prog. PhotoVoltaics: Res. Appl. 2007, 15, 659. (6) Krebs, F. C. Sol. Energy Mater. Sol. Cells 2008, 92, 715. (7) Hauch, J. A.; Schilinsky, P.; Choulis, S. A.; Childers, R.; Biele, M.; Brabec, C. J. Sol. Energy Mater. Sol. Cells 2008, 92, 727.

J. Phys. Chem. C, Vol. 114, No. 18, 2010 8567 (8) Franke, R.; Maennig, B.; Petrich, A.; Pfeiffer, M. Sol. Energy Mater. Sol. Cells 2008, 92, 732. (9) Kuciauskas, D.; Lin, S.; Seely, G. R.; Moore, A. L.; Gust, D.; Drovetskaya, T.; Reed, C. A.; Boyd, P. D. W. J. Phys. Chem. 1996, 100, 15926. (10) Imahori, H.; Fukuzumi, S. AdV. Funct. Mater. 2004, 14, 525. (11) Kesti, T. J.; Tkachenko, N. V.; Vehmanen, V.; Yamada, H.; Imahori, H.; Fukuzumi, S.; Lemmetyinen, H. J. Am. Chem. Soc. 2002, 124, 8067. (12) D’Souza, F.; Chitta, R.; Gadde, S.; McCarty, A. L.; Karr, P. A.; Zandler, M. E.; Sandanayaka, A. S. D.; Araki, Y.; Ito, O. J. Phys. Chem. B 2006, 110, 5905. (13) Vuorinen, T.; Kaunisto, K.; Tkachenko, N. V.; Efimov, A.; Lemmetyinen, H. J. Photoch. and Photob. A 2006, 178, 185. (14) Kaunisto, K.; Vuorinen, T.; Vahasalo, H.; Chukharev, V.; Tkachenko, N. V.; Tolkki, A.; Lehtivuori, H.; Lemmetyinen, H. J. Phys. Chem. C 2008, 112, 10256. (15) Kaunisto, K.; Vahasalo, H.; Chukharev, V.; Tkachenko, N. V.; Vivo, P.; Niemi, M.; Tolkki, A.; Efimov, A.; Lemmetyinen, H. Thin Solid Films 2009, 517, 2988. (16) Kaunisto, K.; Chukharev, V.; Tkachenko, N. V.; Efimov, A.; Lemmetyinen, H. J. Phys. Chem. C 2009, 113, 3819. (17) Dutta, A. K.; Kamada, K.; Ohta, K. Langmuir 1996, 12, 4158. (18) Hussain, S. A.; Paul, P. K.; Bhattacharjee, D. J. Phys. Chem. Solids 2006, 67, 2542. (19) Lehtivuori, H.; Kumpulainen, T.; Hietala, M.; Efimov, A.; Lemmetyinen, H.; Kira, A.; Imahori, H.; Tkachenko, N. V. J. Phys. Chem. C 2009, 113, 1984. (20) Efimov, A.; Vainiotalo, P.; Tkachenko, N. V.; Lemmetyinen, H. J. Porphyrins Phthalocyanines 2003, 7, 610. (21) Isosomppi, M.; Tkachenko, N. V.; Efimov, A.; Vahasalo, H.; Jukola, J.; Vainiotalo, P.; Lemmetyinen, H. Chem. Phys. Lett. 2006, 430, 36. (22) Roberts, G. Langmuir-Blodgett Films; Plenum Press: New York, 1990. (23) Ikonen, M.; Sharonov, A.; Tkachenko, N. V.; Lemmetyinen, H. AdV. Mater. Opt. Electron. 1993, 2, 115. (24) Tkachenko, N. V.; Tauber, A. Y.; Hynninen, P. H.; Sharonov, A. Y.; Lemmetyinen, H. J. Phys. Chem. A 1999, 103, 3657. (25) Tkachenko, N. V.; Vuorimaa, E.; Kesti, T.; Alekseev, A. S.; Tauber, A. Y.; Hynninen, P. H.; Lemmetyinen, H. J. Phys. Chem. B 2000, 104, 6371. (26) Tkachenko, N. V.; Vehmanen, V.; Efimov, A.; Alekseev, A. S.; Lemmetyinen, H. J. Porphyrins Phthalocyanines 2003, 7, 255. (27) Chukharev, V.; Vuorinen, T.; Efimov, A.; Tkachenko, N. V.; Kimura, M.; Fukuzumi, S.; Imahori, H.; Lemmetyinen, H. Langmuir 2005, 21, 6385. (28) Vuorinen, T.; Kaunisto, K.; Tkachenko, N. V.; Efimov, A.; Alekseev, A. S.; Hosomizu, K.; Imahori, H.; Lemmetyinen, H. Langmuir 2005, 21, 5383. (29) Kotiaho, A.; Lahtinen, R. M.; Tkachenko, N. V.; Efimov, A.; Kira, A.; Imahori, H.; Lemmetyinen, H. Langmuir 2007, 23, 13117. (30) Vuorinen, T.; Kaunisto, K.; Chukharev, V.; Tkachenko, N. V.; Efimov, A.; Lemmetyinen, H. J. Phys. Chem. B 2006, 110, 19515. (31) Tkachenko, N. V.; Hynninen, P. H.; Lemmetyinen, H. Chem. Phys. Lett. 1996, 261, 234. (32) Alekseev, A. S.; Tkachenko, N. V.; Tauber, A. Y.; Hynninen, P. H.; Osterbacka, R.; Stubb, H.; Lemmetyinen, H. Chem. Phys. 2002, 275, 243. (33) Lee, Y.-L. Langmuir 1999, 15, 1796. (34) Pommerehne, J.; Vestweber, H.; Guss, W.; Mahrt, R. F.; Ba¨ssler, H.; Porsch, M.; Daub, J. AdV. Mater. 1995, 7, 551. (35) Manoi, A. G.; Narayan, K. S. Opt. Mater. 2002, 21, 417. (36) Yun, J.-J.; Jung, H.-S.; Kim, S.-H.; Han, E.-M.; Vaithianathan, V.; Jenekhe, S. A. Appl. Phys. Lett. 2005, 87, 123102. (37) Imahori, H.; Norieda, H.; Yamada, H.; Nishimura, Y.; Yamazaki, I.; Sakata, Y.; Fukuzumi, S. J. Am. Chem. Soc. 2001, 123, 100. (38) Chukharev, V.; Tkachenko, N. V.; Efimov, A.; Guldi, D. M.; Hirsch, A.; Scheloske, M.; Lemmetyinen, H. J. Phys. Chem. B 2004, 108, 16377.

JP1009862