Intrachain Photoluminescence Dynamics of MEHPPV in the Solid State

to observe the intrinsic intrachain PL property of MEH-PPV in the solid state. The steady-state 0-0 PL band of the blend sample shows a blue-shift of ...
1 downloads 0 Views 92KB Size
Download PDF
J. Phys. Chem. B 2007, 111, 12389-12394

12389

Intrachain Photoluminescence Dynamics of MEH-PPV in the Solid State Katsuichi Kanemoto,* Yoshitaka Imanaka, Ichiro Akai, Mitsuru Sugisaki, Hideki Hashimoto, and Tsutomu Karasawa Department of Physics, Graduate Scholl of Science, Osaka City UniVersity, 3-3-138 Sugimoto, Sumiyoshi-ku, Osaka 558-8585, Japan ReceiVed: May 7, 2007; In Final Form: August 23, 2007

The photoluminescence (PL) dynamics of poly[2-methoxy-5-(2′-ethyl-hexyloxy)-1,4-phenylene vinylene] (MEH-PPV) blended in host polymer (polypropylene, PP) matrix as well as that in the neat film has been studied. The concentration of MEH-PPV in the PP blend is designed to be fairly low (0.01 wt %) in order to observe the intrinsic intrachain PL property of MEH-PPV in the solid state. The steady-state 0-0 PL band of the blend sample shows a blue-shift of 0.12 eV with respect to that of the neat film of MEH-PPV. The PL-excitation (PLE) spectra of the blend sample exhibit definite vibronic structure, and hence we can determine the magnitude of the Stokes shift as 0.06 eV. The blend sample shows a single-exponential PL decay at 4 K with a time constant of 850 ps. We emphasize that this single-exponential-type PL decay is an intrinsic property of the intrachain PL species. Time-resolved PL measurements confirm dynamical red-shift of the PL band in the neat film, whereas this trend is not found in the case of the PP blend. These observations indicate that the energy transfer between finite segments, which can cause exciton migration, is much less efficient within the isolated MEH-PPV polymer chain compared to the case of the interchain transfer. The time-resolved measurements further demonstrate that the Stokes shift identified in the blend sample takes place at the early stage within 50 ps following photoexcitation. We attribute this Stokes shift to the rapid increase of the planarity of the MEH-PPV chain caused by the torsion of some constituent phenyl rings following photoexcitation. Finally, based on an argument on the different magnitudes of Stokes shift between the blend sample and the neat film, we conclude that the PL of MEH-PPV in the neat film predominantly occurs at the site of interchain excitations via the interchain migration of excitons.

I. Introduction Conjugated polymers have been of great interest owing to their potential application to novel light-emitting devices.1 They have been studied vigorously toward realistic applications to display technologies comprised of organic materials.1 Among those polymers, poly(p-phenylene vinylene) (PPV) and its derivatives are the most widely studied luminescent polymers. Within the class of PPVs, poly[2-methoxy-5-(2′-ethyl-hexyloxy)-1,4-phenylene vinylene] (MEH-PPV) is one of the most promising candidates due to its advantages in solubility for common organic solvents as well as in good performance of charge injection.2,3 The luminescence process of MEH-PPV has been extensively studied by using spectroscopy on the basis of photoluminescence (PL).4 However, the interpretation of relaxation processes leading to PL is still under debate due to the coexistence of several competing processes such as lattice or vibrational relaxation that occurs at local sites, exciton migration5-8 and exciton-exciton annihilation9,10 that occur in the intrachain or interchain relaxation process, and so on. Among them the coexistence of the intrachain and interchain processes makes the observed phenomena very complicated. Therefore, the first step to unveil these competing relaxation processes should be to distinguish the intrachain and interchain processes. Indeed, the PL dynamics of solution (diluted) samples of PPV derivatives and that of neat (undiluted) films have been compared for this purpose.10-19 In this present work, we focus * Corresponding author. E-mail: [email protected]. Fax : +816-6605-2522.

on a system where a solid matrix is adopted for dilution, because it can be regarded as an ideal solid dilution system for neat film samples. A number of researches were reported on the PL properties of the PPV derivatives diluted in solid matrixes based on a variety of host polymer materials such as polystyrene (PS),15,19-22 polycarbonate (PC),13,23 poly(methyl methacrylate) (PMMA),16,20 and polyethylene (PE).11,12,24 In these references, blueshifts of the steady-state PL spectra of the diluted samples with respect to those in the neat films were observed. This phenomenon can be commonly regarded as a feature in the PL resulting from intrachain excitations. In contrast, the PL dynamics of the PPV derivatives in the solid matrix has not been well understood. So far, single-exponential-type11,16 or non-singleexponential-type11,13,15,20 behavior of the PL decay has been observed. In this paper, we address the PL dynamics of MEHPPV diluted in a solid matrix in order to observe intrinsic intrachain PL processes. For the observation of the intrachain PL processes, the unnecessary interaction between the guest MEH-PPV and a host material should be minimized. For this purpose we selected polypropylene (PP) as a host material. The PP matrix, consisting of saturated hydrocarbons, is ideal for π-conjugated molecules to reduce unnecessary intermolecular interaction with a support matrix because its hydrocarbons are expected to have negligible interaction with the π-electron backbone in the conjugated polymer chain of MEH-PPV. As one of the expected difficulties in polymer blend samples, it is generally accepted that polymer blends can easily result in phase separation into each polymer phase due to low mixing

10.1021/jp073492b CCC: $37.00 © 2007 American Chemical Society Published on Web 10/10/2007

12390 J. Phys. Chem. B, Vol. 111, No. 43, 2007 entropy.20,25,26 Hence, to increase portions of isolated polymer chains in the solid matrix, the concentration of a guest polymer should be designed as small as possible. In fact, we have recently concluded that observations of PL signals resulting from intrachain photoexcitations are achieved by adopting at least a 1000-fold dilution (per weight) for the PP blend of head-to-tail coupled regio-regular (RR) poly(3-octylthiophene) (P3OT) that has a self-assembling property.27 In this present work, the MEH-PPV/PP blend with a guest concentration of 0.01 wt % is thus adopted in order to realize the observation of the intrachain PL dynamics of the MEH-PPV. It is indeed shown that the choice of this extreme dilution condition enables us to observe some intrachain PL properties in both time-resolved and steady-state measurements. We can therefore make an access to clarify the photophysics leading to PL of the intrachain excitations. Finally, on the basis of the comparison of PL features of isolated MEH-PPV with those of its neat film, the PL dynamics raised from the interchain excitations is distinguishably discussed.

Kanemoto et al.

Figure 1. Absorption spectra at room temperature of poly[2-methoxy5-(2′-ethyl-hexyloxy)-1,4-phenylene vinylene] (MEH-PPV) in chlorobenzene solution, in the neat spin-cast film, and in the polypropylene (PP) blend. The dotted vertical line in the figure indicates a position of the shoulder in the blend sample.

III. Results II. Experimental Section MEH-PPV was purchased from Aldrich and used as received. Dilution of MEH-PPV was carried out using amorphous PP (Aldrich) as a host by dissolving both polymers in xylene with a guest concentration of 0.01 wt %. The details of the procedures were described previously.25 The neat film sample of MEH-PPV was prepared by spin-coating its chlorobenzene solution (4 mg/mL). Absorption spectra were recorded with a Shimazu UV-2400 spectrophotometer. The MEH-PPV/PP blend was semitransparent in the range of absorption measurements due to scattering by PP particles; this is an intrinsic property of PP films. Therefore, the control spectrum of PP was measured and subtracted from the absorption spectra of the blend sample. Steady-state PL measurements were performed using an Ar+ laser (488 nm, Lexell) as a light source (3 mW). PL-excitation (PLE) measurements were performed using a tungsten lamp as a light source attached to a monochromator with a double grating (Spex, 270) with a spectral resolution of 3.6 nm. PL and PLE spectra were corrected for wavelength sensitivity by comparison with a standard tungsten lamp calibrated with NBS (no. EPT-1285). Absorbance of the blend and the neat film samples was designed to be less than 0.25 at most, and hence, self-absorption was assumed to have a negligible influence on the PL and PLE data. The laser system used for the time-resolved PL measurements was composed of a Kerr lens mode-locked Ti:sapphire oscillator pumped by an Nd:YVO4 laser and a Ti:sapphire regenerative amplifier pumped by a Q-switched Nd:YLF laser (Spectra Physics, Hurricane-X) with a reputation rate of 1 kHz at 780 nm (∼100 fs pulse duration). Part of the beam from the Ti: sapphire amplifier (∼0.5 mJ/pulse) was used to pump an optical parametric amplifier (OPA, Spectra Physics, OPA-800C). The frequency-tunable visible pump pulses were produced by sumfrequency mixing the OPA signal with the residual of the 780 nm fundamental beam in a 1 mm thick type II β-BaB2O4 crystal. For the present study, the pump pulses were optimized at 482 nm and used for excitation. Samples were actually irradiated at a fluence of ∼0.05 µJ/cm2. Time-resolved PL signals were detected with a streak camera (Hamamatsu, C2909) triggered by a photodiode that enabled us to record a 90 nm spectral range of time-resolved signals at once by combining it with a monochromator. The time resolution of the system was approximately 50 ps.

A. Steady-State Spectra. Figure 1 shows the absorption spectra at room temperature of MEH-PPV in the chlorobenzene solution, in the neat spin-cast film, and in the PP blend. A small red-shift is identified in the spectrum of the neat film against that of the solution sample. This is usually explained as being due to the presence of some aggregates or to the elongation of the effective conjugation length of the MEH-PPV backbone in the solid state than that in the liquid solution. The PP blend exhibits poorly resolved vibrational structures peaking around 2.3 and 2.5 eV. They presumably originate from the 0-0 and 0-1 peaks, respectively. As depicted with a vertical broken line in Figure 1, the 0-0 peak energy of the blend sample is similar to the peak energy of the shoulder of the neat film. This is a good indication that the effective conjugation length of the species responsible for absorption is mainly close to each other between these two samples. In contrast, the absorption bandwidth of the blend sample is somewhat broader than that of the solution and in the neat film. A possible reason for this is simply the imperfect subtraction of the background in the absorption spectrum, since the contribution of the background was almost as much as the absorption signals in the condition of extreme dilution. However, such difficulty caused by the presence of background can be reduced in the case of PLE experiments. Additionally, thermal motion of the PP chain might somewhat affect chain conformations in MEH-PPV at room temperature. This can cause the absorption of the PP blend to inhomogeneously broaden. Also, there still remains a slight possibility that the PP blend might allow the presence of some aggregates despite extreme dilution (0.01 wt %). Figure 2 shows the PL and PLE spectra of the neat film sample measured at 4 K. The PL spectrum exhibits well-resolved peaks at 2.02 and 1.84 eV and a weak peak around 1.65 eV. They are assigned to the 0-0, 0-1, and 0-2 vibronic peaks of the main emissive species, respectively. This assignment is confirmed based on the PLE measurements. The PLE spectra detected at 2.02 and 1.84 eV are similar to each other (see Figure 2), suggesting that their origin is the same. The PLE spectra have a shoulder at around 2.22 eV and a maximum peak at around 2.40 eV. We assign these as resulting from the 0-0 and 0-1 absorption transitions, respectively. This results in the Stokes shift of ∼0.2 eV in the neat film (see Table 1). We here note that the Stokes shift is defined with respect to a pair of the 0-0 PL and PLE (absorption) peaks for a single segment or over different segments in the polymer sample. Therefore, this

Photoluminescence Dynamics of MEH-PPV

J. Phys. Chem. B, Vol. 111, No. 43, 2007 12391

Figure 2. Photoluminescence (PL) and PL-excitation (PLE) spectra of the neat film of MEH-PPV measured at 4 K. The arrows in the figure indicate the energy positions at which the PLE spectra were recorded. The horizontal solid lines represent the level of the baselines for each PLE spectrum.

Figure 3. Photoluminescence (PL) and PL-excitation (PLE) spectra of the PP blend of MEH-PPV measured at 4 K. The arrows in the figure indicate the energy positions at which the PLE spectra were recorded.

TABLE 1: 0-0 Peaks (eV) at 4 K of the Neat Film and the PP Blend of MEH-PPV Estimated from Their PL and PL-Excitation (PLE) Vibronic Peaks and the Stokes Shift Values (eV) Calculated by the Difference between Those Peaks neat film PP blend

PL

PLE

Stokes shift

2.02 2.14

2.22 2.20

0.20 0.06

parameter can include the information on energy transfer between different sites as well as on relaxation energy at a local site. Figure 3 shows the PL and PLE spectra of the PP blend sample measured at 4 K. The PL spectrum exhibits well-resolved peaks at 2.14 and 1.98 eV and a weak peak at around 1.80 eV. These three peaks with almost constant spacing of ∼0.17 eV are assigned to the 0-0, 0-1, and 0-2 vibronic peaks, respectively. These peaks are found to be blue-shifted by 0.12 eV with respect to the PL peaks in the neat film. Blue-shift in PL accompanied by solid-phase dilution has often been reported for the PPV derivatives not only diluted in polymer matrixes13,19-22 but also in the composites with silica28,29 or almina.30 The observation of the blue-shift in PL can thus be regarded as the signature of the PL from isolated polymer chains produced in the blend sample. In those previous reports, the PL signals with blue-shifted peaks were basically observed as a minor component of the whole spectrum that is composed mainly of the nonisolated portion of polymer samples. In the case of our PP blend sample, its spectral structure around 2.02 and 1.84 eV corresponding to the 0-0 and 0-1 PL peaks in the neat film, respectively, is apparently different from the spectral structure in the same energy range of the neat film. This suggests that the contribution of the PL from the residual amount of aggregate component should be minor. Therefore, the PL spectrum of the

PP blend sample should definitely be given by intrachain photoexcitations. It is found that the relative intensity of the 0-1 peak to the 0-0 peak is smaller in the case of the PP blend sample than that in the neat film (see Figures 2 and 3). This is a good indication that the blend sample has a smaller HuangRhys factor than that in the neat film. This fact indicates that the intrachain excitations responsible for PL have weaker couplings with primary optical phonons (CdC bond) in the case of the blend sample compared with the excitations in the neat film. The same feature has been reported for P3OT as well.25,31 This could hence be a general feature in isolated conjugated polymers. The PLE spectra of the PP blend sample measured at 2.30, 2.14, and 1.98 eV are shown in Figure 3. It is found that the spectra recorded at the 0-0 (2.14 eV) and 0-1 (1.98 eV) PL peaks both exhibit resolved peaks at 2.20 and 2.37 eV. The spacing between these PLE peaks, 0.17 eV, is consistent with that observed in its PL spectrum. The PLE peaks are thus assigned to resulting from the vibronic absorption peaks of the isolated chains in the blend. The peak values of PLE are close to those in the neat film, whereas the PL spectrum in the blend is appreciably blue-shifted against that in the neat film. The blend sample hence exhibits obviously smaller Stokes shift (0.06 eV) than that of the neat film (∼0.20 eV). This is summarized in Table 1. The observation of the different amount of Stokes shift between the blend and the neat film samples is consistent with the previous report by Hagler et al.24 They determined indirectly the Stokes shift value of MEH-PPV from the second derivative of the absorption spectrum in the PE blend with a concentration of 1 wt %. To the best of our knowledge, our result is the first report on the direct observation of the vibronic absorption peaks in the isolated chain of the PPV derivatives. It is here worthwhile to note that the observation of the vibronic peaks in the PLE measurements is difficult even by using a microscope detection technique for a highly diluted MEH-PPV embedded in a PMMA matrix due to spectral diffusion.32 In the PL spectrum of the blend sample in Figure 3, we can find a shoulder around 2.3 eV. Its energy is obviously above the 0-0 absorption peak of the main intrachain species (2.20 eV). This indicates that its origin is not the same as the main intrachain species. In fact, the PLE spectrum monitored at 2.30 eV has a different spectral structure from those at 1.98 and 2.14 eV detection. The origin of this PL shoulder signal is not wholly understood at present. It may be assumed that some domains with distorted chain structures of MEH-PPV exist in the PP blend and they give a somewhat blue-shifted PL band. Such distorted domains can be generated by the mechanical strain caused by PP or be a naturally occurring morphology of isolated MEH-PPV in the solid state. B. Time-Resolved PL Spectra. Figure 4 shows the timeresolved PL spectra restricted around the 0-0 PL peak for the neat film of MEH-PPV measured at 4 K. A set of the spectra displays a dynamic red-shift of the 0-0 peak in the time course of a few hundred picoseconds following photoexcitation. The red-shift occurring in this time scale is typical for exciton migration, as has been previously reported.5-8 Figure 5 shows the time-resolved PL spectra around the 0-0 PL peak for the PP blend sample measured at 4 K. Contrary to the result in the neat film, the spectra in the blend sample demonstrate the absence of an appreciable amount of dynamic red-shift. This difference in the time-resolved PL spectra between the neat film and the blend sample provides direct evidence that the exciton migration observed in the neat film takes place primarily via the interchain transfer, and not via the intrachain one.

12392 J. Phys. Chem. B, Vol. 111, No. 43, 2007

Figure 4. Time-resolved PL spectra around the 0-0 PL peak for the neat film of MEH-PPV measured at 4 K. The spectra were obtained by integrating signals over a given time width. From the top to bottom, the integrated times are 0 ( 35, 100 ( 35, 250 ( 40, 600 ( 50, and 1500 ( 100 ps following photoexcitation. The averaged delay time is shown in the figure for each spectrum. The spectra were normalized to their maximum intensity. The dotted line in the figure indicates the 0-0 peak position in the steady-state PL spectrum.

Figure 5. Time-resolved PL spectra around the 0-0 PL peak for the PP blend of MEH-PPV measured at 4 K. The spectra were obtained by integrating signals over a given time width. From the top to bottom, the integrated times are 0 ( 30, 100 ( 30, 250 ( 40, 600 ( 50, 1500 ( 100, and 2000 ( 150 ps following photoexcitation. The averaged delay time is shown in the figure for each spectrum. The spectra were normalized to their maximum intensity. The dotted line in the figure indicates the 0-0 peak position in the steady-state PL spectrum.

The result of the time-resolved PL in the blend sample also suggests that exciton migration is not so efficient in isolated MEH-PPV chains as seen in the neat film sample. This is somewhat controversial because excitations in conjugated polymers have been believed to migrate efficiently along the conjugated backbone.33,34 In fact, since the presence of some energy gradient is usually expected in the isolated chain, it might be straightforward to expect that efficient exciton migration should take place within the isolated polymer chain. This contradiction leads us to conclude that the intrachain exciton transfer is much less efficient than the case in the interchain transfer. This is not unexpected because dipole transfers between segments parallel to each other are forbidden based on the Fo¨rster mechanism, and such a situation can be true for excitations within a single chain. Our conclusion is supported by the previous report by Nguyen et al.28 They showed through

Kanemoto et al.

Figure 6. Transient PL traces for the neat film and the PP blend measured at 4 K. The traces were obtained by integrating transient PL signals from 2.00 to 2.02 eV for the neat film and from 2.10 to 2.20 eV for the blend. The vertical position of these two traces is intentionally shifted so as to not overlap each other.

schemed optical measurements for MEH-PPV in silica that the migration along the polymer chain is much slower by roughly 2 or 3 orders of magnitude compared with energy transfer between the polymer chains. Similar significance has also been obtained for acceptor-capped polyindenofluorene,35 suggesting that a similar conclusion should be drawn from the investigation of many other conjugated polymers. In the time-resolved spectra of the blend sample at 600, 1500, and 2000 ps shown in Figure 5, a weak shoulder is found to appear around 2.20 eV. We interpret this feature to be ascribable to the transfer from the higher-energy exciton state, which has been found around 2.3 eV in the steady-state PL spectrum. This particular energy-transfer feature could thus be an indication of a kind of intrachain exciton migration. A similar feature has already been found in the blend of RRa P3OT.31 In Figure 6, we present the transient PL traces for the PP blend sample and the neat film measured at 4 K. The PL trace for the blend sample was obtained by integrating transient PL signals from 2.10 to 2.20 eV. This energy range reflects the PL decay around the 0-0 peak in the steady-state PL. In this spectral range, even if there are some residual PL signals from nonisolated portions in the blend, only the main intrachain PL signals can be detected. The PL trace of the blend sample demonstrates that its decay is a single-exponential form. On the contrary, a feature of nonexponential PL decay has been reported for PPV derivatives diluted in polymer matrixes11,13,15,20 and in liquid solutions.14,17,18,36 This discrepancy could solely be due to different measuring temperature; 4 K in our case and room temperature in most of the previous experiments. Particularly, the PL decay of MEH-PPV diluted in PE was previously shown to be nonexponential at room temperature but turn to be single exponential when the sample is cooled down to 80 K.11 We must point out that, for the observation of the intrinsic decay property of intrachain PL in the solid state, sufficient dilution of the guest polymer as well as data acquisition in the select spectral range excluding interchain excitations are necessary in addition to the low-temperature measurement. Experiments under these conditions have never been reported for the PPV derivatives. Our experiment satisfies these severe conditions. Therefore, the results that have been obtained in this study demonstrate that the PL decay of the isolated PPV derivatives should essentially be single exponential. This feature is also expected to hold true for most of conjugated polymers. In fact, PL decays of isolated P3OTs and thiophene oligomers have recently been found to be single exponential under conditions similar to the present work.27,31 The time constant of the PL decay of the blend sample is determined to be 850 ps. It is worth noting that the value of the

Photoluminescence Dynamics of MEH-PPV time constant is close to the radiative lifetime of MEH-PPV (900 ps) estimated from measurements in solution.14 This indicates that a nonradiative process is highly suppressed for the isolated polymer chain at low temperature. Particularly, the PL efficiency can be calculated to be 0.94 using the simple relation of (the PL lifetime)/(the radiative lifetime). This result suggests that the PL of isolated MEH-PPV in the PP blend has considerably high efficiency. To the best of our knowledge, the PL decay time of MEH-PPV estimated in this work is the longest value ever exploited. The kinetic trace for the neat film in Figure 6 was obtained by integrating transient PL signals from 2.00 to 2.02 eV. This spectral range corresponds to PL decay around the 0-0 peak in the steady-state PL. The PL decay can be fitted roughly to a single exponential with a time constant of 700 ps. It should be noted that this decay process includes the transient signals concomitant with the components above 2.02 eV, as readily understood based on the observation in Figure 4. This concomitant contribution is naturally expected to retard the observed PL decay, and hence the actual PL decay of the main emissive component in the neat film should be faster than the observation in Figure 6. This reveals that the decay time constant of the species in the neat film is surely smaller than that in the blend sample. IV. Discussion In the previous section, we obtained several significant features concerning the PL of isolated MEH-PPV in the solid sate. In this section, the PL dynamics for the isolated chain is discussed. Furthermore, the PL dynamics in the neat film is also discussed. The PP blend of MEH-PPV was shown to exhibit the Stokes shift of 0.06 eV by the measurements of the steady-state PL and PLE spectra. In contrast, the time-resolved PL spectra for the blend sample did not show the Stokes shift of the similar amount corresponding to the observation in the steady-state spectra. This indicates that the observed Stokes shift occurs at the early stage of photoexcitation within the limit of time resolution in our system (50 ps). Concerning the origin of the Stokes shift, a rapid intrachain exciton migration could be a candidate. The intrachain migration is presumed to occur due to the presence of the energy gradient between segments within a single chain and hence is expected to be reduced by decreasing a chain length, because the number of segments is likely to increase with the chain length. In contrast, we have recently shown that a series of thiophene oligomers with different chain lengths give rise to the dynamic Stokes shift within 50 ps, regardless of the chain length.31 These previous results suggest that rapid Stokes shift cannot be attributed to the intrachain migration. We therefore focus on a change in the torsional angle around the molecular axis induced by photoexcitation. Previously, Stokes shift in the steady-state spectra for p-phenylene oligomers was shown to be reduced by the introduction of saturated bridges between the phenylene rings.37,38 This suggests that the change in the ring torsional angle after photoexcitation would result in some extent of Stokes shift. Namely, the chain is expected to take a more conjugated configuration after photoexcitation by induced conformational change due to the torsion. This argument can also hold true for the MEH-PPV chain that has no major steric hindrance for the torsional motion. We thus attribute the rapid Stokes shift in the blend of MEHPPV to the change in the ring torsional angle in the MEHPPV chain after photoexcitation. The change in the ring torsional angle is a phenomenon occurring at local sites in the chain. This situation is not a

J. Phys. Chem. B, Vol. 111, No. 43, 2007 12393 peculiar thing only for the intrachain excitation. A similar amount of the ring torsion should also take place when the interchain excitation is generated. Therefore, the interchain photoexcitations are expected to have a similar small degree of Stokes shift to the intrachain ones, ∼0.06 eV. However, in the case of the neat film, a 0.20 eV Stokes shift has been observed. This shows an apparent disagreement with the above consideration for the conformational distortion. This leads us to conclude that the PL in the neat film occurs at minor sites with lower energy, namely, at the interchain excitations. In this view, the dynamic Stokes shift observed in the time-resolved PL for the neat film is concluded to be the exciton migration toward the interchain excitations. The magnitude of migration energy can then be simply estimated to be 0.14 eV from the difference in the Stokes shift between the neat film and the blend sample. This suggests that the migration of the PL 0-0 peak begins at 2.16 eV in the neat film, which is far from the observed 0-0 peak (2.05 eV) at 0 ps in the time-resolved PL. Therefore most of the interchain migration is expected to occur rapidly within the limit of time resolution in our system (50 ps). V. Conclusions The PL features in the time domain as well as in the steadystate spectra have been investigated for the PP blend sample and the neat film of MEH-PPV. The blend, extremely diluted (0.01 wt %), exhibited several PL features that were attributable to the isolation of polymer chains. Those are summarized as below. The steady-state PL spectrum of the blend sample was demonstrated to blue-shift and display a spectral pattern implying a weaker coupling with phonons of the CdC bond, compared with that of the neat film of MEH-PPV. These are concluded to be features of the emissive species in the isolated chains. The PLE spectra measured at the main PL peaks of the intrachain species exhibited appreciable vibronic peaks, which enabled us to determine the magnitude of the Stokes shift as being 0.06 eV. The PP blend sample was shown to exhibit single-exponential PL decay at 4 K. We emphasize that this exponential type of PL decay is an intrinsic feature of the intrachain emissive species. Its decay time (850 ps), larger than that of the neat film, suggests that a nonradiative PL process is highly suppressed under the extremely diluted condition. The time-resolved PL measurements demonstrate that the dynamical red-shift observed for the neat film is not observed in the case of the PP blend. These observations lead us to conclude that the energy transfer causing exciton migration is not so efficient between the segments within the isolated polymer chain as in the neat film. The time-resolved measurements further demonstrated that the Stokes shift identified in the steady-state spectra of the blend sample takes place at the early stage of photoexcitation within 50 ps. We attribute the Stokes shift to a change in the ring torsional angle after photoexcitation. This type of Stokes shift is considered to occur in the interchain excitations as well as in the intrachain ones. The absence of appreciable PLE peak corresponding to the 0-0 absorption in the neat film leads us to conclude that the PL in the neat film occurs at minor sites with lower energy, namely, at the interchain excitations. The PL peak energy in the neat film represents the minimum energy of interchain excitations where excitons can migrate. We thus point out that the difference of the Stokes shift energy between the neat film and the blend sample corresponds to the magnitude of interchain migration.

12394 J. Phys. Chem. B, Vol. 111, No. 43, 2007 Acknowledgment. This work was supported in part by the Grant-in-aid from the Japanese Ministry of Education, Culture, Sports, Science, and Technology (Grant Nos. 17204026, 17654083, 18340091, and 18654074). H.H. also acknowledges PRESTO/JST for financial support. References and Notes (1) Friend, R. H.; Gymer, R. W.; Holmes, A. B.; Burroughes, J. H.; Marks, R. N.; Taliani, C.; Bradley, D. D. C.; Dos Santos, D. A.; Bre´das, J. L.; Logdlund, M.; Salaneck, W. R. Nature 1999, 397, 121. (2) Gustafsson, G.; Treacy, G. M.; Cao, Y.; Klavetter, F.; Colaneri, N.; Heeger, A. J. Nature 1992, 357, 477. (3) Yu, G. Synth. Met. 1996, 80, 143. (4) Samuel, I. D. W.; Rumbles, G.; Friend, R. H. In Primary Excitations in Conjugated Polymers: Molecular Exciton Versus Semiconductor Band Model; Sariciftci, N. S., Ed.; World Scientific: Singapore, 1997; Chapter 7. (5) Kersting, R.; Lemmer, U.; Mahrt, R. F.; Leo, K.; Kurz, H.; Ba¨ssler, H.; Go¨bel, E. O. Phys. ReV. Lett. 1993, 70, 3820. (6) Haynes, G. R.; Samuel, I. D. W.; Phillips, R. T. Phys. ReV. B 1995, 52, 11569. (7) Haynes, G. R.; Samuel, I. D. W.; Phillips, R. T. Phys. ReV. B 1996, 54, 8301. (8) Sperling, J.; Milota, F.; Tortschanoff, A.; Warmuth, C.; Mollay, B.; Ba¨ssler, H.; Kauffmann, H. F. J. Chem. Phys. 2002, 117, 10877. (9) Nguyen, T.-Q.; Martini, I. B.; Liu, J.; Schwartz, B. J. J. Phys. Chem. B 2000, 104, 237. (10) Martini, I. B.; Smith, A. D.; Schwartz, B. J. Phys. ReV. B 2004, 69, 035204. (11) Smilowitz, L.; Hays, A.; Heeger, A. J.; Wang, G.; Bowers, J. E. J. Chem. Phys. 1993, 98, 6504. (12) Smilowitz, L.; Hays, A.; Heeger, A. J.; Wang, G.; Bowers, J. E. Synth. Met. 1993, 55-57, 249. (13) Lemmer, U.; Mahrt, R. F.; Wada, Y.; Greiner, A.; Ba¨ssler, H.; Go¨bel, E. O. Appl. Phys. Lett. 1993, 62, 2827. (14) Samuel, I. D. W.; Rumbles, G.; Collison, C. J. Phys. ReV. B 1995, 52, 11573. (15) Yan, M.; Rothberg, L. J.; Kwock, E. W.; Miller, T. M. Phys. ReV. Lett. 1995, 75, 1992. (16) Heller, C. M.; Campbell, I. H.; Laurich, B. K.; Smith, D. L.; Bradley, D. D. C.; Burn, P. L.; Ferraris, J. P.; Mu¨llen, K. Phys. ReV. B 1996, 54, 5516.

Kanemoto et al. (17) Samuel, I. D. W.; Rumbles, G.; Collison, C. J.; Friend, R. H.; Moratti, S. C.; Holmes, A. B. Synth. Met. 1997, 84, 497. (18) Samuel, I. D. W.; Rumbles, G.; Collison, C. J.; Moratti, S. C.; Holmes, A. B. Chem. Phys. 1998, 227, 75. (19) Marletta, A.; Goncalves, V. C.; Balogh, D. T. J. Lumin. 2006, 116, 87. (20) Yang, G.; Li, Y.; White, O.; Drickamer, H. G. J. Phys. Chem. B 1999, 103, 5181. (21) He, G.; Li, Y.; Liu, J.; Yang, Y. Appl. Phys. Lett. 2002, 80, 4247. (22) Sartori, S. S.; Feyter, S. D.; Hofkens, J.; Brunner, K.; Hofstraat, W. Macromolecules 2003, 36, 500. (23) Lemmer, U.; Mahrt, R. F.; Wada, Y.; Greiner, A.; Ba¨ssler, H.; Go¨bel, E. O. Chem. Phys. Lett. 1993, 209, 243. (24) Hagler, T. W.; Pakbaz, K.; Voss, K. F.; Heeger, A. J. Phys. ReV. B 1991, 44, 8652. (25) Kanemoto, K.; Shishido, M.; Sudo, T.; Akai, I.; Karasawa, T.; Agari, Y. Synth. Met. 2005, 155, 162. (26) Iyengar, N. A.; Harrison, B.; Duran, R. S.; Schanze, K. S.; Reynolds, J. R. Macromolecules 2003, 36, 8978. (27) Kanemoto, K.; Shishido, M.; Sudo, T.; Akai, I.; Hashimoto, H.; Karasawa, T. Chem. Phys. Lett. 2005, 402, 549. (28) Nguyen, T.-Q.; Wu, J.; Liu, J.; Doan, V.; Schwartz, B. J.; Tolbert, S. H. Science 2000, 288, 652. (29) Ho, P. K. H.; Kim, J. S.; Tessler, N.; Friend, R. H. J. Chem. Phys. 2001, 115, 2709. (30) Qi, D.; Kwong, K.; Rademacher, K.; Wolf, M. O.; Young, J. F. Nano Lett. 2003, 3, 1265. (31) Kanemoto, K.; Sudo, T.; Akai, I.; Hashimoto, H.; Karasawa, T.; Aso, Y.; Otsubo, T. Phys. ReV. B 2006, 73, 235203. (32) Feist, F. A.; Tommaseo, G.; Basch, T. Presented at the Conference of Light Harvesting Processes, Banz Monastery, Germany, March 2007. (33) Grage, M. M.-L.; Pullerits, T.; Ruseckas, A.; Theauder, M.; Ingana¨s, O.; Sundstro¨m, V. Chem. Phys. Lett. 2001, 339, 96. (34) Westenhoff, S.; Beenken, W. J. D.; Yartsev, A.; Greenham, N. J. Chem. Phys. 2006, 125, 154903. (35) Beljonne, D.; Pourtois, G.; Silva, C.; Hennebicq, E.; Herz, L. M.; Friend, R. H.; Scholes, G. D.; Setayesh, S.; Mu¨llen, K.; Bre´das, J. L. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 10982.j (36) Samuel, I. D. W.; Crystall, B.; Rumbles, G.; Burn, P. L.; Holmes, A. B.; Friend, R. H. Chem. Phys. Lett. 1993, 213, 472. (37) Grimme, J.; Scherf, U. Macromol. Chem. Phys. 1996, 197, 2297. (38) Pauck, T.; Ba¨ssler, H.; Grimme, J.; Scherf, U.; Mu¨llen, K. Chem. Phys. 1996, 210, 219.