Relating Conformation and Photophysics in Single MEH-PPV Chains

Sep 13, 2008 - The conformations were related to photophysical properties of MEH-. PPV by measuring fluorescence intermittency on the same chains...
0 downloads 0 Views 829KB Size
12575

2008, 112, 12575–12578 Published on Web 09/13/2008

Relating Conformation and Photophysics in Single MEH-PPV Chains Yohei Ebihara and Martin Vacha* Department of Organic and Polymeric Materials, Tokyo Institute of Technology, Ookayama 2-12-1, Meguro-ku, Tokyo 152-8552 Japan ReceiVed: August 5, 2008; ReVised Manuscript ReceiVed: August 24, 2008

We use a novel fluorescence polarization microscope in combination with molecular dynamics calculations to determine the conformation of individual isolated chains of the conjugated polymer MEH-PPV. We found a narrow distribution of defect cylinder conformations in a poor-solvent matrix and two types of defect coil conformations in a good-solvent matrix. The conformations were related to photophysical properties of MEHPPV by measuring fluorescence intermittency on the same chains. We obtained direct evidence that the photophysics is determined by the chain conformations and that small changes in the polymer microscopic structure can qualitatively affect the photophysical properties. Conjugated polymers represent a class of organic materials with optical and electrical properties resembling those of inorganic semiconductors and with a wide potential for applications in optoelectronic devices.1,2 Their optical properties are determined by the microscopic structure of the polymer solid state. The disordered nature of the solid means that, microscopically, each polymer chain is in a unique conformational state. The resulting complex photophysical properties have not been fully understood.3,4 The complexity can be partially lifted by studying individual chains using single-molecule fluorescence detection techniques. Here, we used a novel microscopic method that we theoretically proposed before,5 combined with molecular dynamics calculations, to determine the conformation of individual chains on a true single-molecule level. This enabled us to relate the conformation to the photophysical observations for each molecule individually. We have found a direct relationship between the different types of conformations on the one hand and photophysics such as fluorescence intermittency on the other. Microscopic optical study of the single-chain conformation was reported before in the extensive work of Barbara’s group who measured the polarization spectroscopy of individual chains of the poly(phenylene vinylene) (PPV) family6 and used Monte Carlo simulations to estimate the chain conformation. The method provided conformations for a subensemble of molecules but did not enable assignment of a specific conformation to a concrete molecule. In our approach, we used the fact that the conformation of a polymer molecule is related to its absorption anisotropy, approximated by a rotational absorption ellipsoid (see Figure 1a). The shape of the ellipsoid is determined by the ratio r ) |µb|/|µa|, where µa and µb are transition dipole moment vectors. The ellipsoid results from a sum of the transition dipoles of individual conjugated segments of the chain, which themselves can be straight or bent.7-9 The value r is measured by combined far-field and near-field laser excitation in a fluorescence microscope (Figure 1b). Theoretical analysis gives far* To whom correspondence should be addressed. E-mail: vacha@ op.titech.ac.jp.

10.1021/jp806963u CCC: $40.75

field Ffar and near-field Fnear excited fluorescence as a function of the angle ψ of a linearly polarized excitation beam5

Ffar ∝ sin2θ(1 - r2)cos2(φ - ψ) + r2

(1)

Fnear ∝ (a + dr2) + (b + er2)cos2ψ + c(1 - r2)cos ψ sin ψ (2) Here, φ and θ are the angular coordinates (defined in Figure 1a), and a, b, c, d, and e are functions of φ and θ only. Rotation of the polarization (angle ψ) results in modulated far- and nearfield excited fluorescence traces from individual molecules (Figure 1c), which are fitted simultaneously with eqs 1 and 2 to obtain the spatial orientation angles φ and θ and the ratio r of the ellipsoid (Figure 1d). As a prototype conjugated polymer, we use poly[2-methyloxy-5-(2-ethylhexyl)oxy-1,4-phenylenevinylene] (MEH-PPV) dispersed in thin-film matrixes of the cycloolefin polymer Zeonex and of low-molecular weight polystyrene (PS). Low-Mw PS was used because it simulates well a good solvent for MEH-PPV. A dilute toluene solution of MEH-PPV (Aldrich, Mw 200 000, polydispersity index 5) was mixed with a 1.5 wt % toluene solution of Zeonex (ZEONEX 480, Zeon) or PS (Polymer Standards, Mw 10 000) and spin-coated on cleaned high-refractive index glass substrates (n ) 1.77). High refractive index substrates were used to ensure that total internal reflection occurs between the substrate and the matrix polymer. The final film thickness as measured by a Dektak profiler was on the order of 120 nm. The samples were vacuum-dried at 60 °C for 2 h and used immediately for experiments. The experiments were carried out on an inverted fluorescence microscope10 using a dry objective lens and the 488 nm line of an Ar+ ion laser for excitation. Polarization was modulated by a quarter-wave plate and a linear polarizer rotating at 5 rpm. The laser beam was split by a nonpolarizing beam splitter into an epi-illumination (far-field) beam and a totally internally reflected (TIR, near-field) beam. The TIR beam was coupled into the substrate via a prism, and the total internal reflection occurred at the substrate-polymer matrix interface. The beams were alternated using a mechanical chopper. Images were taken  2008 American Chemical Society

12576 J. Phys. Chem. B, Vol. 112, No. 40, 2008

Figure 1. Principle of the method. (a) Representation of the conformation of the conjugated polymer chain with a rotationally symmetric absorption ellipsoid. The µa and µb are the transition dipole moments, and θ and φ determine the ellipsoid orientation. (b) Schematic principle of the two-beam polarization microscope. Efar represents the epifluorescence excitation beam, Enear is the near-field excitation beam, which undergoes total internal reflection on the substrate-sample interface, and ψ is the linear polarization angle. (c) Typical far-field excited fluorescence image of the MEH-PPV single molecules in the PS matrix. (d) Fluorescence intensity of a single MEH-PPV molecule modulated as a function of ψ. Bottom trace, far-field excitation; top trace, near-field excitation. The traces are fitted with eqs 1 and 2, respectively.

Figure 2. Experiments on single MEH-PPV chains. (a) Distribution of the ratio r measured for 40 molecules in the Zeonex matrix. (b) Distribution of the ratio r measured for 60 molecules in the PS matrix; grey, molecules undergoing blinking and stepwise photobleaching (such as in (d)); red, molecules undergoing continuous photobleaching (such as in (e)). (c) Typical blinking behavior in the Zeonex matrix. (d,e) Different types of blinking and photobleaching in the PS matrix.

with an EM-CCD camera with exposition times of 130 ms. The method and the fitting procedure were checked by measuring the r values for spherical fluorescence beads. The conformations were characterized by measuring distributions of the ratio r for statistical ensembles of single polymer chains. The histogram of r measured in the Zeonex matrix (shown in Figure 2a) has a narrow band centered at 0.5 with a few values scattered up to 2.5, corresponding to an oblong absorption ellipsoid for most (about 70%) of the molecules. In contrast, the histogram in the PS matrix shows a bimodal distribution with bands at 0.5 and 1.6 (Figure 2b), indicating the presence of two different conformations in the same matrix, an oblong ellipsoid for about 40% of the molecules and a disklike ellipsoid for about 60% of the polymer chains.

Letters An important photophysical phenomenon observed in single quantum emitters is fluorescence intermittency (blinking). For conjugated polymers containing tens to hundreds of conjugated segments, the observation of fluorescence blinking implies fast localization of the excitation energy on one or a small number of chromophores.11-19 Blinking due to energy transfer was also observed for other multichromophoric systems.20 In our work, single MEH-PPV molecules in the Zeonex matrix invariably show blinking on a few intensity levels and stepwise photobleaching (Figure 2c). In the PS matrix, the behavior is more complex; both blinking and continuous photobleaching have been detected. Importantly, those molecules that show discrete blinking (such as the one in Figure 2d) have their r values within the 0.5 band (shown in gray in Figure 2b), while the molecules that exhibit continuous photobleaching (such as the one in Figure 2e) are mostly centered in the 1.6 band (shown in red in Figure 2b). This result is direct evidence that the phenomenon of blinking in conjugated polymers is a consequence of the microscopic structure of the polymer chain. The measured absorption ellipsoids can be related to the actual chain conformations by numerical simulations. We used the beads-on-spring model in the coarse-grain molecular dynamics (MD) method to simulate conformations in the two matrixes and used the simulated results to calculate distributions of the ratio r and orientations of the ellipsoids. The simulations were carried out using the software OCTA with the simulation engine COGNAC for a beads-spring model (http://octa.jp/index.html). The conditions were set to closely resemble the real MEH-PPV. The number of beads was 300, corresponding to the actual number of 750 monomers in a MEH-PPV chain with a Mw of 200 000. The stiffness of the chain was simulated by the bending potential EB ) bθ2, with the experimentally determined21 b ) 10 kT rad-2. The solvent effect was simulated by the depth ELJ of an interbead Lennard-Jones potential with ELJ ) 0.6-0.8 kT for the Zeonex matrix and with ELJ ) 0.2 kT for the PS matrix. Each conformation was obtained in 100 000 simulation steps. The simulation results for an ideal chain (not shown) give the expected rod and toroid conformations for poor solvent and random coils for good solvent.22 However, the corresponding r distributions for such ideal chains do not reproduce the measured ones because the inevitable presence of sp3 defects in real chains strongly influences the conformations.23,24 Including the experimentally verified 2.6% of the defects in the simulation parameters25 resulted only in good agreement with the r distribution in the Zeonex matrix (Figure 3a, blue dots). Neither the r distribution in PS (Figure 3b, blue dots) nor the ellipsoid orientations in the two matrixes (Figure 3c, d, blue circles) could be reproduced by the sp3 defects only. The observed orientation of the oblong (r ∼ 0.5) ellipsoids parallel to the substrate (Figure 3c, d, black dots) invokes the presence of a potential normal to the substrate (z-potential) which orients the individual chains. The potential is likely a result of solvent evaporation after the initial formation of the solution layer during the spin-coating process. When this potential was included in the simulation as a near-harmonic potential with a depth of ESP ) 0.06 kT, the ellipsoid orientations were well-reproduced (Figure 3c, d, red circles). The z-potential also affects the conformations. In the Zeonex matrix (Figure 3a), most chains take a collapsed defect cylinder conformation; those molecules with r > 1 correspond to incomplete defect cylinders. In the PS matrix, the original defect coil conformation centered at r ∼ 1 splits into two, disk-like coils with r ∼ 1.6 and oblong coils with r ∼ 0.6, reflecting the experimental results

Letters

J. Phys. Chem. B, Vol. 112, No. 40, 2008 12577

Figure 3. Molecular dynamics simulations. (a) Distribution of the ratio r calculated for 36 molecules in poor-solvent conditions. Blue dots, conditions including 2.6% of sp3 defects; red bars, including the sp3 defects and z-potential to simulate the spin-coating process; black bars, experiment in the Zeonex matrix. Inset: calculated defect cylinder (left) and imperfect defect cylinder (right) conformations. (b) Same as (a) for good solvent and the PS matrix. Inset: calculated oblong defect coil (left) and disk-like defect coil (right) conformations. (c) Correlation between the angle θ and ratio r for poor solvent. Blue circles, with sp3 defects; red circles, with sp3 defects and z-potential; black dots, experiment in the Zeonex matrix. (d) Same as (c) for good solvent and the PS matrix. (e) Calculated radius of gyration Rg with sp3 defects and the z-potential. Blue, poor solvent; black, good solvent. (f) Typical distributions of all interbead distances in poor solvent. Red, oblong defect coil; black, disk-like defect coil. (g) Third moment (skewness S) of interbead distance distributions in poor solvent plotted against the ratio r. Red, oblong defect coil; black, disk-like defect coil.

well (Figure 3b, red bars). The two conformations result from the finite length of the polymer chain. A long defect coil chain would have a spherical symmetry, and the z-potential during spin-coating would only press it into a disk-like coil. In reality, the conformations before spin-coating deviate from the spherical symmetry (as seen in Figure 3b, blue dots at r < 1), and this fact, together with the random chain orientation, leads to the formation of the oblong coil states. The size of the chain (given by the radius of gyration Rg) and the resulting intersegment distances determine localization of the exciton due to Fo¨rster-type energy transfer and the resulting fluorescence blinking. The simulated Rg of the defect cylinders is smaller than that of the defect coils (Figure 3e), and the energy transfer due to smaller intersegment distances explains the blinking process in the Zeonex matrix. The situation is more complicated in the PS matrix where the various photophysical phenomena have so far not been well understood.26 The Rg alone fails to explain the different blinking as both the oblong and disk-like coils are characterized by similar Rg values. The difference is likely caused by microscopic structural inhomogeneities found on the chains with the oblong coil conformations. These inhomogeneities form domains of higher bead density, where the interbead (intersegment) distances are significantly smaller than those in the rest of the chain. The

inhomogeneities are reflected in the shape of calculated distributions of all interbead distances in the chain (see an example in Figure 3f); the larger number of small distances increases the distribution asymmetry. The asymmetry is characterized by the third moment S (skewness) of the distributions in Figure 3g. The correlation between the third moment and the r parameter shows a difference between the two conformations, high S values in the oblong coils and a wide range of S values in the disk-like coils. The domains in the oblong coil chain function as local traps where the energy from part of the chain is localized and which cause the observed blinking in the PS matrix. The existence of such domains in MEH-PPV was recently also suggested in the discussion of time-resolved experiments.27 In summary, we have measured absorption ellipsoids of individual MEH-PPV chains and studied the photophysical behavior on the same chains. The relationship between the chain conformations reconstructed using molecular dynamics simulations and the observed photophysics directly confirms that the photophysics is determined by the chain structure and that even small differences in otherwise similar conformational states can have a strong effect on the resulting photophysical properties. Acknowledgment. We thank Prof. Dr. S. Kawaguchi and Dr. S. Habuchi from the Tokyo Institute of Technology for

12578 J. Phys. Chem. B, Vol. 112, No. 40, 2008 stimulating discussions. We are especially indebted to Dr. T. Ozawa from JRI Solutions, Ltd., for his help with the OCTA simulation software. The work was supported by the Grant-inAid for Scientific Research No. 18340123 of the Japan Society for the Promotion of Science (JSPS). 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.; Bredas, J. L.; Lo¨gdlund, M.; Salaneck, W. R. Nature 1999, 397, 121–128. (2) Huynh, W. U.; Dittmer, J. J.; Alivisatos, A. P. Science 2002, 295, 2425–2427. (3) Rothberg, L. J.; Yan, M.; Papadimitrakopoulos, F.; Galvin, M. E.; Kwock, E. W.; Miller, T. M. Synth. Met. 1996, 80, 41–58. (4) Schwartz, B. J. Annu. ReV. Phys. Chem. 2003, 54, 141–172. (5) Ebihara, Y.; Vacha, M. J. Chem. Phys. 2005, 123, 244710. (6) Barbara, P. F.; Gesquiere, A. J.; Park, S. J.; Lee, Y. J. Acc. Chem. Res. 2005, 38, 602–610. (7) Beenken, W. J. D.; Pullerits, T. J. Phys. Chem. B 2004, 108, 6164– 6169. (8) Becker, K.; Da Como, E.; Feldmann, J.; Scheliga, F.; Thorn Csanyi, E.; Tretiak, S.; Lupton, J. M. J. Phys. Chem. B 2008, 112, 4859–4864. (9) Muls, B.; Uji-i, H.; Melnikov, S.; Moussa, A.; Verheijen, W.; Soumillion, J.-P.; Josemon, J.; Mu¨llen, K.; Hofkens, J. ChemPhysChem 2005, 6, 2286–2294. (10) Vacha, M.; Kotani, M. J. Chem. Phys. 2003, 118, 5279–5282. (11) Hu, D.; Yu, J.; Barbara, P. F. J. Am. Chem. Soc. 1999, 121, 6936– 6937. (12) Huser, T.; Yan, M.; Rothberg, L. J. Proc. Natl. Acad. Sci. U.S.A. 2000, 97, 11187–11191.

Letters (13) Sartori, S. S.; De Feyter, S.; Hofkens, J.; Van der Auweraer, M.; De Schryver, F.; Brunner, K.; Hofstraat, J. W. Macromolecules 2003, 36, 500–507. (14) Sun, W. Y.; Yang, S.-Ch.; White, J. D.; Hsu, J.-H.; Peng, K.-Y.; Chen, S. A.; Fann, W. Macromolecules 2005, 38, 2966–2973. (15) Wang, C. F.; White, J. D.; Lim, T. L.; Hsu, J. H.; Yang, S. C.; Fann, W. S.; Peng, K. Y.; Chen, S. A. Phys. ReV. B 2003, 67, 035202. (16) Schindler, F.; Lupton, J. M.; Feldmann, J.; Scherf, U. Proc. Acad. Natl. Sci. U.S.A. 2004, 101, 14695–14700. (17) Becker, K.; Lupton, J. M.; Feldmann, J.; Setayesh, S.; Grimsdale, A. C.; Mu¨llen, K. J. Am. Chem. Soc. 2006, 128, 680–681. (18) Mirzov, O.; Cichos, F.; Von Borszyskowski, C.; Scheblykin, I. G. Chem. Phys. Lett. 2004, 386, 286–290. (19) Sugimoto, T.; Ebihara, Y.; Ogino, K.; Vacha, M. ChemPhysChem 2007, 8, 1623–1628. (20) Hofkens, J.; Maus, M.; Gensch, T.; Vosch, T.; Cotlet, M.; Ko¨hn, F.; Herrmann, A.; Mu¨llen, K.; De Schryver, F. J. Am. Chem. Soc. 2000, 122, 9278–9288. (21) Orion, I.; Buisson, J. P.; Lefrant, S. Phys. ReV. B 1998, 57, 7050– 7065. (22) Noguchi, H.; Yoshikawa, K. J. Chem. Phys. 1998, 109, 5070–5077. (23) Hu, D.; Yu, J.; Wong, K.; Bagchi, B.; Rossky, P. J.; Barbara, P. F. Nature 2000, 405, 1030–1033. (24) Hu, D.; Yu, J.; Padmanaban, G.; Ramakrishnan, S.; Barbara, P. F. Nano Lett. 2002, 2, 1121–1124. (25) Becker, H.; Spreitzer, H.; Ibrom, K.; Kreuder, W. Macromolecules 1999, 32, 4925–4932. (26) Liang, J. J.; White, J. D.; Chen, Y. C.; Wang, C. F.; Hsiang, J. C.; Lim, T. S.; Sun, W. Y.; Hsu, J. H.; Hsu, C. P.; Hayashi, M.; Fann, W. S.; Peng, K. Y.; Chen, S. A. Phys. ReV. B 2006, 74, 085209. (27) Lin, H.; Tabaei, S. R.; Thomsson, D.; Mirzov, O.; Larsson, P.-O.; Scheblykin, I. G. J. Am. Chem. Soc. 2008, 130, 7042–7051.

JP806963U