Phosphorescence from iridium complexes doped into polymer blends

Transcription

1 JOURNAL OF APPLIED PHYSICS VOLUME 95, NUMBER 3 1 FEBRUARY 2004 Phosphorescence from iridium complexes doped into polymer blends Xiong Gong a) Center for Polymers and Organic Solids, University of California at Santa Barbara, Santa Barbara, California Sang-Hyun Lim Department of Chemistry, University of Illinois at Urbana Champaign, Champaign, Illinois Jacek C. Ostrowski and Daniel Moses Center for Polymers and Organic Solids, University of California at Santa Barbara, Santa Barbara, California Christopher J. Bardeen Department of Chemistry, University of Illinois at Urbana Champaign, Champaign, Illinois Guillermo C. Bazan Center for Polymers and Organic Solids, University of California at Santa Barbara, Santa Barbara, California Received 25 July 2003; accepted 28 October 2003 Energy transfer from the polymer blends, poly vinylcarbazole PVK with 2-tert-butylphenyl-5-biphenyl-1,3,4-oxadiazol PBD, to an organometallic emitter, tris 9,9-dihexyl-2- phenyl-4 - -pyridin-2 -yl fluorene iridium III Ir(DPPF) 3, is investigated by steady-state and time-resolved photoluminescence PL spectroscopy. A redshifted PL and slow fluorescence decay are due to the formation of an exciplex in PVK-PBD blends. A decrease in intensity in polymer blends observed at 425 nm with increasing concentrations of Ir(DPPF) 3 and an evident rising feature observed in films with 1 wt % Ir(DPPF) 3 in the range of 578 to 615 nm within a 200 ns timescale indicate that efficient Förster energy transfer from exciplex to Ir(DPPF) 3 occurs. The electrophosphorescent light-emitting diodes fabricated with PVK-PBD doped with Ir(DPPF) 3 have external quantum efficiency of 8% ph/el, luminous efficiency of 29 cd/a and brightness greater than 3500 cd/m 2 at 1 wt % Ir(DPPF) 3. The devices exhibited no electroluminescence EL emission from PVK or PBD even at a low concentration of Ir(DPPF) wt %, which indicates that the dominant mechanism in EL is charge trapping rather than energy transfer American Institute of Physics. DOI: / I. INTRODUCTION Utilizing triplet-based emitting centers in polymeric and organic light-emitting diodes LEDs, and thereby capturing both singlet and triplet excited states, the internal quantum efficiency can, in principle, be 100%. 1,2 The utilization of triplet emitters to improve efficiency was proposed by several groups. 3 8 Recently, considerable progress has been demonstrated with electrophosphorescent OLEDs based on small molecules and PLEDS based on polymers that employ organometallic emitters as dopants Electrophosphorescent PLEDs are important because they can be fabricated by processing materials from solution, e.g., by spin casting, screen printing or inkjet printing. 4,8,14 16 Several groups have reported phosphorescent PLEDs using poly vinylcarbazole PVK as the host polymer. 7,8,14 19 We also demonstrated high-performance electrophosphorescent PLEDs by employing iridium complexes doped in nonconjugated and conjugated polymer hosts. 8,14 16 However, the mechanism for electrophosphorescent PLEDs is rarely reported. 20,21 In electrophosphorescent PLEDs, electrons and holes are initially injected into polymer host materials and excitation energy is transferred to organometallic emitters. There are three possible excitation processes that need to be considered: 1 energy transfer from the polymer singlet exciton Förster energy transfer, 22,23 2 energy transfer from the polymer triplet exciton Dexter energy transfer, and 3 sequential electron and hole capture by organometallic emitters. Förster energy transfer has been widely used to explain energy transfer phenomena in molecular dye systems, 24 photosynthetic aggregates, 25 polymer polymer systems, 26,27 and more recently it has been applied to the polymerorganometallic emitter systems. 8,14 16 Dexter energy transfer has also been studied in doped and undoped OLEDs. 28 However, the triplet population in the polymer is not affected by the presence of organometallic emitters, indicating Dexter energy transfer processes are very week in organometallic doped polymer systems. 20 Thus, the dominant mechanisms in polymer-organometallic emitters system are Förster energy transfer and/or charge trapping. In this article, we investigate the excitation energy transfer from polymer blends, PVK with 2-tert-butylphenyl-5- biphenyl-1,3,4-oxadiazol PBD, to tris 9,9-dihexyl-2- phenyl-4 - -pyridin-2 -yl fluorene iridium III Ir(DPPF) 3 by steady-state and time-resolved photolumia Electronic mail: xgong@physics.ucsb.edu /2004/95(3)/948/6/$ American Institute of Physics

2 J. Appl. Phys., Vol. 95, No. 3, 1 February 2004 Gong et al. 949 FIG. 2. Absorption spectra of PVK, PBD and PVK-PBD thin films. FIG. 1. a Molecular structures of PVK, PBD and Ir(DPPF) 3 ; b LUMO and HOMO levels of PVK, PBD and Ir(DPPF) 3. nescent spectroscopy. It is seen that efficient Förster energy transfer occurs from exciplexes formed in PVK and PBD to Ir(DPPF) 3. Electrophosphorescent PLEDs fabricated with PVK-PBD with varying concentrations of Ir(DPPF) 3 indicate that the dominant mechanism in electroluminescence is charge trapping rather than energy transfer. II. EXPERIMENT The synthesis and characterization of Ir(DPPF) 3 is reported elsewhere, 29 PVK and PBD were purchased from Aldrich and used without further purification. The molecular structures of PVK, PBD, and Ir(DPPF) 3 with their highest occupied molecular orbital HOMO and lowest unoccupied molecular orbital LUMO levels are shown in Fig. 1. All films were prepared by spin casting from dichloroethane solutions onto quartz substrates for spectroscopy studies. All films had approximately the same thickness for direct comparison of the absorption and emission intensities. The absorption spectra of films were measured on a Shimadzu UV-2401PC UV-vis recording spectrophotometer. Steady-state photoluminescence PL spectra of films were measured on a Spex Fluoromax-2 spectrometer. For time-resolved PL measurements, the films are mounted in a compressed helium cryostat under argon atmosphere. Excitation pulses at 325 nm are generated by using a homebuilt noncollinear optical parametric amplifier NOPA with 800 nm output of a 40 khz regenerative amplifier Spectra Physics, Spitfire. 30 Continuum seed pulses are generated by focusing the 800 nm beam in a 2 mm sapphire plate and amplified in a1mmbbo 31, type I phase matching pumped by the 400 nm pumping beam. In this experiment, 650 nm pulses are selected by an interference filter 10 nm full width at half maximum FWHM, Andover and amplified. Then the 650 nm NOPA output is frequency doubled using 0.1 mm BBO and the remaining 650 nm pulses are removed with two UG11 filters. Picosecond time-resolved fluorescence spectra are measured using a streak camera Hamamatsu, streak scope C The samples are excited with the beam at an angle of about 20 from the sample surface normal. The fluorescence is collected normal to the sample surface, collimated, filtered with a long-wave pass filter 400 nm cutoff, Andover and focused into a spectrometer Spectra Pro-150 with a 150 grooves/mm grating to disperse the spectrum in front of the streak camera. The excitation fluence is kept below 1 J/cm 2 to avoid high fluence effects such as amplified spontaneous emission. 32,33 The instrumental response time of this measurement is about 50 ps within a 5 ns sweep window. For device fabrication, we employed only the single-active-layer configuration with poly 3,4-ethylene dioxythiophene :poly styrene sulfonic acid PEDOT:PSS on indium tin oxide ITO as the hole-injecting bilayer electrode. The device configuration is (ITO)/PEDOT:PSS/PVK-PBD:Ir(DPPF) 3 /Ca/Ag. PEDOT: PSS was first spin cast onto the ITO surface, then the emitting layer was spin cast onto a film with thickness of approximately 100 nm. The Ca/Ag cathode was deposited through a shadow mask by thermal evaporation at Torr. III. RESULTS AND DISCUSSION A. Host polymer: PVK with PBD The absorption spectra of PVK, PBD, and blends of PVK-PBD 40 wt % films are shown in Fig. 2. It can be seen that films of PVK and PVK-PBD 40 wt % have similar absorption spectra with peak maxima at 345 and 325 nm. The film of neat PBD has absorbance at 325 nm. Therefore, PVK, PBD, and PVK-PBD blends can be directly excited

3 950 J. Appl. Phys., Vol. 95, No. 3, 1 February 2004 Gong et al. FIG. 3. a PL spectra of PVK, PBD and PVK-PBD; b PL decay of PVK, and PVK-PBD films. simultaneously at 325 nm. The PL spectra of PVK, PBD, and PVK-PBD 40 wt % films are shown in Fig. 3 a. Excitation of PVK at 325 nm, a maximum PL at 410 nm with a shoulder at 380 nm is observed, which is due to the two different types of excimer formed by PVK. 34,35 The maximum PL for PBD is at 390 nm, slightly shorter wavelength than PVK. However, the maximum PL for PVK-PBD blends, excited at 325 nm, is at 425 nm, which is a level lower in energy than either PVK or PBD. It was observed that the PL intensities of PVK-PBD blends increased with an increase of the weight ratio of PBD. This indicates that PBD is not an electron accepter see energy levels in Fig. 1 b that likes C 60 in MEH-PPV poly 2-methoxy-5 2-ethylhexyloxy phenylenevinylene, and quenches PL emission of MEH-PPV; 36 on the contrary, PBD contributes to the emission at 425 nm. Figure 3 b shows fluorescence decay from films of PVK, PBD, and PVK-PBD blends monitored at 400 nm. The data for PBD are fitted with a single exponential decay with time constant of 1.8 ns. The data for PVK is fitted with biexponential decay with time constants of 0.6 and 8 ns. The fluorescence decay in the PVK-PBD 40 wt % film is similar to that of PVK, but has significant long-lived components, FIG. 4. a Absorption and PL spectra of Ir(DPPF) 3 and PVK-PBD 40 wt % films; b PL spectra of films of PVK-PBD and PVK-PBD doped with different concentrations of Ir(DPPF) 3. with time constants of 1.3 and 35 ns. The slow decay of PVK-PBD blends is due to the formation of an exciplex. 37,38 Therefore, we concluded that the emitting species in PVK- PBD blends is from an exciplex, 39,40 resulting in redshifted PL emission and slow fluorescence decay. B. Blends: PVK-PBD with Ir DPPF 3 The absorption and PL spectra of neat Ir(DPPF) 3 and PVK-PBD 40 wt % films are shown in Fig. 4 a. There is a good overlap between the emission spectrum of PVK-PBD exciplexes and the absorption bands of singlet metal-ligand charge transfer MLCT 410 nm and triplet MLCT 440 nm. Such a guest host system meets the requirement for efficient Förster energy transfer from the singlet-excited state of the host to a singlet MLCT band of the guest, Ir(DPPF) 3. Fast intersystem crossing to the triplet state of the Ir(DPPF) 3 and subsequent emission from this state are built-in features.

4 J. Appl. Phys., Vol. 95, No. 3, 1 February 2004 Gong et al. 951 To test the efficiency of Förster energy transfer, films containing PVK-PBD 40 wt % with varying concentrations of Ir(DPPF) 3 were prepared and examined by optical excitation at 325 nm. Figure 4 b shows the normalized PL spectra of PVK-PBD 40 wt % films and PVK-PBD 40 wt % films with varying concentrations of Ir(DPPF) 3. The PL profile contains two peaks: one is centered 425 nm from emission of the exciplex and the other is at 550 nm with a shoulder at 600 nm from Ir(DPPF) 3 triplet emission. The emission intensity at 425 nm was reduced significantly with an increase in the concentrations of Ir(DPPF) 3. At 8 wt% Ir(DPPF) 3, there is virtually no emission observed from exciplexes, and Ir(DPPF) 3 emission is completely dominant. This indicates that 8 wt % of Ir(DPPF) 3 was required to completely quench the host emission under optical excitation. Therefore, the dominant emission by Ir(DPPF) 3 is evidence that confirms efficient Förster energy transfer from PVK-PBD to Ir(DPPF) 3. The efficiency of Förster energy transfer from PVK-PBD to Ir(DPPF) 3 is given by 21,22 k FET 1 k FET 1/ d 1 R/R 0 6, 1 where K FET is the rate of Förster energy transfer FET from PVK-PBD to Ir(DPPF) 3, which is given by K FET 1 d R/R 0 6, 2 where d is the lifetime of the host in the absence of the guest, R is the distance between the host and the guest, and R 0 is the characteristic Förster radius which is given by R Fd v v v 4 dv, where depends on the relative orientation of the host and the guest dipole moments, the quantum yield of the host in the absence of the guest, and the refraction of the medium; F d (v) and (v) are the fluorescence and extinction spectra of the host and the guest, respectively. The distance between the host and the guest, R, is given by 41 R N 4 1/3 G* 3, 4 where N G is the doping concentration of Ir(DPPF) 3. By combining Eqs. 1 and 4, the efficiency of Förster energy transfer is given by 1 1 R 0 N 4 1/3 G* All films studied were approximately the same thickness and were optically excited under the same conditions. Moreover, the films made by polymer blends with Ir(DPPF) 3 had similar absorption and PL spectra. Thus,, F d (v), and (v) in Eq. 2 can be assumed to be constant, which results in constant R 0 for films with different concentrations of Ir(DPPF) 3. In studies of conjugated polymer blends, a value of R 0 30 Å was obtained from time-resolved studies of Förster energy transfer. 41 However, the precise value of R 0 FIG. 5. Relative energy transfer efficiency as a function of Ir(DPPF) 3 concentrations and the distance between the host and the guest marks represent the calculated values, and lines represent the fitted curves. has not yet been determined for this system. Therefore, the relative energy transfer efficiency can be predicted either as a function of the Ir(DPPF) 3 concentration, N G, or the distance between the host and the guest, R. The results are shown in Fig. 5. At lower doping concentrations, Förster energy transfer from the polymer blends to Ir(DPPF) 3 is incomplete because the average distance from a photoexcited polymer chain to the nearest Ir(DPPF) 3 is too large (R R 0 ). 4 At higher doping concentrations (R R 0 ), all energy was transferred to Ir(DPPF) 3, but the photoemission efficiency was evidently reduced by concentration quenching. We also found the PL of neat Ir(DPPF) 3 film is extremely weak. 29 Thus, these results indicate that there is complete Förster energy transfer from polymer blends to Ir(DPPF) 3 at intermediate doping concentrations as expected for the relatively long-range dipole dipole Förster coupling. 5,7 The exciplex emission decay of PVK-PBD films with varying concentrations of Ir(DPPF) 3 monitored at 440 nm is shown in Fig. 6. The exponential decay observed decreased with an increase of the concentration of Ir(DPPF) 3 of the order of by weight ratio. The decay time of PVK-PBD with 0.1 wt % Ir(DPPF) 3 is much shorter than that of PVK-PBD with 0.01 wt % Ir(DPPF) 3. This indicates that Förster energy transfer in films with 0.1 wt % Ir(DPPF) 3 is more efficient than that with 0.01 wt % Ir(DPPF) 3. Moreover, the decay time of film with 1 wt % Ir(DPPF) 3 is half that with 0.01 wt % Ir(DPPF) 3. This observation indicates a significant increase at least one order of magnitude of Förster energy transfer in films with 1 wt % Ir(DPPF) 3 over films with 0.01 wt % Ir(DPPF) 3. Therefore, higher doping concentrations of Ir(DPPF) 3 in polymer blends induce greater Förster energy transfer from the exciplex to Ir(DPPF) 3. The emission decay of PVK-PBD with varying concentrations of Ir(DPPF) 3 monitored in the range of nm

5 952 J. Appl. Phys., Vol. 95, No. 3, 1 February 2004 Gong et al. FIG. 6. Emission decay of films of PVK-PBD doped with different concentrations of Ir(DPPF) 3 monitored at 440 nm: Ir(DPPF) 3 /PVK-PBD 0.01 wt %; Ir(DPPF) 3 /PVK-PBD 0.1 wt % and Ir(DPPF) 3 /PVK-PBD 1wt%. within a 200 ns timescale window is shown in Fig. 7. The signal from PVK-PBD films with 0.01 wt % Ir(DPPF) 3 decays faster compared to that with films with 0.1 wt % Ir(DPPF) 3. The difference in films of PVK-PBD with 1 and 0.1 wt % Ir(DPPF) 3 is that a rising feature is evidently observed in films with 1 wt % Ir(DPPF) 3. Note the triplet emission lifetime of Ir(DPPF) 3 is around 5 s, 28 and there is an emission tail in the PL spectra of polymer blends in the range of nm. Accordingly, the fast decay of films with 0.01 wt % Ir(DPPF) 3 indicates that the triplet emission signal is too weak to be detected, thereby energy transfer from polymer blends to Ir(DPPF) 3 is not as efficient as that for films with 1 wt % Ir(DPPF) 3. C. Electrophosphorescent PLEDs Figure 8 shows the electroluminescence EL spectrum of the devices made with pure PVK-PBD blends and EL FIG. 7. Emission decay of films of PVK-PBD with different concentrations of Ir(DPPF) 3 from 578 to 615 nm: Ir(DPPF) 3 /PVK-PBD 0.01 wt %, Ir(DPPF) 3 /PVK-PBD 0.1 wt %, and Ir(DPPF) 3 /PVK-PBD 1wt%. FIG. 8. EL spectra of LEDs made from PVK-PBD 40 wt % with varying concentrations of Ir DPPF. spectra of the devices made with PVK-PBD blends with different concentrations of Ir(DPPF) 3. The EL spectrum of the devices with PVK-PBD blends is at 425 nm which is in good agreement with the PL spectrum of films with PVK-PBD blends. The EL data for all devices containing Ir(DPPF) 3 only show the characteristic spectrum of Ir(DPPF) 3, with one peak at 550 nm and another at 600 nm. There is no EL emission of PVK or PBD from any devices containing Ir(DPPF) 3, even at 0.1 wt % guest concentration. Note that the concentration required to completely quench the PL from PVK-PBD 40 wt % was 8 wt % of Ir(DPPF) 3. The absence of PVK or PBD EL emission from Ir(DPPF) 3 doped devices, even at low concentrations of Ir(DPPF) wt %, is consistent with charge trapping of the Ir complex, rather than Förster transfer, as the dominant mechanism in LEDs. We conclude, therefore, that the iridium complex traps both electrons and holes and leads to emission from the Ir complex. If energy transfer were the dominant EL mechanism, the EL emission of PVK or PVK-PBD would be expected to appear when the triplet of Ir(DPPF) 3 becomes saturated. 10,42,43 Similar charge trapping has been observed in PVK-PBD doped with other Ir complexes, 8,9,14 and in a conjugated polymer doped with an Ir complex. 15 The current density versus voltage and brightness versus voltage characteristics of devices made with PVK-PBD 40 wt % with 1 wt % of Ir(DPPF) 3 are shown in Fig. 9. The device turns on at approximately 10 V and has brightness greater than 3500 cd/m 2 at 30 V. A Lambertian intensity profile was assumed in order to calculate the external quantum efficiency QE, QE ext % ph/el, luminous efficiency LE cd/a and power efficiency PE lum/w had the following results: QE ext 8% ph/el at current density J of 15 ma/cm 2, LE 29 cd/a at J 15 ma/cm 2 and PE 3.3 lum/w. The QE ext and LE were observed at 1 wt % Ir(DPPF) 3 and 15 ma/cm 2, i.e., much lower than reported for Ir(ppy) 3 /PVK devices. 7,9

6 J. Appl. Phys., Vol. 95, No. 3, 1 February 2004 Gong et al. 953 FIG. 9. Brightness and current density characteristics of the ITO/PEDOT/ PVK-PBD: Ir(DPPF) 3 /Ca/Ag device with 1 wt % Ir(DPPF) 3 concentration as a function of the voltage applied. IV. CONCLUSION We have investigated the excitation energy transfer in films of PVK-PBD blends with Ir(DPPF) 3 by steady-state and time-resolved photoluminescence spectroscopy studies. Observations of polymer blends demonstrate a redshifted PL and slow fluorescence decay are due to the formation of an exciplex. The PL intensity and PL decay of PVK-PBD blends at 425 nm decreased significantly with an increase of the concentration of Ir(DPPF) 3. A rising feature is observed in films with 1 wt % Ir(DPPF) 3 in the range of nm within a 200 ns timescale. These results indicate that there is efficient Förster energy transfer from the PVK-PBD blends to Ir(DPPF) 3. The electrophosphorescent PLEDs fabricated with PVK-PBD with Ir(DPPF) 3 have external quantum efficiency of 8% ph/el, luminous efficiency of 29 cd/a and brightness greater than 3500 cd/m 2 at 1 wt. % Ir DPPF 3. The devices exhibited no EL emission from PVK or PBD even for a low concentration of Ir DPPF wt. %, which indicates that the dominant mechanism in EL is charge trapping rather than energy transfer. More important, these results demonstrate that understanding the excitation energy transfer, processes in systems of organometallic emitters doped into polymers is important in designing appropriate phosphorescence molecules and polymeric hosts to achieve high-performance electrophosphorescent PLEDs. ACKNOWLEDGMENT This work was supported by Air Force Office of Scientific Research through the Muri Center Polymeric Smart Skins, Program Office, Dr. Charles Lee. 1 Y. Cao, I. D. Parker, G. Yu, C. Zhang, and A. J. Heeger, Nature London 397, M. Wohlgenannt, K. Tandon, S. Mazumdar, S. Ramasesha, and Z. V. Vardeny, Nature London 409, J. Kido, H. Haromatchi, K. Hongawa, K. Nagai, and K. Okuyama, Appl. Phys. Lett. 65, M. D. McGehee, T. Bergstedt, C. Zhang, A. P. Saab, M. B. O Regan, G. C. Bazan, V. I. Srdanov, and A. J. Heeger, Adv. Mater. Weinheim, Ger. 11, M. A. Baldo, M. E. Thompson, and S. R. Forrest, Nature London 403, C. Adachi, M. A. Baldo, and S. R. Forrest, Appl. Phys. Lett. 77, C. L. Lee, K. B. Lee, and J. J. Kim, Appl. Phys. Lett. 77, X. Gong, J. C. Ostrowski, M. R. Robinson, D. Moses, G. C. Bazan, and A. J. Heeger, Adv. Mater. Weinheim, Ger. 14, S. Lamansky, R. C. Kwong, M. Nugent, P. I. Djurovich, and M. E. Thompson, Org. Electron. 2, D. F. O Brien, C. Giebler, R. B. Fletcher, J. Cadlby, L. C. Palilis, D. G. Lidzey, P. A. Lane, D. C. Bradley, and W. Blau, Synth. Met. 116, W. Zhu, Y. Q. Mo, M. Yuan, W. Yang, and Y. Cao, Appl. Phys. Lett. 80, M. Lkai, S. Tokito, Y. Sakamoto, T. Auzuki, and Y. Taga, Appl. Phys. Lett. 79, B. W. D Andrade, M. E. Thompson, and S. R. Forrest, Adv. Mater. Weinheim, Ger. 14, X. Gong, J. C. Ostrowski, D. Moses, G. C. Bazan, and A. J. Heeger, Appl. Phys. Lett. 81, X. Gong, J. C. Ostrowski, D. Moses, G. C. Bazan, A. J. Heeger, M. S. Liu, and A. K-Y. Jen, Adv. Mater. Weinheim, Ger. 15, X. Gong, J. C. Ostrowski, D. Moses, G. C. Bazan, and A. J. Heeger, J. Polym. Sci., Part. B: Poly. Phys., 2003 in press. 17 M. J. Yang and T. Tshutsui, Jpn. J. Appl. Phys., Part 2 39, L Y. Kawamura, S. Yanagida, and S. R. Forrest, J. Appl. Phys. 92, S. Lamansky, P. I. Djurovich, F. Abdel-Razzaq, S. Garon, D. L. Murphy, and M. E. Thompson, J. Appl. Phys. 92, I. H. Campbell, D. L. Smith, S. Tretiak, R. L. Martin, C. J. Neef, and J. P. Ferraris, Phys. Rev. B 65, S. H. Lim, X. Gong, J. C. Ostrowski, G. C. Bazan, D. Moses, and C. J. Bardeen, Chem. Phys. Lett. 376, T. Förster, Discuss. Faraday Soc. 7, D. L. Dexter, J. Chem. Phys. 21, S. A. Azim, H. A. El-Daly, S. A. El-Daly, Kh A. Abou-Zeid, E. M. Ebeid, and J. R. Heldt, J. Chem. Soc., Faraday Trans. 92, P. D. Laible, R. S. Knox, and T. G. Owens, J. Phys. Chem. B 102, D. Vacar, E. S. Maniloff, D. W. McBranch, and A. J. Heeger, Phys. Rev. B 56, D. Vacar, A. Dogariu, and A. J. Heeger, Adv. Mater. Weinheim, Ger. 10, M. A. Baldo, D. F. O Brien, Y. You, A. Shoustikov, S. Sibley, M. E. Thompson, and S. R. Forrest, Nature London 395, J. C. Ostrowski, M. R. Robinson, A. J. Heeger, and G. C. Bazan, Chem. Commun. Cambridge 7, S. H. Lim, T. G. Bjorklund, and C. J. Bardeen, J. Chem. Phys. 117, G. Bjorklund, S.-H. Lim, and C. J. Bardeen, J. Phys. Chem. B 105, G. J. Denton, N. Tessler, M. A. Kalintsev, and R. H. Friend, Adv. Mater. Weinheim, Ger. 9, V. Doan, V. Tran, and B. J. Schwartz, Chem. Phys. Lett. 288, H. F. Kauffmann, B. Mollay, W. D. Weixelbaumer, and J. Bürbaumer, J. Chem. Phys. 85, H. Masuhara, N. Tamai, and N. Mataga, Chem. Phys. Lett. 91, N. S. Sariciftci, L. Smilowitz, A. J. Heeger, and F. Wudl, Science 258, H. Sakai, A. Itaya, and H. Masuhara, J. Phys. Chem. 93, T. J. Kesti, N. V. Tkachenko, V. Vehmanen, H. Yamada, H. Imahori, S. Fukuzumi, and H. Lemmetyinen, J. Am. Chem. Soc. 124, Y. Z. Wang, D. D. Gebler, D. K. Fu, T. M. Swager, A. G. MacDiarmid, and A. J. Epstein, Synth. Met. 85, D. D. Gebler, Y. Z. Wang, S. W. Jessen, J. W. Blatchford, A. G. MacDiarmid, T. M. Swager, D. K. Fu, and A. J. Epstein, Synth. Met. 85, A. Digariu, R. Gupta, A. J. Heeger, and H. Wang, Synth. Met. 100, G. Peter, H. Bässler, W. Schrof, and H. Port, Chem. Phys. 94, M. A. Baldo and S. R. Forrest, Phys. Rev. B 62, ; 62,