Cationic Conjugated Polyelectrolyte Electron Injection Layers: Effect of

Jan 27, 2009 - We examine the influence of halide counteranions on the efficiencies of solution-processed multilayer polymer light-emitting diodes (PL...
3 downloads 11 Views 1MB Size
Download PDF
2950

J. Phys. Chem. C 2009, 113, 2950–2954

Cationic Conjugated Polyelectrolyte Electron Injection Layers: Effect of Halide Counterions Andres Garcia, Jacek Z. Brzezinski, and Thuc-Quyen Nguyen* Mitsubishi Chemical Center for AdVanced Materials, Department of Chemistry and Biochemistry, Center for Polymers and Organic Solids, UniVersity of California, Santa Barbara, California 93106 ReceiVed: July 18, 2008; ReVised Manuscript ReceiVed: NoVember 2, 2008

We examine the influence of halide counteranions on the efficiencies of solution-processed multilayer polymer light-emitting diodes (PLEDs) containing cationic conjugated polyelectrolyte (CPE) electron injection layers (EILs). The parent CPE used in these studies is poly[9,9-bis[6′-(N,N,N-trimethylammonium)hexyl]fluorenealt-co-1,4-phenylene] bromide. Dialysis was used for exchanging counteranions, while X-ray photoelectron spectroscopy (XPS) provides a convenient technique for evaluating the final polymer composition. The luminous efficiencies of PLEDs with a poly(2-methoxy-5-(2′-ethylhexyloxy)-1,4-phenylenevinylene) (MEH-PPV) emissive layer decrease in the order F- > Cl- > Br- > I-. Oxidation of the halide counteranions is likely to occur at the MEH-PPV/CPE interface when the oxidation potential of the counteranion is aligned with the highest occupied molecular orbital of MEH-PPV, that is, Br- and I-. For these heavier halide counteranions, we find that pulsed bias measurements reduce ion migration to the MEH-PPV/CPE interface and result in an increase in the device efficiency. We propose that the oxidation potential of the counterion is a significant factor to consider when selecting CPEs as EILs. Introduction Charge injection and transport play an important role in organic light-emitting diodes (OLEDs),1,2 in which holes are injected from the anode into the highest occupied molecular orbital (HOMO) of the organic semiconductor. Similarly, electrons from the cathode are injected to the lowest unoccupied molecular orbital (LUMO). A balance of both charge carriers is needed to increase the probability of recombination, and hence improve the light output at a given current density. In the absence of interfacial effects, one needs to match the energies of the HOMO and the LUMO with the work function of the anode and cathode, respectively, so to minimize charge injection barriers. Stable metals with high work functions thus typically give rise to larger electron injection barriers when used as cathodes. One therefore needs to rely on multilayer devices, or less stable low work function cathodes such as barium and calcium.3,4 Conjugated polyelectrolytes (CPEs) have been shown recently to function as effective electron injecting layers (EILs) in polymer LEDs (PLEDs).5-11 CPEs are composed of a π-conjugated backbone with pendant groups bearing ionic functionalities. The ionic component allows solubility in polar solvents and thereby fabrication of multilayer devices by deposition atop a neutral emissive layer.12,13 The mechanism for the reduction of electron injection barriers remains under debate, with one model requiring the formation of permanent interfacial dipoles between the cathode and the CPE.5 Another model involves ion migration by the applied electric field and concomitant redistribution of the internal electric field.7 The two functions are independent of each other and may be operating concurrently. A number of CPE structures have been reported with a diversity of backbone structures, appended ionic functionalities, and counterions. How the molecular features of these materials modify the function of the multilayer PLED performances remains to be fully understood.9-11 A recent systematic inves* Corresponding author. E-mail: [email protected].

tigation reported a correlation of the electron mobility in the CPE with device performance.9 Organic counterions have also been shown to lead to orders of magnitude differences in device performance.11 The attractive properties of CPEs with respect to device improvement13 together with the uncertainties in mechanistic function argue in favor of systematic studies with the long-term goal of making concrete structure-function predictions. In this Article, we probe how different halide counterions influence the device function of a cationic CPE as EILs in PLEDs. For testing, we use PLEDs containing the emissive layer poly(2′-methoxy-5-2′-ethylhexyloxy)-1,4-pheylene vinylene) (MEH-PPV) and EILs with poly[9,9-bis[6′-(N,N,N-trimethylammonium)hexyl]fluorene-alt-co-1,4-phenylene] bearing different halide counteranions: F- (PFN+F-), Cl- (PFN+Cl-), Br(PFN+Br-), and I- (PFN+I-). Figure 1 shows the device configuration and relevant molecular structures. Significant differences in devices performances are observed, which correlate with the redox properties of the halide counteranions. This Article is organized as follows. First, optical and X-ray photoelectron spectroscopy (XPS) characterization is provided that confirms ion exchange does not influence greatly electronic properties. Subsequently, we provide the performances of PLEDs that incorporate EILs with different counteranions. We conclude by measuring electron mobilities and PLEDs under pulsed bias conditions and collecting the information in a discussion that provides a mechanism for how structural changes lead to device performance differences. Materials and Methods MEH-PPV14 and PFN+Br-15 were synthesized according to previous reports. PFN+F-, PFN+Cl-, and PFN+I- were acquired from PFN+Br- via ion exchange using dialysis membranes and immersion into solutions containing the potassium salt of the desired halide. Cellulose ester dialysis membranes (Spectrum Laboratories Inc.) with a 1000 molecular weight cutoff were filled with a water solution of PFN+Br-, followed by addition

10.1021/jp806374s CCC: $40.75  2009 American Chemical Society Published on Web 01/27/2009

Cationic Polyelectrolyte Electron Injection Layers

J. Phys. Chem. C, Vol. 113, No. 7, 2009 2951 film. The PEDOT:PSS layer was dried at 150 °C for 1 h. The MEH-PPV solution was spin-coated at ∼1500 rpm for 60 s atop of the PEDOT:PSS layer to yield a ∼80 nm thick film. This step was followed by spin-coating the CPE EIL layers at either 3000 rpm for 60 s (PFN+F-, PFN+Cl-, and PFN+Br-) or 3500 rpm (PFN+I-), yielding ∼25 nm thick films. Devices were completed after drying under a 10-4 torr vacuum overnight by thermal evaporation of Al electrodes at a pressure of 10-8 torr. Two reference devices without the EILs, ITO/PEDOT:PSS/ MEH-PPV/Al and ITO/PEDOT:PSS/MEH-PPV/Ba, were also fabricated for comparison. All devices were fabricated and tested in an inert N2 atmosphere glovebox. Electron-only devices were fabricated by deposition of 100 nm of Al on glass substrates, followed by spin-coating 1.5% w/v CPE solutions at 800 rpm for 60 s yielding films with thicknesses between 100 and 125 nm. The films were then dried under 10-4 torr vacuum overnight before thermal evaporation of ∼5 nm of Ba, followed by 100 nm of Al. Current-voltage (I-V) measurements were recorded with a Keithley 4200 SCS. Voltage measurements were performed with 0.05 V steps and 500 ms delay time for linear increasing/decreasing voltage scans (nonpulsed or continuous bias measurements) and with 500 ms off-times (V ) 0) and 5 ms on-times for step-pulsed voltage scans. Results and Discussion

Figure 1. (a) Chemical structures of MEH-PPV and CPEs used as emissive and electron injection layers, respectively. (b) Schematic drawing of a multilayer PLED structure. (c) Energy levels of the PLED components.

of an excess amount (5:1 weight) of the appropriate potassium halide salt. The dialysis membranes were immersed and stirred in a deionized water bath until the osmotic pressure was reduced (∼4 days) whereupon more salts were added. This procedure was repeated three times, followed by evaporation of water under reduced pressure. Ion exchange was confirmed by using XPS measurements on powdered PFN+X- (X ) F, Cl, Br, or I) samples with a Kratos Axis Ultra XPS system with a base pressure of 1 × 10-10 mbar (UHV), using a monochromated Al KR X-ray source at hν ) 1486 eV. Films were prepared on quartz for spectroscopic measurements or indium tin oxide (ITO) substrates (Thin Film Devices Inc.) for electrical measurements. Substrates were cleaned before use by heating in a 70:30 (v/v) H2SO4:H2O2 solution (H2SO4: H2O2 is extremely exothermic and reactive and must be handled with care) for quartz substrates only, followed by successive rinsing and ultrasonic treatment in water, acetone, isopropyl alcohol, and followed by drying with nitrogen gas for several hours in an oven. The substrates were treated with UV/O3 (UVO Cleaner 42, Jelight Co. Inc.) for an hour prior to polymer deposition. UV-vis absorption and photoluminescence (PL) measurements of polymer films were recorded on a Shimadzu UV-2401 PC diode array spectrometer and a PTI Quantum Master fluorometer. For PLED fabrication, 0.5% w/v solutions of MEH-PPV in toluene, 0.6% w/v methanol solutions of PFN+F-, PFN+Cl-, and PFN+Br-, and 0.6% w/v acetonitrile:methanol (1:4 v/v) solution of PFN+I- were prepared and stirred at 40 °C overnight prior to use. ITO-coated glass substrates were used as the anode to which a solution of poly(3,4-ethylenedioxythiophene):poly(styrenesulfonate), PEDOT:PSS (Baytron 4083, H. C. Stark), was spin-coated at 2000 rpm for 60 s, yielding a ∼40 nm thick

XPS spectra of PFN+X- powders shown in Figure 2 exhibit the correct halide peaks for each sample and no evidence of potassium signals, consistent with greater than 97% exchange of the original bromide ions and the absence of undesired excess salt. As shown in Figure 3, the absorption and PL spectra of PFN+X- films show similar band widths and absorption (∼378 nm) and PL (∼428 nm) maxima. The emission bands exhibit typical vibronic progressions and no additional broad bands at longer wavelengths, indicating lack of π-π aggregates.16-20 The spectroscopic similarities between the CPEs reveal no significant changes in the optical properties by the counteranion or the anion exchange procedure. Figure 3 also shows the electroluminescence (EL) spectrum obtained from a ITO/PEDOT:PSS/MEH-PPV/PFN+F-/Al device, which is typical for all other devices. That only MEHPPV emission is observed indicates that the CPE functions as an electron injection/transport layer. The current density versus voltage (J-V), luminance versus voltage (L-V), and luminous efficiency versus current density (LE-J) characteristics of ITO/PEDOT:PSS/MEH-PPV/PFN+X-/ Al devices and two reference devices (ITO/PEDOT:PSS/MEHPPV/Al and ITO/PEDOT:PSS/MEH-PPV/Ba) are shown in Figure 4. Comparison of Figure 4a and b shows that, at a fixed bias, the differences in luminances for the devices are much larger than in the current densities. Focusing on the PFN+X series, one observes increased luminances (1.8 < 8.9 < 34 < 220 Cd/m2 at 3 V), decreased turn-on voltages (2.4 V > 2.2 V > 1.9 V ) 1.9 V, defined as the point where 0.1 lm of emission is detected), and increased luminous efficiencies (0.03 < 0.05 < 0.24 < 0.60 at 300 mA/cm2) as the size of the anion decreases (PFN+I- f PFN+Br- f PFN+Cl- f PFN+F-). Examination of Figure 4c shows that even with the least effective EIL (PFN+I-), the luminous efficiency of the device is higher than when Al is deposited directly atop MEH-PPV. Furthermore, the PFN+F-/Al cathode leads to devices with efficiencies similar to that of Ba, for which the contact with MEH-PPV is ohmic. The halide effect on PLED efficiency shown in Figure 4c may arise from several possible factors, including dipole

2952 J. Phys. Chem. C, Vol. 113, No. 7, 2009

Garcia et al.

Figure 4. (a) Current density versus bias, (b) luminance versus bias, and (c) luminous efficiency versus current density characteristics of multilayer PLEDs: ITO/PEDOT:PSS/MEH-PPV/Al (b), ITO/PEDOT: PSS/MEH-PPV/Al (O), and ITO/PEDOT:PSS/MEH-PPV/PFN+X-/Al where X ) F (blue curve), Cl (red curve), Br (green curve), and I (purple curve).

Figure 2. XPS measurements of CPEs: PFN+F- (a), PFN+Br- (b), PFN+Cl- (c), and PFN+I- (d).

Figure 3. Normalized absorption and PL spectra of PFN+X- films: PFN+F- (blue), PFN+Cl- (red), PFN+Br- (green), and PFN+I- (purple). Black solid line is the EL spectrum of ITO/PEDOT:PSS/MEH-PPV/ PFN+F-/Al device.

formation at the cathode interface, ion mobility, electron mobility, and electrochemical stability. However, the interfacial dipole at a gold interface has recently been shown to exhibit

little variation for CPEs with different counterions and identical backbones.21 Similarly, ion mobility is unlikely to be a main contributor to the observed trend because CPEs with very different counteranion sizes (PFN+F- and PFN+BIm4-, where BIm4- is tetrakis(imidazolyl)borate) and hence likely different ion mobilities have been shown to exhibit very similar PLED device performance.22 Ion mobility of CPE EILs is believed to have a much greater influence on the response time of PLEDs, relative to overall efficiency.7,22 To examine possible differences in charge transport characteristics, the electron mobilities were measured by using single carrier diode configurations where a PFN+X- layer is sandwiched between two low work function electrodes, that is, Al/PFN+X-/Ba/Al.9,23,24 The large hole injection barrier expected in these devices (>1.5 eV) and smaller energetic difference between the work function of Ba (φBa ≈ 2.7 eV) and the PFN+X- LUMO values (∼2.2 eV)22,25 lead to preferential electron injection. J-V measurements were performed utilizing a step-pulsed voltage technique9,26 that minimizes modifications of the internal electric field and injection barriers due to ion motion. No hysteresis is observed between forward and reverse with voltage sweeps in 0.05 V increments with 500 ms offtime (V ) 0) and 5 ms on-time; see Figure 5. We also note that there is no evidence of light emission, which would necessitate injection of holes and electrons. Electron mobilities were extracted from the data in Figure 5 by fitting to the space-charge-limited-current (SCLC) equation J ) (9/8)εoεrµeV2/L,23,24,27-29 where εo is the vacuum permittivity,

Cationic Polyelectrolyte Electron Injection Layers

J. Phys. Chem. C, Vol. 113, No. 7, 2009 2953

Figure 5. Current density versus bias characteristics of electron only diodes Al/PFN+X-/Ba/Al along with an example of a J ∝ V2 fitted curve (black solid line) where X ) F (blue curve), Cl (red curve), Br (green curve), and I (purple curve).

εr is the relative dielectric constant of the film, µe is the electron mobility, V is the applied voltage, and L is the thickness of the active layer. By this method, µe values of 3.0 × 10-5, 6.7 × 10-5, 1.2 × 10-5, and 5.7 × 10-5 cm2/V · s were determined for PFN+F-, PFN+Cl-, PFN+Br-, and PFN+I-, respectively. There is no obvious correlation of µe with the molecular structure or, more relevant to the discussion, the differences in device performance. The electrochemical stability of counterions has been shown to decrease the performance of light-emitting electrochemical cells (LECs).30 Indeed, we found that there is a correlation between the ease of oxidation of the halide counteranions (φF≈ 7.5 eV, φCl- ≈ 6.0 eV, φBr- ≈ 5.6 eV, and φI- ≈ 5.1 eV)31,32 and the device performance. However, because the CPEs are operating as EILs and not hole-injecting layers (HILs), the oxidation would be expected to occur at the MEH-PPV/CPE interface. Thus, holes are injected into the MEH-PPV layer and accumulate at the MEH-PPV/CPE interface due to the hole blocking properties of the CPEs (HOMOMEH-PPV ≈ 5.3 eV33 versus HOMOCPEs ≈ 5.8 eV21,25). Counteranions of the CPE EIL are anticipated to migrate to the MEH-PPV/CPE interface,7,22 where they are oxidized by the accumulated holes and ultimately lead to deterioration of the semiconducting component. The reactive radical species formed upon oxidation may react with the semiconducting components, possibly leading to alteration of their electronic properties. Pulsed bias conditions were also applied to the PLED measurements to examine the device performance in the absence of ion motion. Figure 6a compares the J-V and L-V characteristics of a device with a PFN+F-/Al cathode under pulsed or continuous bias application. Comparison of the curves shows higher J and L values together with hysteresis between forward and reverse scans under continuous bias conditions. Such features are consistent with the notion of double layer formation at the cathode via ion migration and hole accumulation at the emissive layer interface.7,22 Similar general trends were observed with devices that incorporated PFN+Cl-, PFN+Br-, PFN+I-, and PFN+F-; thus, only the results from the PFN+F- device are shown in Figure 6a. Figure 6b shows the LE response as a function of current density for PLEDs with PFN+F-/Al and PFN+Br-/Al cathodes under continuous or pulsed bias conditions. These data show a drastic reduction in the efficiencies of the devices with the heavier halide. Focusing on the pulsed bias measurements, one observes that the fluoride anion gives rise to higher efficiencies, as compared to bromide. Because these conditions lead to negligible ion motion and given the fact that the electron mobility is similar in both CPE layers, we conclude that the difference in the device efficiency observed is due to different interfacial dipoles.5,34-36

Figure 6. Current density versus bias (black) and luminance versus bias (red) characteristics with (dotted lines) and without (solid lines) step-pulsed bias of ITO/PEDOT:PSS/MEH-PPV/PFN+F-/Al devices (a). (b) Luminous efficiency versus current density of PLED devices with PFN+F- (blue) or PFN+Br- (green) EILs measurement with (dotted lines) and without (solid lines) step-pulsed bias.

Focusing on the PFN+F-/Al device, one observes an increase in LE with ion motion, that is, comparison of continuous and pulsed bias measurements in Figure 6b. The opposite behavior takes place when using the PFN+Br- EIL. Indeed, the performance of the device under continuous bias conditions progressively decreases with increased current density. We propose that with PFN+F- ion motion allows for reduction of the electron injection barrier. However, for PFN+Br-, the accumulation of bromide ions at the MEH-PPV/CPE interface leads to oxidation and therefore device deterioration. These diverging pathways are due to the lower oxidation potential of Br- (5.6 eV) relative to F- (∼7.5 eV). Consistent with this proposal is that PLEDs with PFN+Cl- EILs exhibit behavior similar to that of PFN+F-, while PFN+I- exhibits behavior similar to that of PFN+Br-. Conclusions In summary, a systematic investigation of the influence of the halide counteranion of CPE EILs on multilayer PLED device performance is presented. A general trend in device performance is observed with halide counteranion CPE EILs (PFN+F- > PFN+Cl- > PFN+Br- > PFN+I-), which does not correlate with differences in electron mobilities, as observed in other CPE systems.9 The device efficiency is believed to be influenced by the ease of oxidation of the counteranion. Comparison of device measurements under pulsed bias or continuous modes supports the proposed oxidation process. Higher efficiencies are observed under continuous mode for PFN+F- due to more effective formation of double layers at the interfaces. The opposite is observed with PFN+Br-, which we attribute to bromide migration adjacent to MEH-PPV, where oxidation takes place. The role of the redox properties of the counterions of CPEs in PLED EILs has not received attention previously. As shown here, the chemical nature of this intrinsic part of the CPE structure is essential for determining the ultimate function in an optoelectronic device. More electrochemically stable species should be sought when considering applications as charge injection layers. Acknowledgment. This work is supported by the NSF CAREER Award (DMR# 0547639) and the Mitsubishi Chemical Center for Advanced Materials (MC-CAM) at the University

2954 J. Phys. Chem. C, Vol. 113, No. 7, 2009 of California, Santa Barbara. A.G. thanks the Materials Research Laboratory for the Diversity Fellowship. References and Notes (1) Kulkarni, A. P.; Tonzola, C. J.; Babel, A.; Jenekhe, S. A. Chem. Mater. 2004, 16, 4556. (2) 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.; Lo¨gdlund, M.; Salaneck, W. R. Nature 1999, 397, 121. (3) Scott, J. C.; Kaufman, J. H.; Brock, P. J.; DiPietro, R.; Salem, J.; Goitia, J. A. J. Appl. Phys. 1996, 79, 2745. (4) Cao, Y.; Yu, G.; Parker, I. D.; Heeger, A. J. J. Appl. Phys. 2000, 88, 3618. (5) Wu, H.; Huang, F.; Mo, Y.; Yang, W.; Wang, D.; Peng, J.; Cao, Y. AdV. Mater. 2004, 16, 1826. (6) Wu, H.; Huang, F.; Peng, J.; Cao, Y. Org. Electron. 2005, 6, 118. (7) Hoven, C.; Yang, R.; Garcia, A.; Heeger, A. J.; Nguyen, T.-Q.; Bazan, G. C. J. Am. Chem. Soc. 2007, 129, 10976. (8) Ma, W.; Iyer, P. K.; Gong, X.; Liu, B.; Moses, D.; Bazan, G. C.; Heeger, A. J. AdV. Mater. 2005, 17, 274. (9) Garcia, A.; Yang, R.; Jin, Y.; Walker, B.; Nguyen, T.-Q. Appl. Phys. Lett. 2007, 91, 153502. (10) Niu, X.; Qin, C.; Zhang, B.; Yang, J.; Xie, Z.; Cheng, Y.; Wang, L. Appl. Phys. Lett. 2007, 90, 203513. (11) Yang, R.; Wu, H.; Cao, Y.; Bazan, G. C. J. Am. Chem. Soc. 2006, 128, 14422. (12) Steuerman, D. W.; Garcia, A.; Dante, M.; Yang, R.; Lo¨fvander, J. P.; Nguyen, T.-Q. AdV. Mater. 2008, 20, 528. (13) Zeng, W.; Wu, H.; Zhang, C.; Huang, F.; Peng, J.; Yang, W.; Cao, Y. AdV. Mater. 2007, 19, 810. (14) Neef, C. J.; Ferraris, J. P. Macromolecules 2000, 33, 2311. (15) Yang, R.; Wu, H.; Cao, Y.; Bazan, G. C. J. Am. Chem. Soc. 2006, 129, 14422. (16) Tan, C. Y.; Pinto, M. R.; Schanze, K. S. Chem. Commun. 2002, 5, 446.

Garcia et al. (17) Schnablegger, H.; Antonietti, M.; Go¨ltner; Hartmann, J.; Co¨lfen, H.; Samorı´, P.; Rabe, J. P.; Ha¨ger, H.; Heitz, W. J. Colloid Interface Sci. 1999, 212, 24. (18) Pinto, M. R.; Kristal, B. M.; Schanze, K. S. Langmuir 2003, 19, 6523. (19) Gao, Y.; Wang, C. C.; Wang, L.; Wang, H. L. Langmuir 2007, 23, 7760. (20) Wang, F.; Bazan, G. C. J. Am. Chem. Soc. 2006, 128, 15795. (21) Seo, J. H.; Nguyen, T.-Q. J. Am. Chem. Soc. 2008, 130, 10042. (22) Hoven, C.; Yang, R.; Garcia, A.; Crocett, V.; Heeger, A. J.; Bazan, G. C.; Nguyen, T.-Q. Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 12730. (23) Blom, P. W. M.; de Jong, M. J. M.; Vleggaar, J. J. M. Appl. Phys. Lett. 1996, 68, 3308. (24) Blom, P. W. M.; de Jong, M. J. M.; van Munster, M. G. Phys. ReV. B 1997, 55, R656. (25) Huang, F.; Wu, H.; Wang, D.; Yang, W.; Cao, Y. Chem. Mater. 2004, 16, 708. (26) Garcia, A.; Nguyen, T. Q. J. Phys. Chem. C 2008, 112, 7054. (27) Lampert, M. A.; Mark, P. Current Injection in Solids; Academic: New York, 1970. (28) Bozano, L.; Carter, S. A.; Scott, J. C.; Malliaras, G. G.; Brock, P. J. Appl. Phys. Lett. 1999, 74, 1132. (29) Goh, C.; Kline, R. J.; McGehee, M. D.; Kadnikova, E. N.; Fre´chet, J. M. J. Appl. Phys. Lett. 2005, 86, 122110. (30) Edman, L.; Moses, D.; Heeger, A. J. Synth. Met. 2003, 138, 441. (31) CRC Handbook of Chemistry and Physics, 88th ed.; Lide, D., Ed.; CRC: Cleveland, OH, 2007/08. (32) de Leeuw, D. M.; Simenon, M. M. J.; Brown, A. R.; Einerhand, R. E. F. Synth. Met. 1997, 87, 53. (33) Campbell, I. H.; Hagler, T. W.; Smith, D. L. Phys. ReV. Lett. 1996, 76, 1900. (34) Ishii, H.; Sugiyama, K.; Ito, E.; Seki, K. AdV. Mater. 1999, 11, 605. (35) Brown, T. M.; Millard, I. S.; Lacey, D. J.; Butler, T.; Burrouhnes, J. H.; Friend, R. H. J. Appl. Phys. 2003, 93, 6159. (36) Zhang, F. J.; Vollmer, A.; Zhang, J.; Xu, Z.; Rabe, J. P.; Koch, N. Org. Electron. 2007, 8, 606.

JP806374S