Tuning Intrachain versus Interchain Photophysics via Control of the

Nov 4, 2008 - London Centre for Nanotechnology, and Department of Physics and Astronomy, University College London, Gower Street, London, WC1E 6BT, ...
0 downloads 0 Views 1MB Size
Download PDF
NANO LETTERS

Tuning Intrachain versus Interchain Photophysics via Control of the Threading Ratio of Conjugated Polyrotaxanes

2008 Vol. 8, No. 12 4546-4551

Sergio Brovelli,† Gianluca Latini,† Michael J. Frampton,‡ Shane O. McDonnell,‡ Francine E. Oddy,‡ Oliver Fenwick,† Harry L. Anderson,*,‡ and Franco Cacialli*,† London Centre for Nanotechnology, and Department of Physics and Astronomy, UniVersity College London, Gower Street, London, WC1E 6BT, United Kingdom, and Department of Chemistry, UniVersity of Oxford, Chemistry Research Laboratory, Mansfield Road, Oxford, OX1 3TA, United Kingdom Received September 12, 2008; Revised Manuscript Received October 20, 2008

ABSTRACT Effective nanoscale control of intermolecular interactions in conjugated polymers is needed for the optimal development and exploitation of the latter in low-cost, large-area consumer electronics items, such as light-emitting and photovoltaic diodes, or transistors. Here we report our investigations on insulated molecular wires constituted by conjugated polymers threaded into cyclodextrin rings. Until now, there has been no detailed quantitative understanding of the role of progressive cyclodextrin encapsulation (quantifiable by the so-called “threading ratio”, TR, or number of cyclodextrins per repeat unit) in tailoring the photophysics of the conjugated polymeric wires. We combine spectroscopic, electrical and surface analysis techniques to elucidate how the TR of cyclodextrin-threaded molecular wires controls formation of interchain species and related physical properties (0 < TR e 2.3; the maximum theoretical TR for close-packed CDs is 2.8). Increasing TR enhances the solid-state photoluminescence (PL) and electroluminescence quantum efficiency. To unravel the effect of progressive encapsulation on the intrachain decay kinetics of the polymer backbone, we added an electron-accepting quenching agent, methyl viologen (MV), to the polymer solutions. MV predominantly quenches the aggregate PL, thus enabling measurement of the decay kinetics of the intrinsic exciton even for low-TR polyrotaxanes, for which the different contributions are otherwise difficult to disentangle.

Effective nanoscale control of intermolecular interactions, and in particular of their influence on the photophysics of conjugated polymers, is needed for the optimal development and exploitation of the latter in low-cost, large-area, and even disposable consumer electronics items, such as light-emitting1 and photovoltaic diodes,2 and transistors.3 Low-cost solution processability and tunability of the optical properties are competitive advantages of this class of materials that also benefit from availability of a large variety of synthetic routes for production of purpose-designed, nanoengineered macromolecules. However, intermolecular interactions can lead to formation of interchain species (aggregates and/or excimers) and therefore can be detrimental for the photophysics of such polymer-based devices.4-9 This has therefore become an important issue with significant and increasing attention

* To whom correspondence should be addressed. E-mail: (F.C.) [email protected]; (H.L.A.) [email protected]. † University College London. ‡ University of Oxford. 10.1021/nl802775a CCC: $40.75 Published on Web 11/04/2008

 2008 American Chemical Society

being dedicated to the design, preparation, and characterization of polymeric systems with reduced tendency to aggregate.10-15 An important strategy to achieve such goals is supramolecular assembly16 of which threaded molecular wires (TMWs) are significant and relevant examples.17,18 For example, one class of TMWs consists of conjugated polymers threaded through β-cyclodextrin (β-CD) macrocycles,19,20 which provide molecular spacers with the ability to reduce intermolecular interactions21 and thus the tendency to form aggregates.22 However, there is currently no detailed quantitative understanding of the role of progressive cyclodextrin encapsulation (quantifiable by the so-called “threading ratio”, TR, or number of cyclodextrins per repeat unit) in tailoring the photophysics of the conjugated polymeric wires. Here, we combine spectroscopic, electrical, and surface analysis techniques to elucidate how the TR controls formation of interchain species and related physical properties. In the synthesis of β-cyclodextrin PDV polyrotaxanes, we can vary the TR in the range 0.0-2.3 by varying the reaction conditions; the maximum theoretical TR for close-packed

Figure 1. Top: chemical structure of PDV.Li⊂β-CD and definition of threading ratio (TR ) x/(n +1)). Middle: space-filling model of an oligomer with n ) 3 and TR ) 2. The structure of unthreaded PDV.Li is identical but without the β-CD rings. Bottom: timeintegrated emission spectra of unthreaded poly(4,4′-diphenylene vinylene) (PDV) and PDV polyrotaxane solutions at 0.05 mg/ml concentration of the conjugated moiety for increasing TRs. The average emission energy () extracted from the spectra is displayed in the inset.

CDs is 2.8.22 Each sample has a distribution of species with a range of TRs and chain lengths, as revealed by massspectrometric analysis of closely related polyrotaxanes.23 The width of the distribution of the TRs must become narrower as we approach the theoretical limit for closely packed CDs (2.8).22 For the less highly threaded rotaxanes there will be a broader distribution of threading ratios, but the average threading ratio reflects the most abundant species. We find that increasing the threading ratio enhances the solid-state photoluminescence (PL) and electroluminescence (EL) quantum efficiency and that, surprisingly, the spectroscopic properties of the aggregates are independent of the threading ratio. This is remarkable, since with the assumption of a random distribution of the cyclodextrins along the backbone this implies that the dimension of the aggregates is small compared to the length of the conjugated segments or that the cyclodextrin rings tend to pack together. To unravel the effect of progressive encapsulation on the intrachain decay kinetics of the polymer backbone, we added the electronaccepting quenching agent methyl viologen (MV) to the polymer solutions.24 The time-resolved data demonstrate that MV predominantly quenches the aggregate PL thus enabling measurement of the decay kinetics of the intrinsic exciton even for low-TR polyrotaxanes for which the different contributions are otherwise difficult to disentangle.25 As visible from Figure 1, showing the time-integrated photoluminescence spectra of buffer solutions (concentration of the conjugated backbone: 5 × 10-2 mg/ml) of unthreaded Nano Lett., Vol. 8, No. 12, 2008

and rotaxinated PDV.Li,18 the red tail of the emission profile is progressively reduced with increasing TR with a concomitant growth of the high-energy region that becomes predominant in highly threaded systems (TR > 2). This observation confirms the results of a parallel study on diluted solutions of unthreaded and rotaxinated PDV.Li at increasing concentration,25 that revealed the coexistence of two competitive spectral components, namely an intrachain exciton (responsible for the blue region of the PL spectra) and an interchain aggregate (giving rise to a PL band red-shifted by ∼180 meV with respect to the excitonic emission) and that at high dilution threaded and unthreaded polymers displayed very similar emission properties. Progressive cyclodextrin encapsulation reduces the tendency of the backbones to aggregate, and therefore enables tuning of the relative weight of the two emissions. The average emission-energies extracted through energy-weighted integration (inset of Figure 1) quantify the progressive blueshift of the PL. We also note that all PL spectra present comparable level of structure and spectral width thus suggesting that the vibronic coupling and inhomogeneous broadening are not significantly modified in highly threaded samples. To quantify the effect of TR in terms of supramolecular encapsulation and to clarify the nature of the different emission features in uninsulated and threaded conjugated polymers, we performed time-resolved PL measurements. The decay profiles of all compounds (Figure 2a) can be described by a double-exponential function of the form (I(t) )I0E exp(-t/τE) + I0A exp(-t/τA)), with a fast initial decay (τE ) 780 ps) followed by a long-lived emission (τA ) 2.6 ns), and where I0E and I0A are the initial amplitudes of the two contributions. Interestingly, the time-constants of both PL components do not vary as a function of TR. However, Figure 2c, top panel, shows the evolution of the two relative contributions to the total luminescence with increasing TR. These have been extracted by temporal integration of the single-exponential profiles of the two species and after normalization so as to respect the condition IA + IE ) 1. As expected, the relative contribution of the long-lived component (IA) is dominant for the unthreaded polymer and progressively decreases with respect to the intrachain emission (IE) in insulated systems. This scenario is entirely consistent with the remarkable single exponential decay over more than two and a half decades shown by the polyrotaxanes with TR g 2. To gain further insight into the effects of progressive encapsulation on the photophysics of the conjugated moieties we also looked at how the presence of a MV quencher (see inset of Figure 2b for chemical structure) modifies the temporal decays of the solutions luminescence.26 This approach is relevant since recent reports show that quenching is markedly more efficient for aggregates than for intrachain singlets.27,28 A full discussion of the reasons for which quenching is more effective for such states is beyond the scope of this letter. However, we note that while the role of the more spatially extended, delocalized, nature of these states cannot be ruled out, the higher quenching efficiency 4547

Figure 2. (a) Photoluminescence time decay of unthreaded PDV (circles) and β-cyclodextrin threaded PDV solutions (5 × 10-2 mg/ ml of the conjugated moiety) at different TR (1.0, reversed triangles; 1.4, squares; 2.0, circles; 2.3, triangles). Fits to biexponential functions are reported in solid lines. (b) PL time decay of the same solutions as in panel a above after the addition of MV at a concentration of 1 × 10-5 mol/ml. (c) Top panel: PL intensities of the long-lived (IA, full triangles) and the fast (IE, full circles) components of the fluorescence signal as obtained from timeintegration of the corresponding exponential functions fitted to the decays in Figure 2a. ζ ) ΦMV/Φ where Φ and ΦMV are the quantum yields of pure and quenched solutions, as a function of threading ratio. Bottom panel: PL quantum yield (Φ) of the unthreaded and rotaxinated PDV.Li solutions at increasing TR before and after the addition of MV (ΦMV).

of the MV for the aggregate versus the excitonic emission is likely to result from a higher binding efficiency of the MV dications to the polyanionic aggregates and thus of the ensuing shorter distance from the latter. The electrostatic attraction between the MV dication and the aggregated polyanion is likely to be stronger than that of the nonaggregated polyanion because of the higher local charge around the aggregate. The possibility to tune interchain interactions by control of their threading ratio makes TMWs a useful model system for investigating the influence of aggregation on the PL quenching by MV. The decay curves after the addition of MV are reported in Figure 2b. Here, we observe a strong reduction of the contribution of aggregates to the PL with increasing TR. The long-lived tail is completely suppressed for the systems with TR > 1.0, which in fact exhibit perfectly superimposed decay curves, reminiscent of the typical behavior of single strands in highly diluted (1 × 10-4 mg/ ml) solutions.25 No change of the decay profile is instead observed for unthreaded PDV.Li. Single-exponential fit for TR > 1.0 reveals that the exciton lifetime has remained 4548

remarkably unchanged upon quenching. This observation confirms that the growth of the excitonic contribution to the PL spectra with increasing TR (top-panel of figure 2c) is due to progressive reduction of the concentration of aggregates and not to increased efficiency of the exciton emission. This interpretation is supported by PL quantum yield (Φ) measurements. Figure 2c, bottom panel, reports the efficiency for the whole series of polymers with and without quencher (open circles and open squares, respectively). As expected on the basis of previous results on other insulated systems25,29-31 and in full agreement with the timeresolved analysis reported above, the quantum yield of both quenched and unquenched solutions increases with TR, and the sensitivity to MV is progressively reduced. For example, the quantum yield of unthreaded PDV.Li is reduced by about 98% by the addition of MV, whereas the quenching is limited to a 10%-reduction in the case of PDV.Li⊂β-CD with TR ) 2.3. The quantum yield reduction is due to quenching of both aggregate states and intrachain excitons as both contribute to the emission and are quenched by MV, although more efficiently in the case of aggregates. Nevertheless, it is meaningful to consider, in a zeroth-order approximation, that the PL of quenched solutions is only due to nonaggregated polymer chains, because we can then quantify the contribution of intrachain excitons to the PL efficiency with a term ζ ) ΦMV/Φ, where Φ and ΦMV are the quantum yields of pure and quenched solutions, respectively. Most interestingly, this parameter can then be compared with the exciton contribution to the PL that has been independently extracted from the temporal decays, as discussed above. In Figure 2c (top panel, open circles) we display ζ with IE and IA as a function of the threading ratio, and find that ζ ∼ IE for high TRs, whereas ζ increasingly deviates from IE with smaller TRs and eventually tends to zero at TR ) 0, thus showing that our approximation above is only valid for the most highly threaded polyrotaxanes. Overall, the quenching experiments above allow us to conclude that as the TR is increased the portion of backbone exposed to quenching is reduced and so the sensitivity to MV. In addition, both the quantum yield and the lifetime measurements provide unequivocal evidence that the quenching process is determined mainly by a so-called static mechanism, i.e., one that involves the formation of a bound complex between the quenching agent and the dye molecules that occurs before optical excitation of the dye, rather than after formation of the excited state.32,33 This is consistent with the high value of the Stern-Volmer constant found for these systems.34 Quenching of the chromophores results in a reduced PL quantum yield without affecting the decay kinetics of the emitting species. A detailed study of the PL quenching mechanism and the quantification of the differences in terms of quenching efficiency for various polyrotaxanes requires specific experiments for the determination of the Stern-Volmer constants.24 Such investigation is currently being performed in a parallel study and is beyond the scope of this report.34 It is also interesting to look at the spectral evolution of the PL in the first few nanoseconds after excitation. To this Nano Lett., Vol. 8, No. 12, 2008

Figure 3. (a) Contour plot of the time decay of the PL intensity (arbitrary units, logarithmic scale) excited at 3.3 eV, of PDV.Li⊂β-CD (TR ) 2.0) before (left panel) and after addition of MV (right panel). (b) Top: intensity ratio between unquenched and quenched PL versus time. Bottom: evolution of the fluorescence average energy as function of time, extracted from the spectra in panel a through energyweighted integration.

Figure 4. (a) Photoluminescence quantum efficiency (PLQE) and external quantum efficiency (EQE) of polyrotaxanes-based thin films and LEDs as a function of TR. LEDs were single-layer sandwich structures prepared by spin-coating the polymer solutions onto oxygenplasma-treated ITO substrates38,39 and then by thermally evaporating (10-6 mbar) 150 nm thick aluminum films. For further details, see Supporting Information. (b) Current-voltage luminance curves of polyrotaxanes-based LEDs at increasing threading ratio. Inset: rootmean-squared roughness values calculated from 10 × 10 µm2 AFM images, obtained in tapping mode on spin cast polymer films (100-150 nm thickness) shown in panel c. (c) Tapping mode AFM images (10 × 10 µm2) of unthreaded PDV.Li and PDV.Li⊂β-CD polyrotaxanes with increasing threading ratio on ITO.

end, we report in Figure 3a the contour plots of the photoluminescence decay of PDV.Li⊂β-CD (TR ) 2.0) solutions with and without MV, as measured in the same conditions of excitation power and collection geometry. The two distinct emission regimes discussed previously are also clearly observable in these data. For pure PDV.Li⊂β-CD solutions (Figure 3a, left), we observe a fast emission centered at about 2.8 eV together with a long-lived fluorescence at ∼2.5 eV. Upon addition of MV (Figure 3a, right), the long-lived contribution is suppressed and the emission profile is due to intrachain excitons only. We may now track the quenching process in time by considering the ratio (IPLMV/ IPL) between the spectrally integrated emission intensity with (IPLMV) and without (IPL) methyl viologen (Figure 3b, top panel). IPLMV/IPL is close to unity for the first 4 ns, as expected Nano Lett., Vol. 8, No. 12, 2008

for a quenching process mainly involving the long-lived interchain species, and rapidly decrease after 4 ns to an asymptotic value of IPLMV/IPL ) 0.25 at 10 ns delay, confirming that the long-lived interchain species are more subject to the quenching effect of MV.27,28 The data above are analyzed further in the bottom panel of Figure 3b where we report the time evolution of the average emission energy (), as extracted by energyweighted integration from the contour plots in Figure 3a. For the quenched solution (triangles) we notice that decreases monotonically from 2.57 eV to an asymptotic value of ≈ 2.48 eV after about 10 ns from the excitation pulse. Such red-shift of the PL band is ascribable to exciton migration toward lower energy sites (spectral diffusion) reaching the localization energy (ELOC ≈ 2.48 eV) after 4549

several nanoseconds. This interpretation is supported by theoretical work that demonstrates how excitation migration can extend to the nanosecond time scale in insulated systems of reduced dimensionality such as threaded molecular wires.35,36 The average emission energy of rotaxane solutions without MV (circles) starts from the same initial value, but the time-evolution at longer delay times follows a different trend characterized by a clear change in slope after ∼3 ns and by an asymptotic value of about 2.45 eV after 10 ns. We ascribe this rapid decrease of in unquenched solutions to the progressive increase of interchain emission. With a view to utilization of these materials in lightemitting diodes, LEDs, we report in Figure 4a the solid state photoluminescence efficiency (PLQE, triangles) of solid films and the electroluminescence external quantum efficiency (EQE, solid dots) of single-layer LEDs (ITO/polymer/Al) as a function of the threading ratio. From these we observe that the trends to higher efficiency with TR already observed in solution are also maintained in the solid state, thus indicating a direct correlation that confirms the validity of the strategic approach aimed at improving LED performance by suppressing the formation of aggregates, despite the concomitant effects of such insulation on charge transport. Although the absolute values of EL efficiency do not compare favorably with those of state-of-the-art devices utilizing organic solvent soluble materials, the importance of our results lie in the establishing of the general principles on these model compounds. In Figure 4b we report the current density-luminance versus voltage (IVL) characteristics of unoptimized devices (ITO/polymer/Al) incorporating a 100 nm thick film of PDV.Li and PDV.Li⊂β-CD with increasing TR. The turnon voltage progressively increases with TR. Such dependence is partially explained considering the following two factors: (i) the reduced charge transport properties in threaded molecular wires, and (ii) the slight lowering of the HOMO upon increasing molecular insulation, as shown for PDV.Li (HOMO ) 5.20 eV) and PDV.Li⊂β-CD (TR ) 1.4, HOMO ) 5.36 eV). 18 In fact, because of the electrolytic nature of the polymer backbone (Figure 1), limitation of charge injection by unfavorable interface energetics is expected to occur mainly at the anode. This is a consequence of the accumulation of the mobile cations (Li+) at the cathode during device operation and the subsequent thinning of the energy barrier for electron injection.37 Interestingly, AFM images of the investigated devices (Figure 4c) reveal that progressive cyclodextrin encapsulation also modifies the film morphology: the film roughness decreases with increasing TR, and a significant reduction of morphological defects such as pin-holes is observed. In the inset of Figure 4b, we report the root-mean-square roughness calculated from 10 × 10 µm2 images, obtained in tapping mode on spin cast polymer films (100-150 nm thickness). Therefore, differences between the IVL characteristics might also be partially due to different packing geometries induced by the increasing number of cyclodextrin macrocycles around the conjugated backbones. 4550

In conclusion, we have demonstrated that nanoscale supramolecular engineering of conjugated polymers architectures via tuning of the threading ratio of cyclodextrinbased conjugate polyrotaxanes is a successful approach for controlling conjugated polymers photophysics in solution, solid films, and even when the latter are incorporated in lightemitting devices. We have also demonstrated that the use of an electron-accepting PL quenching agent such as methyl viologen enables further disentangling of the molecular photophysics due to either intrachain singlet excitons or aggregates. Acknowledgment. We thank the EC (EU-contract: MRTNCT-2006-036040) and the EPSRC for financial support. SB acknowledges support from the Fondazione Angelo Della Riccia. Supporting Information Available: This material is available free of charge via the Internet at http://pubs.acs.org. References (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 (6715), 121–128. (2) Hoppe, H.; Sariciftci, N. S. J. Mater. Res. 2004, 19 (7), 1924–1945. (3) Bu¨rgi, L.; Richards, T. J.; Friend, R. H.; Sirringhaus, H. J. Appl. Phys. 2003, 94 (9), 6129–6137. (4) Samuel, I. D. W.; Rumbles, G.; Collison, C. J. Phys. ReV. B 1995, 52 (16), 11573–11576. (5) Rumbles, G.; Samuel, I. D. W.; Collison, C. J.; Miller, P. F.; Moratti, S. C.; Holmes, A. B. Synth. Met. 1999, 101, 158–161. (6) Nguyen, T. Q.; Doan, V.; Schwartz, B. J. J. Chem. Phys. 1999, 110 (8), 4068–4078. (7) Schwartz, B. J. Annu. ReV. Phys. Chem. 2003, 54, 141–172. (8) Schenning, A.; Jonkheijm, P.; Hoeben, F.; van Herrikhuyzen, J.; Meskers, S.; Meijer, E.; Herz, L.; Daniel, C.; Silva, C.; Phillips, R.; Friend, R.; Beljonne, D.; Miura, A.; De Feyter, S.; Zdanowska, M.; Uji-i, H.; De Schryver, F.; Chen, Z.; Wurthner, F.; Mas-Torrent, M.; den Boer, D.; Durkut, M.; Hadley, P. Synth. Met. 2004, 147 (1-3), 43–48. (9) Westenhoff, S.; Abrusci, A.; Feast, W. J.; Henze, O.; Kilbinger, A. F. M.; Schenning, A.; Silva, C. AdV. Mater. 2006, 18 (10), 1281– 1285. (10) Cacialli, F.; Samorı`, P.; Silva, C. Mater. Today 2004, 7 (4), 24–32. (11) Sancho-García, J. C.; Poulsen, L.; Gierschner, J.; Martínez-Alvárez, R.; Hennebicq, E.; Hanack, M.; Egelhaaf, H. J.; Oelkrug, D.; Beljonne, D.; Bre´das, J. L.; Cornil, J. AdV. Mater. 2004, 16 (14), 1193–1197. (12) Poulsen, L.; Jazdzyk, M.; Communal, J. E.; Sancho-Garcia, J. C.; Mura, A.; Bongiovanni, G.; Beljonne, D.; Cornil, J.; Hanack, M.; Egelhaaf, H. J.; Gierschner, J. J. Am. Chem. Soc. 2007, 129 (27), 8585–8593. (13) Gierschner, J.; Ehni, M.; Egelhaaf, H. J.; Medina, B. M.; Beljonne, D.; Benmansour, H.; Bazan, G. C. J. Chem. Phys. 2005, 123 (14), 144914. (14) Calzaferri, G.; Huber, S.; Maas, H.; Minkowski, C. Angew. Chem., Int. Ed. 2003, 32 (42), 3732–3758. (15) Thomas, S.; Joly, G.; Swager, T. Chem. ReV. 2007, 107 (4), 1339– 1386. (16) Lehn, J.-M. Supramolecular Chemistry - Concepts and PerspectiVes; Wiley-WCH: Wienheim, 1995. (17) Taylor, P. N.; O’Connell, M. J.; McNeill, L. A.; Hall, M. J.; Aplin, R. T.; Anderson, H. L. Angew. Chem., Int. Ed. 2000, 39 (19), 3456. (18) Cacialli, F.; Wilson, J. S.; Michels, J. J.; Daniel, C.; Silva, C.; Friend, R. H.; Severin, N.; Samorı`, P.; Rabe, J. P.; O’Connell, M. J.; Taylor, P. N.; Anderson, H. L. Nat. Mater. 2002, 1 (3), 160–164. (19) Harada, A.; Li, J.; Kamachi, M. Nature 1993, 364 (6437), 516–518. (20) Shimomura, T.; Akai, T.; Abe, T.; Ito, K. J. Chem. Phys. 2002, 116 (5), 1753–1756. (21) Gibson, H. W.; Bheda, M. C.; Engen, P. T. Prog. Polym. Sci. 1994, 19 (5), 843–945. (22) Frampton, M. J.; Anderson, H. L. Angew. Chem., Int. Ed. 2007, 46, 1028–1064. (23) Michels, J. J.; O’Connell, M. J.; Taylor, P. N.; Wilson, J.; Cacialli, Nano Lett., Vol. 8, No. 12, 2008

F.; Anderson, H. L. Chem.sEur. J. 2003, 9, 6167–6176. (24) Zhou, Q.; Swager, T. M. J. Am. Chem. Soc. 1995, 117, 12593–12602. (25) Petrozza, A.; Brovelli, S.; Michels, J. J.; Anderson, H. L.; Friend, R. H.; Silva, C.; Cacialli, F. AdV. Mater. 2008, 20, 3218–3223. (26) We also note that to record high-quality fluorescence spectra under sufficiently dilute conditions to avoid aggregation would require very long acquisition times, so adding MV to remove the emission of the aggregate is a much more practical approach. (27) Jiang, H.; Zhao, X.; Schanze, K. S. Langmuir 2007, 23, 9481–9486. (28) Tan, C.; Atas, E.; Muller, J. G.; Pinto, M. R.; Kleiman, V. D.; Schanze, K. S. J. Am. Chem. Soc. 2004, 126 (42), 13685–13694. (29) Frampton, M. J.; Claridge, T. D. W.; Latini, G.; Brovelli, S.; Cacialli, F.; Anderson, H. L. Chem. Comm. 2008, 2797–2799. (30) Latini, G.; Parrot, L.-J.; Brovelli, S.; Frampton, M. J.; Anderson, H. L.; Cacialli, F. AdV. Funct. Mater. 2008, 18 (16), 2419–2427. (31) Wilson, J. S.; Frampton, M. J.; Michels, J. J.; Sardone, L.; Marletta, G.; Friend, R. H.; Samorı`, P.; Anderson, H. L.; Cacialli, F. AdV. Mater. 2005, 17 (22), 2659–2663. (32) Lakowicz, J. R. Principles of Fluorescence Spectroscopy; Springer: New York, 1986.

Nano Lett., Vol. 8, No. 12, 2008

(33) Connors, K. A. Binding Constants: the Measurement of Molecular Complex Stability; Wiley-Interscience: New York, 1987. (34) Stone, M. T.; Oddy, F. E.; Klotz, E. J. F.; Anderson, H. L. (Oxford, U.K.) and Brovelli, S.; Cacialli, F. (London, U.K.)Unpublished work. (35) Beljonne, D.; Pourtois, G.; Shuai, Z.; Hennebicq, E.; Scholes, G. D.; Bre´das, J. L. Synth. Met. 2003, 137 (1-3), 1369–1371. (36) Hennebicq, E.; Pourtois, G.; Scholes, G. D.; Herz, L. M.; Russell, D. M.; Silva, C.; Setayesh, S.; Grimsdale, A. C.; Mullen, K.; Bre´das, J. L.; Beljonne, D. J. Am. Chem. Soc. 2005, 127 (13), 4744–4762. (37) deMello, J. C.; Tessler, N.; Graham, S. C.; Friend, R. H. Phys. ReV. B 1998, 57 (20), 12951–12963. (38) Kim, J. S.; Cacialli, F.; Friend, R. H. Thin Solid Films 2003, 445, 358–366. (39) Kim, J. S.; Cacialli, F.; Granstro¨m, M.; Friend, R. H.; Johansson, N.; Salaneck, W. R.; Daik, R.; Feast, W. J. Synth. Met. 1999, 101, 111. (40) NIST. NIST standard reference material SRM no. 936a; www.nist. gov/srm (accessed Dec 16, 1994).

NL802775A

4551