Photophysical Properties of Dioxolane-Substituted Pentacene

Pentacene Derivatives Dispersed in Tris(quinolin-8-olato)aluminum(III) ..... Christophe P. Bénard, Zhe Geng, Matthew A. Heuft, Kelly VanCrey, and...
5 downloads 0 Views 448KB Size
Download PDF
7928

J. Phys. Chem. B 2006, 110, 7928-7937

Photophysical Properties of Dioxolane-Substituted Pentacene Derivatives Dispersed in Tris(quinolin-8-olato)aluminum(III) Mason A. Wolak,† Joseph S. Melinger,† Paul A. Lane,† Leonidas C. Palilis,† Chad A. Landis,‡ Jared Delcamp,‡ John E. Anthony,‡ and Zakya H. Kafafi*,† U. S. NaVal Research Laboratory, Washington, DC 20375, and Department of Chemistry, UniVersity of Kentucky, Lexington, Kentucky 40506 ReceiVed: March 3, 2005; In Final Form: July 14, 2005

Two novel dioxolane-substituted pentacene derivatives, namely, 6,14-bis-(triisopropylsilylethynyl)-1,3,9,11tetraoxa-dicyclopenta[b,m]pentacene (TP-5) and 2,2,10,10-tetraethyl-6,14-bis-(triisopropylsilylethynyl)-1,3,9,11tetraoxa-dicyclopenta[b,m]pentacene (EtTP-5), have been synthesized and spectroscopically characterized. Here, we examine the steady-state and time-resolved photoluminescence (PL) of solid-state composite films containing these pentacene derivatives dispersed in tris(quinolin-8-olato)aluminum(III) (Alq3). The films show narrow red emission and high absolute photoluminescence quantum yields (φPL ) 59% and 76% for films containing ∼0.25 mol % TP-5 and EtTP-5, respectively). The Fo¨rster transfer radius for both guest-host systems is estimated to be ∼33 Å. The TP-5/Alq3 thin films show a marked decrease in φPL with increasing guest molecule concentrations, accompanied by dramatic changes in the PL spectra, suggesting that intermolecular interactions between pentacene molecules result in the formation of weakly radiative aggregates. In contrast, a lesser degree of fluorescence quenching is observed for EtTP-5/Alq3 films. The measured fluorescence lifetimes of TP-5 and EtTP-5 are similar (∼18 ns) at low concentrations but deviate at higher concentrations as aggregation begins to play a role in the TP-5/Alq3 films. The onset of aggregation in EtTP5/Alq3 films occurs at higher guest molecule concentrations (>1.00 mol %). The addition of ethyl groups on the terminal dioxolane rings leads to an increase in the intermolecular spacing in the solid, thereby reducing the tendency for π-π molecular stacking and aggregation.

Introduction Pentacene derivatives comprise a particularly versatile class of materials that have shown great promise when utilized as the electroactive component of thin film transistors (TFTs),1-3 photovoltaic cells (PVs),4-6 and organic light-emitting diodes (OLEDs).7,8 The electronic and optical properties of pentacene can be easily tuned by altering the substitution pattern on the acene backbone. For example, the introduction of phenyl groups at the 6 and 13 positions yields a bathochromic shift of the fluorescence maximum and a saturated red emission. 6,13Diphenylpentacene was used as the red-emitting center in an OLED structure,7 and electroluminescence quantum efficiencies close to the theoretical limit were achieved.9 When triisopropylsilylethynyl groups are substituted at the 6 and 13 positions, improved solid-state packing is observed,10 and thin film transistors based on this material exhibited field effect hole mobilities on the order of 0.4 cm2/(V s).3 Pentacene can also be converted to an n-type semiconductor by replacing the hydrogen atoms with fluorine atoms; n-channel TFTs based on perfluoropentacene displayed electron mobilities of approximately 0.1 cm2/(V s).11 Indeed, subtle modification of the molecular structure of pentacene can lead to large changes in the bulk electronic and optical properties of the molecular solid. Recently, we reported on several diaryl-substituted pentacene derivatives and their performance as red-emitters when dispersed * Author to whom correspondence should be addressed. Phone: (202) 767-9529. Fax: (202) 404-8114. E-mail: [email protected]. † U. S. Naval Research Laboratory. ‡ University of Kentucky.

in the common electron-transporting host tris(quinolin-8-olato)aluminum(III) (Alq3).8 Organic light-emitting diodes based on this guest-host system feature reasonably high external electroluminescence (EL) quantum efficiencies (ηEL ≈ 1.5%), low turn-on voltages, and excellent color purity. Guest-host emitting layers are often used in OLEDs to hinder crystallization of the emitting material, thereby ensuring that the electroluminescence arises from strong molecular emission rather than weak aggregate emission. In addition, guest-host layers have been used to balance charge injection and transport, optimize device EL quantum efficiency, and improve operational longevity.12-15 Although numerous doped polymer systems have been thoroughly investigated for their energy transfer characteristics,16-20 less work has been done to understand the complex nature of the energy transfer process in doped small molecule thin films.21 We have recently examined the steady-state and time-resolved fluorescence of thin films featuring increasing concentrations of 6,13-bis(2,6-dimethylphenyl)pentacene dispersed in Alq3.22 Using these techniques, we have demonstrated efficient Fo¨rster resonance energy transfer from the host Alq3 molecules to the guest diarylpentacene molecules with energy transfer rates in the range of 0.5-3.0 ns-1.22 There was evidence that the estimated Fo¨rster radius23 may be extended possibly due to exciton migration within the Alq3 host, consistent with the large exciton diffusion length of Alq3 (∼100 ( 40 Å).24 We further noted that strong fluorescence quenching and fast decay dynamics at high guest molecule concentrations (g1.5 mol %) are the result of the formation of nonemissive aggregates, leading to additional nonradiative decay pathways. Developing new pentacene derivatives with modified electronic

10.1021/jp0511045 CCC: $30.25 © 2006 American Chemical Society Published on Web 03/31/2006

Dioxolane-Substituted Pentacene Derivatives CHART 1: Chemical Structures of Guest (TP-5 and EtTP-5) and Host Alq3 Moleculesa

J. Phys. Chem. B, Vol. 110, No. 15, 2006 7929 molecules to the pentacene guest molecules as a function of guest concentration. Experimental Methods

a The nondioxolane-substituted TIPS pentacene derivative is included for purposes of comparison.

and chemical structures leading to improved photoluminescence (PL) quantum yields (while suppressing strong intermolecular interactions in the solid state) remains a challenge. To this end, we have recently designed and synthesized two new highly fluorescent pentacene derivatives, namely, 6,14-bis(triisopropylsilylethynyl)-1,3,9,11-tetraoxa-dicyclopenta[b,m]pentacene (TP-5) and 2,2,10,10-tetraethyl-6,14-bis-(triisopropylsilylethynyl)-1,3,9,11-tetraoxa-dicyclopenta[b,m]pentacene (EtTP-5) (Chart 1). Both compounds feature 1,3-dioxolane moieties fused to the terminal benzenoid rings of the pentacene core, in addition to triisopropylsilylethynyl substituents at the 6,13-position. Chart 1 also shows the chemical structure of 6,13bis(triisopropylsilylethynyl)pentacene (TIPS). The dioxolane units on TP-5 were initially added to lower the oxidation potential of the materials, thereby improving charge injection from gold electrodes in TFTs.25 Spectroscopic characterization of TP-5 showed strong red emission (φPL ) 60% in toluene) relative to that previously reported for diaryl-substituted pentacenes (φPL ≈ 15% in toluene),8 suggesting that TP-5 is a good candidate as a red-emitting center in OLEDs. TP-5 has also been shown to form well-ordered crystals that exhibit significant π-overlap, as shown by the tight stacking between π-faces that are separated by only 3.35 Å.25 Newly synthesized EtTP-5 features two ethyl groups on the sp3-hybridized carbon of the dioxolane rings; this modification was undertaken to disrupt the solid-state packing and hinder interaction between neighboring aromatic rings. Herein, we report on the solid-state photophysical properties of these two new red-emitting dioxolane-substituted pentacene derivatives. We have measured the steady-state PL spectra and the absolute PL quantum yields of guest-host films containing various concentrations of TP-5 or EtTP-5 in Alq3 (∼0.25-2.25 mol %). In addition, we have investigated the dynamics of these films using time-resolved photoluminescence to determine the rate and efficiency of energy transfer from the Alq3 host

Materials. Alq3 was obtained from H. W. Sands. TP-5 was synthesized as previously reported,25 and EtTP-5 was prepared in an analogous manner beginning with condensation of 2,2diethyl-1,3-benzodioxolane-6,7-dicarboxaldehyde with 1,4-cyclohexanedione.26 Each compound was purified via duplicate train sublimation (270-290 °C, 5.0 × 10-6 Torr) prior to use.8 Crystals of TP-5 were grown from toluene solution at -20 °C. Crystals of EtTP-5 were grown by the slow evaporation of a 1,2-dichloroethane solution at room temperature. Preparation of Thin Films. Fused silica substrates were cleaned with organic solvents and subjected to UV-O3 treatment prior to use (30 min, 4:1 N2/O2, flow rate ) 1.5 SCFH). Neat and composite films (∼750-nm-thick) were prepared by vacuum sublimation (3.0 × 10-7 Torr) from resistive heating furnaces. Quartz crystal microbalances were used to monitor the rate of deposition and determine the film composition and thickness. The guest and host materials were codeposited at a combined rate of 2-4 Å/s to yield films with varying guest molecule concentrations (0.2-2.5 mol %). Photoluminescence Measurements. The PL spectra were measured using a Cary Eclipse fluorimeter (λex ) 350 nm). The PL quantum yields of dilute toluene solutions of TP-5 and EtTP-5 were measured relative to a rubrene standard (φPL ≈ 100% in benzene) using our previously reported method.8 Absorption spectra were measured using a Hewlett-Packard 8423 spectrophotometer and used to confirm guest molecule concentrations in the composite films. All spectra were collected under ambient conditions. Absolute PL quantum yields of the solid films were measured using an integrating sphere and a HeCd laser (λex ) 325 nm) in dry N2.21 The UV laser line was filtered at the silicon photodiode with a Kodak Wratton 2B filter (λ < 385 nm cutoff). Time-resolved photoluminescence spectra were measured under ambient conditions using the time-correlated singlephoton-counting technique.27 Samples were irradiated at λ ) 300 nm from a frequency-doubled, synchronously pumped rhodamine dye laser at an energy fluence of 10-20 nJ/cm2 per pulse and a repetition rate of 1 MHz. The pump intensity per pulse (∼1015 photons/cm3) was well below the number density of the host (1021 molecules/cm3) and guest (1018-1019 molecules/ cm3). The resulting emission was dispersed in a monochromator and detected using a cooled microchannel plate Hamamatsu photomultiplier tube. The monochromator was set at or near the peak fluorescence wavelength, and the transient signals were collected at the magic angle of 54.7°. An instrument response function (full width at half-maximum) of ∼50 ps was measured by collecting scattered laser light off a ground glass surface. Transient fluorescence decays were collected in 75 or 100 ns time windows. In a typical experiment, approximately 104 photons were collected at the peak channel of the fluorescence transient. Fluorescence decay data were deconvoluted with the instrument response function then normalized by setting the maximum photon count equal to 1. The fluorescence decays of guest molecules were fit to a sum of exponential decay functions: I(t) ) Σi(Ai exp(-t/τi)) where Ai is the preexponential factor and τi is the time constant. Weighting factors were determined for each decay component relative to the total decay as wi ) Aiτi/Σi(Aiτi). The average time constant was calculated from the individual decay components as 〈τ〉 ) Σi(Aiτi)/ΣiAi. The

7930 J. Phys. Chem. B, Vol. 110, No. 15, 2006

Figure 1. Normalized emission of Alq3 host (bold line) and extinction spectra of pentacene guests (TP-5, solid line; EtTP-5, dotted line).

goodness of the fits was assessed using visual inspection and the χ2 parameter. Satisfactory fits were obtained when χ2 was 1.3 or less. Time constants were determined as an average of 3 or 4 measurements per sample. The standard deviation of the average was approximately (4% for the low concentration films (e1.00 mol %) and (9% for the high-concentration films (>1.00 mol %). Fluorescence decay of the Alq3 host is analyzed in terms of Fo¨rster theory with a random distribution of dopants. This analysis is described in the Discussion section. X-ray Crystallography. The crystal structure of TP-5 was determined as previously reported.25 The crystal structure of EtTP-5 was determined using a Bru¨ker-Nonius X8 Proteum diffractometer at 90 K, solved by direct methods using SHELXS-97, and refined using SHELXL-97. Experimental Results Absorption and Photoluminescence of Pentacene Solutions. Both TP-5 and EtTP-5 were designed for use in guesthost composite films with Alq3 serving as the host matrix. Alq3 was chosen as the host due to the large overlap between its emission spectrum and the absorption of the pentacene guests, a necessary prerequisite to achieve efficient Fo¨rster resonance energy transfer from the host to the guest molecules. Figure 1 shows both the emission spectrum of a neat Alq3 film and the extinction spectra of TP-5 and EtTP-5 in dilute toluene solutions. The broad, featureless emission from Alq3 is centered at ∼535 nm and extends beyond 700 nm, making good overlap with the π-π* absorption bands of the pentacene derivatives. The two pentacene compounds display very similar absorption spectra in the visible region (λmax ≈ 620 nm); the only appreciable difference occurs at the additional absorption peak centered at ∼460 nm, where EtTP-5 has a somewhat stronger absorption than TP-5. This transition is barely noticeable in the absorption spectrum of TIPS28 and is thought to arise from the introduction of the terminal dioxolane rings of TP-5 and EtTP-5. The fluorescence spectra of dilute toluene solutions (1 µM) of TP-5 and EtTP-5 are shown in Figure 2. The spectral features of the two molecules are quite similar with λmax ≈ 625-630 nm, yielding very small Stokes shifts of ∼5-6 nm. The emission maximum of EtTP-5 is slightly red-shifted relative to that of TP-5 as a result of the minor electron-donating nature of the ethyl substituents on the terminal dioxolane rings. A photoluminescence quantum yield of 72% was measured for EtTP-5 in toluene, a nearly 5-fold enhancement over previously studied diarylpentacenes8,22 and a moderate improvement over TP-5 (φPL ) 60%).

Wolak et al.

Figure 2. Normalized fluorescence spectra of dilute toluene solutions of TP-5 (solid line) and EtTP-5 (dashed line).

Photoluminescence of Composite Films. The PL spectra of TP-5 and EtTP-5 dispersed in Alq3 as a function of the guest molecule concentration are illustrated in Figure 3. Red emission dominates the spectra of the composite films, indicating efficient energy transfer from the host Alq3 molecules to the guest pentacene molecules. The strongest emission peak is centered at λmax ) 651 nm for TP-5/Alq3 composite films and slightly red-shifted relative to those featuring the alkylated analogue EtTP-5 (λmax ) 647 nm). Note that the photoluminescence from the composite films is red-shifted relative to that of the solutions by approximately 15-20 nm. Low concentration films (1.0 mol %; a quantum yield of nearly 30% ( 3% is measured for the film containing ∼2.0 mol % EtTP-5 while the corresponding TP-5/Alq3 film shows minimal emission (φPL < 5.0% ( 0.5%). Surprisingly, vacuum-deposited neat films of TP-5 and EtTP-5 on fused silica substrates were nonemissive. However, we were able to detect emission from the purified powders of both pentacene derivatives (Figure 3), showing large bathochromic shifts relative to their solution spectra. Emission from TP-5 powder is extremely weak (λmax ) 721 nm) whereas emission from EtTP-5 powder is readily visible (λmax ) 693 nm). Host (Alq3) PL Decay Dynamics. Although the dominant feature of the PL spectra of the composite films is emission from the dioxolane-substituted pentacene guest, Alq3 host emission persists even at the highest guest molecule concentrations. Key information pertaining to the energy transfer rates

and efficiencies of our guest-host system can be obtained by analysis of the PL decay dynamics of this emission. The temporal profiles of Alq3 emission (probed at 530 nm) of films doped with TP-5 and EtTP-5 are illustrated in Figure 5. The figures also show the single-exponential time-resolved PL decay of a pristine Alq3 film for comparative purposes. The temporal profiles quickly change upon the introduction of a small amount of pentacene derivative to the films; the Alq3 decay becomes more and more rapid as the guest molecule concentration increases. The decay in the EtTP-5-doped films appears slightly faster. Guest (TP-5 and EtTP-5) PL Decay Dynamics. Figure 6 depicts the transient radiative decay profiles of TP-5/Alq3 and EtTP-5/Alq3 composite films (probed at 650 and 645 nm, respectively) as a function of guest concentration. The films were excited at λ ) 300 nm so as to minimize direct excitation of the guest pentacene derivatives, which have very strong absorption bands at ∼325 nm ( ) 1.8-2.0 × 105 M-1 cm-1 in toluene). Although the extinction coefficients of TP-5 and EtTP-5 at 300 nm are approximately a factor of 6 lower than the values measured at 325 nm, a small amount of directly excited pentacene molecules may exist (particularly at high concentrations). The transient signal shows an initial rise followed by a short plateau at the maximum PL intensity, whereupon the pentacene excited state starts its natural decay. The initial rise is attributed to the energy transfer process from the Alq3 host to the pentacene guest, resulting in an increasing population of guest molecules in the excited state. After the initial population buildup, the guest molecules undergo natural radiative and nonradiative decay. The length of time required for buildup of guest molecules in the excited state (referred to as the rise time) decreases as the guest concentration increases. In general, the rise times of EtTP-5/Alq3 films are slightly shorter than the rise times of TP-5/Alq3 films at comparable concentrations. Significant differences are noted between the decay profiles of TP-5 and EtTP-5 for high-concentration films. At guest molecule concentrations g1.00 mol %, the TP-5/Alq3 films display much more rapid decay than the EtTP-5/Alq3 films. Furthermore, the TP-5 decay begins to show multiexponential character at ∼1.00 mol % while the EtTP-5 decay retains a single-exponential character (after the initial rise time). Note that at low concentrations (e0.50 mol %) the TP-5 and EtTP-5 decay profiles are nearly identical. Figure 7 shows the time-resolved PL of TP-5/Alq3 films probed at 700 nm. This investigation was undertaken to gain

7932 J. Phys. Chem. B, Vol. 110, No. 15, 2006

Figure 5. (a) Time-resolved PL decays of Alq3 (probed at 530 nm) in neat and TP-5-doped composite films at various TP-5 concentrations (0.23-2.25 mol %). A continuous decrease in decay time is observed as the TP-5 concentration increases. The decays are normalized for clarity, and the solid lines represent the calculated fits as a monoexponential decay function (for neat Alq3) or using eq 8 as detailed in Table 1. (b) Time-resolved PL decays of Alq3 (probed at 530 nm) in EtTP-5-doped composite films at various EtTP-5 concentrations (0.261.97 mol %).

further insight into the nature of the long-wavelength emission observed in high-concentration TP-5/Alq3 (not in EtTP-5/Alq3) films. The transient decay curves bear a great deal of similarity to the ones shown in Figure 6a, particularly at guest concentrations e0.93 mol %. However, there are notable differences between the temporal profiles for films with guest concentrations g1.45 mol %. The initial decline in PL intensity probed at 700 nm appears considerably faster than the corresponding decline when probed at 650 nm. In addition, there is a more noticeable long-lived state that is observed in the decays probed at 700 nm compared to those probed at 650 nm. Discussion Steady-State Fluorescence Spectra. Excitation of guest molecules may occur in more than one way: (1) direct excitation, (2) reabsorption of host emission (radiative transfer), (3) Fo¨rster resonance energy transfer (FRET), and (4) exciton diffusion within the host followed by energy transfer (diffusionassisted FRET). The first process can be neglected as the dopant absorbs a negligible fraction (1 mol %) for which there are only a few data points that can be fit. The energy transfer rate can be related to the rise time through the rate constants in eq 7: 〈kET〉 ) 1/τrise - 1/τH. The energy transfer rates calculated from the buildup of guest fluorescence predict much higher transfer efficiencies than those actually observed. For example, consider the film containing 0.23 mol % TP-5 in Alq3. The energy transfer rate calculated from the rise time of TP-5 fluorescence is 〈kET〉 ) 0.18 ( 0.03 GHz. The corresponding energy transfer efficiency from this rate constant, ηET ) 63 ( 4%, is significantly higher than that determined from the host decay, ηET ) 51 ( 4%. Energy transfer between the closest-lying donor-acceptor pairs dominates the dynamics at the earliest times following excitation. The rise time reflects the dynamics of these pairs, whereas the transfer efficiency reflects the overall distribution of the ensemble. The transfer efficiencies calculated from the decay of Alq3 fluorescence are consistent with the relative contributions of host and guest to the steady-state PL spectra. The rise times of the EtTP-5/Alq3 films are slightly shorter than those of the TP-5/Alq3 films at comparable concentrations, consistent with the estimated average lifetimes of the Alq3 host decay and the differences in the Alq3 contribution seen in the PL spectra. The dynamics of the excited state decays of both guest and host in the composite films indicate more efficient energy transfer from Alq3 to EtTP-5 than to TP-5 at very low concentrations. Given the identical Fo¨rster transfer radii of TP-5 and EtTP-5, we suggest that this discrepancy is due to a better dispersion of EtTP-5 than TP-5 in Alq3. TP-5 shows a greater tendency to aggregate than EtTP-5, even at the lowest concentrations. Individually dispersed molecules will occupy a greater volume within the host, increasing the opportunity for energy transfer. Changes in the absorption spectrum of the guest due to aggregation may reduce the spectral overlap between donor and acceptor and, hence, the transfer efficiency. Following the initial population buildup (τrise) of the excited guest molecules, the transient PL signal is represented by a

single-exponential decay (τ1) or the sum of two exponential decays (τ1 and τ2). As the rise time is much shorter than the decay time, the decay constant should approximately represent the molecular singlet excited-state lifetimes in these composite films. TP-5 and EtTP-5 have nearly the same PL lifetime at very low concentrations: 〈τTP-5〉 ) 18.4 ( 0.7 ns and 〈τEtTP-5〉 ) 18.3 ( 0.7 ns at ≈ 0.25 mol % doping and 〈τTP-5〉 ) 16.2 ( 0.6 ns and 〈τEtTP-5〉 ) 16.4 ( 0.7 ns at ≈ 0.48 mol % doping. This similarity is striking, given the difference in molecular fluorescence lifetimes determined from dilute solutions of TP-5 and EtTP-5 (10-6 M in toluene): 〈τTP-5〉 ) 14.5 ( 0.8 ns and 〈τEtTP-5〉 ) 7.2 ( 0.4 ns. Natural radiative lifetimes τr ) 24 ns for TP-5 and τr ) 10 ns for EtTP-5 are extracted from the PL quantum yields of these solutions (τr ) 〈τ〉/φPL).27 Aggregation Effects. As concentrations increase, the PL decay rate of the guest molecules increases and deviates from monoexponential decay. Faster PL decay at higher concentrations, coupled with a second slower time constant component, suggests that additional nonradiative decay pathways of the pentacene excited state now compete with emission from isolated molecules. The PL lifetime of TP-5 decreases from 18.4 ( 0.7 ns at 0.23 mol % to 0.5 ( 0.1 ns at 2.25 mol %. The decrease in the PL lifetime of EtTP-5 is less drastic, decreasing from 18.3 ( 0.7 ns at 0.26 mol % to 3.4 ( 0.3 ns at 1.97 mol % concentration. The TP-5/Alq3 films are modeled with τ1 and τ2 components at concentrations g0.93 mol % while EtTP-5/ Alq3 films are modeled with τ1 and τ2 components at concentrations g1.45 mol %. Emission from isolated molecules is inferred where the dynamics can be modeled by a single-exponential decay (τ1) whereas an additional decay channel is added when the dynamics are modeled with both τ1 and τ2 components. Analogous behavior has been reported for the PL decay of tetraphenylchlorin dispersed in Alq3 at similar dopant concentrations.40 We find that EtTP-5 is less susceptible to aggregation than TP-5. The PL lifetime of EtTP-5 is 3 and 7 times longer than the corresponding TP-5 lifetimes at equivalent concentrations of ∼1.5 and ∼2.0 mol %, respectively. The longer excitedstate lifetimes are consistent with the higher φPL of EtTP-5/ Alq3 films relative to TP-5/Alq3 films at high guest molecule concentrations. Further, the PL dynamics of EtTP-5 can be fit to monoexponential decay up to concentrations of approximately 1 mol %. This illustrates that distinct changes in the dynamics

7936 J. Phys. Chem. B, Vol. 110, No. 15, 2006

Wolak et al.

TABLE 3: Summary of Time-Resolved TP-5 PL Decay Fitting Parameters for TP-5/Alq3 Composite Films as a Function of Pentacene Concentration (TP-5 Emission Monitored at 700 nm)a TP-5 concentrationb (mol %) 0.23 0.48 0.93 1.45 2.25

τrise (ns)

τ1 (ns)

w1c

τ2 (ns)

w2c

〈τ〉 (ns)

4.1 ( 0.2 18.9 ( 0.8 1 18.9 ( 0.8 2.7 ( 0.1 16.5 ( 0.7 1 16.5 ( 0.7 0.6 ( 0.1 8.4 ( 0.8 0.86 17.9 ( 1.6 0.14 9.0 ( 0.7 0.3 ( 0.03 1.5 ( 0.1 0.48 19.3 ( 1.7 0.52 2.9 ( 0.2 0.2 ( 0.02 0.8 ( 0.1 0.38 12.7 ( 1.1 0.62 1.8 ( 0.1

a τrise is the rise time constant, and 〈kET〉 is the Fo¨rster transfer rate calculated from τrise. τi and wi are time constants and relative weights of the decay components, respectively, while 〈τ〉 is the average time constant. Solutions and films were excited at 300 nm, and the PL decay was monitored at 650 nm (TP-5). b The uncertainty in the guest molecule concentration is approximately (0.1 mol %. c The error in the relative weight w of each PL decay component is approximately (10%.

occur at different guest molecule concentrations for the two pentacene derivatives. The significant shortening of the overall average lifetime of TP-5 with increasing concentration coincides with the relative increase of emission centered at ∼725 nm observed at the same concentration. There is no comparable increase in the long-wavelength PL peak of EtTP-5 at high concentrations. Steric hindrance effects that lead to isolation of the chromophoric pentacene core of EtTP-5 may discourage aggregation and lessen the PL concentration quenching. To further investigate possible aggregation effects in TP-5/ Alq3 films, time-resolved photoluminescence of the longwavelength peak of TP-5 (probed at 700 nm) was measured (Figure 7). In general, the time constants, rise times, and average lifetimes measured at 700 nm are nearly identical to those measured at 650 nm for TP-5/Alq3 films with concentrations below 1 mol % (Table 3). However, the higher-concentration films show a marked difference with a much longer-lived τ2 component (τ2 ) 19.3 ( 1.7 ns for 1.45 mol % and 12.7 ( 1.1 ns for 2.25 mol %). The slow component results in a lengthening of the 〈τ〉 for the higher-concentration films (g1.45 mol %) relative to the values obtained when the PL was collected at 650 nm. The distinct differences in decay dynamics of the peaks at 650 and 700 nm suggest that a different species is contributing to the 700 nm peak. The data show that the radiative lifetime (τr) of this species is considerably longer than that of the monomer. The increased long-wavelength emission (λmax ≈ 725 nm) in the g0.93 mol % TP-5/Alq3 films is a result of the formation of a weakly emissive species that shows a slight red shift relative to the original vibronic peak. Other solid-state guest-host organic systems show similar emission behavior at comparable guest molecule concentrations attributed to aggregate formation.41,42 The shape and location of the long-wavelength emission in the high-concentration TP-5/Alq3 films bears considerable resemblance to the emission observed from the TP-5 powder (Figure 3a). The very weak emission of TP-5 powder can be explained in terms of the close and well-ordered π-stacking of TP-5 molecules in the solid state that lead to strong intermolecular interactions and fluorescence self-quenching. These same interactions lead to aggregation in TP-5/Alq3 films, even at relatively low guest molecule concentrations. The additional ethyl groups on the terminal dioxolane rings of EtTP-5 may provide some steric hindrance and lead to less close π-stacking, explaining why quenching effects and changes in the PL spectra are suppressed relative to TP-5. This is clearly seen in the X-ray crystal structures of TP-5 and EtTP-5, which illustrate the

Figure 9. (a) X-ray crystal structure of TP-5 (viewed along the acene long-axis) depicting π-stacking of the aromatic core between two separate molecules with a π-face separation of 3.35 Å. (b) X-ray crystal structure of EtTP-5 (viewed along the acene long-axis) depicting π-stacking of the aromatic core between two separate molecules with a π-face separation of 5.30 Å. Note the lateral shift in the location of acene cores relative to the TP-5 structure, thereby precluding direct π-π stacking. Hydrogen atoms have been omitted for clarity in both images.

differences between the π-stacking and the intermolecular interactions of the two molecules. The structure of TP-5 reveals that the terminal dioxolane rings are situated above and below the central ring of the nearest pentacene neighbor and the centerto-center distance between the aromatic rings is only 3.35 Å (Figure 9a).25 The close spatial proximity of the aromatic units in the crystal structure implies strong π-π interactions that may lead to dimer formation of TP-5 in an Alq3 matrix, consistent with the observed PL quenching and concurrent changes in the PL spectra of TP-5/Alq3 films at high concentrations (g0.93 mol %). The TIPS pentacene (not shown) features a slightly longer center-to-center distance of 3.43 Å, but the crystal structure reveals significantly more π-overlap than in TP-5, which also leads to strong intermolecular interactions. Corresponding TIPS/Alq3 films show even stronger PL quenching as a function of guest molecule concentration than do TP-5/ Alq3 films (Figure 4).28 The crystal structure of EtTP-5 is considerably different.43 Given the size and orientation of the ethyl groups attached to the sp3-hybridized carbon of the dioxolane rings on EtTP-5, it is not surprising that EtTP-5 will not assume the same close π-stacked geometry that is observed for TP-5. The structure of EtTP-5 depicted in Figure 9b shows a much larger interplanar spacing (5.30 Å) relative to that in TP-5. Furthermore, the addition of the ethyl groups, which must be held perpendicular to the aromatic plane due to the rigidity of the five-membered dioxolane ring, precludes any significant π-face interaction. Thus, the aromatic backbone undergoes a lateral shift to place itself between the solubilizing triisopropylsilylethynyl groups of the molecules above and below it, effectively insulating the chromophore within the crystal.44 It is reasonable to assume that the intermolecular interactions that lead to aggregation in TP-5/Alq3 films will not play as large a role as those in the corresponding EtTP-5/Alq3 films. Indeed, the PL quenching in the EtTP-5/Alq3 films is considerably lessened, and little to no

Dioxolane-Substituted Pentacene Derivatives evidence of increased long-wavelength emission with increasing guest molecule concentration is seen in comparison to the TP5/Alq3 films (Figure 3). Conclusion We have demonstrated efficient energy transfer in composite films featuring Alq3 host molecules and dioxolane-substituted pentacene guest molecules using a combination of steady-state and time-resolved PL spectroscopy. The two pentacene derivatives, TP-5 and EtTP-5, show characteristic concentrationdependent differences in their PL spectra, absolute PL quantum yields, and transient PL decay profiles, which are traceable to differences in their molecular structures. The transient PL decay of the host reveals that energy transfer from Alq3 to EtTP-5 is slightly more efficient than energy transfer from Alq3 to TP-5. Both steady-state and time-resolved PL clearly identify the presence of aggregates at high concentrations (>1.00 mol %) in the TP-5/Alq3 films. The time-resolved PL data for EtTP-5/ Alq3 films also show some evidence of aggregation, but it is not as prominent and occurs at higher guest molecule concentrations. Aggregation in the TP-5/Alq3 films is further confirmed by changes in the PL spectra and strong quenching of the absolute PL quantum yields. Comparison of the X-ray crystal structures of the two molecules reveals that the ethyl groups on the terminal dioxolane rings of EtTP-5 result in extending the distance between π-faces, thereby lessening intermolecular interactions that lead to aggregation. Acknowledgment. This work was supported by the Office of Naval Research. M.W. acknowledges the National Research Council for administering the NRC postdoctoral fellowship program at the Naval Research Laboratory. References and Notes (1) Meng, H.; Bendikov, M.; Mitchell, G.; Helgeson, R.; Wudl, F.; Bao, Z.; Siegrist, T.; Kloc, C.; Chen, C.-H. AdV. Mater. 2003, 15, 1090. (2) Dimitrakopoulos, C. D.; Purushothaman, S.; Kymissis, J.; Callegari, A.; Shaw, J. M. Science 1999, 283, 822. (3) Sheraw, C. D.; Jackson, T. N.; Eaton, D. L.; Anthony, J. E. AdV. Mater. 2003, 23, 2009. (4) Videlot, C.; Fichou, D.; Garnier, F. J. Chim. Phys. 1998, 95, 1335. (5) Mayer, A. C.; Lloyd, M. T.; Herman, D. J.; Kasen, T. G.; Malliaras, G. G. Appl. Phys. Lett. 2004, 85, 6272. (6) Yoo, S.; Domercq, B.; Kippelen, B. Appl. Phys. Lett. 2004, 85, 5427. (7) Picciolo, L. C.; Murata, H.; Kafafi, Z. H. Appl. Phys. Lett. 2001, 78, 2378. (8) Wolak, M. A.; Jang, B. B.; Palilis, L. C.; Kafafi, Z. H. J. Phys. Chem. B. 2004, 108, 5492. (9) The theoretical limit to the external EL quantum efficiency may be expressed by the equation ηEL ) RγηrφPL, where R is the light output coupling factor (R ) 1/(2n)2, n ) refractive index), γ is the probability of carrier recombination, ηr is the production efficiency of excitons, and φPL is the absolute PL quantum yield of the emitter. Further information on ηEL for pentacene-based OLEDs can be found in refs 7 and 8. (10) Anthony, J. E.; Brooks, J. S.; Eaton, D. L.; Parkin, S. R. J. Am. Chem. Soc. 2001, 123, 9482. (11) Sakamoto, Y.; Suzuki, T.; Kobayashi, M.; Gao, Y.; Yasushi, F.; Inoue, Y.; Sato, F.; Tokito, S. J. Am. Chem. Soc. 2004, 126, 8138. (12) Baldo, M. A.; O’Brien, D. F.; You, Y.; Shoustikov, A.; Sibley, S.; Thompson, M. E.; Forrest, S. R. Nature 1998, 395, 151. (13) Chwang, A. B.; Kwong, R. C.; Brown, J. J. Appl. Phys. Lett. 2002, 80, 725.

J. Phys. Chem. B, Vol. 110, No. 15, 2006 7937 (14) Sakamoto, G.; Adachi, C.; Koyama, T.; Taniguchi, Y.; Merritt, C. D.; Murata, H.; Kafafi, Z. H. Appl. Phys. Lett. 1999, 75, 766. (15) Murata, H.; Merritt, C. D.; Kafafi, Z. H. IEEE J. Sel. Top. Quantum Electron. 1998, 4, 119. (16) Cleave, V.; Yahioglu, G.; LeBarny, P.; Hwang, D.-H.; Holmes, A. B.; Friend, R. H.; Tessler, N. AdV. Mater. 2003, 13, 44. (17) Virgili, T.; Lidzey, D. G.; Bradley, D. D. C. AdV. Mater. 2000, 12, 58. (18) Dogariu, A.; Gupta, R.; Heeger, A. J.; Wang, H. Synth. Met. 1999, 100, 95. (19) Liu, L.-I.; Barashkov, N. N.; Palsule, C. P.; Gangopadhyay, S.; Borst, W. L. J. Appl. Phys. 2000, 88, 4860. (20) Lane, P. A.; Palilis, L. C.; O’Brien, D. F.; Giebeler, C.; Cadby, A. J.; Lidzey, D. G.; Campbell, A. J.; Blau, W.; Bradley, D. D. C. Phys. ReV. B 2001, 63, 235206. (21) (a) Mattoussi, H.; Murata, H.; Merritt, C. D.; Iizumi, Y.; Kido, J.; Kafafi, Z. H. J. Appl. Phys. 1999, 86, 2642. (b) Shoustikov, A.; You, Y.; Burrows, P. E.; Thompson, M. E.; Forrest, S. R. Synth. Met. 1997, 91, 217. (22) Palilis, L. C.; Melinger, J. S.; Wolak, M. A.; Kafafi, Z. H. J. Phys. Chem. B. 2005, 109, 5456. (23) The initial reported value of R0 ) 26 Å from ref 22 was based on an incorrect form of the equation used to calculate the Fo¨rster radius (eq 1) in which π6, instead of π5, was given in the original manuscripts by Fo¨rster (refs 29 and 30). This extra π factor was later pointed out by Fo¨rster himself in Modern Quantum Chemistry; Sinanoglu, P., Ed.; Academic Press: London, 1965; p 135. (24) Sokolik, I.; Priestley, R.; Walser, A. D.; Dorsinville, R.; Tang, C. W. Appl. Phys. Lett. 1996, 69, 4168. (25) Payne, M. M.; Delcamp, J. H.; Parkin, S. R.; Anthony, J. E. Org. Lett. 2004, 6, 1609. (26) Reid, W.; Anthofer, F. Angew. Chem. 1953, 65, 601. (27) Lakowicz, J. R. Principles of Fluorescence Spectroscopy, 2nd ed.; Plenum Publishers: New York, 1999. (28) Picciolo, L. C.; Murata, H.; Gondarenko, A.; Noda, T.; Shirota, Y.; Eaton, D. L.; Anthony, J. E.; Kafafi, Z. H. Proc. SPIE-Int. Soc. Opt. Eng. 2002, 4464, 383. (29) Fo¨rster, T. Ann. Phys. 1948, 2, 55. (30) Fo¨rster, T. Discuss. Faraday Soc. 1959, 27, 7. (31) The extinction spectrum of 6,13-bis(2,6-dimethylphenyl)pentacene (λmax ≈ 600 nm) displays better overlap with the emission of Alq3 relative to the dioxolane-substituted pentacenes examined in this study. However, the overall overlap integral is actually smaller due to the weaker oscillator strengths of the diarylpentacene’s absorbances in the overlap region ( ≈ 2.2 × 104 M-1 cm-1). (32) Scholes, G. D. Annu. ReV. Phys. Chem. 2003, 54, 57. (33) Scholes, G. D.; Jordanides, X. J.; Fleming, G. R. J. Phys. Chem. B. 2001, 105, 1640. (34) Sokolik, I.; Walser, A. D.; Priestley, R.; Tang, C. W.; Dorsinville, R. Synth. Met. 1997, 84, 921. (35) Kalinowski, J.; Fattori, V.; Di Marco, P. Chem. Phys. 2001, 266, 85. (36) Mezyk, J.; Kalinowski, J.; Meinardi, F.; Tubino, R. Chem. Phys. Lett. 2004, 395, 321. (37) Powell, R. C. Phys. ReV. B 1970, 2, 2090. (38) Tang, C. W.; VanSlyke, S. A.; Chen, C. H. J. Appl. Phys. 1989, 65, 3610. (39) Priestley, R.; Sokolik, I.; Walser, A. D.; Tang, C. W.; Dorsinville, R. Synth. Met. 1997, 84, 915. (40) Bera, R. N.; Sakakibara, Y.; Tokumoto, M.; Yase, K. Jpn. J. Appl. Phys. 2003, 42, 7379. (41) Sakakibara, Y.; Okutsu, S.; Enokida, T.; Tani, T. Appl. Phys. Lett. 1999, 74, 2587. (42) Ohno, O.; Kalzu, Y.; Kobayashi, H. J. Chem. Phys. 1993, 99, 4128. (43) Crystal data for EtTP-5 (C54H70O4Si2): M ) 839.28, crystal size 0.4 × 0.3 × 0.04 mm3; triclinic, space group P1h, a ) 9.896(2), b ) 10.861(2), c ) 12.926(3) Å, R ) 75.87(3)°, β ) 77.93(3)°, γ ) 64.82(3)°, Z ) 1, V ) 1210.5(4) Å3, density (calcd) ) 1.151 g cm-3, T ) 90.0(2) K, 4296 measured reflections, 3475 were unique; R1 ) 0.0582, wR2 ) 0.1600, R1 ) 0.0714, wR2 ) 0.1696 (for all data); data-to-parameter ratio 15.39; Cu KR radiation (0.71073 Å), graphite monochromated. (44) EtTP-5 is the first pentacene derivative to show emission from solution-grown crystals.