Efficient Förster energy transfer from phosphorescent organic molecules to J-aggregate thin films


We demonstrate efficient Förster resonance energy transfer (FRET) from a thin film of phosphorescent dye, fac tris(2-phenylpyridine) iridium (Ir(ppy)3), to a thin film of J-aggregated cyanine dye, 5,6dichloro-2-[3-[5,6-dichloro-1-ethyl-3-(3-sulfopropyl)-2(3H)-benzimidazolidene]-1-propenyl]-1-ethyl-3(3-sulfopropyl) benzimidazolium hydroxide (TDBC). The measurement is performed on a planar sandwich structure with the layer of Ir(ppy)3 and the layer of J-aggregates separated by a uniform optically inert spacer layer. Quenching of Ir(ppy)3 photoluminescence due to FRET of Ir(ppy)3 excitons to J-aggregates enables us to calculate the experimentally-determined Förster radius of 3.8 nm, which is in good agreement with the theoretically calculated value. 2009 Published by Elsevier B.V. Due to their record high absorption constant and narrow photoluminescence (PL) linewidth [1,2], thin films of J-aggregated cyanine dyes have been extensively studied for their potential applications in novel opto-electronic devices, such as organic light emitting diodes (OLEDs), optical switches, and lasers [3–5]. Jaggregates’ unique optical properties also enabled the first demonstration of strong QED coupling between light and matter in room temperature microcavity structures containing J-aggregates under optical and electrical excitation [6–8]. Most recent studies demonstrated J-aggregate – quantum dot (QD) composite structures where the optical properties of J-aggregated molecular crystal or films are integrated with the optical response of QD nanostructures, where the coupling is activated via Förster resonant energy transfer [9,10]. Similarly, J-aggregates can be used as efficient exciton acceptors for a variety of organic dye donors including the technologically interesting phosphorescent molecules. In this Letter, we demonstrate and measure Förster resonant energy transfer (FRET) from an organic phosphorescent dye, fac tris(2-phenylpyridine) iridium (Ir(ppy)3), to a J-aggregate thin film of 5,6-dichloro-2-[3-[5,6-dichloro-1-ethyl-3-(3-sulfopropyl)-2(3H)benzimidazolidene]-1-propenyl]-1-ethyl-3-(3-sulfopropyl) benzimidazolium hydroxide, inner salt, sodium salt (TDBC) cyanine dye molecules. Organic phosphorescent materials have been extensively investigated due to their applications in high efficiency phosphorescent organic LEDs (OLEDs) [11,12]. In phosphorescentmaterials with high photoluminescence (PL) efficiency, strong spin–orbit coupling allows for mixing of the spin-singlet and spin-triplet states, leading to fast phosphorescent decay (<1 ls) and efficient phosphorescence at room temperature [13,14]. The considerable overlap between Elsevier B.V. the Ir(ppy)3 PL spectrum and the J-aggregate thin film absorption spectrum (Fig. 1) suggests the possibility of FRET from the excited Ir(ppy)3 molecules to the TDBC J-aggregates. The Ir(ppy)3 to TDBC exciton energy transfer was tested on structures consisting of three layers of organic materials on glass substrates. The bottom layer in all of our structures is a 5 nm thick J-aggregate thin film, comprised of bi-layers of TDBC and poly(diallyldimethylammonium chloride) (PDAC), deposited on glass using the layer-by-layer (LBL) method [15]. We followed the growth procedure for the J-aggregate film described in [2] with two minor modifications to the procedure: (1) No sodium hydroxide was used so that the pH of all the dipping solutions used was 5.5 (corresponding to the de-ionized water). (2) The polyelectrolyte-adsorption step consisted of immersing the glass substrates in the polyelectrolyte solution for 5 min instead of previously reported 15 min. The deposited TDBC/PDAC film is covered with a spacer layer of p-bis(triphenylsilyl)benzene (UGH-2) which is thermally evaporated. Three UGH-2 film thicknesses (0 nm, 4 nm, and 8 nm) were used to quantify the dependence of the energy transfer rates on the film-to-film distance. UGH-2 is a wide bandgap material (Eg 4.5 eV), used as a host for blue phosphorescent materials in efficient OLEDs [16]. The wide bandgap is a necessary property for the spacer layer in order to avoid perturbing the energy transfer between the donor and acceptor materials. The top layer in all of our structures is thermally evaporated 5 nm thick film of UGH-2 doped with Ir(ppy)3 at a concentration of 10% by volume. The surface of the J-aggregate thin film was chemically treated to improve the morphological uniformity of the spacer layer deposited on top. The symmetric chemical structure of UGH-2molecule suggests its hydrophobicity while, in contrast, TDBC and PDAC in the J-aggregate films are water soluble. The chemical dissimilarity causes UGH-2 to form rough films upon deposition onto Fig. 1. Spectral overlap of Ir(ppy)3 PL (solid line) and J-aggregate absorption (dashdot) suggests that energy transfer should be possible from Ir(ppy)3 to J-aggregates. Inset: chemical structure of Ir(ppy)3, UGH-2 and TDBC. 244 Y. Shirasaki et al. / Chemical Physics Letters 485 (2010) 243–246 an untreated surface of J-aggregates. To facilitate the formation of uniform UGH-2 films, the glass substrates with the J-aggregate films are treated with a surfactant by first immersing them in 0.03 M de-ionized (DI) water solution of sodium polystyrene sulfonate (SPS), a strong anionic polyelectrolyte, for 1 min and then rinsing them with DI water. The resulting films are then immersed in 0.01 M solution of cetyl trimethylammonium bromide (CTAB), a cationic surfactant, for 10 min and then rinsed with DI water. Finally, the films are rinsed with chloroform. The hydrophobicity of the surface treated films is tested using contact angle measurements with DI water, with which we find a contact angle of 80 , confirming the hydrophobic nature of the resulting surface. (We note that hydrophilic surfaces, such as oxygen–plasma-treated glass substrate and the untreated J-aggregate film, yield contact angles of 5 and 15 , respectively.) The atomic force microscope images in Fig. 2a and b show that our surface treatment slightly increases the surface roughness of J-aggregate films (root-meansquare (RMS) roughness changes from 1.0 nm to 1.8 nm). However, Fig. 2. AFM images of (a) 5 nm thick J-aggregate thin film on glass, (b) surface treated 5 nm thick J-aggregate thin film on glass, (c) 5 nm thick UGH-2 evaporated onto (a), and (d) 5 nm thick UGH-2 evaporated onto (b). The surface treatment of the J-aggregate thin film changes the surface from hydrophilic to hydrophobic and allows for uniform deposition of UGH-2 on top. Fig. 2d shows significantly lower roughness for a 5 nm thick UGH-2 film deposited onto a surface treated J-aggregate film (RMS roughness 1.9 nm) compared to a 5 nm thick UGH-2 film deposited onto an untreated J-aggregate film in Fig. 2c (RMS roughness 12.5 nm). The results indicate that the alteration of the J-aggregate film surface has made the deposition of uniform UGH-2 spacer layer possible. The thickness of the surface treatment layer was measured to be 1.7 nm using an ellipsometer (Gaertner Scientific) with k = 830 nmwavelength light source, which is roughly the thickness expected from two additional steps in a LBL process. We note that the immersion of the J-aggregate films in SPS solutions results in fading color of the J-aggregate films. We attribute this fading to the partial replacement of TDBC molecules in the film by the SPS molecules due to their strong anionic polyelectrolyte nature [17]. TDBC cyanine dye was obtained from Nippon Kankoh Shikiso Kenkyusho Co., Ltd. (CAS 28272-54-0). PDAC polyelectrolyte with molecular weight Mw = 400 000–500 000 was obtained from Sigma–Aldrich (CAS 26062-79-3). Ir(ppy)3 was purchased from Universal Display Corp. The chemical structures of the materials used in this Letter are shown in the inset of Fig. 1. To prevent exposure of our structures to ambient oxygen and moisture during measurements, the samples were packaged inside a nitrogen glovebox using glass coverslips sealed to the substrates with UVcuring epoxy. The finished three-layer structures were optically excited with k = 395 nm wavelength laser from the Ir(ppy)3 side with 200 fs pulse width and 100 kHz repetition rate obtained by frequency doubling the output of a Ti-Sapphire laser (Mira, Coherent). Time-resolved PL was measured using picosecond fluorescence lifetime system (Hamamatsu C4780) consisting of a streak camera (Hamamatsu C4334) and a spectrograph (Hamamatsu C5094). Fig. 3 shows the photoluminescence (PL) spectra of a 5 nm thick film of 10% Ir(ppy)3 in UGH-2 (by volume), a 5 nm thick surface treated J-aggregate thin film deposited onto glass substrates, as well as the PL spectra of the three hybrid structures (shown in the inset of Fig. 3) in which the Ir(ppy)3:UGH-2 films are separated from the J-aggregate films by the UGH-2 spacer layers of different thicknesses (0 nm, 4 nm and 8 nm). The PL spectra of the hybrid structures exhibit pronounced PL peaks at k = 515 nm and k = 595 nm, corresponding to Ir(ppy)3 and J-aggregate, respectively. We find that the Ir(ppy)3 emission component decreases and the J-aggregate emission component increases with decreasing Fig. 3. PL spectra of a surface treated 5 nm thick J-aggregate thin film, a neat 5 nm thick 10% (by volume) Ir(ppy)3 in UGH-2 film and the three hybrid structures with UGH-2 spacer layers (0 nm, 4 nm and 8 nm thick) excited at k = 395 nm. Ir(ppy)3 PL decreases with thinner spacer layer for the hybrid structures suggesting energy transfer. Inset: schematic drawing of the hybrid structure. Y. Shirasaki et al. / Chemical Physics Letters 485 (2010) 243–246 245 thickness of the UGH-2 spacer layer. Quenching of Ir(ppy)3 PL and a concomitant J-aggregate PL enhancement indicate energy transfer of Ir(ppy)3 triplet excitons to the J-aggregate thin films. To further confirm the energy transfer origin of the Ir(ppy)3 PL quenching and the J-aggregate PL enhancement, we performed time-resolved PL measurements. Fig. 4 shows the normalized PL decay curves for the neat Ir(ppy)3:UGH-2 film, the surface treated J-aggregate film, and the three hybrid samples over a 1 ls time window. From the trace in Fig. 4, we estimate Ir(ppy)3 PL relaxation time of 520 ns, consistent with previously reported value [18]. Since the J-aggregate PL relaxation time of 70 ps [19] is short compared to the resolution of the streak camera, we are unable to resolve the J-aggregate PL decay dynamics in this experiment. Consequently, the direct laser excitation of J-aggregates in the hybrid samples will result in a rapid PL decay of the J-aggregate excitons. We attribute the remaining PL of J-aggregates that persists over many nanoseconds after the excitation to energy transfer from Ir(ppy)3 to J-aggregates that is limited by the exciton diffusion rate in the Ir(ppy)3 film. From the time-resolved plots of the three hybrid structures we subtract the time-resolved PL of the surface treated J-aggregate film in order to monitor the remaining PL that is due to the energy transfer from Ir(ppy)3 to J-aggregates. We integrate the time-resolved PL signal over the 20 nm window around the J-aggregate emission peak (k = 595 nm) and find the J-aggregate PL relaxation time to be 45 ns, 135 ns and 165 ns for the hybrid samples with 0 nm, 4 nm and 8 nm thick UGH-2 spacers, respectively. The PL relaxation times of the hybrid samples decrease with thinner spacer layer due to an accelerated energy transfer rate via the long range (of up to several nm) FRET mechanism. FRET is a consequence of the dipole–dipole interaction between the molecules of the donor (e.g., Ir(ppy)3) and the acceptor (e.g., J-aggregate) in which exciton energy is transferred non-radiatively over distances exceeding the size of individual donor or acceptor molecules. An alternative mechanism of energy transfer is a direct electron exchange between a donor and an acceptor (Dexter energy transfer) which is a short range energy transfer (with characteristic distance of 1 nm) and therefore is unlikely in our system due to comparatively wide spacer layers that include the J-aggregate surface treatment layer. Lastly, radiative energy transfer (an acceptor molecule absorbing a photon emitted by a donor molecule) is a longer range energy transfer than FRET, but the PL quenching or the PL relaxation time of the donor is independent of the donor–acceptor Fig. 4. PL decay of a surface treated 5 nm thick J-aggregate thin film, a neat 5 nm thick 10% (by volume) Ir(ppy)3 in UGH-2 film and the three hybrid structures with UGH-2 spacer layers (0 nm, 4 nm and 8 nm thick) excited at k = 395 nm. Ir(ppy)3 PL decay rate increases with thinner spacer layer for the hybrid structures suggesting FRET. separation. Our hybrid samples show PL dependence on the spacer layer thickness, indicating that the FRET is the dominant energy transfer mechanism in our samples. Within the Förster theory, FRET rate between a donor molecule and an acceptor molecule is given by [20]:

4 Figures and Tables

Cite this paper

@inproceedings{Shirasaki2009EfficientFE, title={Efficient Förster energy transfer from phosphorescent organic molecules to J-aggregate thin films}, author={Yasuhiro Shirasaki and Polina O. Anikeeva and Jonathan Randall Tischler and Michael Scott Bradley and Vladimir Bulovic}, year={2009} }