Analyzing nanostructures in mesogenic host–guest systems for polarized phosphorescence

Organic Electronics 15 (2014) 311–321

Contents lists available at ScienceDirect

Organic Electronics journal homepage: www.elsevier.com/locate/orgel

Analyzing nanostructures in mesogenic host–guest systems for polarized phosphorescence Yu-Tang Tsai a, Chien-Yu Chen a, Li-Yin Chen a, Su-Hao Liu a, Chung-Chih Wu a,⇑, Yun Chi b, Shaw H. Chen c, Hsiu-Fu Hsu d, Jey-Jau Lee e a Department of Electrical Engineering, Graduate Institute of Electronics Engineering, Graduate Institute of Photonics and Optoelectronics, and Innovative Photonics Advanced Research Center (i-PARC), National Taiwan University, Taipei 10617, Taiwan, ROC b Chemistry Department, National Tsing-Hua University, Hsin-Chu 30013, Taiwan, ROC c Chemical Engineering Department, University of Rochester, Rochester, NY 14623-1212, USA d Chemistry Department, Tamkang University, Taipei 25137, Taiwan, ROC e National Synchrotron Radiation Research Center, Hsin-Chu 30076, Taiwan, ROC

a r t i c l e

i n f o

Article history: Received 20 October 2013 Received in revised form 5 November 2013 Accepted 12 November 2013 Available online 25 November 2013 Keywords: OLEDs Phosphorescence Polarized emission Liquid crystal GIXS

a b s t r a c t Feasibility of polarized phosphorescent organic light-emitting devices (OLEDs) had been previously demonstrated by combining a discotic Pt(II) complex with a glassy-nematic oligofluorene host to form a mesogenic host–guest phosphorescent emitting system. Previous photophysical studies suggested that in the host–guest film, the Pt(II) complex tended to aggregate into columnar stacks, exhibiting metal–metal-to-ligand charge transfer (MMLCT) emission. Both host molecules and guest aggregates in the host–guest films could be oriented by a conductive alignment layer, giving rise to polarized phosphorescence from the Pt(II) aggregates. Nevertheless, film morphologies and nanostructures of the mesogenic host–guest systems have remained to be elucidated. In this work, grazing incidence X-ray scattering (GIXS) was carried out to analyze nanostructures in both neat films of the discotic Pt(II) complex and mesogenic host–guest mixture films. In addition, confocal laser scanning microscopy (CLSM) was also utilized for visualization of the morphologies of mesogenic host–guest systems. The columnar axes of nanostructured Pt(II) stacks lying on the alignment-treated surfaces were found to be preferentially oriented perpendicular to the rubbing direction, which is responsible for the observed linearly polarized phosphorescence. Ó 2013 Elsevier B.V. All rights reserved.

1. Introduction The ability to organize nanostructures in ordered orientations over large areas is of primary importance for the bottom-up fabrication of nanostructure-based devices. By controlling the spatial arrangement and the degree of ordered nanostructures, it is possible to control polarization of light for optical information processing, such as displays, optical communication, optical storage, and stereoscopic (3D) imaging systems etc. [1,2]. For instance, ⇑ Corresponding author. Tel.: +886 2 33663636. E-mail address: [email protected] (C.-C. Wu). 1566-1199/$ - see front matter Ó 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.orgel.2013.11.025

polarized electroluminescence devices would be useful for backlights of liquid–crystal displays (LCDs) to make them more power efficient and for pixels of 3D displays to simplify their configurations [3–5]. As such, there have been substantial efforts in developing polarized organic light-emitting devices (OLEDs) [6–19]. With the strong intrinsic anisotropy in polymer chains, conjugated polymers that can form well aligned thin films [13–18,20], such as mesogenic polyfluorenes [21–24], represent a common class of active materials for polarized OLEDs. Meanwhile, with better control of molecular structures and material purity, mesogenic conjugated oligomers that can form well aligned films are another promising

312

Y.-T. Tsai et al. / Organic Electronics 15 (2014) 311–321

class of active materials for polarized OLEDs [6–12]. Liquid–crystalline (LC) oligofluorenes have been reported as hosts to guide molecular orientation of guest emitters, forming host–guest emitting systems for highly polarized and efficient OLEDs spanning the whole visible spectrum and white-light emission [8–11]. Most of previous efforts in polarized OLEDs, however, were mainly focused on fluorescence mechanisms. Yet, the development of OLEDs in recent years has largely shifted toward phosphorescent OLEDs [25,26], since phosphorescent OLEDs could effectively utilize both singlet and triplet excitons and realize essentially 100% internal quantum efficiencies [27,28]. As such, it is of both scientific and technical importance to explore the possibility of achieving polarized phosphorescent OLEDs. In a recent publication [29], we have reported an initial attempt to realize workable and functional polarized phosphorescent OLEDs by mixing a discotic mesogenic phosphorescent metal (Pt(II)) complex N200 (Fig. 1) with a glassy-nematic oligofluorene host (F(MB)5, Fig. 1) to form the corresponding mesogenic host–guest (phosphorescent) emitting system. Spectral properties of the neat N200 spin-cast films suggest N200 molecules exhibit strong ground-state intermolecular interactions (instead of simply excited-state interaction, like excimers) and they tend to self-assemble to form aggregates in films [29]. Previous luminescence characterization of bulk samples of these Pt(II) complexes revealed that they exhibited red emission only in the columnar mesophase, while monomer-like green emission was observed in isotropic liquid state or dilute solution [30]. Thus the appearance of red emission in N200 samples is a good indication of columnar-like mesophase packing. Due to the columnar mesomorphic nature of N200 and the resemblance of photophysical properties in spin-cast films to those of bulk samples in the liquid crystal phase [30], presumably the square-planar N200 molecules also pack into a one-dimensional columnar stacking arrangement in spin-cast films.

Fig. 1. The mesogenic materials, the Pt(II) complex N200 and the fluorene oligomer F(MB)5, used in this work.

To account for the observed mesomorphism, photophysical characteristics, and polarized transitions of these materials in various states and compositions, we proposed that N200 self-assemble into columnar stacking in the host–guest film, exhibiting metal–metal-to-ligand charge transfer (MMLCT) emission of the Pt(II) complex N200 [29,30]. Both the host molecules and guest aggregates in the host–guest films were successfully aligned on the rubbed conducting polymer alignment layer. With such alignment and effective host-to-guest energy transfer, polarized red phosphorescence and electrophosphorescence from the phosphorescent Pt(II) complex as aggregates were observed. Although this proposed scenario appeared consistent with all the physical and photophysical characterizations reported earlier, experimental observations are desired for film morphologies and nanostructures of the N200-oligofluorene host–guest systems. In this work, the grazing incidence X-ray scattering (GIXS) was performed to investigate morphologies and nanostructures in both neat N200 and its mixture with F(MB)5 films. In addition, the confocal laser scanning microscopy (CLSM) was also utilized for the visualization of morphologies of mesogenic host–guest systems.

2. Experiment 2.1. Materials Fig. 1 shows the mesogenic materials, the Pt(II) complex N200 and the fluorene oligomer F(MB)5 used in this work. N200 is a Pt(N^N)2 complex, in which N^N = 2-(3-(3,4,5trihexoxyphenyl)-1H-pyrazol-5-yl) pyridine. It adopts the distinctive pyridyl azolate as ligands, which are known to form strong chelate bonding with Pt ions and render the molecule a square-plannar geometry, affording distinctive emission (phosphorescent) properties [30–32]. Indeed, such Pt(II) complexes exhibit rather efficient green phosphorescence at room temperature in dilute solutions [30]. Further attaching alkyloxyphenyl groups with the alkyl chain (i.e., 3,4,5-trihexoxyphenyl) onto the ligands imparts certain flexibility to the molecular core and mesomorphic characteristics, thereby yielding luminescent metallomesogens [30,33–36]. As studied by polarizing optical microscopy and differential scanning calorimetry, N200 shows liquid–crystalline properties across a wide temperature range. Upon heating the crystalline samples of these Pt(II) complexes, a transition from the solid to the columnar mesophase (as verified by X-ray diffraction [30]) and a transition from the mesophase to the isotropic liquid occurs at 98 °C and 342 °C, respectively. The details of syntheses and basic material properties of the Pt(II) complexes had been reported elsewhere [30]. Employed as the host, the fluorene oligomer F(MB)5 consists of five fluorene units on the backbone and two 2-methylbutyl substituents at C9 atoms [8,9]. F(MB)5 belongs to a class of glass-forming nematic oligofluorenes, that is, materials that exhibit the nematic mesophase at elevated temperatures and yet also the stable glass phase at room temperature [8–11]. As studied by polarizing optical microscopy and differential scanning calorimetry,

Y.-T. Tsai et al. / Organic Electronics 15 (2014) 311–321

313

F(MB)5 exhibits a glass transition (Tg) at 89 °C and a nematic-to-isotropic transition at 167 °C. It had been demonstrated that liquid–crystal glass of F(MB)5 could be formed by freezing its mesophase down to room temperature without crystallization, and that the aligned glassy nematic film of F(MB)5 could be formed by utilizing mesophase-mediated alignment (e.g., assisted by the rubbed alignment layer of a polyimide or a conducting polymer) at elevated temperatures [8,9]. F(MB)5 is also an efficient fluorescent blue emitter on its own with a thin-film quantum yield of >50% [8,9]. Detailed syntheses and characterizations of F(MB)5 could be found in previous publications [8]. 2.2. Molecular Alignment To conduct and study molecular alignment of the mesogenic oligofluorene/mesogenic Pt(II) complex host–guest mixtures, different sample structures and alignment approaches were tested. In scheme I, the active thin films under investigation were spin-coated onto the quartz or Si/ SiO2 substrates and then uniaxially rubbed with a cloth for surface alignment. In scheme II, a conducting polymer, poly(3,4-ethylene-dioxythiophene):poly(styrenesulfonate) (PEDOT:PSS), was used as the alignment layer. PEDOT:PSS was spin-coated onto quartz or Si/SiO2 substrates, dried at 130 °C in air, and uniaxially rubbed with the cloth. The active thin films under investigation were then spin-coated onto the rubbed PEDOT:PSS for surface alignment. In both schemes I and II, the active layer to be aligned could be a neat film of F(MB)5, a neat film of the Pt(II) complex N200, or a host–guest mixture film of F(MB)5 and Pt(II) complex N200 (with 25 wt.% of N200). To induce mesophase-mediated molecular alignment, all samples were further annealed at elevated temperatures for 1 h: 150 °C for neatN200 films and 120 °C for the neat F(MB)5 or the F(MB)5:N200 mixture films. With an aim to construct polarized phosphorescent OLEDs using the mesogenic Pt(II) complexes, we first tried to use the rubbed PEDOT:PSS film to align all material layers (i.e., scheme II), since PEDOT:PSS was to be used as the conductive alignment layer in polarized OLED. However, since scheme II (the one preferred for device applications) failed to orient the neat Pt(II) complex films, scheme I was followed to investigate fundamental properties of neat Pt(II) complex films [29]. Scheme II was successfully implemented to accomplish orientation in neat F(MB)5 and the F(MB)5:N200 mixture films for both fundamental studies and device applications. 2.3. Grazing incidence X-ray scattering GIXS was conducted for the analysis of host–guest systems. Compared to other conventional techniques for morphological characterizations (e.g. atomic force microscopy, scanning or transmission electron microscopy etc.), GIXS has the particular advantage of being able to provide structural/morphological information of a thin film at different scales [37–44], instead of being limited to local observation. Fig. 2 illustrates the configuration of the GIXS measurement, which was conducted at the BL17A end

Fig. 2. Schematic illustration of the GIXS measurement. hIN denotes the incident angle, 2h the diffraction angle (relative to the sample-detector axis). The propagation direction of the incident X-ray is defined as the Y axis, the out-of-plane (OOP) direction as the Z axis, and in-plane (IP) direction perpendicular to the Y axis as the X axis. QX and QZ are the components of the scattering vector along the IP (X axis) direction and the OOP (Z axis) direction, respectively. For samples subjected to rubbing and annealing treatments and thus having anisotropic in-plane characteristics, GIXS was conducted in two different configurations: (i) the propagation direction of the incident X-ray was parallel (aligned) with the rubbing direction (x = 0°); (ii), the rubbing direction was rotated away from the propagation direction of the incident X-ray by 90° (i.e., x = 90°), thus perpendicular to the incident beam.

station of the National Synchrotron Radiation Research Center, Hsinchu, Taiwan. The 9.3-keV (k = 1.33 Å) beam having a 1 mm beam diameter was directed at a 0.1° incidence angle (hIN). The 2D scattering images were collected by a MAR345 CCD detector array 210 mm away. To avoid the reflection of the incident beam from the substrate that was strong relative to the scattering signals, aluminum foils were applied as the semitransparent beam stop. The thin-film samples were kept at room temperature in air during irradiation and GIXS image collection. In Fig. 2, hIN denotes the incident angle, and 2h is the diffraction angle (relative to the sample-detector axis). The propagation direction of the incident X-ray is defined as the Y axis, the out-of-plane (OOP) direction relative to the sample surface as the Z axis, and in-plane (IP) direction perpendicular to the Y axis as the X axis. QX and QZ are the components of the scattering vector Q along the IP (X axis) direction and the OOP (Z axis) direction, respectively, where Q = 4psin(h)/k and k = the X-ray wavelength. For samples subjected to rubbing and annealing treatments and thus having anisotropic in-plane characteristics, GIXS was conducted in two different configurations. In the first configuration, the propagation direction of the incident X-ray was parallel to the rubbing direction (i.e., the angle between these two directions x = 0°). In the second configuration, the rubbing direction was rotated away from the propagation direction of the incident X-ray by 90° (i.e., the angle between these two directions x = 90°), thus perpendicular to the incident beam. 2.4. Confocal laser scanning microscopy (CLSM) CLSM was used as one of the techniques for visualization of morphologies and nanostructures of thin films [45,46]. CLSM of thin-film samples was conducted on a

314

Y.-T. Tsai et al. / Organic Electronics 15 (2014) 311–321

Leica TCS SP5 confocal laser scanning microscope system (Leica Microsystems) using an 100X oil-immersion objective. The system used a linearly polarized diode UV laser at 405 nm as the excitation source and a spectral PMT as the detector. The scanning rate is 400 Hz, and the step size is 40 nm. The optical resolution of CLSM can be optimized to 200 nm. Three thin-film samples (either before or after rubbing and annealing treatments) were investigated: neat films of F(MB)5 (host), neat films of Pt(II) complexes (guest), and mixture films of F(MB)5:Pt(II) complexes. The spectral window of the PMT was set at 420–460 nm for collecting the (blue) luminescence CLSM image of the host material [F(MB)5], and at 600–700 nm for collecting the (red) luminescence CLSM image of the guest material [Pt(II) complex aggregates]. For oriented thin-film samples, polarization-dependent CLSM was studied by aligning the polarization of the excitation laser either parallel (//) or perpendicular (\) to the rubbing direction. 2.5. Variable-angle spectroscopic ellipsometry Variable-angle spectroscopic ellipsometry (VASE) in the reflection mode was used to study the optical constants of thin films of neat N200 for verifying the in-plane and outof-plane anisotropy absorption of the columnar stacking. Ellipsometry measures the change in polarization of light as a function of incident angle and wavelength [6,47,48]. The experimentally determined ellipsometric parameters are W and D, which are related to the ratio of Fresnel reflection coefficients Rp and Rs for p- and s-polarized light, respectively, by Rp/Rs = tan(W) eiD. Optical constants of the materials are then determined by first constructing an optical model of the sample with physically meaningful structural and optical parameters, and then by iteratively adjusting these parameters to obtain the best fit to the measured ellipsometric data by minimizing the mean square error (MSE) [47]. Ellipsometry over a wavelength range of 270–1100 nm in 5 nm steps was performed in air using the J.A. Woollam V-VASE spectroscopic ellipsometer equipped with a xenon lamp source. The angles of incidence used were between 65° and 75° relative to the surface normal in steps of 5°. The nonlinear regression analysis of the measured ellipsometric data was performed using the J.A. Woollam WVASE32 software. 3. Results and discussions 3.1. GIXS and ellipsometric analyses of neat films of Pt(II) complexes Fig. 3 presents the 2D GIXS images of the neat N200 films under three different conditions: (a) as-deposited, without rubbing and annealing treatments, (b) after rubbing and annealing treatments, measured with the incident X-ray parallel to the rubbing direction (x = 0°), and (c) after rubbing and annealing treatments, measured with the incident X-ray perpendicular to the rubbing direction (x = 90°). As can be seen in Fig. 3(a), even for the asdeposited film (without rubbing and annealing treatments), there are already clear X-ray scattering patterns,

Fig. 3. 2D GIXS images of the neat N200 films under three different conditions: (a) as-deposited, without rubbing and annealing treatments, (b) after rubbing and annealing treatments, measured with the incident X-ray parallel with the rubbing direction (x = 0°), and (c) after rubbing and annealing treatments, measured with the incident X-ray perpendicular to the rubbing direction (x = 90°).

confirming the formation of N200 aggregates in the asdeposited neat film. Although not very strong, in Fig. 3(a), one can observe a pattern representing the Pt– Pt stacking around Qx = 19 Å1. It indicates that the N200 Pt complexes tend to self-assemble into columnar stacks in the as-cast film. In addition, the Pt–Pt signal only

Y.-T. Tsai et al. / Organic Electronics 15 (2014) 311–321

appears in the in-plane direction (i.e., along Qx), not observed in the out-of-plane direction (i.e., along Qz). It suggests that the columnar axis lies on the substrate surface. In Fig. 3(a), one also observes GIXS patterns at smaller Q’s, which represent the inter-column assembly structure of the aggregates. The more ring-like GIXS patterns at smaller Q’s, however, indicate a somewhat random distribution of the inter-column structures. As seen in Fig. 3(b) and (c), upon rubbing and thermal annealing, the signals representing Pt–Pt stacking become sharper and stronger, indicating growth in aggregate size. Again, the Pt–Pt signals still appear only in the in-plane direction (i.e., at Qx  19 Å1), not observed in the out-ofplane direction (i.e., along Qz). As Fig. 3(b) is compared with Fig. 3(c), GIXS signals at smaller Q’s converge into specific peaks (dots) and are significantly enhanced with rubbing and annealing. The observed GIXS patterns indicate that aggregates grow in size and adopt better organized inter-column structures and orientation as a result of rubbing/annealing. Moreover the better organized structures are preserved upon subsequent cooling. Examining the GIXS patterns in Fig. 3(b) and (c) reveals that the assembly of the Pt complexes has the tetragonal columnar structure (Colt), whose packing structure and assignment of Miller indices are illustrated in Fig. 4. Note that Liao et al. [30] had reported that similar Pt complexes, although exhibiting the hexagonal columnar mesophase at higher temperatures (>240 °C), did exhibit the rectangular/tetragonal columnar mesophase at lower temperatures (<240 °C, like the case here). For convenience, the [1 0 0] is assigned along the Pt–Pt stacking direction, while [0 1 0] and [0 0 1] are assigned for the two inter-column packing directions. With the emerging assembly structure and indexing system, the GIXS peaks in Fig. 3(b) and (c) can now be unambiguously assigned as such. Reflections occurring along the outof-plane orientation that represent the out-of-plane intercolumn packing can be assigned [0 0 1] and higher-order ones, while reflections occurring along the in-plane orientation that represent in-plane inter-column packing can be assigned [0 1 0] and higher-order ones. Anisotropic characteristics of the aggregates of Pt complexes in rubbed and annealed samples can be further analyzed by considering different packing orientations through comparing GIXS images detected with the incident X-ray parallel to the rubbing direction (i.e., GIXS at x = 0°, Fig. 3(b)) to those detected with the incident

Fig. 4. The packing structure of the tetragonal columnar structure (Colt) and the assignment of Miller indices.

315

X-ray perpendicular to the rubbing direction (i.e., GIXS at x = 90°, Fig. 3(c)). Fig. 5 shows three possible columnar packing orientations relative to the surface and the rubbing direction. In case A, the columns (or intra-column Pt–Pt stacking) are vertical to the surface and the inter-column packing is along the surface. In such a case, ideally in GIXS one would only observe the Pt–Pt stacking signal along the OOP direction (i.e., along Qz) but not along the IP direction (i.e., along Qx). In addition, one would also observe no inter-column packing signal (i.e., [0 0 1] or [0 1 0]) along the OOP direction. In case B, columns (or intra-column Pt–Pt stacking) are parallel to the surface and the rubbing direction, while one inter-column packing direction (i.e., [0 0 1]) is along the OOP direction and another (i.e., [0 1 0]) is perpendicular to the rubbing direction. In such a case, ideally in 2D GIXS one would observe no Pt–Pt stacking signal along the OOP direction, and whether the Pt–Pt stacking signal along the IP direction can be observed depends on x (0° or 90°). At x = 0°, one would observe inter-column packing signals [0 0 1]/[0 1 0] along the OOP/IP direction, respectively, but no Pt–Pt signal along the IP direction (i.e., along Qx) since the Pt–Pt packing is along the Y direction under such a measurement configuration (thus cannot be easily resolved by the 2D GIXS). On the other hand, by rotating the sample to make x = 90°, one shall now observe (0 0 1)/Pt–Pt stacking signal along the OOP/IP direction, respectively, but no [0 1 0] signal along the IP direction since the [0 1 0] orientation is now along the Y direction. In case C, columns (or intra-column Pt–Pt packing) are parallel with the surface but perpendicular to the rubbing direction, while one inter-column packing direction (i.e., [0 0 1]) is along the OOP direction and another (i.e., [0 1 0]) is parallel with the rubbing direction. In such a case, ideally in 2D GIXS one would observe no Pt–Pt stacking signal along the OOP direction, while whether the Pt–Pt stacking signal can be observed again depends on x (0° or 90°). At x = 0°, one would observe [0 0 1]/Pt–Pt stacking signal along the OOP/IP direction, respectively, but no [0 1 0] signal along the IP direction since the [0 1 0] orientation is along the Y direction. On the other hand, at x = 90°, one shall now observe inter-column packing signals [0 0 1]/[0 1 0] along the OOP/IP direction, respectively, but no Pt–Pt signal along the IP direction since the Pt–Pt packing is now along the Y direction. Since in GIXS images of Fig. 3(b) and (c) one sees clear inter-column packing signal [0 0 1] but no intra-column Pt–Pt stacking signal along the OOP direction (i.e., along Qz), the packing orientation case A can be excluded. Observation of the Pt–Pt stacking signal mainly along the IP direction suggests packing orientations of cases B and C. Cases B and C can be further distinguished by comparing GIXS images collected under x = 0° and x = 90° configurations. Fig. 6 shows the in-plane profiles (i.e., along Qx) of the three GIXS images in Fig. 3(a)–(c), with the Pt–Pt stacking signal (i.e., at Qx  19 Å1) being enlarged and shown in the inset. In the inset of Fig. 6, one sees that for the sample subjected to the rubbing and annealing treatments, the Pt– Pt signal detected with x = 0° is stronger than that detected with x = 90°. In addition, for the sample subjected to the rubbing and annealing treatments, the (0 1 0) signal

316

Y.-T. Tsai et al. / Organic Electronics 15 (2014) 311–321

Fig. 5. Three possible/representative column packing orientations (of the tetragonal columnar assembly) relative to the surface and the rubbing direction.

Fig. 6. In-plane profiles (i.e., along Qx) of the three GIXS images in Fig. 3(a)–(c), with the Pt–Pt stacking signal (i.e., around Qx  19 Å1) being enlarged and shown in the inset.

detected with x = 0° is significantly weaker than that detected with x = 90°. These results indicate that, with the rubbing and annealing treatments: (i) the stacking columns of the Pt(II) complexes appear to have the preferential orientation perpendicular to the rubbing direction; (ii) the in-plane inter-column packing (i.e., [0 1 0] orientation) is preferentially along the rubbing direction. That is, among the three possible column packing orientations relative to the surface and the rubbing direction illustrated in Fig. 5, case C represents the preferential column packing orientation, although the intra-column packing (i.e., Pt–Pt) and inter-column packing (i.e. [0 1 0]) perhaps are not perfectly aligned relative to the rubbing direction and there is some distribution between cases B and C. Such results are consistent with previous results of polarized optical spectroscopy (i.e., polarized photoluminescence and absorption) of neat films of Pt complexes subjected to rubbing and annealing treatments [29]. Previous results of polarized optical spectroscopy showed that the polarization of the MMLCT absorption and emission (and thus the columnar Pt–Pt stacking) were more along the direction perpendicular to the rubbing direction, while the absorption associated with the pyridyl azolate ligands (and thus the molecular plane of the discotic Pt complex) was more along the rubbing direction. From GIXS patterns of the rubbed and annealed samples shown in Fig. 3(b) and (c), the d spacings along the intra-column Pt–Pt stacking direction (dPt–Pt), along the

inter-column [0 0 1] direction (d0 0 1), and along the intercolumn [0 1 0] direction (d0 1 0) are calculated based on Bragg’s law [49] and are listed in Table 1. Furthermore, the average dimensions of the aggregates/grains along the above different packing directions (i.e., LPt–Pt, L0 0 1, L0 1 0) are also estimated using the Scherrer grain-size analysis [50] and are also listed in Table 1. Note that the calculated d0 1 0 of 22.22 Å is p close to that of d0 0 1 of 22.62 Å, ffiffiffi and that both are about 2 times the d spacing along the [0 1 1] direction (15.48 Å, not listed in Table 1), consistent with the assignment of the packing structure to the tetragonal columnar structure. The calculated inter-column spacing of 22.4 ± 0.2 Å is shorter than the molecular length (29.19 Å), but longer than the rigid core length of 16.24 Å, suggesting the discotic Pt complexes are assembled into the columnar organization with a tilted format relative to the inter-column [0 0 1] and [0 1 0] orientations. The intra-column Pt–Pt stacking distance (dPt–Pt) is calculated to be 3.33 Å, which is well within the Pt–Pt distances allowing strong MMLCT transitions to occur [51,52]. On the other hand, with LPt–Pt = 43.33 Å, L0 0 1 = 274.47 Å, and L0 1 0 = 206.48 Å in the rubbed and annealed neat films, one estimates the average size of the aggregates/grains to be 13 Pt complexes along the Pt–Pt stacking direction, 12 columns along the inter-column [0 0 1] direction and 9–10 columns along the inter-column [0 1 0] direction. The Scherrer grain-size analysis also shows that the assembly size of Pt complexes had grown with rubbing and annealing treatment. For instance, LPt–Pt had grown from 26 Å of the as-deposited neat film to 43.33 Å of the rubbed and annealed film. From the above GIXS analyses, it has been established that in the rubbed and annealed neat films of N200, the structure and orientational alignment of the molecular assembly is inclined toward case C described in Fig. 5. In addition, it is deduced that the discotic Pt complexes are assembled into the columnar organization with a tilted format relative to the inter-column [0 0 1] and [0 1 0] orientations. Such anisotropic molecular assembly can be further verified by characterizing the anisotropic optical properties of rubbed and annealed neat films originating from the MMLCT transitions and ligand-centered transitions (utilizing the variable-angle spectroscopic ellipsometry-VASE). Good fittings to the experimentally measured ellipsometric values W and D could be obtained by using a biaxial model for refractive indices and extinction coefficients, which in turn were constructed by a Kramers–Kronig consistent

317

Y.-T. Tsai et al. / Organic Electronics 15 (2014) 311–321

Table 1 Packing parameters of N200 assemblies/aggregates in films with different compositions and processing/treatment conditions, including the d spacings along the intra-column Pt–Pt stacking direction (dPt–Pt), along the inter-column [0 0 1] direction (d0 0 1), and along the inter-column [0 1 0] direction (d0 1 0), and the average dimensions of the aggregates/grains along the above different packing directions (i.e., LPt–Pt, L0 0 1, L0 1 0).

N200 no treatments N200 with treatments F(MB)5:N200 no treatments F(MB)5:N200 with treatments

dPt–Pt (Å)

LPt–Pt (Å)

d0 0 1 (Å)

L0 0 1 (Å)

d0 1 0 (Å)

L0 1 0 (Å)

3.34 3.33 – 3.35

25.97 43.33 – 18.96

22.36 22.62 – 23.68

183.19 274.47 – 143.21

22.15 22.22 – 22.10

184.82 206.48 – 121.62

model using a combination of one Cauchy background and several Gaussian oscillators. Fig. 7 shows the extracted absorption coefficients converted from the extracted anisotropic extinction coefficients for the rubbed and annealed neat film of N200. a//,IP, a\,IP, and aOOP represent the absorption coefficients for the in-plane absorption coefficient along the rubbing direction, the in-plane absorption coefficient perpendicular to the rubbing direction, and the outof-plane absorption coefficient, respectively. One notes that a//,IP and a\,IP agree well with the normal-incidence absorption spectra of the uniaxially rubbed N200 film measured with the polarization parallel (//) and perpendicular to (\) the rubbing direction (inset of Fig. 7, [29]), respectively, inspiring additional confidence in the anisotropic optical constants obtained. For the low-energy MMLCT transition region (e.g., 520–600 nm), the strongest and clear absorption peak in a\,IP, weaker a//,IP, and nearly zero aOOP are consistent with the foregoing interpretation that the Pt–Pt stacking orientation is most likely case C of depicted in Fig. 5, mingled with a bit of case B, and hardly any case A, since the MMLCT transition is typically along the chain of the metal cores [53]. For the higher-energy ligand-centered transition region (e.g., 300 nm), both a//,IP and aOOP are larger than a\,IP, with the dichroic ratio of >2, which suggests that in the aligned neat films of Pt(II) complexes, the pyridyl azolate ligands with alkyloxyphenyl groups are roughly along the surface expanded by the rubbing

Fig. 7. Extracted absorption coefficients converted from the extracted anisotropic extinction coefficients for the rubbed and annealed neat film of N200. a//,IP, a\,IP, and aOOP represent the absorption coefficients for the in-plane absorption coefficient along the rubbing direction, the in-plane absorption coefficient perpendicular to the rubbing direction, and the out-of-plane absorption coefficient, respectively. The inset shows the normal-incidence absorption spectra of the uniaxially rubbed N200 film measured with the polarization parallel with (//) and perpendicular to (\) the rubbing direction.

and the OOP directions (with certain tilting relative to the surface). Again, the residual a\,IP suggests the alignment is not perfect and there is certain mixture of case B and case C (Fig. 5).

3.2. GIXS analyses of host: guest mixture films Fig. 8 shows the 2D GIXS images of the F(MB)5:N200 mixture films coated over PEDOT:PSS under three different conditions: (a) without alignment treatment, (b) with alignment treatment, measured with the incident X-ray parallel with the rubbing direction (x = 0°), and (c) with alignment treatment, measured with the incident X-ray perpendicular to the rubbing direction (x = 90°). In Fig. 8(a), for the F(MB)5:N200 mixture film on PEDOT:PSS without alignment treatment, no specific GIXS patterns (e.g., peaks, rings etc.) associated with N200 assembly can be detected, except for the cloudy background. Although previous photoluminescence studies indicate there is Pt–Pt stacking and MMCLT emission of N200 even in the pristine F(MB)5:N200 mixture films subjected to no alignment, the sizes and/or quantities of the N200 assemblies are probably not sufficient and the orientations are probably too random to be detected by GIXS. Comparison of Fig. 8(a) and Fig. 3(a) also suggests the presence of the unoriented F(MB)5 host significantly disturbs the assembly/stacking of the Pt complex N200. As seen in Fig. 8(b) and (c), for the mixture film subjected to alignment, the signal representing Pt–Pt stacking (at Qx  19 Å1) of N200 appears, indicating growth of N200 aggregates. Again, the Pt–Pt signals still appear only in the in-plane direction (i.e., at Qx  19 Å1), not observed in the out-of-plane direction (i.e., along Qz). Moreover, alignment produced signals representing the inter-column packing (i.e., peaks at smaller Q’s) that suggest growth of N200 aggregates and adoption of more organized inter-column structures and orientation within the oriented F(MB)5 host. Overall, the GIXS pattern of the aligned F(MB)5:N200 mixture film is similar to that of the rubbed/annealed neat N200 film (Fig. 3(b) and (c)), indicating that N200 assemblies in both cases have similar Pt–Pt stacking, inter-column structure (i.e., tetragonal columnar structure, Colt), and orientation/alignment. As such, Miller indices of GIXS peaks in Fig. 8(b) and (c) can be assigned in a way similar to that in Fig. 3(b) and (c), with the [1 0 0] being for the Pt– Pt stacking direction, [0 0 1] and higher-order ones being for reflections and inter-column packing along the outof-plane orientation, [0 1 0] and higher-order ones being for reflections and inter-column packing along the in-plane orientation.

318

Y.-T. Tsai et al. / Organic Electronics 15 (2014) 311–321

Fig. 9. In-plane profiles (i.e., along Qx) of the three GIXS images in Fig. 8(a)–(c), with the Pt–Pt stacking signal (i.e., around Qx  19 Å1) being enlarged and shown in the inset.

Fig. 8. 2D GIXS images of the F(MB)5:N200 mixture films coated over PEDOT:PSS under three different conditions: (a) without no alignment treatments, (b) with alignment treatments, measured with the incident X-ray parallel with the rubbing direction (x = 0°), and (c) with alignment treatment, measured with the incident X-ray perpendicular to the rubbing direction (x = 90°).

Fig. 9 shows the in-plane profiles (i.e., along Qx) of the three GIXS images in Fig. 8(a)–(c), with the Pt–Pt stacking signal (i.e., around Qx  19 Å1) being enlarged and shown in the inset. As in the case of the neat N200 film, for the

sample subjected to the rubbing and annealing treatments, the Pt–Pt signal detected with x = 0° is stronger than that detected with x = 90°. In addition, the inter-column [0 1 0] signal detected with x = 0° is significantly weaker than that detected with x = 90°. Thus in the mixture film, with the alignment treatment, the N200 assemblies in the F(MB)5 host also preferentially adopt the column packing orientation (relative to the rubbing direction and the surface) like case C in Fig. 5, although the alignment is imperfect (i.e., there may be some distribution between case B and case C). Such results are consistent with previous results of polarized optical spectroscopy (i.e., polarized photoluminescence) of the F(MB)5:N200 mixture films subjected to alignment treatments [29]. Previous results of polarized optical spectroscopy showed that the polarization of the MMLCT emission (and thus the columnar Pt–Pt stacking) of N200 guests were more perpendicular to the rubbing direction, while the F(MB)5 host emission was more along the rubbing direction (i.e., F(MB)5 host molecules are aligned along the rubbing direction). Although F(MB)5 host molecules are oriented on the alignment-treated surface, yet no GIXS signals associated with F(MB)5 can be detected. It is perhaps because F(MB)5 possess a very short intermolecular correlation length not readily detectable by GIXS. From GIXS patterns of the rubbed and annealed samples shown in Fig. 8(b) and (c), the d spacings dPt–Pt, d0 0 1, and d0 1 0 for the aligned mixture film are calculated and are listed in Table 1. One notices that these d spacing values are similar to those in the N200 neat films, confirming similar assembly structures in both cases. The average dimensions of the N200 aggregates in mixture films along the above different packing directions (i.e., LPt–Pt, L0 0 1, L0 1 0) are also estimated and are also listed in Table 1. On the other hand, with LPt–Pt = 18.96 Å, L0 0 1 = 143.21 Å, and L0 1 0 = 121.62 Å in the aligned mixture films, one estimates the average size of the aggregates at 5–6 Pt complexes along the Pt–Pt stacking direction, 6 columns along the inter-column [0 0 1] direction and 5–6 columns along the inter-column [0 1 0] direction. From results of Table 1, one might speculate that although the nematic F(MB)5

Y.-T. Tsai et al. / Organic Electronics 15 (2014) 311–321

host can guide the alignment of the N200 aggregates in the mixture film, the presence of the F(MB)5 hosts disturbs the assembly of the Pt complex N200, thereby reducing the sizes of N200 aggregates to nearly half.

319

Although the resolution of CLSM used might not be sufficient to quantify the nanostructures encountered in this study, the CLSM images were collected as supportive information for aggregates of N200 and the nanoscale phase separation in the F(MB)5:N200 mixture films. Fig. 10(a) and (b) show the luminescence CLSM images of the neat F(MB)5 film and the neat N200 film, respectively, either without (panel (i)) or with (panels (ii) and (iii)) rubbing/annealing treatments. For the aligned thinfilm samples (i.e., with rubbing/annealing), polarizationdependent CLSM images were taken by aligning the polarization of the excitation laser either perpendicular (\, panel (ii)) or parallel to (//, panel (iii)) the rubbing direction. In panel (i) of Fig. 10(a), one sees that CLSM image of the neat F(MB)5 film without rubbing/annealing is homogeneous, as expected of an amorphous/glassy organic film. With rubbing and annealing, the image of the // polarization (panel (iii) of Fig. 10(a)) becomes significantly stronger than that of \ polarization in panel (ii) of Fig. 10(a) (as expected), and yet the CLSM images are still generally homogeneous (except for a few spots associated with non-ideal rubbing/alignment). In contrast, in panel (i) of Fig. 10(b), one observes a grainy CLSM image even in the

neat N200 film without rubbing/annealing, indicating formation of aggregates/grains as coated. With rubbing and annealing (panels (ii) and (iii) of Fig. 10(b)), in addition to polarization dependence of CLSM images (due to alignment), one clearly sees substantial growth of domains. Fig. 11(a) and (b) show the luminescence CLSM images of the F(MB)5 emission and the N200 MMLCT emission taken from a same F(MB)5:N200 mixture film, respectively, either without (panel (i)) or with (panels (ii) and (iii)) rubbing/annealing treatments. For the aligned thin-film samples (i.e., with rubbing/annealing), polarizationdependent CLSM images were taken by aligning the polarization of the excitation laser either perpendicular (\, panel (ii)) or parallel to (//, panel (iii)) the rubbing direction. In panel (i) of Fig. 11(a) and (b), one observes grainy CLSM images for both F(MB)5 emission and N200 MMLCT emission even in mixture films without rubbing/ annealing, indicating certain extent of phase separation. With rubbing and annealing (panels (ii) and (iii) of Fig. 11(a) and (b)), in addition to polarization dependence of CLSM images (due to alignment), again one sees significant growth of phase-separated domains and N200 aggregates/grains, which is consistent with the observation in GIXS. In CLSM images of the aligned host–guest mixture films, //-polarization N200 MMLCT emission (panel (iii) of Fig. 11(b)) can be somewhat detected, although weaker than the \-polarization emission (panel (ii) of Fig. 11(b)). It is perhaps due to imperfect alignment of guest aggregates as suggested by GIXS.

Fig. 10. Luminescence CLSM images of (a) the neat F(MB)5 film and (b) the neat N200 film, either without (panel (i)) or with (panels (ii) and (iii)) rubbing/annealing treatments. For the aligned thin-film samples (i.e., with rubbing/annealing), polarization-dependent CLSM images were taken by aligning the polarization of the excitation laser either perpendicular to (\, panel (ii)) or parallel with (//, panel (iii)) the rubbing direction.

Fig. 11. Luminescence CLSM images of (a) the F(MB)5 emission and (b) the N200 MMLCT emission taken from a same F(MB)5:N200 mixture film, either without (panel (i)) or with (panels (ii) and (iii)) rubbing/annealing treatments. For the aligned thin-film samples (i.e., with rubbing/annealing), polarization-dependent CLSM images were taken by aligning the polarization of the excitation laser either perpendicular to (\, panel (ii)) or parallel with (//, panel (iii)) the rubbing direction.

3.3. CLSM

320

Y.-T. Tsai et al. / Organic Electronics 15 (2014) 311–321

(a) Neat N200, no treatments

(b) Neat N200, with treatments

(c) F(MB)5:N200, no treatments

(d) F(MB)5:N200, with treatments

Fig. 12. Schematic representations of the morphologies and molecular alignments of: (a) N200 neat film without rubbing/annealing treatments, (b) N200 neat film with rubbing/annealing treatments, (c) F(MB)5:N200 mixture film without rubbing/annealing treatments, and (d) F(MB)5:N200 mixture film with rubbing/annealing treatments.

4. Conclusions The GIXS and CLSM analyses were conducted to furnish new insight into molecular self-assembly, orientation and the resultant nanostructures in four distinct cases: (a) N200 neat film without rubbing/annealing treatments, (b) N200 neat film with rubbing/annealing treatments, (c) F(MB)5:N200 mixture film without rubbing/annealing treatments, and (d) F(MB)5:N200 mixture film with rubbing/annealing treatments. These combined experimental observations led to the film morphologies as depicted in Fig. 12. In the as-cast neat N200 film (Fig. 12(a)), GIXS patterns suggest aggregates of N200 have readily formed, although their sizes are relatively smaller and orientations are somewhat random. Rubbing and annealing induce growth and better organized orientation of the N200 aggregates (Fig. 12(b)). Examining the GIXS patterns of rubbed/annealed N200 neat films reveals that the assembly of the Pt complexes has the tetragonal columnar structure (Colt). With the rubbing and annealing treatments, the columnar stacks of the Pt(II) complexes appear largely oriented perpendicular to the rubbing direction with the in-plane inter-column packing (i.e., [0 1 0] orientation) preferentially along the rubbing direction. Such results are consistent with those previously deduced from polarized optical spectroscopy of N200 films. For the F(MB)5:N200 mixture film without

rubbing and annealing treatments (Fig. 12(c)), although previous photophysical studies indicate Pt–Pt stacking and MMCLT emission of N200, no specific signals associated with N200 assembly can be detected in its GIXS patterns. It suggests the presence of the (un-oriented) F(MB)5 host significantly disturbs the assembly of the Pt(II) complex, so that the sizes and/or quantities of the N200 assemblies are probably not sufficient and the orientations are probably too random for GIXS to detect. For the mixture film subjected to alignment (Fig. 12(d)), again the signals associated with assemblies of N200 (similar to those observed for rubbed/annealed neat N200 films) appear, indicating the growth of N200 aggregates and adoption of more organized inter-column structures and orientation within the aligned F(MB)5 host. The organization and alignment are presumably induced by the oriented F(MB)5 host oligomers on the rubbed PEDOT:PSS alignment layer. Overall, the GIXS pattern of the aligned F(MB)5:N200 mixture film is similar to that of the rubbed/annealed neat N200 film. It indicates that N200 assemblies in aligned mixture films have Pt–Pt stacking, inter-column structure (i.e., tetragonal columnar structure, Colt), and orientation/alignment similar to those in aligned neat N200 films, although the Scherrer grain-size analysis reveals that the aggregate sizes in the mixture films are smaller than those in N200 neat films.

Y.-T. Tsai et al. / Organic Electronics 15 (2014) 311–321

Acknowledgements The authors gratefully acknowledge the financial support from National Science Council and Ministry of Education of Taiwan (under the Grant NSC 98-2120-M-002-005, NSC 99-2221-E-002-118-MY3, NSC 102-2221-E-002-203MY3, and NTU-CESRP-102R7607-2). References [1] J.Y. Son, B. Javidi, IEEE J. Disp. Technol. 1 (2005) 125. [2] S. Jung, J.H. Park, H. Choi, B. Lee, Appl. Opt. 42 (2003) 2513. [3] R.H. Friend, R.W. Gymer, A.B. Holmes, J.H. Burroughes, R.N. Marks, C. Taliani, D.D.C. Bradley, D.A. Dos Santos, J.L. Bredas, M. Logdlund, W.R. Salaneck, Nature 397 (1999) 121. [4] S. Pastoor, M. Wopking, Displays 17 (1997) 100. [5] J.H. Park, S. Jung, H. Choi, B. Lee, Opt. Express 11 (2003) 1862. [6] M.C. Quiles, P.G. Etchegoin, D.D.C. Bradley, Phys. Rev. B 72 (2005) 045209. [7] M. Grell, D.D.C. Bradley, Adv. Mater. 11 (1999) 895. [8] Y.H. Geng, S.W. Culligan, A. Trajkovska, J.U. Wallace, S.H. Chen, Chem. Mater. 15 (2003) 542. [9] S.W. Culligan, Y.H. Geng, S.H. Chen, K. Klubek, K.M. Vaeth, C.W. Tang, Adv. Mater. 15 (2003) 1176. [10] Y.H. Geng, A.C.A. Chen, J.J. Ou, S.H. Chen, K. Klubek, K.M. Vaeth, C.W. Tang, Chem. Mater. 15 (2003) 4352. [11] A.C.A. Chen, S.W. Culligan, Y.H. Geng, S.H. Chen, K.P. Klubek, K.M. Vaeth, C.W. Tang, Adv. Mater. 16 (2004) 783. [12] M. Jandke, D. Hanft, P. Strohriegl, K.S. Whitehead, M. Grell, D.D.C. Bradley, Proc. SPIE 4105 (2001) 338. [13] A. Bolognesi, G. Bajo, J. Paloheimo, T. Ostergard, H. Stubb, Adv. Mater. 9 (1997) 121. [14] V. Cimrova, M. Remmers, D. Neher, G. Wegner, Adv. Mater. 8 (1996) 146. [15] M. Hamaguchi, K. Yoshino, Polym. Adv. Technol. 8 (1997) 399. [16] M. Jandke, P. Strohriegl, J. Gmeiner, W. Brutting, M. Schwoerer, Adv. Mater. 11 (1999) 1518. [17] P. Dyreklev, M. Berggren, O. Inganas, M.R. Andersson, O. Wennerstrom, T. Hjertberg, Adv. Mater. 7 (1995) 43. [18] M. Grell, W. Knoll, D. Lupo, A. Meisel, T. Miteva, D. Neher, H.G. Nothofer, U. Scherf, A. Yasuda, Adv. Mater. 11 (1999) 671. [19] T. Yasuda, H. Ooi, J. Morita, Y. Akama, K. Minoura, M. Funahashi, T. Shimomuro, T. Kato, Adv. Funct. Mater. 19 (2009) 411. [20] A. Bolognesi, F. Bertini, G. Bajo, A. Provasoli, D. Villa, O. Ahumada, Thin Solid Films 289 (1996) 129. [21] H. Thiem, M. Jandke, D. Hanft, P. Strohriegl, Macromol. Chem. Phys. 207 (2006) 370. [22] A.W. Grice, D.D.C. Bradley, M.T. Bernius, M. Inbasekaran, W.W. Wu, E.P. Woo, Appl. Phys. Lett. 73 (1998) 629. [23] M.P. Aldred, A.E.A. Contoret, S.R. Farrar, S.M. Kelly, D. Mathieson, M. O’Neill, W.C. Tsoi, P. Vlachos, Adv. Mater. 17 (2005) 1368. [24] M.P. Aldred, A.J. Eastwood, S.M. Kelly, P. Vlachos, A.E.A. Contoret, S.R. Farrar, B. Mansoor, M. O’Neill, W.C. Tsoi, Chem. Mater. 16 (2004) 4928.

321

[25] M.A. Baldo, D.F. O’Brien, Y. You, A. Shoustikov, S. Sibley, M.E. Thompson, S.R. Forrest, Nature 395 (1998) 151. [26] S. Reineke, F. Lindner, G. Schwartz, N. Seidler, K. Walzer, B. Luessem, K. Leo, Nature 459 (2009) 234. [27] C. Adachi, M.A. Baldo, M.E. Thompson, S.R. Forrest, J. Appl. Phys. 90 (2001) 5048. [28] Y. Kawamura, K. Goushi, J. Brooks, J.J. Brown, H. Sasabe, C. Adachi, Appl. Phys. Lett. 86 (2005). [29] S.H. Liu, M.S. Lin, L.Y. Chen, Y.H. Hong, C.H. Tsai, C.C. Wu, A. Poloek, Y. Chi, C.A. Chen, S.H. Chen, Org. Electron. 12 (2011) 15. [30] C.T. Liao, H.H. Chen, H.F. Hsu, A. Poloek, H.H. Yeh, Y. Chi, K.W. Wang, C.H. Lai, G.H. Lee, C.W. Shih, P.T. Chou, Chem. Eur. J. 17 (2011) 546. [31] J. Kavitha, S.Y. Chang, Y. Chi, J.K. Yu, Y.H. Hu, P.T. Chou, S.M. Peng, G.H. Lee, Y.T. Tao, C.H. Chien, A.J. Carty, Adv. Func. Mater. 15 (2005) 223. [32] S.Y. Chang, J. Kavitha, S.W. Li, C.S. Hsu, Y. Chi, Y.S. Yeh, P.T. Chou, G.H. Lee, A.J. Carty, Y.T. Tao, C.H. Chien, Inorg. Chem. 45 (2006) 137. [33] M.H. Qi, G.F. Liu, J. Phys. Chem. B 107 (2003) 7640. [34] E. Terazzi, S. Suarez, S. Torelli, H. Nozary, D. Imbert, O. Mamula, J.P. Rivera, E. Guillet, J.M. Benech, G. Bernardinelli, R. Scopelliti, B. Donnio, D. Guillon, J.C.G. Bunzli, C. Piguet, Adv. Funct. Mater. 16 (2006) 157. [35] K. Binnemans, J. Mater. Chem. 19 (2009) 448. [36] V.N. Kozhevnikov, B. Donnio, D.W. Bruce, Angew. Chem. Int. Ed. 47 (2008) 6286. [37] R.M. Beal, A. Stavrinadis, J.H. Warner, J.M. Smith, H.E. Assender, A.A.R. Watt, Macromolecules 43 (2010) 2343. [38] R. Giridharagopal, D.S. Ginger, J. Phys. Chem. Lett. 1 (2010) 1160. [39] M. Reyes-Reyes, K. Kim, J. Dewald, R. López-Sandoval, A. Avadhanula, S. Curran, D.L. Carroll, Org. Lett. 7 (2005) 5749. [40] M. Sanyal, B. Schmidt-Hansberg, M.F.G. Klein, A. Colsmann, C. Munuera, A. Vorobiev, U. Lemmer, W. Schabel, H. Dosch, E. Barrena, Adv. Energy Mater. 1 (2011) 363. [41] T. Agostinelli, S. Lilliu, J.G. Labram, M. Campoy-Quiles, M. Hampton, E. Pires, J. Rawle, O. Bikondoa, D.D.C. Bradley, T.D. Anthopoulos, Adv. Funct. Mater. 21 (2011) 1701. [42] E. Verploegen, R. Mondal, C.J. Bettinger, S. Sok, M.F. Toney, Z. Bao, Adv. Funct. Mater. 20 (2010) 3519. [43] M. Sanyal, B. Schmidt-Hansberg, M.F.G. Klein, C. Munuera, A. Vorobiev, A. Colsmann, P. Scharfer, U. Lemmer, W. Schabel, H. Dosch, Macromolecular 44 (2011) 3795. [44] W.H. Baek, T.S. Yoon, H.H. Lee, Y.S. Kim, Org. Electron. 11 (2010) 933. [45] N.S. Claxton, T.J. Fellers, M.W. Davidson, Olympus, 2006. [46] H. Hasegawa, High Polym. Jpn. 55 (2006) 961. [47] J.G.E. Jellison, Thin Solid Films 313–314 (1998) 33. [48] H.W. Lin, C.L. Lin, C.C. Wu, T.C. Chao, K.T. Wong, Org. Electron. 8 (2007) 189. [49] M. Birkholz, Thin Film Analysis by X-ray Scattering, Wiley.com, 2006. [50] D.M. Smilgies, J. Appl. Crystallogr. 42 (2009) 1030. [51] S.C.F. Kui, S.S.Y. Chui, C.M. Che, N. Zhu, J. Am. Chem. Soc. 128 (2006) 8297. [52] S.C.F. Kui, I.H.T. Sham, C.C.C. Cheung, C.W. Ma, B. Yan, N. Zhu, C.M. Che, W.F. Fu, Chem. Eur. J. 13 (2007) 417. [53] D.W. Breiby, F. Hansteen, W. Pisula, O. Bunk, U. Kolb, J.W. Andreasen, K. Müllen, M.M. Nielsen, J. Phys. Chem. 109 (2005) 22319.