We selected the 9,9-propyldiarylfluorene dimer (DDC3F) as the model building block to examine the role of crystallization in improving photostability (Figs. S1 and S2) The bulk-shaped single crystal was luckily obtained via the facile volatile process under the conditions of THF/ethanol mixed solvents (1:3 volume ratio) (Fig. S3). According to the single-crystal data, the DDC3F crystal exhibits the monoclinic space group of P 21/c, with corresponding lattice parameters of a = 9.034 (3) Å, b = 16.378 (5) Å, c =10.570 (4) Å, α = 90 °, β = 103.037 (7) °, γ = 90 ° (Supplementary Table S1). By further study the difluorenyl backbone exhibits anti-conformation, where the torsional angle between two fluorenyl moieties is approximate to 180°. The propyl chains are oriented on two sides of difluorene plane, which shows massive sites of Van der Waals force in symmetric distribution (Fig. S6). Based on above conformational feature, the intermolecular packing of DDC3F building blocks were performed in a herringbone manner with the angle of 58.57°. In such packing architecture, multiple C-H…π intermolecular interactions (in the distance of 2.895 Å) between two sterically adjacent molecules connecting the alkyl chains and the fluorene planes. In addition, compared with other pendant groups, the propyl chains also possess the role of steric hindrance to increase the distance (8.29 Å and 8.98 Å) between two adjacent difluorene plane (Fig. S7).30,35
Then, we investigated the surfactant assisted self-assembly of DDC3F into micro/nanocrystals to explore its crystalline features. Figs. 2A, 2B and 2D show the scanning electron microscopy (SEM) image, transmission electron microscopy (TEM) image and fluorescence microscopy of the self-assembled 1D nanorods with smooth surfaces and sharp edges. The length (L) of nanorods is about 20-40 µm and the width (W) are around 600-1200 nm, with the length: width ratio as ≈ 100: 3. And atomic force microscope (AFM) image reveals that the average thickness of nanorods is about 90 nm (Fig. 2E). Besides, XRD spectra of the nanorods show a series of sharp diffraction peaks corresponding to the {011} crystal plane (Fig. 2F). The observation of high order diffraction peaks of (011), (022) indicate the high crystallinity of these assembled nanorods that adopt a lamellar nanostructure with the (011) crystal plane adhered to the Si substrate. Meanwhile, the corresponding selected area electron diffraction (SAED) pattern were further used to corroborate the crystalline property of nanorod architectures, where the detected directions are parallel to [011 crystal direction and perpendicular to the flat nanorod surface of (011) single plane, respectively (Fig. 2C). Along the [100] crystal direction, the SAED spots confirm the single crystal state of nanorod architectures, where the red rounded sets of SAED spots are assigned to (100) and (0\(\stackrel{-}{1}\)1) crystal planes with d-spacing values of 8.80 Å and 8.72 Å, respectively. These SAED patterns are consistent with the XRD feature and are further reconfirmed by the simulation results (inset of Fig. 2F). In addition, the crystal planes (011) and (0 \(\stackrel{-}{1}\stackrel{-}{1}\)) on the top and bottom surfaces, the (100), (\(\stackrel{-}{1}\)00), (0\(\stackrel{-}{1}\)1) and (01\(\stackrel{-}{1}\)) crystal planes on the lateral surfaces, which are completely correspondence with theoretically predicted growth morphology (Fig. S8). On this basis, we investigated the crystalline effect on the photophysical properties and photo-stability of DDC3F. Fig. 3A shows the normalized steady-state absorption and PL spectra of DDC3F in different states. Two absorption band centered at 270 and 330 nm were observed on the dilute solution and the spin-coated amorphous film, the latter of which is assigned to the π-π* transition on difluorenyl moiety. For PL spectra, DDC3F in the dilute solution exhibits a clearly vibronic progression of 0-0, 0-1, 0-2 and 0-3 transitions at 373, 388, 408 and 438 nm, respectively. The DDC3F-based amorphous film exhibits a slightly red-shifted with the vibronic progression of 0-0, 0-1, 0-2 and 0-3 transitions at 375, 390, 410 and 440 nm, respectively, where the relative intensity of 0-0 emission is diminished. However, the DDC3F-based nanorods film only shows 0-1, 0-2 and 0-3 emission peaks at 392, 412 and 443 nm, respectively, according to the fitting of Lorentzian line shapes (Fig. S9), while the 0-0 emission is completely vanished. This deeply-blue emission feature is probably consistent with the highly ordered H-aggregate emission.36 Deeply, the photoluminescence quantum yield (PLQY) of DDC3F is 97% in dilute solution, 57% on amorphous film and 87% on crystalline nanorods film (Fig. 3B), respectively. These results reveal that such nanorod architectures possess the crystallization-induced luminescence enhancement (CIEE) effect as approximately the highest level of PLQY.37 Meanwhile, the corresponding photophysical data were shown in Table 1 and Figs. S10-12. According to the kr = ΦF /τFL equation, the calculated radiative deactivation rate (kr= 1.23× 109 S−1) of nanorods film is enhanced, while the non-radiative decay rate (knr=1.83× 108 S−1) is reduced compared to that of DDC3F in amorphous film. This is probably because of the inhabitation of quenching effect in crystal through the introduction of pendant propyl-alkyl chains.
The thermostability and photostability were examined for spin-coating amorphous film and drop-casting nanorods film. Initially, DDC3F powder displays excellent thermostability with a high thermal decomposition temperature of 320°C (Fig. S5). After thermal annealing at 100°C under ambient conditions for 10 min ~ 3 h, the PL spectra of both film states are maintained without the addition of green emission band, indicating the excellent spectra stability. It is noted that, probably owing to the presence of crystallization behaviors, the intensity of 388 nm (0-0 band) on amorphous film were reduced (Fig. S13). Further, phtotostability of films was investigated under UV irradiation (room temperature, RH: 70%, light: 80 W). In Fig. 3C, after UV exposure for 10 min, the amorphous film generates a new green emission band at 500-600 nm. Further exposed for 1~3 h, such green emission at 530 nm is gradually enhanced, with the green index Igreen/Iblue increasing from 1 to 4. These results indicate the presence of photooxidation process on amorphous film, which affords oxidative defects and results in the spectra instability (Top of Fig. 3C). In contrast, the nanorods film is capable of maintaining deep-blue emission (with Igreen/Iblue = 0), almost without the generation of green-emission band even under the same or longer UV irradiation (Fig. 3D). The excellent photostability was also confirmed by the patterning experiment through brushing process, as shown in Fig. S14. Under the same UV exposure, the crystalline nanorod region “CMSOD” can still maintain the blue-emission, while the amorphous background region shows the yellow-green luminescence.
To gain insight into the photostability mechanism of nanorods film, FT-IR spectra and transient spectra were used to detect the oxidative defects (Fig. 3E). We observed that the FT-IR spectrum of amorphous film (undergoing UV irradiation) generates an infrared absorption at 1720 cm−1, assigned to the stretching vibration of C=O on ketone moieties. Meanwhile, the lifetime at 530 nm emission was measured as ~ 5.4 ns, confirming the ketone emission with partly forbidden transition feature (Fig. 3F).38 In contrast, no C=O stretching vibration (at ~1720 cm−1) was observed on nanorod films after the same UV exposure, along with no emission behaviors with the lifetime of ~ 5 ns, both of which indicate the nonexistence of ketone defects. These results reveal that the ketone defects during the exudation process are majorly formed on the amorphous film, while the nanorod film has a high-efficient influence on suppressing the oxidation process for the formation of ketone defects (Fig. S15). Therefore, the closely molecular packing in crystalline nanorod architectures, analogous to molecular wall guard, allows the excellent ability to isolate oxygen (in air) and protect the 9-position of fluorenyl moieties from undergoing exudation. It is noted that the high photostability from such crystalline molecular wall strategy is significant to efficiency, reproducibility and color purity of photoelectronic devices.
In order to further the verify the strategy of crystalization-enhanced stability, the deep blue emission of DDC3F nanorod as Fabry-Pe´rot (FP) optical resonator was also tentatively examined through micro/nanocrystals nanosecond lasing. In Figs. 4A and S17, the lasing spectra of nanorod architecture with the length of 42.8 µm possesses the violet emission at 392 (0-1) and 412 (0-2) nm under the pump density of 187 to 234 mW/cm2. However, the shorter nanorod structure (~16 µm length) only shows a stimulation emission at 392 nm (Fig. 4B). According to the variation of output intensity as a function of pump density (Fig. 4C), the lasing threshold was evaluated as ~102 mW/cm2, and the maximum width at half intensity (FWHM) was calculated as FWHM = 0.14 nm, along with the quality factor Q of 2800 (Fig. S18A). Moreover, if increasing the nanorod length excessing ~ 70 µm, it is surprising that the simulation emission switches into the 412 nm (0-2 channel) (Fig. 4D), along with the lasing threshold of 216 mW/cm2 (Fig. 4E), FWHM = 0.46 nm and the quality factor Q = 896 (Fig. S18B). In contrast, the amorphous film only exhibited amplified self-emission (ASE) feature at 390 nm, with the pump energy threshold of ~ 99 mW/cm2 (Fig. S19). These results reveal that that such simulation emission switching is closely associated with the nanorod architectures of size-dependent from reabsorption effect,35,39 which provide the stable platform of efficiently propagating and manipulating miniaturized optics chip on subwavelength scale or microscale.
Finally, we tentatively examined the electroluminescent feature of self-assembled crystalline film based on the inverted OLED configuration of ITO/ZnO (10 nm)/Cs2CO3 (20 nm)/Crystals film (150 nm)
Table 1
the photophysical properties of DDC3F in dilute THF solution, amorphous film and nanorods film.
DDC3F
|
λabs a
|
λem b
|
ФF c
|
τFLd
|
kr e
|
Knr f
|
THF solution
|
279, 284, 330
|
373, 388, 408, 438
|
0.95
|
0.79
|
1.20× 109
|
6.33× 107
|
Amorphous film
|
300, 330
|
375, 390, 410,440
|
0.57
|
0.67
|
8.51× 108
|
6.42× 108
|
Nanorods film
|
330, 348, 366
|
392, 412, 443
|
0.87
|
0.71
|
1.23× 109
|
1.83× 108
|
a Absorption maximum of DDC3F in three different states (dilute THF solution: 10−5 mol/L, spin-coating amorphous film and nanorods film). b Emission maximum. c Fluorescence quantum yield (ФF) determined by a calibrated integrating sphere. d Fluorescence lifetime. e Radiative deactivation rate calculated according to the kr = ΦF /τFL equation. f Nonradiative deactivation rate calculated according to the knr = 1/τFL − kr equation. |
/NPB (40 nm)/ MoO3 (8nm)/LiF (1 nm) /Al (100 nm) in Figs. 5A and S20. As the nanorods film is too rough to exhibit the electroluminescence performance, we used thermal annealing method to afford the dendritic-like crystalline film (Figs. S21 and S22). The turn-on voltage (VT) was measured as 3.7 V, indicating efficient charge injection into the crystalline device. When the driving voltage is 7.8 V, the luminance of crystalline device is up to 15 cd /cm2 (Fig. 5B). Besides, the maximum current efficiency of device display is 0.013 cd/A at 4.5 V (Fig. 5C). Fig. 5D shows the corresponding EL spectra under different driving voltages and the emission peak is centered at 445 nm. It is worth mentioning that three layers were processed by the solution method based on crystalline OLED, which totally different from previous reported.40 Therefore, the crystalization strategy demonstrated it is a unique technology for electroluminescent properties, which can be exploited for the whole solution OLED displays.