Polymeric Planar Microcavities Doped with a Europium Complex

: Organo-metallic europium complex tetrakis (dibenzoyl methide) triethylammonium (EuD 4 TEA) shows a sharp emission spectrum, which makes it interesting for photonic applications. In this work, we embedded it into all-polymeric planar microcavities and investigated the e ﬀ ect of the photonic environment on its emission spectrum. To this end, submicron-sized EuD 4 TEA crystals were loaded into a blend of polystyrene and carboxylic terminated polystyrene matrix, which served to stabilize the emitter in the polymer and to make the composite processable. The new composite was then casted by spin-coating as a defect layer in a polymeric planar microcavity. Spectroscopic studies demonstrate that ﬁne spectral tuning of the cavity mode on the sharp organometal luminescence is possible and produces spectral redistribution of the ﬂuorophore emission, along with a remarkable cavity quality factor.


Introduction
Organometallic transition metal compounds are particularly attractive for the development of optical devices owing to their large quantum yield and chemical stability [1][2][3]. Among several of these compounds, europium tetrakis (dibenzoyl methide) triethylammonium (EuD 4 TEA) shows one of the most efficient emission processes [4][5][6]. EuD 4 TEA is indeed an organometallic triboluminescent [7] crystal showing an intense and sharp red emission, which is allowed by the europium complexation [8,9]. Indeed, its structure consists of four coordinated electron donating groups (1,3-diphenyl-1,3-propanedione), which function as light absorbers; one is triethylamine, which serves to stabilize the complex, and the complexed emissive metal [8,9] (Figure 1). As mentioned above, the material shows an intense and sharp near-band-edge exciton emission at 614 nm, where the triboluminescence also occurs [3,10,11]. In spite of its outstanding emissive properties, EuD 4 TEA applications are currently limited to the sensing of mechanical stimulations using the pristine material, or embedding the material into polymer matrices such as poly methyl methacrylate [10]. These applications are allowed by the triboluminescence process, which is the emission of photons resulting from chemical bond breakage within crystalline particles after the application of a mechanical force [12,13]. On the other hand, the sharp emission of EuD 4 TEA together with the crystal robustness makes this compound highly attractive for integration in plastic photonic crystals, a viable approach to produce devices with engineered fluorescence. Since their first description in the 1980s [14,15], inorganic photonic crystals have been largely employed to enhance intensity and to obtain directional control of photoluminescence, as well as for lasing, switching, and sensing [16][17][18][19][20][21][22]. Only in the last decades, when polymers have entered the spotlight for photovoltaics and emitting devices [23][24][25], have these materials also become interesting for photonic applications [26][27][28][29][30][31][32]. Among polymer photonic structures, monodimensional multilayered photonic crystals, namely distributed Bragg reflectors (DBRs) and planar microcavities, are the simplest from a fabrication and modelling point of view [33]. Moreover, they are already fabricated industrially via melt processing [34,35]. In addition, these systems are highly versatile. Indeed, all-polymer structures have already been demonstrated in coupling with several kinds of emitters including, besides conjugated molecules and polymers [36][37][38], inorganic nanocrystals [39][40][41][42], J-aggregates [43] and perovskites [44,45].
In this work, as a proof of principle, we demonstrate polymer planar microcavities doped with EuD4TEA embedded into a blend of polystyrene (PS) and PS carboxi-terminated (PSCT). While the pristine PS serves to increase the viscosity of the starting solution to enhance its processability, PSCT serves to stabilize EuD4TEA through the -OH group of the carboxylic PS termination [10]. We study the EuD4TEA photoluminescence directional and spectral redistribution. Emission lifetime is also discussed to assess radiative enhancement effects.

Materials and Methods
The synthesis of EuD4TEA was carried out by the hot injection method [10]. An aliquot (50 mL) of the solution of 1,3-diphenyl-1,3-dipronedione (Sigma-Aldrich; 98%) in ethanol with a concentration of 58.2 mg/mL was heated until reflux. Subsequently, the preheated solution of Eu(NO3)3 (52.8 mg/mL) was injected into a vial under vigorous stirring. The 1.42 g (14 mmol) of triethylamine (Sigma-Aldrich, ≥99.5) was added to the boiling mixture and the reaction was maintained for 1 h. For the crystallization, the solution was first allowed to slowly cool down to room temperature. Afterwards, it was further crystallized at 4 ℃ for 24 h. The crystals were filtrated and washed with cold ethanol, thereafter dried for 24 h in a vacuum chamber at 40 ℃. Figure 1 displays the structure of EuD4TEA crystals. The colorless crystals of EuD4TEA were composed of an anionic europium complex and a protonated triethylammonium molecule. EuD4TEA crystallizes in the monoclinic crystal system with non-centrosymmetric space group P21. The structure of EuD4TEA has europium (III) with eight oxygen atoms coordinated to the 1,3-diphenyl-1,3-propanedianato ligands.
In this work, as a proof of principle, we demonstrate polymer planar microcavities doped with EuD 4 TEA embedded into a blend of polystyrene (PS) and PS carboxi-terminated (PSCT). While the pristine PS serves to increase the viscosity of the starting solution to enhance its processability, PSCT serves to stabilize EuD 4 TEA through the -OH group of the carboxylic PS termination [10]. We study the EuD 4 TEA photoluminescence directional and spectral redistribution. Emission lifetime is also discussed to assess radiative enhancement effects.

Materials and Methods
The synthesis of EuD 4 TEA was carried out by the hot injection method [10]. An aliquot (50 mL) of the solution of 1,3-diphenyl-1,3-dipronedione (Sigma-Aldrich; 98%) in ethanol with a concentration of 58.2 mg/mL was heated until reflux. Subsequently, the preheated solution of Eu(NO 3 ) 3 (52.8 mg/mL) was injected into a vial under vigorous stirring. The 1.42 g (14 mmol) of triethylamine (Sigma-Aldrich, ≥99.5) was added to the boiling mixture and the reaction was maintained for 1 h. For the crystallization, the solution was first allowed to slowly cool down to room temperature. Afterwards, it was further crystallized at 4°C for 24 h. The crystals were filtrated and washed with cold ethanol, thereafter dried for 24 h in a vacuum chamber at 40°C. Figure 1 displays the structure of EuD 4 TEA crystals. The colorless crystals of EuD 4 TEA were composed of an anionic europium complex and a protonated triethylammonium molecule. EuD 4 TEA crystallizes in the monoclinic crystal system with non-centrosymmetric space group P2 1 . The structure of EuD 4 TEA has europium (III) with eight oxygen atoms coordinated to the 1,3-diphenyl-1,3-propanedianato ligands.
The polymer microcavities were fabricated on glass substrates by alternate spin-coating deposition of a toluene solution (35 mg/mL) of poly (N-vinylcarbazole) (PVK; Across Organic; M W = 136,600 g/mol; refractive index, n = 1.68) [43,46,47] and polyacrylic acid (PAA; Sigma-Aldrich; M W = 1800 g/mol; n = 1.44) [43] solubilized in 2-methyl-2-pentanol (28 mg/mL). The rotation speed was kept at 66 rounds per second. No thermal annealing was performed during the fabrication process. The microcavity layer was placed on 25 bilayers of the two polymers by spin-coating of EuD 4 TEA dissolved in a PS (20 mg/mL):PSCT (10 mg/mL) solution in toluene (PS, Sigma Aldrich, MW = 92 kg/mol; PSCT, Sigma-Aldrich M W = 200 kg/mol). The cavity layer was casted at 35 RPS. After the deposition of the photoactive material, the microcavity was completed with the deposition of the other 25 bilayers of PAA and PVK. The reference sample was instead fabricated under the same condition casting a layer of EuD 4 TEA:PS on a thin film of PAA.
Reflectance, transmittance and emission spectra were collected using setups based on optical fibers connected to a Charge-Coupled Device spectrometer (AvaSpec-2048, 200−1150 nm with resolution 1.4 nm from Avantes) Apeldoorn, Netherlands. Reflectance data were collected using an aluminum mirror as a reference. The light source employed for reflectance and transmittance measurements was a deuterium−halogen Micropak DH2000BAL (Ocean Insight, Largo, FL, USA), while emission spectra were collected by exciting the sample with a continuous wave laser, Oxxius 405 nm, with power 50 mW. Reflectance and angular resolved transmittance data were modelled by employing a transfer matrix method formalism as described in [33]. The layer thicknesses were employed as fitting parameter for the reflectance data, while the optical functions of PVK, PAA and PS were previously measured and reported in other publications [43,48]. The retrieved geometrical thicknesses and the refractive index just mentioned were then employed to model the angular dispersion of the microcavity transmittance spectrum and the local density of photonic states (LPDOS) as described in [49].
Photoluminescence decay was recorded with an angle-resolved setup coupled with a PicoQuant Time-Correlated Single-Photon Counting system (Time Harp 260 PICO board with a temporal resolution of 25 ps, a PMA Hybrid 40 detector, a response time of 250 ps, and a 405 nm LDH-P-C-405 laser driven (PicoQuant, Berlin, Germany) with a PDL 800B driver with a 5-80 MHz repetition rate as the excitation source) equipped with a Solar Laser Systems monochromator.

Results and Discussion
Figure 2a reports a schematic and the reflectance spectrum of a microcavity containing the EuD 4 TEA:PS composite. The photonic structure consists of two DBRs made of 25 alternating bilayers of PAA and PVK sandwiching a thin film of the emissive composite. The reflectance spectrum is characterized by an intense peak centered at 625 nm that is assigned to the photonic band gap (PBG) of the microcavity with a reflectance value approaching 100% and full width half maximum (FWHM) of 70 nm (marked with # in Figure 2b). These values agree with previous reports for polymeric microcavities [33,43]. The peak is characterized by a minimum positioned at 632 nm, which corresponds to the cavity mode (marked with * in Figure 2b). The spectrum background is instead characterized by a rather complex Fabry-Perot patter, which arises mainly from the constructive and disruptive light interference arising from beams partially reflected from sample external interfaces (marked with • in Figure 2b). The presence of this fringe pattern, together with the large reflectance intensity and spectral width of the PBG, confirm the excellent optical quality of the structure. Indeed, these characteristics would not be detectable in the spectrum in the presence of scattering phenomena in the composite defect layer. The good quality of the microcavity is also testified by the good agreement between the experimental reflectance spectrum and the modelled data, reported as a black dotted line in Figure 2b (see experimental section). The calculation allowed us to retrieve geometrical thicknesses of 77 nm for PVK, 118 nm for PAA and 266 nm for the defect layer. Figure 2 also reports the calculated (panel b) and experimental (panel c) angular resolved transmittance spectra for the microcavity as contour plots. In these plots, the incidence angle for light is reported on the y-scale, while the x-scale shows the spectral wavelength. The intensity is instead reported in false colors so that larger transmittances appear in blue and green tones, while lower values appear in red tones. Note that the y-scales for the two plots are reported symmetrically. We note that the two plots are in good agreement. For both, at normal incidence (0 • ) the PBG and the cavity mode (marked respectively with # and * in Figure 2d) are observed at the same spectral wavelengths detected in the reflectance data of Figure 2b. Increasing the collection angle, the spectral position of the two features moves to the short wavelength side of the spectrum, in agreement with the band structure for planar photonic crystals [33].
Crystals 2019, 9, x FOR PEER REVIEW 4 of 10 reflectance data of Figure 2b. Increasing the collection angle, the spectral position of the two features moves to the short wavelength side of the spectrum, in agreement with the band structure for planar photonic crystals [33]. The effect of the photonic crystals on the EuD4TEA emission spectrum was investigated trough measurements of steady-state photoluminescence exciting the emitter with a continuous wave laser source at 405 nm. Figure 3 compares the normalized emission of the bare EuD4TEA:PS composite casted on a PAA thin film on a glass substrate with the LPDOS and the microcavity emission (see experimental section). As mentioned in the introduction, the emission of the europium complex is sharp and centered in the red part of the visible spectrum. It is characterized by an intense peak at 615 nm with FWHM of 6 nm (Figure 3a). Other structures with much lower intensities are also detected at 582 nm, 595 nm and 703 nm. The calculated LPDOS of the microcavity shows two maxima positioned at 605 nm and 645 nm, which are assigned to the short wavelength edge of the PBG and to the cavity mode, respectively (Figure 3b). At the PBG frequencies (610-630 nm), the value of LPDOS is low, but, in agreement with the relatively low dielectric contrast typical of polymer multilayered structures, the density is larger than zero. [50][51][52][53] Then, suppression of the europium complex emission intensity is expected in this spectral region. In turn, the emission should be redistributed in the spectral regions where the LPDOS is larger. When observing the emission spectrum of the microcavity collected at normal incidence, it is clear that the photonic environment has a strong effect on the emission linewidth of the EuD4TEA (Figure 3c). In agreement with predictions, the main emission peak of the europium complex is suppressed and redistributed at 595 nm and 645 nm. These wavelengths correspond to the PBG edge and to the cavity mode, respectively (compare with panel b). At this angle, the cavity mode is indeed detuned with respect to the main EuD4TEA emission (compare with Figure 2b). The effect of the photonic crystals on the EuD 4 TEA emission spectrum was investigated trough measurements of steady-state photoluminescence exciting the emitter with a continuous wave laser source at 405 nm. Figure 3 compares the normalized emission of the bare EuD 4 TEA:PS composite casted on a PAA thin film on a glass substrate with the LPDOS and the microcavity emission (see experimental section). As mentioned in the introduction, the emission of the europium complex is sharp and centered in the red part of the visible spectrum. It is characterized by an intense peak at 615 nm with FWHM of 6 nm (Figure 3a). Other structures with much lower intensities are also detected at 582 nm, 595 nm and 703 nm. The calculated LPDOS of the microcavity shows two maxima positioned at 605 nm and 645 nm, which are assigned to the short wavelength edge of the PBG and to the cavity mode, respectively (Figure 3b). At the PBG frequencies (610-630 nm), the value of LPDOS is low, but, in agreement with the relatively low dielectric contrast typical of polymer multilayered structures, the density is larger than zero [50][51][52][53]. Then, suppression of the europium complex emission intensity is expected in this spectral region. In turn, the emission should be redistributed in the spectral regions where the LPDOS is larger. When observing the emission spectrum of the microcavity collected at normal incidence, it is clear that the photonic environment has a strong effect on the emission linewidth of the EuD 4 TEA (Figure 3c). In agreement with predictions, the main emission peak of the europium complex is suppressed and redistributed at 595 nm and 645 nm. These wavelengths correspond to the PBG edge and to the cavity mode, respectively (compare with panel b). At this angle, the cavity mode is indeed detuned with respect to the main EuD 4 TEA emission (compare with Figure 2b).  The behavior of the sample at normal incidence collection was described in Figure 3a and is reported here for comparison. Because the PBG spectral position is angularly dependent (photon momentum, see Figure 2c and d) at collection angles larger than 0°, the cavity mode is detected at shorter wavelengths with respect to lower collection angles. In agreement, at 28°, the europium emission is spectrally superimposed to the cavity mode ( Figure 4b). In this case, the photoluminescence of the microcavity results centered at 613 nm, having a FWHM of 3.7 nm, which correspond to a twofold reduction and to a quality factor λ0/Δλ = 166, which agrees with previous reports for similar polymer systems [33]. For larger angles, the PBG is again detuned (Figure 4c) and the emission spectra of the reference and of the microcavity are very similar. The overall behavior is reported in Figure 4d, where the emission spectra of the cavity are stacked. At 0° (first spectrum from the bottom), we clearly notice three main features positioned at 595, 613 and 645 nm marked in the same order by a red dashed line (595 nm at 0°), a gray dotted line (613 nm at 0°) and a black dashed line (645 nm at 0°). As described above, the features detected at 595 nm correspond to the short wavelength PBG edge, the one at 613 to the emission peak of the Eu complex, while the one at 645 corresponds to the cavity mode (compare with Figure 3). In agreement with the angular dispersion of the PBG reported in Figure 2c, the spectral position of the peak and of the cavity mode moves toward the shorter wavelengths, and so do the two optical features assigned to the PBG edge and to the cavity mode. Then, increasing the collection angle, the PBG shifts, and suppresses almost entirely the emission of the EuD4TEA in the range between 8° and 24°. At 24°, the cavity mode is spectrally overlapped to the main emission peak. Indeed, the lines tracking the cavity mode and the peak position cross in this region in Figure 4d. At this point, a sharp peak arises at 613 nm (black spectrum, 28°), and blueshift increases the collection angle up to 32°. At larger angles, the emission peak of the Eu complex results overlap the long wavelength edge of the PBG and it is not affected anymore by the photonic environment. Moreover, starting from 38°, the spectral position of the cavity mode (black dashed line) overlaps the weak emission signal initially positioned at 595 nm. These peak results enhance in intensity with respect to a shorter collection angle. Increasing the angle, this signal is angularly redistributed. Indeed, it moves  The behavior of the sample at normal incidence collection was described in Figure 3a and is reported here for comparison. Because the PBG spectral position is angularly dependent (photon momentum, see Figure 2c,d) at collection angles larger than 0 • , the cavity mode is detected at shorter wavelengths with respect to lower collection angles. In agreement, at 28 • , the europium emission is spectrally superimposed to the cavity mode (Figure 4b). In this case, the photoluminescence of the microcavity results centered at 613 nm, having a FWHM of 3.7 nm, which correspond to a twofold reduction and to a quality factor λ 0 /∆λ = 166, which agrees with previous reports for similar polymer systems [33]. For larger angles, the PBG is again detuned (Figure 4c) and the emission spectra of the reference and of the microcavity are very similar. The overall behavior is reported in Figure 4d, where the emission spectra of the cavity are stacked. At 0 • (first spectrum from the bottom), we clearly notice three main features positioned at 595, 613 and 645 nm marked in the same order by a red dashed line (595 nm at 0 • ), a gray dotted line (613 nm at 0 • ) and a black dashed line (645 nm at 0 • ). As described above, the features detected at 595 nm correspond to the short wavelength PBG edge, the one at 613 to the emission peak of the Eu complex, while the one at 645 corresponds to the cavity mode (compare with Figure 3). In agreement with the angular dispersion of the PBG reported in Figure 2c, the spectral position of the peak and of the cavity mode moves toward the shorter wavelengths, and so do the two optical features assigned to the PBG edge and to the cavity mode. Then, increasing the collection angle, the PBG shifts, and suppresses almost entirely the emission of the EuD 4 TEA in the range between 8 • and 24 • . At 24 • , the cavity mode is spectrally overlapped to the main emission peak. Indeed, the lines tracking the cavity mode and the peak position cross in this region in Figure 4d. At this point, a sharp peak arises at 613 nm (black spectrum, 28 • ), and blueshift increases the collection angle up to 32 • . At larger angles, the emission peak of the Eu complex results overlap the long wavelength edge of the PBG and it is not affected anymore by the photonic environment. Moreover, starting from 38 • , the spectral position of the cavity mode (black dashed line) overlaps the weak emission signal initially positioned at 595 nm. These peak results enhance in intensity with respect to a shorter collection angle. Increasing the angle, this signal is angularly redistributed. Indeed, it moves from about 605 nm in the spectrum collected at 32 • to 581 nm in the spectrum collected at 44 • . Then, at larger angles, the effect of the PBG on the emission fades and the emission profile of the microcavity results is similar to the one of the reference sample (see Figure 4c). To estimate the effect of the microcavity on the overall emission intensity, Figure 4e shows the values of the integrated intensities in the spectral range between 450 nm and 850 nm for collection angles ranging from 0 • to 60 • . The data show that in the entire angular range, the overall microcavity intensity is one order of magnitude larger than for the reference sample. This data confirms that the microcavity induces an overall intensity enhancement. On the other hand, it is not clear whether this effect arises from the modified LPDOS or from the better light extraction operated by the multilayered structure.
Crystals 2019, 9, x FOR PEER REVIEW 6 of 10 from about 605 nm in the spectrum collected at 32° to 581 nm in the spectrum collected at 44°. Then, at larger angles, the effect of the PBG on the emission fades and the emission profile of the microcavity results is similar to the one of the reference sample (see Figure 4c). To estimate the effect of the microcavity on the overall emission intensity, Figure 4e shows the values of the integrated intensities in the spectral range between 450 nm and 850 nm for collection angles ranging from 0° to 60°. The data show that in the entire angular range, the overall microcavity intensity is one order of magnitude larger than for the reference sample. This data confirms that the microcavity induces an overall intensity enhancement. On the other hand, it is not clear whether this effect arises from the modified LPDOS or from the better light extraction operated by the multilayered structure. The larger emission intensity from the microcavity suggests that radiative rate enhancement of EuD4TEA might take place in the photonic structure. To evaluate the effects, we also registered its emission decay at a collection angle of 28°, where the cavity mode superimposes the main fluorescence peak of the complex. Indeed, theory foresees faster decay rates for radiative rate The larger emission intensity from the microcavity suggests that radiative rate enhancement of EuD 4 TEA might take place in the photonic structure. To evaluate the effects, we also registered its emission decay at a collection angle of 28 • , where the cavity mode superimposes the main fluorescence peak of the complex. Indeed, theory foresees faster decay rates for radiative rate enhancement. Then, when the emission occurs within the cavity mode, where the density of photonic states is larger, the emission should be faster than for the reference, where the density of photonic states in unmodified [43,52,54,55]. Figure 5 shows the emission decays for the two samples collected at 615 nm (compare with Figure 4b). The experimental data (red dots for the microcavity and black dots for the reference sample) were fitted with a biexponential decay. The lifetimes retrieved were 1.2 ns and 8.9 nm for the reference sample, while for the microcavity sample we obtained 1.4 ns and 10.6 ns, respectively. While the difference between the two shorter lifetimes is not significant, as it is within the instrumental resolution, the difference observed for the two longer values is not negligible. The increase in emission lifetime is counterintuitive in a microcavity [43,[56][57][58]. On the other hand, it has been demonstrated that the spontaneous lifetime can be increased by more than a factor of 10 if the emitter is placed in the antinode of the microcavity standing wave or in structures without guided modes [54,55]. Then, even though it is not possible to claim radiative rate enhancement without a knowledge of both radiative and non-radiative rates, this is the first report of such a variation in emission lifetime operated by polymer planar photonic crystals.
Crystals 2019, 9, x FOR PEER REVIEW 7 of 10 enhancement. Then, when the emission occurs within the cavity mode, where the density of photonic states is larger, the emission should be faster than for the reference, where the density of photonic states in unmodified [43,52,54,55]. Figure 5 shows the emission decays for the two samples collected at 615 nm (compare with Figure 4b). The experimental data (red dots for the microcavity and black dots for the reference sample) were fitted with a biexponential decay. The lifetimes retrieved were 1.2 ns and 8.9 nm for the reference sample, while for the microcavity sample we obtained 1.4 ns and 10.6 ns, respectively. While the difference between the two shorter lifetimes is not significant, as it is within the instrumental resolution, the difference observed for the two longer values is not negligible. The increase in emission lifetime is counterintuitive in a microcavity [43,[56][57][58]. On the other hand, it has been demonstrated that the spontaneous lifetime can be increased by more than a factor of 10 if the emitter is placed in the antinode of the microcavity standing wave or in structures without guided modes [54,55]. Then, even though it is not possible to claim radiative rate enhancement without a knowledge of both radiative and non-radiative rates, this is the first report of such a variation in emission lifetime operated by polymer planar photonic crystals. Figure 5. Emission decay collected at 615 nm for the reference sample (black lines) and for the microcavity (red line). All the data were collected at 28° from the normal incidence to the sample plane.

Conclusions
We demonstrated that EuD4TEA can be efficiently implemented into polystyrene matrices and employed as an emitting medium in polymeric planar microcavities. A characterization of the emission intensity enhancement generated by polymer microcavities demonstrated spectral and directional redistribution of EuD4TEA emission assigned to the diverse density of photonic states in the microcavity. The photoluminescence decay of the microcavity suggested that the microcavity plays an active role in the radiative processes occurring in the microcavity. On the other hand, further studies are necessary to investigate possible radiative rate enhancement processes.

Conclusions
We demonstrated that EuD 4 TEA can be efficiently implemented into polystyrene matrices and employed as an emitting medium in polymeric planar microcavities. A characterization of the emission intensity enhancement generated by polymer microcavities demonstrated spectral and directional redistribution of EuD 4 TEA emission assigned to the diverse density of photonic states in the microcavity. The photoluminescence decay of the microcavity suggested that the microcavity plays an active role in the radiative processes occurring in the microcavity. On the other hand, further studies are necessary to investigate possible radiative rate enhancement processes.