Multinuclear Iridium Complex Encapsulated by Oligocarbazole Dendrons for Enhanced Nondoped Device Efficiency.

A dendritic multinuclear Ir complex, namely Cz-3IrB-IrG, has been designed and synthesized by introducing the second-generation oligocarbazole dendrons into its periphery. Because of the characteristic encapsulation, the intermolecular interactions could be effectively alleviated to prevent the unwanted triplet-triplet annihilation stemmed from the outer blue Ir complexes. Compared with 3IrB-IrG in the absence of dendrons, the film photoluminescence quantum yield of Cz-3IrB-IrG is greatly increased from 0.46 to 0.82 together with a small blue-shifted emission from 524 to 520 nm. On the basis of Cz-3IrB-IrG as the emitting layer alone, the nondoped device realizes a promising luminous efficiency of 40.9 cd/A (12.0%), much higher than that of 3IrB-IrG (32.6 cd/A, 9.7%). The obtained improvement clearly indicates that further dendronization toward multinuclear Ir complex will provide an alternative strategy to construct highly efficient phosphors used for nondoped phosphorescent organic light-emitting diodes.

Introduction

Nondoped (n class="Chemical">also known as host-free) phosphorescent organic light-emitting diodes (PhOLEDs) have simultaneously received much academic interest besides doped ones, where the emitting layer (EML) is composed of a single phosphor without any additional host.[1−3] In this case, the delicate and exact control of the ratio between different components is not required to constitute the EML any longer, beneficial for the device reliability and reproducibility.[4] Furthermore, the potential phase separation existing in doped devices could be avoided to improve the device stability during long-term operation.[5] Given these advantages, many efforts should be paid for the development of efficient phosphors capable of nondoped PhOLEDs.

As for the conventional phosphors, the severe triplet–triplet annihilation (n class="Chemical">TTA) in solid states is believed to mainly limit their applications in nondoped PhOLEDs.[6] Therefore, nowadays several molecular design rules have already been exploited to dampen the TTA effect and increase the nondoped device efficiency: (i) incorporating the electron-withdrawing substituents (e.g., −F and −CF3) or bulky groups (e.g., tert-butyl, phenyl, triphenylsilyl, fluorenyl, and carbazolyl) into the ligand to modulate intermolecular interactions;[7−10] (ii) encapsulating a phosphorescent core with functional dendrons to form a so-called self-host feature;[11,12] and (iii) utilizing specific emissive mechanisms including metal–metal-to-ligand charge transfer and excimer.[13−17] Among them, phosphorescent dendrimers seem to be more attractive because of easy functionality of core and peripheral dendrons independently as well as intrinsic solution processability. In fact, based on such phosphorescent dendrimers, highly efficient nondoped PhOLEDs are realized showing emission colors in the whole visible region.[18−22]

Recently, we have put forward an alternative strategy, namely multinucleation, toward nondoped phosphors.[23] The issued tetranuclear n class="Chemical">iridium (Ir) complex 3IrB–IrG possesses a characteristic self-host feature similar to phosphorescent dendrimers (Figure ), where the three outer blue Ir complexes can act as the host for the one inner green Ir core. In addition, the corresponding nondoped device achieves an appealing luminous efficiency as high as 32.6 cd/A. It is noteworthy that, in 3IrB–IrG, TTA still exists between the outer blue complexes although they can protect the inner green complex from interacting with each other. Thereby, an unwanted energy loss may occur during the outside-in energy transfer, leading to the relatively low photoluminescence quantum yield (PLQY) of 3IrB–IrG. To solve this problem, here, we report a dendritic multinuclear Ir complex Cz–3IrB–IrG by introducing the second-generation oligocarbazole dendrons into the surrounding of 3IrB–IrG. Benefiting from the encapsulation effect, the intermolecular interactions and thus TTA stemmed from the outer blue Ir complexes could be prohibited indeed, so that the PLQY is significantly enhanced from 0.46 of 3IrB–IrG to 0.82 of Cz–3IrB–IrG. As a consequence, the nondoped device with Cz–3IrB–IrG as the EML achieves an enhanced luminous efficiency of 40.9 cd/A (12.0%) compared with that of 3IrB–IrG (32.6 cd/A, 9.7%).

Figure 1

Molecular structures of multinuclear Ir complexes without and with oligocarbazole dendrons.

Results and Discussion

Synthesis and Characterization

The synthetic route for Cz–3IrB–IrG is presented in Scheme . Similar to our previous work,[23] a postn class="Chemical">dendronization method was adopted here to convergently construct the dendritic multinuclear Ir complex. Starting from the bromo-substituted oligocarbazole dendron (Cz–Br) and blue Ir complex containing three reactive hydroxyl groups (IrB–OH), a quantitative Williamson reaction was performed to afford the key intermediate (Cz–IrB–OH) by controlling the feed ratio. There was one remaining hydroxyl group in Cz–IrB–OH, which could be further brominated in the next step and finally contributed to the target complex Cz–3IrB–IrG through reacting with the three-hydroxyl-containing green Ir complex (IrG–OH). The molecular structure of Cz–3IrB–IrG was fully characterized by 1H NMR, 19F NMR, matrix-assisted laser desorption ionization–time-of-flight (MALDI–TOF), and element analysis. Also, it is well-soluble in common organic solvents, such as dichloromethane, tetrahydrofuran, ethyl acetate, toluene, and so on, ensuring the good film-forming capability via solution processing.

Scheme 1

Synthesis of the Dendritic Multinuclear Ir Complex Cz–3IrB–IrG

Reagents and conditions: (i) Cs2CO3, DMF, 70 °C and (ii) 1,4-dibromobutane, Cs2CO3, DMF, 70 °C.

Reagents and conditions: (i) Cs2CO3, n class="Chemical">DMF, 70 °C and (ii) 1,4-dibromobutane, Cs2CO3, DMF, 70 °C.

Photophysical Properties

Figure a depicts the absorption and photoluminescence (PL) spectra in solutions for Cz–n class="Chemical">3IrB–IrG with 3IrB–IrG as the control. Consistent with the literature,[23] the absorption bands below 325 nm are ascribed to the spin-allowed ligand-centered transitions, whereas the weak bands in the range of 325–500 nm are assigned to the metal-to-ligand and ligand-to-ligand charge-transfer transitions. Compared with the 3IrB–IrG, the sharp enhancement of the absorption coefficient below 325 nm for Cz–3IrB–IrG is reasonable from the incorporation of the peripheral oligocarbazole dendrons. Meanwhile, both Cz–3IrB–IrG and 3IrB–IrG exhibit almost the same emission mainly from the inner green Ir complex (λem = 512 nm) accompanied by a residue from the outer blue Ir complex (λem = 467 nm). This is due to the nonconjugated linkage between oligocarbazole dendrons and multinuclear Ir complexes, guaranteeing the independent electronic behaviors of each fragment. On going from solutions to films, the residual blue emission is completely quenched (Figure b), indicative of both intra- and intermolecular energy transfer. In addition, the emission maximum moves toward a shorter wavelength from 524 nm of 3IrB–IrG to 520 nm of Cz–3IrB–IrG. The observed hypsochromic shift can be explained when considering the reduced intermolecular interactions after the introduction of the surrounding oligocarbazole dendrons.

Figure 2

(a) UV/vis spectra in dichloromethane and PL spectra in toluene; (b) PL spectra in neat films for Cz–3IrB–IrG compared with 3IrB–IrG.

(a) UV/vis spectra in dichloromethane and PL spectra in n class="Chemical">toluene; (b) PL spectra in neat films for Cz–3IrB–IrG compared with 3IrB–IrG.

To investigate the shielding effect on TTA, the transient PL spectra in films were n class="Chemical">also measured by exciting the samples with a 355 nm light pulses and detecting the emission peaks of the outer blue and inner green complexes, respectively. As can be seen in Figure , the lifetime of the inner green Ir complex in Cz–3IrB–IrG is prolonged to 0.80 μs compared to that in 3IrB–IrG (0.35 μs), confirming the reduced intermolecular interactions and thus luminescence quenching. On the other hand, the outer blue Ir complexes in Cz–3IrB–IrG are found to decay much slower (τav = 0.13 μs) than that in 3IrB–IrG (0.05 μs). Because of the dendrons’ encapsulation, TTA between the outer blue Ir complexes is anticipated to be prevented effectively. Consequently, the energy loss induced by TTA could be avoided, which contributes to the enhancement of the inner core’s emission. For example, the PLQY of Cz–3IrB–IrG in thin film is determined to be 0.82, about 1.8-fold that of 3IrB–IrG (0.46). Finally, the PL spectra and PLQYs of the doping films of Cz–3IrB–IrG in the host H2[24] were also investigated. When the doping concentration is up from 5 to 100 wt % (Figure S4 and Table S1), a slight red shift of 6 nm is observed for the emission maximum together with the small variation of the corresponding PLQYs in the range of 0.82–0.92. The observation indicates again that the peripheral oligocarbazole dendrons could provide an effective encapsulation toward the multinuclear Ir complex (Table ).

Figure 3

Transient PL spectra for Cz–3IrB–IrG compared with 3IrB–IrG in thin films. The above plot is aimed to study the emissive behavior of the inner green Ir complex, whereas the below plot is drawn to analyze the emissive behavior of the outer blue Ir complex.

Table 1

Photophysical and Electrochemical Properties of Cz–3IrB–IrG Compared with 3IrB–IrG

	λabsa (nm)	λemb (nm)	λemc (nm)	ΦPc	τavc (μs)	HOMOd (eV)	LUMOd (eV)
3IrB–IrG	265, 299, 313, 351, 379, 414, 447, 490	467, 512	524	0.46	0.05, 0.35e	–4.91	–2.24
Cz–3IrB–IrG	270, 297, 313, 348, 377, 415, 446, 490	467, 512	520	0.82	0.13, 0.80f	–4.91	–2.24

Measured in CH2Cl2 at a concentration of 10–5 M.

Measured in toluene at a concentration of 10–5 M.

Measured in neat films under N2.

HOMO = −(Eox + 4.8) eV, LUMO = −(Ered + 4.8) eV, where Eox and Ered are the onset values from the first oxidation and reduction waves, respectively.

Detected at 484 and 524 nm, respectively.

Detected at 484 and 520 nm, respectively.

Transient PL spectra for Cz–3IrB–IrG compared with n class="Chemical">3IrB–IrG in thin films. The above plot is aimed to study the emissive behavior of the inner green Ir complex, whereas the below plot is drawn to analyze the emissive behavior of the outer blue Ir complex.

Electrochemical Properties

The electrochemical properties of n class="Chemical">3IrB–IrG and Cz–3IrB–IrG were studied by cyclic voltammetry (CV). As shown in Figure , three quasi-reversible oxidation processes are observed for Cz–3IrB–IrG during the anodic sweep. With respect to 3IrB–IrG, the first and second waves are from the inner and outer Ir complexes, respectively, whereas the third one is aroused by the oxidation of the oligocarbazole dendrons. However, during the cathodic sweep, similar reduction processes occur for both 3IrB–IrG and Cz–3IrB–IrG because the reduction of the oligocarbazole dendrons is unable to be detected within the testing range. The highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) energy levels of Cz–3IrB–IrG are estimated to be the same as those of 3IrB–IrG (−4.91 and −2.24 eV, respectively). This unravels again that the involvement of oligocarbazole dendrons via a nonconjugated linkage does not affect the electronic nature of multinuclear Ir complex.

Figure 4

CV curves for Cz–3IrB–IrG and 3IrB–IrG.

CV curves for Cz–3IrB–IrG and n class="Chemical">3IrB–IrG.

Electroluminescence Properties

To explore the potential of Cz–n class="Chemical">3IrB–IrG used as the EML alone, the nondoped device was fabricated with a configuration of ITO/PEDOT:PSS/EML/SPPO13/LiF/Al (Figure a), where SPPO13 (2,7-bis(diphenylphosphoryl)-9,9′-spirobifluorene) acted as the electron-transporting material.[25] For comparison, a control device based on 3IrB–IrG instead of Cz–3IrB–IrG was also prepared under the same conditions. Thanks to efficient outside-in energy transfer, they both show only bright green electroluminescence (EL) from the central core without any remnant emission from the outer blue Ir complexes or oligocarbazole dendrons (Figure b). In well-agreement with the film PL counterparts, the EL spectrum of Cz–3IrB–IrG is blue-shifted by about 4 nm relative to that of 3IrB–IrG. The corresponding Commission Internationale de L’Eclairage (CIE) coordinates are (0.38, 0.58) and (0.36, 0.59) for 3IrB–IrG and Cz–3IrB–IrG, respectively.

Figure 5

Nondoped device performance of Cz–3IrB–IrG compared with 3IrB–IrG: (a) device configuration; (b) EL spectra at a driving voltage of 6 V; (c) current density–voltage–luminance characteristics; and (d) luminous efficiency and EQE as a function of luminance.

Nondoped device performance of Cz–n class="Chemical">3IrB–IrG compared with 3IrB–IrG: (a) device configuration; (b) EL spectra at a driving voltage of 6 V; (c) current density–voltage–luminance characteristics; and (d) luminous efficiency and EQE as a function of luminance.

Figure c,d plots the current density–voltage–luminance characteristics together with luminous efficiency and external quantum efficiency (EQE) as a function of luminescence. As one can see, the current density as well as luminescence at the same driving voltage turns out to be decreased from n class="Chemical">3IrB–IrG to Cz–3IrB–IrG. In addition, the turn-on voltage at 1 cd/m2 is correspondingly up from 2.8 to 4.2 V. According to the literature,[26] a cascade charge transport from dendron wedge to blue Ir complex and then to green Ir core is within our expectation in Cz–3IrB–IrG. Therefore, the introduction of oligocarbazole dendrons with deeper energy levels than Ir complexes (as demonstrated in Figure ) would inevitably sacrifice the charge transport to some degree, resulting in the observed higher driving voltage and lower current density. Albeit this, because of the inhibited intermolecular interactions and TTA, the peak luminous efficiency and EQE are obviously elevated from 32.6 cd/A and 9.7% of 3IrB–IrG to 40.9 cd/A and 12.0% of Cz–3IrB–IrG (Table ). In comparison with the 3IrB–IrG, the EQE of Cz–3IrB–IrG is increased by only 25%, which is not consistent with the PLQY improvement (80%). We first investigated the film morphology of Cz–3IrB–IrG by using atomic force microscopy. The film is found to be quite uniform and homogenous with a very small root-mean-square value of 0.34 nm (Figure S5). Therefore, the influence of the film quality on the device performance could be reasonably excluded. Subsequently, we note that there exists a considerable hole injection barrier between poly(3,4-ethylenedioxythiophene):poly(styrenesulfonate) (PEDOT:PSS) and the EML caused by the dendritic wedge, which may be tentatively responsible for the mismatch between EQE and PLQY. To further optimize the nondoped device performance, hole injection should be facilitated by inserting an additional hole-transporting layer between PEDOT:PSS and the EML[27] or tuning the work function of PEDOT:PSS.[28] Also, the related work is under way.

Table 2

Nondoped Device Performance of Cz–3IrB–IrG Compared with 3IrB–IrG

EML	Vona (V)	Lb(cd/m2)	ηcc (cd/A)	ηpc (lm/W)	EQEd (%)	CIEe (x, y)
3IrB–IrG	2.8	11 400	32.6	34.2	9.7, 8.8, 7.9	(0.38, 0.58)
Cz–3IrB–IrG	4.2	13 706	40.9	28.5	12.0, 9.7, 8.3	(0.36, 0.59)

Turn-on voltage at 1 cd/m2.

Maximum luminance (L).

Maximum values for luminous efficiency (ηc), power efficiency (ηp), and EQE, respectively.

Maximum values, values at 100 and 500 cd/m2 for EQE.

CIE coordinates at 10 V.

Maximum values, vn class="Chemical">alues at 100 and 500 cd/m2 for EQE.

Conclusions

In summary, we present the design and synthesis of a novel dendritic multinuclear Ir complex Cz–3IrB–IrG encapsulated by n class="Chemical">oligocarbazole dendrons. The unwanted intermolecular interactions and thus TTA induced by the outer blue Ir complexes could be effectively inhibited in Cz–3IrB–IrG, owing to the characteristic shielding effect. As a result, Cz–3IrB–IrG reveals an improved PLQY and nondoped device efficiency relative to 3IrB–IrG without dendrons. The result clearly indicates that further dendronization toward multinuclear Ir complex is a promising strategy to construct highly efficient phosphors used for nondoped PhOLEDs.[29,30]

Experimental Section

General Information

1H and 19F n class="Chemical">NMR spectra were recorded with Bruker AVANCE nuclear magnetic resonance (NMR) spectrometer. The elemental analysis was performed using a Bio-Rad elemental analysis system. MALDI–TOF mass spectra were performed on an AXIMA CFR MS apparatus (COMPACT). The UV/vis absorption in dichloromethane and PL spectra in both toluene and film were measured by a PerkinElmer Lambda 35 UV/vis spectrometer and a PerkinElmer LS 50B spectrofluorometer, respectively. In addition, thin films on quartz for spectroscopic measurements were prepared by drop-coating. The lifetimes of phosphorescence from the film samples were measured in argon atmosphere by exciting the samples with 355 nm light pulses with ca. 3 ns pulse width from a Quanty-Ray DCR-2 pulsed Nd:YAG laser. Moreover, following a biexponential fitting, the average lifetimes were calculated according to the equation: τav = (A1τ12 + A2τ22)/(A1τ1 + A2τ2). CV experiments were performed on an EG&G 283 (Princeton Applied Research) potentiostat/galvanostat system. The measurements were carried out in dichloromethane for anodic sweeping and in dimethylformamide (DMF) for cathodic sweeping with a conventional three-electrode system consisting of a platinum working electrode, a platinum counter electrode, and an Ag/AgCl reference electrode. The supporting electrolyte was 0.1 M tetrabutylammonium perchlorate (n-Bu4NClO4). All potentials were calibrated against the ferrocene/ferrocenium couple. The HOMO and LUMO levels were calculated according to the equation HOMO = −e[Eox + 4.8 V] and LUMO = −e[Ered + 4.8 V], respectively, where Eox and Ered were the initial oxidation peak value and reduction peak value, respectively.

Device Fabrication and Testing

To fabricate nondoped PhOLEDs, a 40 nm thick PEDOT:PSS film was first deposited on the precleaned indium tin oxide (n class="Chemical">ITO)-glass substrates (20 Ω per square) and subsequently baked at 120 °C for 40 min. Then, the EML composed of the multinuclear Ir complex alone was spin-coated from a chlorobenzene solution and annealed at 100 °C for 30 min to remove the residual solvent in a N2 atmosphere. The thickness of the EML was about 40 nm. In successive processes, a 65 nm thick film of SPPO13 was thermally evaporated on top of the EML at a base pressure less than 4 × 10–4 Pa. Finally, a 0.5 nm thick LiF and 100 nm thick Al were evaporated as the cathode. The active area of all of the devices was 14 mm2. The EL spectra and CIE coordinates were measured using a PR650 spectra colorimeter. The current density–voltage and luminance–voltage curves of the devices were measured using a Keithley 2400 source meter and a calibrated silicon photodiode. All of the measurements were carried out at room temperature under ambient conditions. The EQE was calculated from the luminance, current density, and EL spectrum, assuming a Lambertian distribution.

Synthetic Procedures

All chemicn class="Chemical">als and reagents were used as received from commercial sources without further purification. Solvents for chemical synthesis were purified according to the standard procedures. Cz–Br,[19] IrB–OH,[31] and IrG–OH[20] were prepared according to the literature procedures.

Cz–IrB–OH

A mixture of IrB–OH (1.05 g, 1.30 mmol), Cz–Br (2.23 g, 2.60 mmol), and Cs2CO3 (0.89 g, 2.73 mmol) was added to n class="Chemical">DMF (20 mL) and heated at 70 °C for 9 h. After that, the mixture was poured into water, filtered and washed with water, and dried in vacuum. The pure product (1.2 g) was obtained by column chromatography using petroleum ether/dichloromethane = 1:1 as the eluent in a yield of 41%. 1H NMR (400 MHz, CDCl3): δ (ppm) 8.19 (br, 4H), 8.15 (d, J = 1.6 Hz, 8H), 7.88–7.73 (m, 2H), 7.70–7.64 (m, 3H), 7.62 (d, J = 5.7 Hz, 6H), 7.42 (dd, J = 8.7, 1.8 Hz, 8H), 7.35 (br, 3H), 7.30–7.27 (m, 9H), 7.21 (d, J = 6.2 Hz, 1H), 6.47–6.43 (m, 2H), 6.38–6.31 (m, 3H), 6.29–6.21 (m, 3H), 4.56 (t, J = 6.8 Hz, 4H), 4.13–4.10 (m, 4H), 2.25–2.20 (m, 4H), 2.04–1.98 (m, 4H), 1.45 (s, 72H). 19F NMR [376 MHz, CDCl3, δ (vs fluorobenzene)]: 4.14 (d, J = 9.4 Hz, 2F), 3.98 (d, J = 9.1 Hz, 1F), 2.04 (m, 3F). MALDI–TOF MS: calcd for C145H140F6IrN9O3, 2362.1; found, 2362.0 [M+]. Anal. Calcd for C145H140F6IrN9O3: C, 73.70; H, 5.97; N, 5.33. Found: C, 73.75; H, 6.00; N, 5.30.

Cz–IrB–Br

A mixture of 1,4-dibromobutane (60 μL, 0.5 mmol), n class="Chemical">Cs2CO3 (89.6 mg, 0.27 mmol), and DMF (6 mL) was heated at 70 °C. Then, a solution of Cz–IrB–OH (590 mg, 0.25 mmol) in DMF (10 mL) was added dropwise. The mixture was poured into water after reacting for 9 h and extracted with dichloromethane and dried with Na2SO4. The pure product (170 mg) was obtained by column chromatography using petroleum ether/dichloromethane = 4:1 as the eluent in a yield of 30%. 1H NMR (400 MHz, CDCl3): δ (ppm) 8.19 (br, 4H), 8.15 (d, J = 1.2 Hz, 8H), 7.86–7.77 (m, 2H), 7.75–7.73 (m, 1H), 7.64–7.63 (m, 8H), 7.42 (d, J = 8.7 Hz, 8H), 7.30–7.27 (m, 11H), 6.46 (dd, J = 6.4, 1.8 Hz, 2H), 6.39–6.30 (m, 4H), 6.27–6.22 (m, 3H), 4.57–4.55 (m, 4H), 4.13–4.11 (m, 4H), 4.02 (t, J = 5.8 Hz, 2H), 3.44 (t, J = 6.4 Hz, 2H), 2.26–2.23 (m, 4H), 2.04–2.02 (m, 6H), 1.94–1.91 (m, 2H), 1.45 (s, 72H). 19F NMR [376 MHz, CDCl3, δ (vs fluorobenzene)]: 4.19 (dd, J = 9.7, 4.0 Hz, 2F), 4.05 (d, J = 9.7 Hz, 1F), 2.10–2.03 (m, 3F). MALDI–TOF MS: calcd for C149H147BrF6IrN9O3, 2496.0; found, 2496.0 [M+]. Anal. Calcd for C149H147BrF6IrN9O3: C, 71.64; H, 5.93; N, 5.05. Found: C, 71.68; H, 6.00; N, 5.00.

Cz–3IrB–IrG

A mixture of IrG–OH (54 mg, 0.05 mmol), Cz–IrB–Br (402 mg, 0.16 mmol), and Cs2CO3 (56 mg, 0.17 mmol) was added to n class="Chemical">DMF (10 mL) and heated at 70 °C for 12 h. After that, the mixture was poured into water, filtered and washed with water, and dried in vacuum. The pure product (300 mg) was obtained by column chromatography using petroleum ether/dichloromethane = 1:1 as the eluent in a yield of 69%. 1H NMR (400 MHz, DMSO-d6): δ (ppm) 8.41 (s, 12H), 8.20 (s, 24H), 7.87–7.83 (m, 12H), 7.69 (br, 9H), 7.59 (br, 12H), 7.36–7.34 (m, 33H), 7.22–7.15 (m, 27H), 7.06–6.98 (m, 9H), 6.79–6.71 (m, 15H), 6.66–6.54 (m, 9H), 6.48–6.42 (m, 12H), 6.27–6.25 (m, 6H), 6.13–6.11 (m, 9H), 4.60–4.58 (m, 12H), 4.19–4.12 (m, 24H), 2.12 (br, 12H), 1.94–1.88 (m, 24H), 1.38 (s, 216H). 19F NMR [376 MHz, CDCl3, δ (vs fluorobenzene)]: 4.39–4.00 (m, 9F), 2.19–2.10 (m, 9F). MALDI–TOF MS: calcd for C504H477F18Ir4N33O12, 8299.42; found: 8299.2 [M+]. Anal. Calcd for C504H477F18Ir4N33O12: C, 72.94; H, 5.79; N, 5.57. Found: C, 72.70; H, 5.88; N, 5.40.