Ligand Design and Preparation, Photophysical Properties, and Device Performance of an Encapsulated-Type Pseudo-Tris(heteroleptic) Iridium(III) Emitter

The organic molecule 2-(1-phenyl-1-(pyridin-2-yl)ethyl)-6-(3-(1-phenyl-1-(pyridin-2-yl)ethyl)phenyl)pyridine (H3L) has been designed, prepared, and employed to synthesize the encapsulated-type pseudo-tris(heteroleptic) iridium(III) derivative Ir(κ6-fac-C,C′,C″-fac-N,N′,N″-L). Its formation takes place as a result of the coordination of the heterocycles to the iridium center and the ortho-CH bond activation of the phenyl groups. Dimer [Ir(μ-Cl)(η4-COD)]2 is suitable for the preparation of this compound of class [Ir(9h)] (9h = 9-electron donor hexadentate ligand), but Ir(acac)3 is a more appropriate starting material. Reactions were carried out in 1-phenylethanol. In contrast to the latter, 2-ethoxyethanol promotes the metal carbonylation, inhibiting the full coordination of H3L. Complex Ir(κ6-fac-C,C′,C″-fac-N,N′,N″-L) is a phosphorescent emitter upon photoexcitation, which has been employed to fabricate four yellow emitting devices with 1931 CIE (x:y) ∼ (0.52:0.48) and a maximum wavelength at 576 nm. These devices display luminous efficacies, external quantum efficiencies, and power efficacies at 600 cd m–2, which lie in the ranges 21.4–31.3 cd A–1, 7.8–11.3%, and 10.2–14.1 lm W1–, respectively, depending on the device configuration.


■ INTRODUCTION
The OLED devices based on phosphorescent emitters (PHOLEDs) show better performance than those employing fluorescent emissive compounds. 1Among the phosphorescent emitters, iridium(III) complexes occupy a prominent position; 2 their emissions depend on both the metal center and the ligands.As a consequence of ligand dependence, the photophysical properties of emitters of a particular metal can be governed by controlling the arrangement of the donor atoms of its coordination sphere, the reason why the ligand design is of great relevance. 3A finer adjustment of the photophysical characteristics is achieved with heteroleptic systems, which can be generated in two alternative ways: by combining different ligands or by mixing different electron donating moieties within the same polydentate ligand.As a consequence of having alternative methodologies, today, the dream of assembling emitters tailored to a certain requirement is closer to being realized. 2,3ctahedral iridium(III) compounds bearing different bidentate ligands (b) with the ability of donating 3 electrons, 3b, are of special interest, in particular [3b+3b′+3b″]  complexes that coordinate three different 3-electron donor bidentate groups.The reason for the significance of these tris(heteroleptic) compounds is that the presence of three different ligands in the metal allows a finer adjustment of the photophysical characteristics of the emitter.However, they are also difficult to obtain and purify.In addition to having a large number of stereoisomers, these complexes are often involved in redistribution equilibria. 4One way to avoid the problem, which arouses interest, is to reduce the number of ligands in the coordination sphere of the metal by increasing the number of donor atoms of some of them.In addition, it is believed that a stronger metal−ligand interaction should increase the efficiency of the emitter, despite the distortions that are generated as a consequence of the greater coordination rigidity. 5A first approximation was the use of pincer ligands, 6 in line with the impact of these groups on transition metal chemistry. 7As a consequence, during the past decade a remarkable number of iridium(III) emitters with two different tridentate ligands have been prepared. 8In recent years, efforts have been directed to the search for ligands with higher denticity; mainly the nonplanar tetradentates. 9Those formed by two different bidentate units (tt′) have been pursued with particular zeal, 10 although they are still very scarce.
A further step in the development of ligands with high denticity is the design of hexadentates (h).In principle, they should increase the strength of the metal−ligand interaction, decrease the number of stereoisomers, and avoid the formation of mixtures by ligand redistribution.However, their use in organometallics and coordination chemistry is even less frequent than the utilization of tetradentate ones.This class of ligands is of great interest because they allow encapsulating metal ions. 11Hexadentate ligands have not been employed in PHOLED technology, although some iridium(III) emitters have been prepared with these groups.They are based on tripodal structures formed by flexible arms attached to the orthometalated phenyl substituent or the heterocycle of three independent and equal units based on 2-phenylpyridine. 12hese emitters are therefore homoleptic compounds that coordinate a formally hexadentate ligand.
Our interest in the development of emitters for PHOLED technology 4g,8i,10d,13 prompted us to design the molecule 2-(1phenyl-1-(pyridin-2-yl)ethyl)-6- (3-(1-phenyl-1-(pyridin-2-yl)ethyl)phenyl)pyridine (H 3 L), as precursor of the first pseudotris(heteroleptic) iridium(III) emitter based on a hexadentate ligand.This molecule is formed by a 2-phenylpyridine moiety and two slightly different 2-benzylpyridine units.The difference between the 2-benzylpyridine units is the junction with the 2-phenylpyridine moiety through the respective methylene groups, which act as linkers; a 2-benzylpyridine ties to the phenyl substituent of the 2-phenylpyridine moiety whereas the other connects to the heterocycle.Although subtle, the asymmetry should be enough to promote different contributions of the two 2-benzylpyridine units to the frontier orbitals of the emitter.A nonenantioselective synthesis of this molecule should provide the racemic mixture of the diastereoisomers This paper describes the preparation of the designed molecule, its coordination to iridium, the photophysical properties of the resulting complexes, and the first OLED devices based on an emitter bearing a hexadentate ligand.

■ RESULTS AND DISCUSSION
Preparation of H 3 L.The designed molecule was prepared by the procedure shown in Scheme 1.It consisted of six steps, which can be grouped into three stages: introduction of a phenyl fragment into the methylene unit of a 2-benzylpyridine, coupling of said phenyl with a pyridyl group, and linkage of the latter with a second 2-benzylpyridine through its methylene unit.
The first stage is formed by three steps.In the first one, 2benzoylpyridine (1) was used as the synthon for the 2benzylpyridine that was employed as a support to introduce the phenyl group.The coupling was achieved with an organolithium reagent.The ketone dissolved in tetrahydrofuran was treated with 3-chlorophenyllitium, which was generated in situ.The nucleophilic addition of the organometallic reagent to the carbonyl group followed by the hydrolysis of the resulting alcoholate 14 led to (3-chlorophenyl)(phenyl)-(pyridine-2-yl)methanol (2), which was isolated as an orange oil in almost quantitative yield.The second step was the direct deoxygenation of the alcohol to form 2-((3-chlorophenyl)-(phenyl)methyl)pyridine ( 3), as a brown oil in 83% yield.The reaction was carried out in acetic acid, at 100 °C, using an aqueous solution of hydroiodic acid as deoxygenating reagent.The use of this procedure merits some additional comments.Alcohol deoxygenation processes are usually mediated by metals.The Barton−McCombie methodology is probably the most representative, although it involves several steps and the utilization of toxic tin hydride limits its application from an industrial point of view. 15Titanium(III) derivatives are gaining popularity in recent years as an alternative method, since they allow deoxygenation to be carried out in one step. 16Single-step metal-free deoxygenation, such as that employed here, is more challenging, 17 due to the high stability of the C−O bond and kinetic inertia, 18 and as a consequence it has been used less.In this context, it should be pointed out that Bro̷ nsted acids, as hydroiodic, are promising deoxygenating agents due to their versatility. 19Once 3 was formed, the C(sp 3 )−H hydrogen atom was subsequently replaced by a methyl group, in the third step, to prevent the formation of trityl-type radicals.The replacement was executed in tetrahydrofuran, at −78 °C, by proton extraction with phenyllithium and subsequent capture of the resulting anion with methyl iodide.The methyl counterpart of 3, 2-(1-(3-chlorophenyl)-1-phenylethyl)pyridine (4), was also obtained as a brown oil in almost quantitative yield.
Steps four and five constitute the second stage.Having generated in the first stage a phenyl with two substituents in 1,3-positions, a 2-benzylpyridine linked by the methylene unit and a chlorine atom, we approached the formation of the 2phenylpyridine-type compound taking advantage of the presence of the chloride substituent.The latter was replaced by a pinacolboryl group (Bpin) in the fourth step, to subsequently perform a Suzuki−Miyaura cross-coupling reaction with 2-bromo-6-fluoropyridine in the fifth one.The borylation of 4, which afforded 2-(1-phenyl-1-(3pinacolborylphenyl)ethyl)pyridine ( 5), was executed with pinB-Bpin in the presence of 4 equiv of potassium acetate, at 130 °C, using 5 mol% of complex Pd(OMs){κ 2 -C,N-(C 6 H 4 -NH 2 )}(XPhos) (XPhos-Pd-G3; OMs = methanesulfonate, XPhos = 2-(dicyclohexylphosphino)-2′,4′,6′-triisopropyl-1,1′biphenyl) as catalyst precursor, 20 and dimethylformamide as solvent.Buchwald's XPhos-Pd-G3 complex had previously proved to be efficient for the direct borylation of a variety of aryl halides. 21Although a black solid was formed under these conditions, probably due to decomposition of the catalyst precursor to palladium(0), the Miyaura-borylation of the aryl halide 22 took place in an efficient manner after 3 h.Thus, the borylated product 5 was isolated as a light-yellow oleaginous gum in 61% yield, with previous purification of the reaction crude by silica column chromatography.The Suzuki−Miyaura cross-coupling 23 between 5 and 2-bromo-6-fluoropyridine was Chart 1. Diastereomers of H 3 L Inorganic Chemistry carried out in a dioxane/water mixture as solvent, at 95 °C, using 10 mol% of the palladium derivative Pd(PPh 3 ) 4 as catalyst precursor, and 3 equiv of potassium carbonate as base.Under these conditions, the coupling was complete within 3 h.Thus, after purification of the reaction crude by silica column chromatography, the 2-phenypyridine-type compound 2fluoro-6- (3-(1-phenyl-1-(pyridine-2-yl)ethyl)phenyl)pyridine (6) was isolated in 86% yield as a pale yellow oil.
The presence of the fluorine substituent at the pyridyl group of the 2-phenylpyridine moiety of 6 facilitated the attachment of a second 2-benzylpyridine in the last stage, which consists of only one step, the sixth.The nucleophilic aromatic substitution of this substituent by the anion 1-phenyl-1-(pyridin-2-yl)ethan-1-ide, resulting from the deprotonation of the tertiary 7), in agreement with that previously observed for 2-fluoro-6-phenyl-pyridine.10c The reaction was carried out in tetrahydrofuran, and the desired compound was obtained as a white solid in 41% yield, about 17% with regard to 1, after a laborious workup including purification by silica column chromatography.Proligand 7 was formed as the mixture of diastereoisomers shown in Chart 1, which were found to be indistinguishable by NMR spectroscopy.
Coordination of H 3 L to Iridium.As a first option, we tested the well-known dimer [Ir(μ-Cl)(η 4 -COD)] 2 (8, COD = 1,5-cyclooctadiene) as an iridium precursor.This compound Scheme 1. Synthesis of the Proligand H 3 L Scheme 2. Synthesis of Complexes 9 and 10 Inorganic Chemistry had previously been shown to coordinate the heterocycles and promote the activation of an ortho-CH bond of the phenyl groups of the proligands 2-phenyl-6-(1-phenyl-1-(pyridin-2yl)ethyl)pyridine 10c and 1-phenyl-3-(1-phenyl-1-(pyridin-2yl)ethyl))isoquinoline.10d In both cases, the products resulting from the assembly process presented the expected tetradentate 6-electron donor ligands (6tt′).However, the generated complexes depended on the reaction conditions and the primary or secondary character of the alcohol used as solvent.Mononuclear carbonyl derivatives [Ir(6tt′)Cl(CO)] were obtained with the primary 2-ethoxyethanol, at reflux, while the secondary character of 1-phenylethanol prevented the decarbonylation of the alcohol 24 and allowed the formation of dimers [Ir(μ-Cl)(6tt′)] 2 .The nature of the alcohol used as solvent in the reaction of 8 with the new organic proligand 7 has a marked influence not only on the type of resulting product but also on the class of ligand generated, since two different coordination modes arise (Scheme 2).
Treatment of suspensions of 8, in 2-ethoxyethanol, with 2.0 equiv of 7, under reflux, for 72 h afforded the carbonyl derivative Ir(κ 4 -cis-C,C′-cis-N,N′-HL)Cl(CO) (9), the [Ir-(6tt′)Cl(CO)]-counterpart resulting from 7. In a consistent manner with the previous [Ir(6tt′)Cl(CO)] complexes, its formation involves the selective orthometalation of the benzylpyridine moiety attached to the pyridyl group of the 2-phenylpyridine unit, in addition to the orthometalation of the latter and the alcohol decarbonylation.Complex 9 was separated as a yellow solid, in 36% yield, from the reaction crude, which contained a significant amount of decomposition products.The separation was performed by neutral alumina column chromatography.Figures S27 and S28 show the 1 H and 13 C{ 1 H} NMR spectra of the solid, which reveal that the diastereoisomers resulting from the reaction 8 with both enantiomeric pairs H 3 L RR -H 3 L SS and H 3 L RS -H 3 L SR are also indistinguishable by NMR in this case.In addition, the complex was characterized by X-ray diffraction analysis.The structure (Figure 1) proves the preference of the iridium center by the benzylpyridine moiety attached to the pyridyl group.The coordination around the metal center is the expected octahedral with the heterocycles of the 6tt′ ligand mutually cisdisposed (N(1)−Ir−N(2) = 90.54(17)°).The phenyl group of the 2-phenylipyridine unit is situated trans to the pyridyl ring of the 2-benzylpyridine moiety (C(24)−Ir−N(1) = 168.42(18)°),whereas the phenyl group of the latter locates trans to the chloride anion (C(1)−Ir−Cl = 167.87(15)°.For its part, the carbonyl ligand lies trans to the heterocycle of the 2-phenylpyridine unit (C(38)−Ir−N(2) = 170.7(2)°).In accordance with the presence of a carbonyl ligand in 9, its IR spectrum displays a characteristic ν(CO) band at 2017 cm −1 and the 13 C{ 1 H} NMR spectrum in dichloromethane-d 2 contains a singlet at 172.3 ppm.
The use of 1-phenylethanol instead of 2-ethoxyethanol allowed the orthometalation of both benzylpyridine substituents of the 2-phenylpyridine core of 7. Given the presence of two chiral centers in each isomer of the racemic mixture, enantiomeric pairs of four different diasteroisomers are possible in the resulting complex [Ir(9h)] (9h = 9-electron donor hexadentate ligand), one of them with a fac disposition of carbons and heteroatoms and the others displaying a mer arrangement (Chart S1).The former should be a consequence of the selective reaction of H 3 L RR -H 3 L SS with 8, while the other three would result from reactions of both H 3 L RR -H 3 L SS and H 3 L RS -H 3 L SR .DFT calculations (B3LYPG-D3//SDD(f)-6−31G**) revealed that the fac isomer is the most stable.In agreement with this, treatment of suspensions of 8, in 1phenylethanol, with 2.0 equiv of 7, under reflux, for 72 h led to a brown solid, from which the calculated fac isomer Ir(κ 6 -fac-C,C′,C″-fac-N,N′,N″-L) (10) was separated as an orange solid, in 12% yield, by neutral alumina column chromatography.Its formation was supported by the 1 H and 13 C{ 1 H} NMR spectra of the obtained solid (Figures S29−S30) and X-ray diffraction analysis.The structure, which contains both crystallographically independent enantiomers (Λ-Ir(κ 6 -fac-C,C′,C″-fac-N,N′,N″-L RR ) and Δ-Ir(κ 6 -fac-C,C′,C″-fac-N,N′,N″-L SS )) in the asymmetric unit, demonstrates the coordination of the three heterocycles to the iridium center along with the orthometalacion of the three phenyl groups, as well as the fac dispositions of both types of atoms, carbon and nitrogen.Figure 2 gives a view of the enantiomer Λ-Ir(κ 6 -fac-C,C′,C″-fac-N,N′,N″-L RR ).The angles N−Ir−C, which lie in the range 166−175°, are consistent with a high molecular stability.
Having established that the proligand 7 is able to generate complexes bearing 6tt′ and 9h ligands, depending upon the reaction solvent, we decided to address the improvement of the yields for the preparation of the complexes stabilized by such ligands.We had observed that the use of the cyclooctene (COE) derivative [Ir(μ-Cl)(η 2 -COE)] 2 (11) instead of 8 facilitates the coordination of the proligand 1-phenyl-3-(1phenyl-1-(pyridin-2-yl)ethyl))isoquinoline to the iridium center, increasing the yield for the preparation of the corresponding dimer [Ir(μ-Cl)(6tt′)] 2 from 34% to 84%.10d Such surprising improvement prompted us to change the precursor for the reactions shown in Scheme 2. Unfortunately, none of them improved.Using precursor 11 as starting material, the yield in the formation of 9 decreased to 22%, whereas complex 10 was not formed.An alternative precursor sometimes successfully used to synthesize iridium(III) compounds bearing a hexadentate ligand is the tris- (acetylacetonate) derivative Ir(acac) 3 (12).12a,c,e Its use doubled the yield of the preparation of 10, which reached 29% (Scheme 2).The reason for the relatively low yields obtained in the formation of 10, with the investigated precursors, appears to be related to the fact that only half of 7, H 3 L RR -H 3 L SS , is capable of forming the product.
Photophysical and Electrochemical Properties of 9 and 10.The ultraviolet−visible (UV−vis) spectra of 2methyltetrahydrofuran (2-MeTHF) solutions of 9 and 10 are typical for six-coordinate iridium(III) species (Figure 3), showing the usual three energy regions: <300, 350−450, and >450 nm.Table 1 points out some characteristic transitions, which were assigned based on DFT (TD-DFT) calculations (B3LYP-D3//SDD(f)/6−31G**) in THF (Figures S1 and S2, Tables S1 and S2).Higher energy bands (<300 nm) are due to 1 π−π* transitions in the ligand, which take place without metal participation.Spin allowed charge transfers from the iridium center to the heterocycles combined with transitions from the phenyl groups to the heterocycles are observed between 350 and 450 nm, whereas formally spin forbidden transitions are also evident after 450 nm.They result from the large spin− orbit coupling produced by the iridium presence and mainly occur from the HOMO to the LUMO.The HOMO of both compounds delocalizes between the metal center and the phenyl groups, while the LUMO covers the heterocycles (Figures S5 and S6, Tables S3 and S4).
DFT-calculated HOMO energy levels agree with those experimentally obtained from the electrochemical study of both complexes (Table 2).Figure S4 depicts the voltammograms, which were measured in acetonitrile, under argon, using [Bu 4 N]PF 6 as supporting electrolyte (0.1 M).They display an Ir(III) to Ir(IV) oxidation, which appears at 1.10 V for 9 and at 0.14 V for 10, versus Fc/Fc + .10d The notable difference between both values can be associated with the presence of the carbonyl ligand in 9, which significantly stabilizes the HOMO.The oxidation of 9 is irreversible, while that of 10 is quasi-reversible.The irreversible character of the oxidation of 9 is not surprising.In this context, it should be noted that the carbonyl ligand is acidic and therefore should destabilize the oxidized species.Reductions were not detected within the acetonitrile window, between −2.9 and 1.6 V.
Complexes 9 and 10 are phosphorescent emitters upon photoexcitation.The measurements were performed in a doped poly(methyl methacrylate) (PMMA) film at 5 wt%, at room temperature, and in 2-MeTHF at room temperature and at 77 K. Figure 4 collects the spectra recorded under such conditions.Emissions occur from the respective T 1 excited states, as is suggested by excellent agreement between the experimental wavelengths and the values calculated in THF for the difference in energy between the optimized triplet states T 1 and the singlet states S 0 .
The shape of the emission bands is broadly similar in both compounds.Full width at half-maximum (fwhm) values depend on the medium and lie in the range 2150−3526 cm −1 for 9 and 2088−2675 cm −1 for 10.The main difference between the spectra is observed in the energy of the emission maxima.Complex 9 is a green emitter with maxima between 515 and 549 nm.The presence of the carbonyl group in the latter produces an increase of the HOMO−LUMO gap with respect to 10 (Table 2).Thus, complex 10 with a smaller gap between frontier orbitals emits in the lower energy yellow region of the spectrum, with maxima between 552 and 587 nm.The lifetimes are short, lying in the range 0.6−8.4μs.Quantum yields are moderate, about 0.40, and similar for both compounds in PMMA.While for 10 the value is maintained in 2-MeTHF at room temperature, for 9 it is reduced by half.Such a decrease is associated with a decrease in the radiative rate constant (k r ) for 9 of about 1 order of magnitude with respect to the value observed for 10.This suggests significant differences in the solvation of both compounds, which tentatively could be related to the different nature of the polydentate ligands.As a consequence of this situation, the ratio between the radiative and nonradiative (k nr ) rate constants is the same for both compounds in PMMA and also in solution for 10 (0.7).In contrast to 10, the ratio decreases to 0.2 for 9 in 2-MeTHF (Table 3).

Electroluminescence of OLED Devices Based on 10.
To investigate the applicability of encapsulated-type pseudotris(heteroleptic) iridium(III) compounds in OLED device fabrication, we studied the behavior of complex 10 in four bottom-emission OLED structures, as an example of a yellow phosphorescent emitter.Figure 5 outlines the configurations of the devices d 1 −d 4 , including the energy levels and the thickness of the layers, as well as the chemical nature of the materials used.
The devices were built by high vacuum (<10 −7 Torr) thermal evaporation.Immediately after their fabrication, they were encapsulated within a nitrogen glovebox (<1 ppm of H 2 O and O 2 ) and a moisture getter was incorporated inside the package, which was closed with a glass lid that was subsequently sealed with an epoxy resin.The anode and cathode electrodes of the four devices were set up with the same components.The anode consisted of 750 Å of indium tin oxide (ITO), while the cathode was formed by a 10 Å LiF electron injection layer followed by another 1000 Å Al layer.The simplest device structure (d 1 ) consisted of the following layers sequentially disposed from the ITO surface to the cathode: 100 Å of HAT-CN as the hole injection layer (HIL), 450 Å of NPD as a hole transporting layer (HTL), 300 Å of an emissive layer (EML) containing host (H1) doped with complex 10 as an emitter at 9%, and 450 Å of Alq 3 as an electron transporting layer (ETL).In the search for improvement to the d 1 features, we also fabricated devices d 2 −d 4 , resulting from modifications of the d 1 structure (Figure 5).The d 2 device was made with the aim of preventing holes and excitons from leaking into the low triplet Alq 3 electron transport layer, quenching them.With this goal, a 50 Å layer of the hole-blocking material BL (T 1 = 2.56 eV) was assembled between the emissive and Alq 3 layers.At the same time, the thickness of the latter was reduced to 400 Å.The d 3 device displays the most complete structure.It incorporates, between the hole transporting and emissive layers of d 2 , a 50 Å electron and excitons blocking layer of the high triplet energy TCTA compound (T 1 = 2.76 eV), simultaneously reducing the thickness of the hole transporting layer to 400 Å.This   additional layer (EBL) should prevent exciton leakage to the low triplet NPD compound, which would improve the device EQE.To assess the relative relevance of both introduced layers, we finally built d 4 removing the BL layer from d 3 and increasing the thickness of the electron transporting layer to 450 Å.The total thickness of the organic stack remained constant across all four devices.Such design was carried out to eliminate possible distortions in the efficiency measurement due to changes in the outcoupling related to the thickness of the organic layers.Table 4 summarizes the performance of the devices, including emission features and values of turn-on voltage, luminous efficacy (LE), external quantum efficiency (EQE), and power efficacy (PE) at a luminance of 600 cd m −2 Complex 10 provided a yellow emission with 1931 CIE (x:y) ∼ (0.52:0.48), wavelength maximum at 576 nm, full width half-maximum of 84 nm, and emission offset below 525 nm (Figure 6a).The finding corresponds to a triplet emission energy of the emitter of more than 2.35 eV.To efficiently confine these high triplet excitons, layers of high triplet material are required around the emissive one.However, both the NPD compound that acts as a hole transporting (T 1 = 2.29 eV) and the Alq 3 derivative that proceeds as an electron carrier (T 1 ∼2.0−2.1 eV) have a lower triplet, which cannot efficiently confine the excitons within the emissive layer.As a consequence, the simplest device structure d 1 , without the protecting layers of BL and TCTA compounds, has the lowest EQE of 7.8%.As expected, the introduction of the BL hole blocking material between the emissive and Alq 3 layers prevents excitons leaking to the latter.As a consequence, the d 2 configuration improves the device EQE by about 40% with regard to d 1 , reaching a value of 11.0%.A further slight improvement is achieved with the exciton and electron blocking layer of TCTA, between the emissive and hole transporting layers.Thus, the d 3 device shows an EQE of 11.3%, the highest value of the four.The BL effect is significantly greater than the TCTA effect in EQE improving, as demonstrated by the d 4 device.The latter, bearing the electron and exciton blocking TCTA compound as a unique additional layer, achieves an EQE of 8.4%, a significantly poorer value than that of the device d 2 containing the BL hole blocking layer.This could be explained by the location of the recombination zone in the emissive layer.As follows from the energy levels shown in Figure 5, holes are being transported via emissive layer by the emitter, while the host transports electrons.Thus, the most likely recombination zone is in the proximity to the ELT side of the emissive layer rather than to HTL side.Figure 6b gives additional evidence of the marked difference in effect between the layers of BL and TCTA compounds on the EQE of devices, as a consequence of the localization of the recombination region of the emissive layer close to the ETL side.At low luminance (10−100 cd m −2 ) the efficiency of devices d 1 and d 4 , which do not bear BL layer, is very low and significantly increases with luminance increase, whereas the efficiency of devices d 2 and d 3 , containing BL layer, remains very high even at low luminance.However, the recombination zone appears to propagate toward the HTL side of the emissive layer with current density and luminance increase.As a result, the effect of BL and TCTA compounds on the device efficiency is approximately the same, at luminance level above 10000 cd m −2 .The four devices display very similar profiles for current density versus voltage showing turn-on voltages between 6.6 and 7.1 V (Figure 6c).

■ CONCLUDING REMARKS
This study reveals that encapsulated-type pseudo-tris-(heteroleptic) iridium(III) emitters of class [Ir(9h)], with a fac-disposition of carbon and nitrogen atoms, are accessible when the organic molecule responsible of the formation of the hexadentate ligand h is properly designed and can be prepared.Once the organic proligand is isolated, several details should be taken into account for its coordination to the iridium center, the process that gives rise to the synthesis of the emitter.Such details are related to the metal precursor and the reaction solvent.Although the well-known dimer [Ir(μ-Cl)(η 4 -COD)] 2 is suitable as an iridium precursor, the tris(acetylacetonate) derivative Ir(acac) 3 is a more appropriate starting material.Secondary alcohols of high boiling point are preferred solvents over primary alcohols, since the former prevent a possible metal carbonylation that could inhibit the full coordination of the proligand.As here demonstrated, these emitters have applicability in the fabrication of OLED devices, as the device performance is more than reasonable.
In summary, we here report the overall process up to the fabrication of a yellow emitting device, which bears the first encapsulated-type pseudo-trisheteroleptic iridium(III) emitter and displays a maximum wavelength of 576 nm, an external quantum efficiency of 11.3%, and a luminous efficacy of 31.3 cd A −1 at 600 cd m −2 .

Inorganic Chemistry
added.The anion was stirred for 3 h at −78 °C forming a red solution/suspension.At this temperature a solution of 6 (500 mg, 1.41 mmol) in THF (4 mL) was added forming a dark red/brown solution, which was stirred at −78 °C for 1 h and then allowed to warm to rt.During this time a red/burgundy solution forms, and the reaction was held at rt for 30 min and then heated to 70 °C overnight.The reaction was cooled to rt, and an additional amount of 2-(1phenylethyl)pyridine (342 mg, 1.87 mmol) and lithium chloride (795.6 mg, 18.8 mmol) were added.The reaction mixture was then cooled back down to −78 °C.Then, more LDA (1 M in THF, 3.52 mmol, 3.52 mL) was added, and the reaction was stirred at −78 °C for 3 h before being allowed to warm to rt.The reaction was stirred at rt for 30 min and then heated to 70 °C for 2 days, forming a red/ brown solution with a yellow precipitate.The reaction was cooled to rt and concentrated in vacuo to ca. 10% volume, and then water and EtOAc were added followed by heating to 60 °C to ensure full dissolution of the product.The biphasic mixture was filtered through celite to remove some flocculants solids.The layers were separated, and the aqueous phase was extracted with EtOAc.The organic extracts were combined and washed with water, and then brine, and finally dried over MgSO 4 before filtering and concentrating in vacuo to provide a brown oil.It was loaded onto silica and purified by column chromatography, eluting with EtOAc/pentane increasing the polarity from 5% to 30% of EtOAc, to afford a white solid (298 mg, 41%  5, 167.3, 166.2, 155.9, 149.3 (all q ), 149.2, 148.9 (both CH), 148.7, 139.4 (both C q ), 137.0, 136.2, 136.0,  129.7, 129.1 (4C), 128.5, 128.3 (4C), 127.9, 126.5, 126.4,124.8,  124.1, 123.8, 123.8, 122.3, 121.5, 121.4,117.8 (all CH), 58.3, 55.7  (both C q Me), 29.7, 28.4 (both Me).
Preparation of Ir(κ 6 -C,C′,C′′,N,N′,N′′-L) (10).Route a: A mixture of 8 (75 mg, 0.111 mmol) and H 3 L (7) (224 mg, 0.444 mmol) in 5 mL of 1-phenyethanol was heated under reflux for 72 h, and a dark red solution was formed.Afterward, the solution was cooled to rt and the solvent was removed under vacuum.The crude was washed with diethyl ether (10 × 10 mL) to remove all the possible traces of 1-phenylethanol.The brown solid crude was purified by neutral alumina column chromatography using diethyl ether as eluent and finishing with a mixture of 3:1 diethyl ether/ dichloromethane.An orange solid was obtained.Yield: 20 mg (12%).
■ ASSOCIATED CONTENT * sı Supporting Information

Figure 5 .
Figure 5. Energy levels and materials used in the different layers of the devices.