3.1 Platinum Phosphors with N^N^C-coordinating Ligands

Much work on cyclometalated Pt(II) complexes based on N^N^C-coordinating ligands (PtNNC-1 to PtNNC-9, Figure 3) and their emitting characteristics have been reported by Che et al. (19, 20). In solution, tridentate Pt(II) complexes bearing σ-alkynyl auxiliary ligands can emit a variety of colours from 550 nm to 630 nm, induced by ligands with different steric and electronic properties. Although their EL performance is poor, these complexes have provided valuable information on tuning the emission colour of this type of phosphor.

Importantly, their performance was markedly improved by extending the π-conjugation of the ligands, for example PtNNCCl-1 (Figure 4) had a Φ_P of 0.68 and a maximum CE of 20.2 cd A⁻¹ (21). The performance of this device was superior to those based on tetradentate Pt(II) Schiff base phosphors (15, 16).

Extending the π-conjugation of the N^N^C ligands with a fluorene unit has been shown to play a critical role in increasing phosphorescent emission and EL efficiency, in devices based on complexes PtNNCCl-2 to PtNNCCl-6 (22). The Φ_P for PtNNCCl-2 was 0.07 and that for PtNNCCl-3 was 0.16, while PtNNCCl-4 and PtNNCCl-5 exhibited ФP values of 0.35 and 0.73, respectively. Owing to its high ФP, PtNNCCl-5 was used in a vacuum deposition fabricated device which exhibited very attractive EL efficiencies with an EQE of 5.5%, a CE of 14.7 cd A⁻¹ and a PE of 9.2 lm W⁻¹. The moderate EL performance of PtNNCCl-6 is thought to be due to the method of solution-processed device fabrication.

3.2 Platinum Phosphors with N^C^N-Coordinating Ligands

As noted earlier, higher rigidity of the molecular skeleton generally leads to higher Φ_P. J. A. G. Williams et al. reported that the platinum–carbon bond lengths in Pt(N^C^N) complexes are around 1.90 Å, about 0.14 Å shorter than those in typical Pt(N^N^C) complexes (23). Furthermore, the shorter Pt–C bond length is expected to deactivate the d-d states by raising their energy, and hence lead to superior performance. This can be seen in a new series of Pt(N^C^N) complexes bearing various aryl substituents (Figure 5) (24). All of the new phosphors display Φ_P values from 0.46 to 0.65 with emission maxima from 481 nm to 588 nm. Additionally, the introduction of 4-(dimethylamino)phenyl to N^C^N ligands can suppress the self-quenching and aggregation/excimer formation which would otherwise hamper the EL efficiencies. OLEDs employing complexes PtL¹Cl to PtL⁹Cl as emitters (25) can achieve maximum EQE values from 4 to 16% and CE values from 15 to 40 cd A⁻¹, accompanying extremely low EL efficiency roll-off with increased driving voltage.

The EL performance of these phosphors can also be optimised by changing the auxiliary ligand and the host materials in devices. Replacing the ancillary chloride (–Cl) ligand in PtL³Cl with a phenoxy (–OPh) group gives PtL³OPh (26), which can provide very admirable green-emitting OLEDs with a maximum EQE of 16.5% and a CE of 67.1 cd A⁻¹. By adopting a two-host system to optimise the energy levels between charge-transport layers and confining the emissive zone, the EL efficiencies for PtL¹⁰Cl-based devices have been almost doubled with respect to those of their single-host counterparts, to give an EQE of 11.5%, a CE of 38.9 cd A⁻¹ and a PE of 27.2 lm W⁻¹ (27).

Jabbour et al. constructed a highly efficient blue phosphor PtL¹¹Cl (28) with a maximum Φ_P of 0.8 in degassed dichloromethane (CH₂Cl₂) (29), which was ascribed to both its rigid triplet state configuration and very high ligand-field strength giving a strong destabilisation to the metal-centred d-d excited states. A deep-blue device based on PtL¹¹Cl doped in a co-host gave a maximum EQE of 16% and a PE of 20 lm W⁻¹. A WOLED based on 8 wt% PtL¹¹Cl doped in 1,3-bis(N-carbazolyl)benzene (mCP) exhibited a maximum EQE of 9.3%, a PE of 8.2 lm W⁻¹ and Commission Internationale de l’Éclairage (CIE) coordinates of (0.33, 0.35) at 1300 cd m⁻², very close to the pure white point at (0.33, 0.33).

Another highly efficient phosphor PtL¹²Cl (Figure 5) with a Φ_P of 0.87 was incorporated into OLEDs at doping levels of 10, 15, 20 and 25 wt%, respectively. By varying the doping level, the EL colour of the devices could be tuned from bluish-green to red with very high EQE values up to 18.3% (30). It is worth mentioning that the electron-donating groups attached to the lateral pyridyl rings of the Pt complexes resulted in a blue-shift in the monomer emission, as well as excimer emission resulting from the interactions between emitter molecules (29). Accordingly, PtL¹³Cl (31) showed a blue-shift emission compared with that of PtL¹²Cl due to the stronger electron-donating ability of the methoxy group (–OCH₃) compared to the methyl group (–CH₃). By tuning the doping concentration from 5 wt% to 35 wt%, OLEDs based on PtL¹³Cl could be made to emit nearly any colour from blue to yellowish red. The dependence of emission colour on doping concentration offers a very simple and practical way to regulate OLED emission colour by adjusting the contributions of monomer and excimer emission.

3.3 Platinum Phosphors with N^N^N-Coordinating Ligands

Chen et al. introduced a trifluoromethyl-pyrazolyl group to 2,2′-bipyridine to form a neutral tridentate cyclometalated ligand, 6-(5-trifluoromethyl-pyrazol-3-yl)-2,2′-bipyridine, and investigated the photophysical properties of the resulting complexes (Figure 6) (32). These complexes emit phosphorescence between 524 nm for PtNNN-1 and 604 nm for PtNNN-2. The bathochromic effect in emission may lie in the stronger donor ability of acetamide and the larger π-conjugation involving the nitrogen and the carbonyl group of acetamide in PtNNN-2. An optimised device doped with 28 wt% PtNNN-1 exhibited a maximum EQE of 8.5%, corresponding to a CE of 18.5 cd A⁻¹ and a maximum brightness of 47,543 cd m⁻² at 18.5 V.

Cola et al. also reported several Pt(N^N^N)-type complexes displaying Φ_P values of up to 0.73 in deaerated chloroform solution (33), which are among the highest found for tridentate chelates (21, 22, 27, 34). A multilayer vapour deposited device doped with 6% PtNNN-3 (Figure 6) generated a green emission peak at 510 nm with a second peak at 544 nm and gave a maximum CE of 15.2 cd A⁻¹ and a PE of 6.8 lm W⁻¹, which is comparable with the performance of devices based on the well-established green emitter, Ir(ppy)₃, in a similar architecture.

3.4 C^N^C-Coordinating Ligands Based Platinum Phosphors

Though several Pt(II) complexes with C^N^C type ligands were reported in 2002 (35), they drew little attention due to their poor emission (35, 36). Recently, Kui et al. reported the first examples of complexes of these ligands with intense phosphorescence in solution at room temperature (Figure 7) (37). By functionalising with carbazole, fluorene, or thiophene heterocyclic unit(s) at the periphery, they can show Φ_P values up to 0.26 in solution at room temperature. This emission performance has been ascribed to both the σ-donating ability of the C≡NR ancillary ligand and the rigidity of the triplet state due to the carbazole, fluorene or thiophene heterocyclic units in the main ligand. Red-emitting OLEDs based on this ligand system can give very decent EL performance. For example, a device based on PtCNC-1 displayed a maximum EQE of 12%, a PE of 10.9 lm W⁻¹ and a CE of 13.9 cd A⁻¹, while a device doped with 6 wt% PtCNC-2 achieved a maximum EQE, PE and CE of 12.6%, 10.5 lm W⁻¹ and 13.4 cd A⁻¹, respectively.

4.1 Conventional (N^C)Pt(O^O) Phosphors

Complexes with one N^C main ligand (2-phenylpyridine-type or ppy-type) and one ancillary β-diketonato ligand (acetyl acetone or its derivatives) represent the first and most developed type of EL phosphors, with good emission and tunable colour.

J. Brooks et al. reported a series of this type of phosphor, Ptppy-1 to Ptppy-8 (Figure 8), which show intense phosphorescence at room temperature (38). They also show distinct emission colour tuning properties. Introducing electron-withdrawing groups to the phenyl moiety and electron-donating groups to the pyridyl ring in the ppy ligand can induce a blue-shift in the emission maxima (λ_max = 486 nm for Ptppy-1, λ_max = 484 nm for Ptppy-3 and λ_max = 466 nm for Ptppy-2, λ_max = 456 nm for Ptppy-6 and λ_max = 440 nm for Ptppy-7), while electron-donating groups on the phenyl ring of the ligand will cause a red-shift (λ_max = 525 nm for Ptppy-4 and λ_max = 480 nm for Ptppy-5). Theoretical calculations indicate that this colour tuning behaviour can be explained in terms of the highest occupied molecular orbitals (HOMOs), contributed mainly by the π orbitals of the phenyl ring and the d_π orbital of the Pt centre, and the lowest unoccupied molecular orbitals (LUMOs), located mainly on the π orbitals of the pyridyl ring in the ligands.

The excimer emission in the orange or red region caused by interaction among molecules means that blue-emitting phosphors can be used to fabricate simple WOLEDs by a blue-orange ‘colour complementary’ strategy. Co-doping Ptppy-2 and FIrpic (Figure 8) in a single layer allows the blue monomer emission from FIrpic and the orange excimer emission from Ptppy-2 to be combined, providing white light for simple WOLEDs with a single emissive layer (EML) (39). Based on the same strategy, the single phosphor Ptppy-8 can also be used to create simple WOLEDs with just one EML (40).

Importantly, the performance of these simple WOLEDs can easily be improved by inserting an electron blocking layer (EBL), selecting the proper host materials, and optimising the energy levels of the functional layers, thus balancing the carrier injection in the device. In single EML WOLEDs with Ptppy-2 as the emitter, the EL efficiencies could be improved to a peak EQE of 6.4%, a CE of 17.0 cd A⁻¹ and a PE of 12.2 lm W⁻¹ by inserting an Irppz EBL. The host 2,6-bis(N-carbazolyl)pyridine (mCPy) could guarantee a more efficient energy transfer to Ptppy-2 than conventional mCP. Simple WOLEDs with Ptppy-2 doped into mCPy could thereby furnish better device efficiencies with an EQE of 6.9% and a CE of 15.7 cd A⁻¹ at 500 cd m⁻². Taking poly(3,4-ethylenedioxythiophene):poly(4-styrenesulfonate) (PEDOT:PSS) as the hole injecting layer (HIL) and poly(N-vinylcarbazole) (PVK) as the hole transport layer (HTL) as well as the EBL, the efficiencies of simple WOLEDs could be maximised to an EQE of 18%, a PE of 29 lm W⁻¹ and a CE of 42.5 cd A⁻¹ (41), corresponding to an IQE of nearly 100% (42). These dramatic enhancements may be attributed to the smoother interface of PEDOT:PSS/PVK compared to that of PEDOT:indium tin oxide (ITO)/PVK, resulting in a smaller interfacial surface area and lower leakage current. Furthermore, ITO/PVK has an injection barrier 0.6 eV larger than that of ITO/N,N′-di-[(1-naphthyl)-N,N′-diphenyl]-1,1′-biphenyl)-4,4′-diamine (NPD) (the HOMOs for PVK and NPD are 5.8 eV and 5.2 eV, respectively) and PVK has a much lower hole mobility than NPD, which can effectively reduce the number of holes injecting to the emissive layer, leading to a more balanced ratio of holes and electrons.

Some phosphors with substituted β-diketonato ligands (Ptppy-9 and Ptppy-10) (Figure 9) were also obtained (43). Although the substituents can exert only limited electronic effects on the complex cores, they can evidently influence the molecular packing in the crystals and the thin films to facilitate the fabrication of simple non-doped devices. The attractive EL performance of non-doped devices based on Ptppy-10, with a peak EQE of 9.76%, a CE of 28.1 cd A⁻¹ and a PE of 9.91 lm W⁻¹ at 500 cd m⁻² indicates the great potential of these phosphors for simplifying device fabrication, since traditional phosphorescent OLEDs generally possess doped EMLs which are much more complicated to construct.

Ptppy-2 can be attached to a random terpolymer backbone containing hole-transport (HT) and electron-transport (ET) moieties, leading to a range of phosphorescent terpolymers (44) (Figure 10). The polymeric phosphor Ptppy-P was employed to prepare a simple solution-processed OLED, with a maximum EQE of 4.6% and CIE coordinates of (0.33, 0.50). Ptppy-2 has also been attached to a polyhedral oligomeric silsesquioxane (POSS) core (45) (Figure 10). The carbazole moieties provided efficient HT character as well as diluting the concentration of the Pt complex bonded to the POSS, thus controlling the emission contributions from the monomer and excimer states. The maximum EQE was 8.4% for a Ptppy-POSS based solution-processed WOLED. Although its CIE coordinates of (0.46, 0.44) deviate slightly from the ideal white point at (0.33, 0.33), the EL spectrum of the device was voltage-independent, indicating good colour stability.

4.2 Functionalised (N^C)Pt(O^O) Phosphors

Besides the Pt phosphors described above, functionalised analogues have also been developed to address some critical issues in OLEDs, including carrier injection/transport, emission colour tuning and T–T annihilation.

A series of Pt phosphors functionalised with main group elements have been designed and synthesised (Figure 11) (46–49). Some of these might represent the highest Φ_P values ever reported for Pt phosphors, at 0.93 for Pt-PO and 0.95 for Pt-SO₂. The diphenylamino group (NPh₂) has been shown to add HI/HT functionality to Pt-N, while dimesitylboron B(Mes)₂, diphenylphosphoryl (POPh₂) and phenylsulfonyl (SO₂Ph) moieties can provide EI/ET functionality to Pt-B, Pt-PO and Pt-SO₂, respectively (48). It is well known that carrier injection/transport are very important to the EL process. Hence, Pt-B, Pt-N, Pt-PO and Pt-SO₂ exhibit much better EL efficiencies (with CE values of 30.00 cd A⁻¹, 29.74 cd A⁻¹, 22.06 cd A⁻¹ and 19.59 cd A⁻¹, respectively) than do other congeners (8.47 cd A⁻¹ for Pt-Si, 8.49 cd A⁻¹ for Pt-Ge, 11.42 cd A⁻¹ for Pt-O and 16.77 cd A⁻¹ for Pt-S, for example), indicating the great potential of such functionalisation for enhancing the EL performance of Pt phosphors.

Inspired by the success of Pt-B, other efficient Pt phosphors bearing B(Mes)₂ were also prepared (Figure 11). Pt-pyB can show outstanding EL performance with an EQE of 20.9%, a CE of 64.8 cd A⁻¹ and a PE of 79.3 lm W⁻¹ (50). This complex also shows comparable HI/HT and EI/ET properties to those of the analogous Pt-NO and Pt-NB, and has decent efficiencies with an EQE of 10.6%, a CE of 33.2 cd A⁻¹ and a PE of 34.8 lm W⁻¹ (51).

Additionally, these phosphors can exhibit unique colour tuning behaviours. Pt-B (λ_em = 542 nm), Pt-PO (λ_em = 500 nm) and Pt-SO₂ (λ_em = 503 nm) bearing electron-withdrawing main group moieties on the phenyl ring of their ppy ligands show a bathochromic (red-shift) effect in their emission maxima with respect to that of their parent complex Ptppy-1 (λ_em = 486 nm). These results are in conflict with traditional colour tuning theory, which predicts a hypsochromic (blue-shift) effect in the emission wavelength when electron-withdrawing moieties are introduced to the phenyl ring of the ppy ligands (38). Time-dependent density functional theory (TD-DFT) calculations show that the electrons in the metal-to-ligand charge transfer (MLCT) process mainly go to the main group moieties B(Mes)₂, POPh₂ and SO₂Ph in Pt-B, Pt-PO and Pt-SO₂ rather than to the pyridyl ring in the ligand as in their conventional counterparts (38). This means that the electron-withdrawing main group moieties have changed the direction of the MLCT process compared to the unfunctionalised pyridyl ring complexes, an effect which is not seen with other electron-withdrawing groups such as F or CN which cannot host electrons. The stronger electron-withdrawing properties of the main group moities compared to the pyridyl ring in the ligand allows them to stabilise the MLCT states. The phosphorescence from triplet MLCT states in Pt-B, Pt-PO and Pt-SO₂ can therefore be expected to show red-shift. Similar colour tuning behaviour was also observed in fluorenone-based Pt phosphors, indicating the universality of this effect (52). The importance of this new colour tuning strategy lies in that it makes possible the design and synthesis of long wavelength phosphors with EI/ET properties.

The high LUMO levels for Pt-Si, Pt-Ge, Pt-O and Pt-S mean that electrons can leak into the 4,4′-bis-[N-(1-naphthyl)-N-phenylamino]biphenyl (NPB) layer, resulting in blue emission which provides a chance to generate white light by mixing with the green monomeric emission and orange/red excimer emission from the Pt phosphors. This could provide another route towards simple WOLEDs emitting good quality white light. Pt-Ge and Pt-O have been employed in OLEDs with the simple configuration indium tin oxide/NPB /x% Pt: 4,4′-bis(N-carbazolyl)-1,1′-biphenyl/bathocuproine/tris(8-hydroxyquinolinato)aluminium/lithium fluoride/aluminium (ITO/NPB/x% Pt:CBP/BCP/Alq₃/LiF/Al) (49). As expected, these simple devices can emit high quality white light with a blue component from NPB and green and orange/red components from the monomer and excimer of the Pt phosphor, respectively. Pt-Ge based devices can reach an exceptionally high colour rendering index (CRI) of up to ca. 97 at brightness > 15,000 cd m⁻², a correlated colour temperature (CCT) of 4719 K and CIE coordinates of (0.354, 0.360) at a doping level of 10 wt%. A 6 wt% Pt-O doped WOLED produced white light with a CRI of ca. 94, a CCT of 6606 K and a CIE of (0.320, 0.340), perhaps the best quality white light ever produced by WOLEDs, comparable with natural sunlight and therefore most suitable for artificial lighting applications.

Typically, Pt phosphors suffer from efficiency roll-off due to T–T annihilation induced by their inherent square-planar structure geometry favouring strong molecular interaction, as well as their relatively long τ_P (13, 30). This undesired T–T annihilation has been successfully hindered by attaching bulky triphenylamine moieties to the Pt phosphor ligands (Figure 12) (53). Pt-F2TPA has a τ_P of 8.2 μs, and T–T annihilation can only occur at very high current densities of 360 mA cm⁻². Triphenylamine can also enhance its HI/HT characteristics.

4.3 N^N-Coordinating Ligands Based Platinum Phosphors

Chi et al. have synthesised a series of Pt(N^N)₂ phosphors bearing N-heterocycle substituted pyrazole ligands (Figure 13). A device doped with 20 wt% PtNN-1 showed a maximum brightness of 40,973 cd m⁻² at 15 V, a peak EQE of 5.96%, a CE of 19.70 cd A⁻¹ and a PE of 6.44 lm W⁻¹ (54). The optimal device based on PtNN-2 achieved a maximum luminescence of 20,296 cd m⁻² at 16 V, a peak EQE of 5.78%, a CE of 12.19 cd A⁻¹ and a PE of 6.12 lm W⁻¹ (55). Similarly to other congeners, they also undergo self-aggregation at high doping levels in devices.

M. A. Omary et al. reported a turquoise-blue Pt(N^N)₂ complex, PtNN-3 (Figure 13), in which the N^N represents a pyridyltriazolate derivative (56, 57). This complex features intense Pt^…Pt intermolecular interactions due to the strong polarity induced by pyridyltriazolate as well as its square-planar geometry. Generally, interactions among the phosphor molecules will decrease EL efficiency. However, EL efficiencies could be enhanced by increasing the doping level of PtNN-3 and a maximum EQE of 19.7%, a PE of 44.7 lm W⁻¹ and a CE of 62.5 cd A⁻¹ were achieved at a doping level of 65%. This interesting result may be attributed to energy-level matching among the functional layers within the devices to balance the hole/electron ratio and confine the recombination zone in the EML. Furthermore, the short lifetime of PtNN-3 in the solid film may play a critical role in reducing T–T annihilation and thus enhancing the device efficiencies. These results might provide valuable information on how to cope with the problem of intermolecular interactions associated with Pt phosphors.

4.4 Platinum Phosphors with C^C-Coordinating Ligands

Carbene-type ligands have also been employed to prepare novel Pt phosphors. Y. Unger et al. prepared some homoleptic Pt(II) biscarbene complexes which were emissive in the deep blue region (58). The complex PtCC-1 (Figure 14) gave intense photoluminescence at 386 nm with a Φ_P of 0.45 under a nitrogen atmosphere. Due to its emission energy being in the ultraviolet (UV) region, the PtCC-1 phosphor is less suitable for fabricating OLEDs. However, a new series of carbene-type Pt(II) phosphors were subsequently constructed using modified structures (59). The complex PtCC-2 (Figure 14) had a Φ_P of 0.9 under a nitrogen atmosphere, qualifying it as a triplet emitter for EL devices. A device based on PtCC-2 had a maximum brightness of 6750 cd m⁻² and a peak EQE of 6.2% at 13.2 V.

4.5 Platinum Phosphors with Bridged N^C-Coordinating Ligands

Recently, K. Feng et al. reported some interesting Pt phosphors with bridged ligands (PtbNC-1 and PtbNC-2) and their copolymers (PtbNC-P1 and PtbNC-P2) (Figure 15) (60). These Pt complexes exhibit peak emissions around 525 nm with a maximum Φ_P around 0.55 in degassed CH₂Cl₂ and their corresponding polymers display almost identical photoluminescence spectra in PVK films. The investigation of the EL properties of these green-lighting polymers shows that the novel phosphorescent polymers can show a maximum brightness of ca. 1000 cd m⁻² and a peak EQE of 2.5% (61).

Vezzu et al. also synthesised some Pt(N^C)₂ complexes (Figure 15) emitting between 474 nm and 613 nm with Φ_P values ranging from 0.14 to 0.75 (62). Doping 4% PtbNC-3 into a 2,2′,2″-(1,3,5-benzenetriyl)tris(1-phenyl-1-H-benzimidazole) (TPBI)-4,4′,4″-tris(N-carbazolyl)triphenylamine (TCTA) co-host produced a bright green emitting device (λ_max = 512 nm) with a peak EQE of 14.7% and a maximum CE of 50.0 cd A⁻¹. When the current density was increased to 10 mA cm⁻², the brightness rose to 3698 cd m⁻² with significant efficiency roll-off (EQE of 10.6% and CE of 37 cd A⁻²), implying the occurrence of T–T annihilation in the device.

Acknowledgements

The authors thank all coworkers, PhD and BSc students involved in the research. This work was financially supported by Xi’an Jiaotong University (Tengfei Project), a Research Grant from Shaanxi Province (No. 2009JQ2008), the National Natural Science Foundation of China (NSFC) (No. 20902072), the Program for New Century Excellent Talents in University, and the Ministry of Education of China (NECT-09-0651).