Permissions for Reuse

Phosphorescent organic light-emitting diodes (PHOLEDs) for light-emitting sources have been actively investigated due to their ability to efficiently utilize both singlet and triplet excitons.^1,2 Although internal quantum efficiency of the green or the red organic electrophosphorescent devices reached almost 100%, there are still several obstacles for PHOLEDs to result in high efficiency.^{3,4,5,6,7,8,9} There have been many reports on the efficient blue PHOLEDs. An endothermic or exothermic host-guest energy transfer by using wide band gap host was introduced to get high efficiency in blue PHOLEDs.^3,4,5,6 The efficient blue dopants such as FIrtaz, FIrN4, and Ir(pmb)₃ were also reported in blue PHOLEDs.^7,8,9 On the other side, Adamovich et al. noted the use of carrier blocking layer in OLEDs to prevent electron-hole recombination in the adjacent carrier transporting layer.¹⁰ There have also been a few reports on the effect of doping profile and emission mechanism of green or red OLED structures.^11,12 However, more intensive studies on the blue OLEDs are in need as they required wider band gap materials, both host and dopant, than those of green or red OLEDs.

In this work, important factors for the improvement of blue PHOLED performance have been investigated with stepwise doping profile within the emissive layer (EML). We have clarified the relationship between the doping profile and the blue PHOLED characteristics based on systematic device studies. We obtained highly improved blue OLEDs with a peak efficiency of 10.4% external quantum efficiency (EQE) and 13.3  lm/W power efficiency by controlling doping profile within the EML. These results represent a significant enhancement over the previous reports for the device based on mCP and FIrpic, which has 7.5% EQE and 8.9  lm/W power efficiency.⁴

A series of organic light-emitting diode was made with the structure of indium tin oxide (ITO)/2-TNATA (10  nm)/NPB (40  nm)/EML (30  nm)/Bphen (50  nm)/LiF (1  nm)/Al (120  nm). ITO was cleaned by the standard oxygen plasma treatment. 4,4,4-tris[N-(2-naphthyl)-N-phenyl-amino]triphenylamine (2-TNATA), 4,4-bis[N-(1-nathyl)-N-phenylamino]biphenyl (NPB), and bathocuproine (2, 9-dimethyl-4, 7-diphenyl-1, 10-phenanthroline) (Bphen) were used as the hole injection layer, the hole-transporting layer (HTL), and the electron transporting layer (ETL), respectively. FIrpic was doped in a wide band gap host mCP, which has a large triplet energy of 2.9  eV.⁴

We have investigated the characteristics of uniformly doped device (Device A) by analyzing partially doped devices (devices B–D). Devices A–D were fabricated, as shown in Fig. 1. The blue emitting FIrpic was uniformly doped with a doping ratio of 7% in the whole EML (device A), partially doped in the EML close to the HTL (device B), close to the ETL (device D), and center of the EML (device C). The performances of uniformly or partially doped blue PHOLEDs (devices A–D) are shown in Fig. 2.

Figure 1.

Figure 2.

Electroluminescence spectra of four devices showed similar maximum luminescence wavelength at 462  nm, while devices B and C relatively exhibited low intensity and additional broad peak around 420  nm [Fig. 2(a)]. The hole transporting material, NPB, has much higher carrier (hole) mobility (µ_h=8.8×10⁻⁴  cm²/V  s) than electron mobility of ETL, Bphen (µ_n=3.9×10⁻⁴  cm²/V  s).¹³ Moreover, mCP which is a carbazole based material¹⁴ is also hole-transport-type host. Therefore, most of hole/electron recombination probably occurred at the EML close to the ETL in devices A–D. Hence, devices B and C, which have only mCP with no FIrpic dopant at the ELM close to the ETL, showed additional mCP emission around 420  nm, which may degrade the blue PHOLED performances. We have further confirmed that this additional peak is originated from the mCP emission by examining electroluminescence of dopant-free devices (not shown here). As expected from the above results, device D which has FIrpic dopant at the EML close to the ETL showed very similar device performance to device A and resulted in the highest EQE among partially doped devices B–D [inset of Fig. 2(a)]. By gathering up the results of devices A–D, we brought into conclusion the following: (1) NPB/mCP:FIrpic/Bphen based PHOLEDs are hole carrier dominant emitting system: (2) hole/electron recombination zone of uniformly doped standard device A is positioned in the EML close to the ETL.

We have also found the substantial difference of current density characteristics od devices A–D and that was shown in Fig. 2(b). Compared to the same driving voltage, uniformly doped (A) and partially EML/ETL side doped (D) devices showed higher current density than that of others. By understanding of carrier transport and direct trapping mechanism in these mCP-FIrpic based blue PHOLEDs, higher current density of devices A and D can be explained and a strategy to improve device performance will be proposed based on this explanation. Commonly, carrier injection in OLEDs depends on injection barrier difference from the HTL (or the ETL) and the EML. As shown in Fig. 1, the highest occupied molecular orbital (HOMO) level and the lowest unoccupied molecular orbital (LUMO) level of mCP are 5.9 and 2.4  eV, respectively.⁵ The energy gap of HOMO between NPB and mCP is 0.5  eV, while that of LUMO between mCP and Bphen is 0.6  eV.¹⁵ Therefore, hole and electron carriers have encountered large energy barrier for injection to the EML. On the other hand, direct carrier injections from the HTL and the ETL to the dopant in the EML could also be considered.^12,16 Seeing that the HOMO and the LUMO of the FIrpic dopant are 5.9 and 3.0  eV, respectively, therefore, the carrier injection barriers are 0.5  eV for hole and only 0  eV for electron. From the hole carrier point of view, hole injection barrier from NPB to both mCP and FIrpic is almost the same, so that FIrpic doping ratio does not affect hole injection and transport. However, from the electron carrier point of view, there is an interesting aspect. Due to the lower injection barrier between Bphen and FIrpic than that between Bphen and mCP, most of the electrons might be directly injected to FIrpic dopant from the ETL rather than transferred from Bphen to mCP.¹⁷ This is why doping position and concentration within the EML, in particular, at the EML close to the ETL, greatly influence current density–voltage characteristics. The EML/ETL side doped device D considerably showed higher current density than that of devices B and C and exhibited almost the same as that of standard device A.

In general, the performances of OLEDs could be improved when the hole/electron recombination zone spreads over the whole EML region.¹⁸ Unfortunately, the recombination zone of standard device A is positioned at the EML/ETL side. Therefore, the modification of the hole/electron balance in device A by increase in injection from the ETL to the EML could be one possible approach to enhance the device performance. Moreover, it has been well known that the increase in doping concentration in the EML leads to high current density in EML because of hopping transport between dopants.¹⁷ Hence, the increase in electron injection/transport by controlling a doping profile in the EML can lead effective hole/electron balance and recombination zone shift to result in highly improved efficiency could be expected.

We have examined the device characteristics of uniformly doped devices with varying dopant concentration. The device with the doping concentration of 10% showed higher current density than that of device A with the doping concentration of 7% (not shown here). However, highly doped (10%) device exhibited relatively low EQE than that of the control device, as shown in the inset of Fig. 3. This was attributed to the triplet-triplet annihilation,^19,20,21 the triplet-polaron quenching,²¹ and electric field induced dissociation of excitons,²¹ which occur in the heavily doped EML. Triplet exciton escape from the EML to the adjacent ETL (Bphen) would also be a key quenching process because we did not applied the exciton confinement layer at the EML/ETL interface. Commonly, holes are more mobile than electrons in many OLED materials and the host material mCP is also hole-transport-type material; therefore, exciton quenching at the Bphen layer in our blue OLEDs should be considered.

Figure 3.

To simultaneously overcome these detrimental phosphorescent quenching processes with enhancement of electron injection/transport in blue PHOLEDs, we have introduced stepwise doping profile into the EML (device E). The devices, in which two-thirds of the EML close to the HTL (20  nm) was doped with doping concentration of 7% and one-third of the EML close to the ETL (10  nm) was doped with various doping concentration (x=4%, 10%, and 13 %, respectively), have been prepared. Due to the limited electron injection/transport, lightly doped device at EML/ETL side (x=4%) showed the lowest current density and the worst device performance among stepwise doped devices. However, heavily doped devices at the EML/ETL side (x=10% and 13%) showed enhanced current density and efficiencies than those of 7% doped device A which shows the highest performance among uniformly doped devices, as shown in Fig. 3. High doping concentration at the EML/ETL side enhanced the electron injection by direct injection to the dopant and the mobility by hopping mechanism, and therefore exciton recombination is well balanced. Moreover, increment of electron injection/transport in the EML would lead exciton recombination zone shifts from the EML/ETL side to the center of the EML, which reduces detrimental phosphorescent quenching processes in the heavily doped EML region and/or in the adjacent Bphen layer. Here, the devices with well controlled stepwise doping profile (device E: x=10% and 13%) showed greatly enhanced EL performances. Compared to not well-tuned device E (x=4%), peak EQE of device E (x=13%) is more than 50% higher. [EQEs of device E are (x=4%) are 6.85% and (x=13%) 10.4%]. Moreover, device E (x=13%) showed almost 40% improved power efficiency (13.3  lm/W) compared to that of conventional device A with uniformly 7% doping concentration (9.67  lm/W).

In summary, the charge carrier injection and transport in blue PHOLEDs could be improved by device engineering with doping profile within EML. Stepwise doping profile with high concentration at the EML close to the ETL offered efficient hole/electron recombination balance with reduced triplet quenching processes, therefore gave highly improved OLED performances compared to uniformly doped standard device. We believe that this study provides an in-depth understanding of PHOLEDs and a promising method to enhance the device performances.

This work was supported by the future technology development program of MOCIE/ITEP (2006-10028439, OLED lighting).