Study On Magneto-electroluminescence Of Charge Transfer States In Organic Light Emitting Diodes

Organic light-emitting diodes (OLED) have great potential in the flat-paneldisplays and solid-state lighting. Although many research institutions and companieshave invested heavily on the development of OLED, the industrializations are farbeyond our expectations because of some unresolved issues, e.g., the lowpower-conversion efficiency and the poor device stability. Understandings ofelectronic structure characteristics in organic semiconductors and excitonic processesin organic optoelectronic devices should have important meanings for furtherenhancing the power-conversion efficiency and stability of OLED.Organic magnetic field effects (OMFE) in ordinary OLED without any magneticcomponent have recently become a hot spot in physics of organic semiconductors.When the organic optoelectronic devices are submitted to the external magnetic field(about tens of millitesla), the electroluminescence (EL) intensity and current can belargely changed (from several percents to tens percents). These features of OMFE notonly make it valuable in applications such as Information Storage, Sensor, andPen-based Device, but also to be an effective tool in investigating the charge transfer,excitonic processes, and spin polarization in organic semiconductors. Despite greatprogress has been made in OMFE in recent years, there are still many scientific issues needed to be clarified. First, the underlying mechanisms of OMFE are quite debatable.Second, the exploration of OMFE in revealing the “useful” and “useless” processesfor power-conversion efficiency does not arouse enough attention.Currently, the underlying mechanisms of OMFE are mainly based on three models:“electron-hole pair model”,“bipolaron pair model”, and “exciton-polaron interactionmodel”. One of biggest arguments among these three models is, what does themagnetic field affect: the charge carrier mobility or excitonic processes? The holesand electrons of exciplex (or called intermolecular charge-transfer state) localize ontwo neighboring molecules, and its emission comes from the direct recombination ofintermolecular electron-hole pairs. Thus, the exciplex becomes an ideal object for“electron-hole pair model”. Results show that the magneto-electroluminescence(MEL) of exciplex-based device is larger than that of exciton-based device by a factorof2.2, and its line-shape can be fitted by non-Lorentzian law very well, providingdirect evidence for “electron-hole pair model”. Moreover, the transient EL with andwithout magnetic fields show that the onset of the fast rising edges of EL pulsesoverlap perfectly, while the falling edge of EL pulses separated, confirming themagnetic field has no effect on the charge mobility but on the charge recombinationprocess, implying the charge mobility-related mechanisms may be less dominantabove the turn-on voltage.Based on the clarification of physical mechanisms, we can use the MEL as a toolto investigate the interconversion between singlet and triplet in OLED, and thenreveal the factors limiting the EL efficiency of exciplex. In exciplex, the distancebetween holes and electrons is relative large, leading to very small spin-exchangeenergy between singlet and triplet. Thus, the conversion from singlet to triplet (S�'T)should be efficient. By comparing the MEL between m-MTDATA: Alq3exciplex-based device and NPB/Alq3exciton-based device, m-MTDATA: Alq3exciplex exhibits a positive MEL effect, and its amplitude is larger than that ofNPB/Alq3exciton-based device by a factor of3.2, indicating more efficient S�'Tconversion in m-MTDATA: Alq3exciplex. This S�'T conversion may cause thereduction of singlet exciton, which is an important factor lowering the EL efficiency of exciplex. The MEL results also tell us suppressing S�'T conversion is veryimportant for further improving the OLED power conversion efficiency.Besides suppressing the S�'T conversion, realizing the up-conversion of T�'S isalso critical in utilization of triplet energy in fluorescent OLED. Generally, there aretwo approaches for T�'S up-conversion: Reverse Inter-system Crossing (RISC) andTriplet-Triplet Annihilation (TTA). Because both RISC and TTA could generatedelayed fluorescence, it is not an easy task to discern them merely from the traditionaltime-resolved spectra. We note that, RISC and TTA are highly spin-dependentprocesses that can generate distinct MEL responses. In DCJTB device where the CTspecies dominate the light emission, the RISC-mediated MEL shows a rapid decreasein the low-field regime (<40mT) and then tend to be saturated in the high-fieldregime (>40mT). But for the Rubrene device, the TTA-mediated MEL is comprisedof a small increase within the field range of20mT followed by a remarkable decreaseat higher field. Our studies indicate the MEL could serve as a “finger-print” forexploiting the RISC and TTA, and other spin-dependent mutual conversion in OLED.In the pursuit of high power conversion efficiency of OLED, reducing theefficiency roll-off is another key point researchers are thinking about. The balancedcharge carrier injection and transport is previously considered to suppressing theefficiency roll-off. To achieve this goal, optimizing the device structure andimproving the fabrication technique were generally proposed. Different to these twomethods, here we report a new type of thermally activated delay fluorescence-basedof dithienylbenzothiadiazole (red-1b) which combines RISC and TTA up-conversion.In its device performance, high exciton utilization of31.25%and relative lowefficiency roll-off (only~27%efficiency drops when brightness reaches to~8000cd/m2) were realized. The Gaussian09calculated frontier orbital energy levels andthe solvent effects suggest the intra-molecular charge-transfer (ICT) characteristic ofred-1b which is favorable for RISC. In addition, the calculated T1-Tn transition showsthe singlet and triplet energetically satisfy S1<2T1<Tn, indicating efficient singletproduction via TTA. Finally, the MEL measurements demonstrate the co-existence ofRISC and TTA, and the contributions of RISC to high exciton utilization under low current density, and TTA to suppression of efficiency roll-off under high currentdensity, respectively.