| To date, organic light-emitting device (OLED) has achieved commercialmanufacture, and study of OLED has entered the stage of “how to ensure highefficiency simultaneously reducing production costs” from the stage of “onlyconcentrating on increasing device efficiency”. However, the study of magnetic fieldeffects (MFEs) in OLED is still on the initial stage. MFEs refer to the changes of thecurrent and electroluminescence (EL) of the non-magnetic OLEDs under the externalmagnetic field. It has aroused extensive research interest since the first report in2003.It has to be mentioned that the MFEs in OLED cannot be explained by any physicalmechanism of MFEs in the inorganic devices. However, clarifying the physicalmechanism behind it will certainly deepen our understanding of organicsemiconductor materials and devices, especially various kinds of spin correlatedelectronic processes. Moreover, it is anticipated that we could develop some new kindof devices that integrate optical, electronic and magnetic properties, after we knowinghow to make full use of the MFEsIn the past several years, for explaining the MFEs, researchers have observedvariety of MFE phenomena, and proposed many models based on their own data, e.g.the electron-hole pair (EHP) model, the bipolaron model, the triplet-carrier (polaron) interaction (TPI) model, and the triplet-triplet annihilation (TTA) model. However,each model can only explain a portion of the MFE phenomena, leading to theextensive debate of the underlying mechanism of MFEs. Noteworthy, the electronicstructures of organic semiconductors are very different with their inorganiccounterpart. In addition, there are so many organic compounds, differing in structuresand elements, various in morphology when they combined to thin solid films.Obviously, any one of these differences could result in the inconformity of theelectronic processes, leading to the dissimilarity of the MFEs.We note that all the previous researches were carried out in the steady-statecondition, i.e., the devices are driven by constant voltage or current. The transientmethod, i.e., the devices are driven by pulsed voltage, had been ignored in the MFEsstudies, even though it is a convenient way to study the kinetic processes of excitonsand carriers in OLED.In this thesis, for the first time, we use the transient method to investigate theMFEs on the EL (Magneto-electroluminescence, MEL) of OLEDs. We fabricatedNPB/Alq3OLEDs, and explored the MFEs on different period of the transient EL. Itwas shown that, in the presence and absence of the external magnetic field, the raisingedge of the transient ELs was overlapped, demonstrating that the mobility of thecarriers was not changed by the magnetic field, for the delayed time (td) from thepulsed voltage arrived the devices to the onset of the EL is the transporting time of thecarriers in the OLED. Thus, the bipolaron model cannot be dominant when both of theelectron and hole are balanced injected into the OLEDs. Furthermore, we added anoffset voltage (larger than the turn-on voltage of the OLEDs) to the pulsed voltage todrive the devices, in which experiment we found that the transient EL raising edges,in the absence and presence of the external magnetic field, were still overlapped. Thisresult revealing that the TPI model was not dominant when the OLED was underbipolar injection, because the triplets already existed in the OLED before the pulsedvoltage arrived. It is worth to note that under the two kinds of driving voltage, the flatperiods and falling edges of transient EL do show MFEs, indicating that the MEL isvery relevant to the existence of the excitons in OLEDs. In other words, the EHP model and TTA model could be applicative for explaining the MEL of the OLEDunder bipolar injection. For this purpose, we investigated the time-resolved MELs ofthe OLED in the pulse-duration period, and modeled the kinetic process of carriersand excitons in the OLED to simulate the experimental data. When we consideredboth of the EHP model and TTA model in the simulation, the calculated results fittedwell with the experimental data, demonstrating that both the EHP model and TTAmodel are accountable for the MELs.When we used the transient method to study the MELs, we also carried outsteady-state experiments to investigate the MELs, and then comparatively studied theMELs in the transient and steady-state experiments. We explored the origin of theMEL and magneto-current (MC). In this work, we fabricated hole-only devices,electron-only devices, and bipolar devices. There was no MC could be detected in thehole-only devices, and very small MC was detected in the electron-only devices.However, the bipolar devices showed common MEL and MC. These results revealedthat whether the MFEs can be detected depends on the bipolar injection of the device,and the MC is originated from the MEL. In addition, we found that the curves of theMELs versus magnetic field in the transient measurements and steady-statemeasurements showed much different behaviors, indicating that the difference indriving condition could affects the MELs. Thus, we changed the duty cycle of thepulsed voltage by turning the pulse width and frequency of the pulsed voltage, andmeasured the MELs in different duty cycle. We found that the MEL increases withthe increasing of duty cycle. When the duty cycle of the pulse voltage approached unit,the value of the MEL in the transient experiments was similar with that in thesteady-state experiments. Analyzing the results we found that, once again, both theEHP model and TTA model were the underlying mechanisms of the MEL, when theOLEDs were under bipolar injection. The differences of the MEL in the two drivingconditions were originated from the different trapping and de-trapping processes.As we have clarified the underlying mechanisms of the MEL, we utilized theMEL as a research tool to explore the physical processes in OLEDs. Firstly, weinvestigated the inter-conversion between singlets and triplets in the charge-transfer (CT) fluorescence-based OLED. It is said that in the CT materials, triplet can beconverted to singlet by the reverse intersystem crossing (RISC), owing to the smallenergy gap between singlet and triplet (ΔEST). However, there is very few evidence ofthe occurrence of RISC in the CT materials. In this work, we used an intramolecularCT molecule, TPA-NZP, to fabricate OLED, and tested the MELs. The MEL isnegative, demonstrating that the RISC process indeed occurred in the OLED.However, according to the non-radiative decay theory, when the ΔESTis small, boththe RISC and intersystem crossing (ISC) can take place efficiently. So the direction ofthe inter-conversion between singlet and triplet depends on which process (the RISCor ISC) is dominant. We investigated the direction of the inter-conversion bymeasuring the MELs as we increased the driving voltage and time. With theincreasing of the driving voltage and time, the MELs increased and turned fromnegative to positive, indicating that the direction of the inter-conversion changed frombackward to forward. The results in this work demonstrated that, in CTfluorescence-based OLED, the inter-conversion between singlet and triplet is adynamic-balance process, for enhancing the triplet harvesting, one should increase theradiative decay rate of singlet and decrease the non-radiative decay rate of the triplet.Using the MEL as a tool, we also studied some other physical processes in OLED.We explored the generation paths of the excited states of the dopant in dye-dopedfluorescent OLED: energy transfer and charge trapping. We selected TPA-NZP as thedopant and Alq3and mCP as the host to fabricate doped OLED, and TPA-NZP-basedand Alq3-based non-doped OLED were also fabricated. Comprehensively analyzingthe MELs of the doped OLEDs and non-doped OLEDs, we found that:1) as theenergy gap of HOMO or LUMO between host and dopant increases, the chargetrapping becomes more dominant,2) even though the energy gap of HOMO orLUMO between host and dopant is small and the absorption spectrum of the dopantoverlapped well with the emission spectrum of the host, the charge trapping processcannot be neglected. We studied the spin configuration of the excited states in the“open shell” molecule based OLED. We used a neutral π radical as emitter tofabricate a new kind of OLED. There is only one electron in the single highest occupied molecular orbit (SOMO), when the molecule is excited, the SOMO is empty.Thus the radiative decay of the excited states of the open shell molecule is totallyspin-allowed, making the up-limited internal quantum efficiency of the open shellmolecule-based OLED be100%. We measured the MEL of the OLED, but no MELcan be detected, revealing that there is neither triplet nor singlet in the OLED,indicating that the spin configuration of the excited states of the open shell moleculeis doublet. This work propose a novel method for achieving100%exciton utilizationin OLED. |
PDF Full Text Request