High‐Performance Hot‐Exciton OLEDs via Fully Harvesting Triplet Excited States from Both the Exciplex Co‐Host and the TBRb Emitter

Abstract The high‐level reverse intersystem crossing (HL‐RISC, T2 → S1) process from triplet to singlet exciton, namely the “hot exciton” channel, has recently been demonstrated in the traditional fluorescent emitter of TBRb. Although it is a potential pathway to improve the utilization of non‐radiative triplet exciton energy, highly efficient fluorescent organic light emitting diodes (FOLEDs) based on this “hot exciton” channel have not been developed. Herein, high‐efficiency and low‐efficiency roll‐off FOLEDs are achieved through doping TBRb molecules into an energy‐level matched exciplex co‐host. Combining the low‐level RISC (LL‐RISC, EX3 → EX1) process in the exciplex co‐host with the HL‐RISC process of hot excitons in TBRb to fully harvest the triplet energy, a record‐high external quantum efficiency (EQE) of 20.4% is obtained via a proper Dexter energy transfer of triplet excitons, realizing the efficiency breakthrough from fully fluorescent material‐based OLEDs with TBRb as an end emitter. Furthermore, the fingerprint Magneto‐electroluminescence (MEL) as a sensitive measuring tool is employed to visualize the “hot exciton” channel in TBRb, which also directly verifies the effective energy confinement and the full utilization of hot excitons. Obviously, this work paves a promising way for further fabricating high‐efficiency TBRb‐based FOLEDs for lighting and flat‐panel display applications.


Supporting Texts
Text S1: The strength relationship of channels I-III in TBRb-doped devices.
In the cohost-guest system of this work, three channels are co-existing but channels II and III are comparable and both are stronger than channel I, qualitatively.The specific reasons are described as follows.As we can see from Figure 3d, for pure exciplex-based OLEDs (devices A1-A7), the maximum EQE of the exciplex co-host OLEDs is around 10%, that is, channels I and II produce 10% EQE together.Theoretically, channel I has the maximum EQE of 5% when EX1 and EX3 have equal formation rates because the quantity of EX3 is three times that of EX1 according to the spin statistic rule.Thus, we can conclude that channel I and II have 5.0% EQE, respectively, when there is no TBRb dopant in the pure exciplex-based OLEDs.When TBRb is doped into the exciplex co-host of DMAC-DPS: PO-T2T, from Figure 1b we can see that the TBRb-doped OLEDs has the maximum EQE of near 20%.Thus, the EQE enhancement due to the occurrence of channel III of the TBRb dopant is 10% if channel II (LL-RISC, EX1←EX3) is not changed after the TBRb dopant is doped into the exciplex co-host of DMAC-DPS: PO-T2T.In this case, channel III is the strongest one among channels I-III.However, as reported in the literature, channel II will become stronger after TBRb is doped into the exciplex co-host because EX1 states will quickly evolve onto by the S1 excitons via a fast FRET process occurring from the exciplex co-host to the TBRb dopant. [1]Therefore, channel III reduces because the HL-DET process decreases.As a result, channels II and III are comparable and both are stronger than channel I. Some work focus on the dynamics of excitons in "hot exciton" materials when they are used as purely emitting layer or host. [1,2] or "hot exciton" materials, the HL-RISC process from Tn to S1 states is in competition with the IC process from Tn to Tn-1 states.Their rate constants (kHL-RISC and kIC) are inversely proportional to the energy difference (ΔE) between their respective initial and final states.Theoretically, the generation of Tn-1 will be negligible when kHL-RISC is much larger than kIC.However, this condition is difficult to meet, as reported in these articles.On one hand, the Tn-1 (n= 2) excitons can be generated by the IC process due to the relatively large kIC.On the other hand, the generation of T1 excitons is also promoted because of the aggregation of many host matrixes in the emitting layer.In this case, the T1 exciton density will be increased, especially at high current densities.Therefore, it is reasonable that the T1T1A process will occur, as demonstrated in these works.Nevertheless, in our work, the TTA signal of the MEL curves from TBRb-doped devices originates from T2, TBRb instead of T1, TBRb.Firstly, for TBRb, a very small Δ S 1 T 2 (− 0.11ev) and a large Δ T 2 T 1 (1.32 eV) lead to a very large kHL-RISC / kIC.This causes almost all T2, TBRb excitons to be converted into S1, TBRb by the HL-RISC process, rather than to T1, TBRb via the IC channel.Secondly, we doped TBRb with low concentration into a host as an emitting layer, which can further suppress the IC process as demonstrated in the literature. [3]In this case, the generated T1, TBRb excitons is negligibe.Therefore, the TTA signal of the MEL curves does not come from the T1, TBRb but T2, TBRb states.

Text S3: Origin of the EQE decline for the TBRb-doped devices.
As can be seen in Figure S13, the PLQYs of the corresponding films decrease slightly with increasing doping concentration.We guess the weak reduction of PLQY is attributed to the concentration quenching effect. [4]That is, the aggregation of TBRb molecules at relatively large doping concentrations will lead to a trivial decrease in the EL efficiency of the device.However, this is not the major factor leading to the decline in the device EQE.The EQE of an OLED can be expressed as follows: [5] EQE =  ×   ×   ×   , whereγis the balance factor of injected hole and electron charges (γ= 1 for balanced devices),   is the PLQY of film materials,   is the efficiency of radiative exciton production, and   is the light-outcoupling efficiency (20~30%). [6]As the TBRb doping concentration increased from 0.5 to 5 wt%, the PLQY values of the corresponding films decreased by about 9%.Thus, the reduction in EQE caused by the TBRb aggregation should be theoretically less than 2.7%.However, the EQE maximum declined by about 11% from device B1 to B4.Therefore, there are other factors for the EQE decline of the devices under an electrical excitation.We believe that the major cause is the triplet annihilation of TBRb (T2T2A) by analyzing the high-field MEL components from TBRb-doped devices (Figure 4).This is because, as compared to the HL-RISC channel (S1←T2, i.e., one hot exciton T2 produces one emissive singlet S1), T2T2A (T2+T2→S1+S0) is the process in which two hot excitons generate one singlet S1.Thus, in competition with the HL-RISC of hot excitons, the T2T2A process reduces the yield of radiative excitons leading to lower EQE values of the TBRb-doped devices.Taken together, the weak concentration quenching and strong triplet annihilation effects jointly lead to the decline in dopantconcentration-dependent EQEs from TBRb-doped devices.

Text S2 :
Source of the TTA signal for the MEL curves of the devices B 3 -B 4 .

.
Figure S1.The detailed chemical information of organic materials used in the main text.(a) The molecular structures of materials.(b) The HOMO and LUOM energy levels of materials.

Figure S3 .
Figure S3.Schematic diagrams of fingerprint MEL line-shapes corresponding to various microscopic evolution process (including ISC, RISC, SF, and TTA process) of spin-pair states.

Figure S5 .
Figure S5.The MEL response values (B= 9 mT) of exciplex-based devices A1-A7 with different weight ratio at a bias-current of 100 µA.

Figure S7 .
Figure S7.Room temperature current-dependent MEL traces of exciplex-based device A6, and its inset shows the MEL details within the B range of − 9 mT to 9 mT.

Figure S8 .
Figure S8.Superimposed diagram of ISC and RISC line shapes with different linewidths.The MEL curves in this figure are acquired from devices A6 (red), B5 (green) and B2 (blue) at 100 µA, respectively.

Figure S9 .
Figure S9.(a-d) The MEL response curves acquired from TBRb molecules in TBRb-doped devices B1-B4.All curves were obtained by subtracting the MEL data of non-doped devices A6 from the MEL data of TBRb-doped devices B1-B4.

Figure S10 .
Figure S10.Temperature-dependent MEL traces from the non-doped device A6 at a biascurrent of 100 µA.

Figure S12 .
Figure S12.Schematic illustration of energy transfer mechanisms in the TBRb-based devices B3-B4 with higher TBRb dopant contents.