Effect of a twin-emitter design strategy on a previously reported thermally activated delayed fluorescence organic light-emitting diode

  1. ‡,1 ,
  2. ‡,2 ,
  3. 3 ,
  4. 3 ,
  5. 3 ORCID Logo ,
  6. 4 ,
  7. 4 ORCID Logo ,
  8. 1 ,
  9. 1,5 ,
  10. 5 ,
  11. 4 ORCID Logo ,
  12. 3 ,
  13. 3 ORCID Logo ,
  14. 2,6,§ ORCID Logo and
  15. 1,¶ ORCID Logo
1Organic Semiconductor Centre, EaStCHEM School of Chemistry, University of St Andrews, St Andrews, Fife, KY16 9ST, UK
2Institute of Organic Chemistry, Karlsruhe Institute of Technology (KIT), Fritz-Haber-Weg 6, 76131 Karlsruhe, Germany
3Institute for Chemical Research, Kyoto University, Uji, Kyoto 611-0011, Japan
  1. Corresponding author email
  2. ‡ Equal contributors
§ Fax: (+49)-721-6084-8581; Tel: (+49)-721-6084-2903
¶ Web: http://www.zysman-colman.com; Tel: +44 (0)1334 463826
Associate Editor: P. J. Skabara
Beilstein J. Org. Chem. 2021, 17, 2894–2905. https://doi.org/10.3762/bjoc.17.197
Received 02 Sep 2021, Accepted 19 Nov 2021, Published 08 Dec 2021
A non-peer-reviewed version of this article has been posted as a preprint https://doi.org/10.33774/chemrxiv-2021-sz2gb
Full Research Paper
cc by logo


In this work we showcase the emitter DICzTRZ in which we employed a twin-emitter design of our previously reported material, ICzTRZ. This new system presented a red-shifted emission at 488 nm compared to that of ICzTRZ at 475 nm and showed a comparable photoluminescence quantum yield of 57.1% in a 20 wt % CzSi film versus 63.3% for ICzTRZ. The emitter was then incorporated within a solution-processed organic light-emitting diode that showed a maximum external quantum efficiency of 8.4%, with Commission Internationale de l’Éclairage coordinate of (0.22, 0.47), at 1 mA cm−2.


Organic thermally activated delayed fluorescence (TADF) materials have elicited tremendous excitement as an alternative to phosphorescent complexes in organic light-emitting diodes (OLEDs) because these organic compounds can also achieve a theoretical 100% internal quantum efficiency (IQE) but do not require the use of scarce, noble metals [1,2]. Since the luminescence in an OLED is achieved through the radiative decay of electrically generated excitons, high-efficiency devices must be able to harvest both the 25% singlet and 75% triplet excitons to produce light [3]. Distinct from phosphorescent compounds, TADF molecules harvest triplet excitons by converting them into emissive singlets via a reverse intersystem crossing (RISC) mechanism. This mechanism is operational when the energy gap (ΔEST) between the lowest-lying singlet and triplet excited states (S1 and T1) is sufficiently small and spin-orbit coupling (SOC) is non-negligible [4-7]. This small ΔEST can be achieved by spatially separating the highest occupied molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO), thereby reducing the exchange integral of these two orbitals determining the energies of the S1 and T1 states relative to the ground state. The spatial separation of the HOMO and LUMO on donor and acceptor, respectively, will result in an S1→S0 transition with predominantly charge transfer (CT) character. Highly twisted donor–acceptor architectures are typically employed to realize small ΔEST [4,8]. SOC can be enhanced by ensuring that the nature of the S1 and T1 states is different, for example by additionally involving a third (local) triplet state with different symmetry, because otherwise SOC vanishes when the orbital types for these two states are the same, according to El-Sayed’s rule [9].

Designing a molecule able to achieve RISC and the desired 100% IQE is just the first step toward an efficient OLED since the light needs to escape the device. A device is composed of a stack of several layers of organic semiconductor materials, each possessing different refractive indices, sandwiched between two electrodes. Depending on the angle of emission of the light with respect to the plane of the device, total internal reflection at the organic-glass as well as the glass-air interfaces can occur as can coupling to surface plasmon polaritons (SPP) at the interface with the cathode, all contributing to decreasing the external quantum efficiency (EQE) of the device. A compound will emit light perpendicular to its transition dipole moment (TDM), quantified by the anisotropy factor, a. Controlling the orientation of the TDM to lie horizontally in the film (where 1 − a is the fraction of horizontally aligned TDMs) will lead to a maximized amount of light exiting the device. However, when the transition dipoles of the emitter are randomly oriented then only around 20% of the light can escape the device [10].

Indolocarbazole (ICz)-based emitters have been recently employed in several high-performance and highly horizontally oriented materials. ICz acts as a weak, planar, and rigid donor [11-14]. Examples of compounds incorporating an ICz unit include reports from Xiang et al. with the emitters IndCzpTr-1 and IndCzpTr-2 [11], and Maeng et al. with the emitter TRZ-TPDICz [12] (see Figure 1). In the doped film, IndCzpTr-1 and IndCzpTr-2 present high photoluminescence quantum yields, ΦPL, of 75.2% and 71.9%, respectively, and delayed fluorescence lifetimes, τd, of 25.48 μs and 34.31 μs, respectively. The devices produced with these materials reached maximum external quantum efficiencies (EQEmax) values of 14.5% and 30% at low brightness, but efficiency roll-off was significant, with EQE at 100 cd m−2, EQE100, of 11.0% and 15.3% for the OLEDs with IndCzpTr-1 and IndCzpTr-2, respectively. The addition of two phenyl units on the ICz in TRZ-TPDICz increased the donor strength and led to ΦPL of near unity (96%) and a much shorter τd of 8.57 μs in 20 wt % DBFPO film (DBFPO = 2,8-bis(diphenylphosphine oxide)dibenzofuran). The device made from this material has a very high EQEmax of 30.3%, which decreases to 18.4% at 1000 cd m−2; the use of a stronger donor in TRZ-TPDICz results in a red-shift of the electroluminescence, compared to IndCzpTr-1 and IndCzpTr-2 (the electroluminescence maximum wavelength, λEL of 472 nm and 496 nm for IndCzpTr-1 and IndCzpTr-2, respectively, against λEL of 509 nm for TRZ-TPDICz). In our previous work, we presented the first example of a di-functionalized ICz-based emitter ICzTRZ [13,14], that presented nearly complete horizontal orientation in a wide number of host matrices. The best combination of properties was obtained in mCBP as a host, with the photoluminescence maximum wavelength, λPL of 479 nm, ΦPL of 70%, and a τd of 121.1 μs for the vacuum-deposited doped film. The anisotropy factor (a) in 5 wt % mCBP film is 0.09, indicating a very high degree of horizontal orientation (91%), which together with the high ΦPL led to a high-performing device with EQEmax of 22.1% (Figure 1).


Figure 1: Molecular structures of emitters.

It has been documented in the literature that some multichromophore emitters show enhanced molar extinction coefficients of absorption and high ΦPL [15-18]. This led to OLEDs employing dual or multi emitter-designed compounds to show much improved EQEmax compared to devices with their single-emitter counterparts (Figure 1), albeit with a red-shifted emission [16-19]. The advantages of the dual-emitter design are best illustrated by the cross-comparison of CzTRZ [20,21], a molecule that did not present any TADF and thus the OLED showed a low EQEmax of 5.8%, while the emitter, 33TCzTTrz [22], is TADF and the OLED showed a much superior EQEmax of 25.0%. There is a significant red-shift of the electroluminescence, with λEL going from 449 nm for CzTRZ to 490 nm for 33TCzTTrz.

In this work, we utilized a similar strategy to assess the change in optoelectronic properties and device performance of the compound DICzTRZ (Figure 2) compared to our recently reported ICzTRZ study [14]. We note that the effective doubling of the molecular weight necessitates that we fabricate solution-processed devices. Importantly, solution-processed films tend to present isotropic orientation [10] due to the slower deposition times coupled with higher degree of freedom of movement in the solution, unlike the orientation of the emitter in vacuum-deposited films, which occurs only at the surface of the film where the emitter orientation is then “frozen” into place once additional layers of material have covered it. While this loss of controlled orientation in the solution-processed film is true for small molecules, polymers and other high molecular weight emitters can show at least some degree of orientation in solution-processed films. For instance, Senes et al. [23,24]. showed that the OPVn series of polymers exhibited higher horizontal orientation by increasing the length of the polymer chain, and by extension the molecule. Considering the high degree of horizontal orientation that ICzTRZ already showed in vacuum-deposited films (anisotropy factor of 0.09 in 10 wt % film of mCBP) and the high molecular weight of DICzTRZ, we hypothesized that DICzTRZ may also present horizontal orientation in the film and subsequently improve light outcoupling in the device.


Figure 2: a) Molecular structure and b) optimized DFT-calculated geometry of DICzTRZ. Hydrogen atoms are omitted for clarity.

Results and Discussion


The oxidative coupling conditions for the synthesis of carbazole dimers were initially applied to access the dimer of ICzTRZ [25,26]. Treating ICzTRZ with FeCl3 in dichloromethane (DCM) at room temperature for 12 hours did not lead to any product formation. However, when the temperature was increased to 40 °C, DICzTRZ was formed and was isolated in a yield of 20%, while increasing the temperature to 60 °C resulted in complete consumption of the starting material and DICzTRZ was isolated in 66% yield. The identity and purity of DICzTRZ were determined by a combination of NMR spectroscopy, mass spectrometry, and IR spectroscopy.

Theoretical calculations

Density functional theory (DFT) and time-dependent DFT (TD-DFT) calculations in the gas phase at the PBE0/6-31G(d,p) level reveal the potential of DICzTRZ as a TADF material. The nature of the S1 and T1 states and their corresponding energies were then obtained using the Tamm–Dancoff approximation [27] to TD-DFT (TDA-DFT). DICzTRZ possesses a ΔEST of 0.19 eV, comparable to 0.22 eV obtained for ICzTRZ at the same level of theory. We can observe a slightly stabilized S1 energy of 2.83 eV (2.92 eV for ICzTRZ) and T1 energy of 2.64 eV (2.70 eV for ICzTRZ) [14] compared to those of ICzTRZ. Compared to ICzTRZ, there is a much higher density of intermediate triplet states between S1 and T1, which is expected to enhance the efficiency of the RISC process due to the presence of increased spin-vibronic coupling [28-33]. The permanent dipole moment (PDM) of DICzTRZ is substantial increased to 2.1 Debye compared to that in ICzTRZ (0.3 Debye); however, both the transition dipole moment (TDM) and oscillator strength (f) are slightly smaller (TDM = 7.6 Debye and f = 0.62) than the values calculated for ICzTRZ (TDM = 7.9 Debye and f = 0.72). DICzTRZ shows a shallower HOMO at −5.03 eV, reflective of a certain degree of conjugation between the two indolocarbazole moieties, compared to the HOMO of ICzTRZ (−5.19 eV). The LUMO level remains essentially unchanged (−1.76 eV for DICzTRZ vs −1.75 eV for ICzTRZ) since the electronic environment surrounding the t-Bu-triazine remains essentially unperturbed (Figure 3).


Figure 3: HOMO, HOMO–1 (H–1), LUMO, and LUMO+1 (L+1) electron density distributions (isovalue: 0.02) and energy levels, excited state energy levels.

Optoelectronic properties

The electrochemical properties of the two materials were studied in degassed DCM with tetra-n-butylammonium hexafluorophosphate as the electrolyte and Fc/Fc+ as the internal reference, data are reported versus a saturated calomel electrode (SCE). In both DICzTRZ and ICzTRZ [14] we observed a reversible oxidation wave with respective oxidation potential (Eox) at 0.87 V and 0.96 V vs SCE (Figure 4a). Both compounds also present a second oxidation wave that is more prominent and cathodically shifted for DICzTRZ at 1.05 V, compared to 1.14 V for ICzTRZ. No reduction wave is observed for DICzTRZ. The HOMO value calculated from the oxidation potential obtained from differential pulse voltammetry (DPV), is −5.21 eV, which is stabilized compared to that predicted from DFT (EHOMO: −5.03 eV); however, the less positive oxidation potentials in DICzTRZ versus ICzTRZ does align with the predictions obtained by DFT.


Figure 4: a) Cyclic voltammetry (CV) and differential pulse voltammetry (DPV) of DICzTRZ in DCM (scan rate = 100 mV/s). b) UV–vis absorption spectrum of DICzTRZ in 10−5 M toluene solution. c) Ground and excited state solvatochromism study of DICzTRZ (excitation wavelength, λexc, = 340 nm, * = second harmonic of the excitation source); d) emission spectra of DICzTRZ in 10−5 M toluene solution (cyan), 20 wt % CzSi film (blue), and 10 wt % PMMA film (red), (λexc = 340 nm).

The UV–vis absorption spectrum of DICzTRZ, while slightly red-shifted and with higher molar absorptivity (as was the case for previously published multichromophore materials) [15-17,34,35], coincides closely with the one from ICzTRZ [14] (Figure 4b) and also with other indolocarbazole-based compounds [11]. The nearly identical profile leads us to conclude that the character of the transitions is likely to be very similar to those associated with ICzTRZ. The two absorption bands located between 330 and 350 nm are ascribed to the LE transitions within the diindolocarbazole donor. The two lower energy and lower absorptivity bands at 390 nm and 410 nm are both assigned to CT-type transitions (Figure 4b).

Solvatochromic studies for DICzTRZ show that the PDM of the ground state structure is small and so the absorption spectrum is essentially not affected by changes in polarity, while the excited state shows the characteristic positive solvatochromism associated with an emission from a CT state (λPL going from 462 nm in the least polar methylcyclohexane to 548 nm in the most polar dichloromethane). From the previously calculated HOMO level determined from DPV and the optical gap obtained from the intersection of the normalized absorption and emission spectra in DCM (Egap = 2.71 eV), we were able to obtain a LUMO energy value of −2.50 eV (Figure 4c).

The emission of DICzTRZ in degassed toluene is red-shifted at 477 nm compared to ICzTRZ [14], at 462 nm (Figure 4d). The excitation spectrum mirrors the profile of the UV–vis absorption (Supporting Information File 1, Figure S4a). Transient PL measurements in degassed toluene show mono-exponential prompt and delayed fluorescence decays at 8.94 ns and 28.83 µs, respectively (Supporting Information File 1, Figure S4c,d). After exposure to oxygen, the delayed fluorescence disappears while the prompt decay lifetime, τp, is slightly reduced to 6.80 ns, implying the involvement of triplet states in the emission. When compared to ICzTRZ in degassed toluene, DICzTRZ presents comparable τp (9.0 ns for ICzTRZ), while we observe a substantial one order of magnitude decrease in the delayed lifetime, τd, (229.2 µs for ICzTRZ [14]), reflective of a more efficient RISC process. DICzTRZ is less emissive than ICzTRZPL of 72% [14]), with ΦPL in degassed toluene of 60% that decreases to 44% once exposed to oxygen. This reduction in ΦPL is in part due to the decrease in the radiative decay rate given the smaller calculated oscillator strength for the emissive S1 state for this compound compared to ICzTRZ. The ΔEST of DICzTRZ in toluene glass at 77 K is 0.21 eV (see Supporting Information File 1, Figure S4, which is significantly smaller than the 0.32 eV obtained for ICzTRZ under the same conditions. The T1 levels of both DICzTRZ and ICzTRZ are comparable at 2.59 eV and 2.62 eV, respectively, while the S1 level for DICzTRZ is more stabilized at 2.80 eV vs 2.94 eV for ICzTRZ). We can clearly observe that the phosphorescence spectrum presents a well-defined structure, typical for transitions coming from a local excited (LE) type state on the diindolocarbazole. TDA-DFT calculations in the gas phase predict that the T1 state is CT in nature while the lowest-lying triplet states with LE character are T3 and T4 (T3 and T4 are at 2.73 eV and 2.75 eV, respectively, while T1 is at 2.64 eV, see Supporting Information File 1, Table S1 and Figure S3). The character of the different transitions was also evaluated by analysis of the natural transition orbitals (NTOs) (see Supporting Information File 1, Table S2). The T1 and T2 HONTO and LUNTO (highest occupied and lowest unoccupied natural Transition orbitals) are localized on the central diindolocarbazole and adjacent triazine, respectively, showing a clear CT between donor and acceptor moieties in the molecule. As previously mentioned, T3 and T4 present LE character, with the NTOs localized mainly the central di-indolocarbazole. The character of each of T5 and T6 is more difficult to assign as the electron density of the transition is localized on one of the indolocarbazole-triazine fragments and showing a high degree of overlap between the HONTO and LUNTO, which indicates a transition with a mixed CT and LE character. S1 also presents a clear CT transition from the diindolocarbazole to the triazine.

With a view to incorporating DICzTRZ as the emitter in a solution-processed OLED, we next investigated the photophysical behavior of this compound in solid host matrices. We began with 10 wt % doped film of DICzTRZ in PMMA as the polarity of PMMA emulates well that of toluene [36]. The emission maximum in PMMA is 514 nm (Supporting Information File 1, Figure S5a) with a corresponding ΦPL of 29% under N2. The significantly red-shifted emission in the PMMA film compared to that in toluene coupled with a significantly lower ΦPL is suggestive that aggregation-caused quenching is prevalent in this host matrix. Transient PL measurement (Supporting Information File 1, Figure S5b,c) showed multiexponential decay kinetics and lifetimes with an average τp of 8.6 ns [τ1 = 3.5 ns (37.5%), τ2 = 11.6 ns (62.5%)] and an average τd of 156.1 µs [τ1 = 27.98 µs (39.5%), τ2 = 239.7 µs (60.5%)], respectively. The average prompt fluorescence lifetimes are of a similar magnitude to that of ICzTRZp = 11.5 ns) [14] while the average delayed fluorescence decays much faster for DICzTRZd = 252.8 µs for ICzTRZ) [14]. We next focused on the photophysical study in a suitably high triplet energy small molecule host material, CzSi (9-(4-tert-butylphenyl)-3,6-bis(triphenylsilyl)-9H-carbazole). The emission in CzSi at 488 nm, is only slightly red-shifted compared to that in toluene (Figure 4d). Gratifyingly, the ΦPL is substantially higher at 57% in 20 wt % doped CzSi film, compared to that in the 10 wt % PMMA films (Table 1). In this host, transient PL measurements show the presence of both prompt and delayed fluorescence (Figure 5a,b) with respective average lifetimes of τp of 7.7 ns [τ1 = 3.8 ns (42.5%), τ2 = 10.6 ns (57.6%)] and τd of 69.49 µs [τ1 = 23.07 µs (49.6%), τ2 = 115.2 µs (50.4%)]. While the ΦPL largely benefits from the change in the host, the lifetimes of the prompt fluorescence remain largely unchanged while we observe a much shorter delayed fluorescence. Both prompt and delayed lifetimes of DICzTRZ in CzSi are shorter than those of ICzTRZ in the same host (τp 9.5 ns, τd of 147.3 µs, Supporting Information File 1, Figure S5f,g). The ΔEST values in CzSi (Figure 5d) and PMMA (Supporting Information File 1, Figure S5d), are 0.19 eV and 0.03 eV, respectively. From a cross-comparison of the state energies (Table 1) we can see that the T1 state remains essentially the same regardless of the environment, this due to the LE nature of this excited state. The energy of the S1 state varies with the environments (with energies of 2.94 eV, 2.72 eV, and 2.75 eV for toluene solution, CzSi film and PMMA film, respectively for ICzTRZ), characteristic of a CT type state, but the shape of the spectra in all media adopt a structured profile, typical for LE-type states, suggesting a state of mixed CT and LE character (Supporting Information File 1, Figures S4d and S5d). DICzTRZ and ICzTRZ possess comparable ΔEST in CzSi, at 0.19 eV and 0.16 eV respectively. Temperature-dependent time-resolved PL decays (Figure 5c) reveal the clear increase in the intensity of the delayed emission with higher temperature, a hallmark of TADF.


Figure 5: a) Prompt and b) delayed time-resolved decay in spin-coated 20 wt % CzSi film of DICzTRZexc = 378 nm); c) delayed fluorescence decay data measured at different temperatures in spin-coated 20 wt % CzSi film of DICzTRZexc = 378 nm); d) prompt fluorescence and phosphorescence spectra at 77 K in drop-casted 20 wt % CzSi film (λexc = 343 nm, prompt and delayed fluorescence spectra were obtained in the 1–100 ns and 1–10 ms time range, respectively).

Table 1: Photophysical properties of ICzTRZ[15] and DICzTRZ.

Material Environment λPL / nma ΦPL N2 (air)b / % τp, τdc / ns; μs S1d / eV T1e / eV ΔESTf / eV
ICzTRZg toluene (10−5 M)[15] 462 72 (56)h 9.0; 229.2 2.94 2.62 0.32
  CzSi 20 wt %i 475 63 (50)j 9.5; 147.3 2.72 2.56 0.16
  PMMA 10 wt %i [15] 470 31 (28)j 115; 252.8 2.75 2.64 0.11
DICzTRZk toluene (10−5 M) 477 60 (44)h 8.9; 28.83 2.80 2.59 0.21
  CzSi 20 wt %i 488 57 (42)j 7.7; 69.49 2.63 2.44 0.19
  PMMA 10 wt %i 514 29 (22)j 8.6; 156.1 2.61 2.58 0.03

aMeasured at room temperature; bλexc = 340 nm; cτp (prompt lifetime) and τd (delayed lifetime) were obtained from the transient PL decay of degassed solution/doped film, λexc = 378 nm; dS1 was obtained from the onset of the prompt emission measured at 77 K; eT1 was obtained from the onset of the phosphorescence spectrum measured at 77 K; fΔEST = S1 – T1. gprevious work [14]; hobtained via the optically dilute method [37] (see Supporting Information File 1), quinine sulfate (0.5 M) in H2SO4 (aq) was used as the reference, ΦPL: 54.6% [38], λexc = 360 nm; ispin-coated films; jobtained via integrating sphere; kthis work.

In our previous work, we investigated the orientation of ICzTRZ in a variety of guest–host systems prepared by co-evaporation [13,14]. In all of these systems ICzTRZ presented nearly-completely horizontal orientation with anisotropy values in the range 0.06 to 0.12, depending on the host materials. However, in going from vacuum deposition, which was possible for the low-molecular weight emitter ICzTRZ, toward solution processing required for the much bigger DICzTRZ, one can expect significant changes of the orientation behaviour. It was shown, for example that phosphorescent iridium complexes like Ir(ppy)2(acac) display horizontal orientation (a ≅ 0.25) after vacuum co-evaporation, while the orientation changed toward isotropic in spin-coated films with PMMA as the host [39]. Moreover, upon solution processing with a low-Tg host like CBP, which is prone to crystallization, the obtained emitter orientation even turned vertical with the a factor approaching 0.40 [39].

Thus, measurements of the anisotropy factor were accordingly carried out for both ICzTRZ and DICzTRZ. Polarization and angle dependent luminescence spectroscopy was used to measure a for solution processed films of 20 wt % DICzTRZ in CzSi. The data were then analyzed via optical simulation to yield and anisotropy factor of 0.53, which disappointingly implies that the emitter presents a strongly vertical orientation (Figure 6); the corresponding measurement of 20 wt % ICzTRZ in CzSi is shown in Supporting Information File 1, Figure S6. It also shows vertical emitter orientation (a = 0.50; see fit in Supporting Information File 1, Figure S6).


Figure 6: Angle-resolved photoluminescence measurement of a solution-processed film of 20 wt % DICzTRZ in CzSi. The red line shows a fit using the dipole emission model as described in detail in Supporting Information File 1, yielding an anisotropy factor, a, of 0.53 (data taken at λem = 500 nm).

Clearly, both emitters exhibit unfavourable orientation of their TDMs when processed from solution. As stated above, this change of orientation in relation to the used processing conditions is not unexpected and confirms – once more – that an important driving force for non-isotropic emitter orientation upon vacuum deposition is the non-equilibrium situation at the surface of a growing film, as suggested by the Ediger group [40]. This is not the case for solution processing where molecules in the liquid film can almost freely rotate and adopt a more or less random orientation before the solvent evaporates and their orientation is fixed in the solid film. There may also be some effect of the host on the resulting orientation as well, which seems to be the case here for CzSi where we observe pronounced vertical orientation of both emitters. There is a difficulty in designing host molecules that lead emitters to orient horizontally without sacrificing other preferable properties of the host; for example, high triplet energy, good film-forming ability.

OLED devices

Finally, DICzTRZ and ICzTRZ-based OLEDs were fabricated using the following device structure: ITO (indium tin oxide) (50 nm)/PEDOT:PSS (poly(3,4-ethylenedioxythiophene) polystyrene sulfonate) (35 nm)/PVK (poly(9-vinylcarbazole)) (10 nm)/X wt % DICzTRZ or ICzTRZ: CzSi (20 nm)/PPF (2,8-bis(diphenylphosphoryl)dibenzo[b,d]furan) (5 nm)/TPBi (1,3,5-tris(1-phenyl-1H-benzo[d]imidazol-2-yl)benzene) (50 nm)/Liq (lithium quinolin-8-olate) (1 nm)/Al (80 nm), where X is 20 or 30. The PVK layer is applied to facilitate hole injection from PEDOT:PSS to the emitting layer. Besides, PVK and PPF, possessing high T1 energies of 3.0 eV [41] and 3.1 eV [42], respectively, were inserted to confine the excitons in the emitting layer. PEDOT:PSS, PVK and the emitting layer were fabricated by spin-coating, and the other layers were vacuum-deposited. Device characteristics are shown in Figure 7 for DICzTRZ, Figure S7 (Supporting Information File 1) for ICzTRZ, and the device performance is summarized in Table 2. As shown in Table 2, 20 wt % DICzTRZ-based OLEDs achieved EQEmax of 8.4% and λEL of 494 nm with CIE coordinates (x, y) of (0.22, 0.47) at 1 mA cm−2. The 20 wt % ICzTRZ-based OLEDs exhibited a slightly higher EQEmax of 11.6% and blue-shifted emission with λEL of 485 nm. This result is consistent with that of the photophysical measurements for 20 wt % TADF emitter:CzSi films (ΦPL = 57% and λPL = 488 nm for DICzTRZ, ΦPL = 63% and λPL = 475 nm for ICzTRZ, respectively).


Figure 7: Device characteristics of 20 and 30 wt % DICzTRZ-based OLEDs, which are represented by red and blue, respectively. a) EQE-current density, b) EL spectra and c) current density-voltage-luminance properties.

Table 2: Device performances of X wt % ICzTRZ- and DICzTRZ-based OLEDs (where X = 20, 30).

Emitter Concentration / % EQEmax / % λEL / nma CIE (x, y)
ICzTRZ 20 11.6 485 (0.19, 0.37)
  30 6.6 485 (0.20, 0.39)
DICzTRZ 20 8.4 494 (0.22, 0.47)
  30 5.4 498 (0.22, 0.49)

aDetermined from EL spectrum at 1 mA cm−2.

We next simulated the device EQE (Supporting Information File 1, Figure S8). As shown in Figure 8, with the pre-determined parameters (ΦPL and a) along with the optical constants of the different materials in the OLED stack, we predict the DICzTRZ device to show an EQEmax of between 9–10%, which aligns well with the measured EQEmax, whereas the corresponding solution-processed ICzTRZ OLED (see Supporting Information File 1, Figure S7 for experimental data) is expected to show an EQEmax of about 11%, again in fairly good agreement with the measured results. The simulation also demonstrates that for ICzTRZ with the typically obtained alignment factor of about 0.1 in an evaporated device (and a slightly higher ΦPL of 70% as documented in ref. [14], a device EQE of about 22% can be expected as a result of the horizontal orientation of the emitter within an evaporated EML. Thus, vacuum deposition of this family of TADF emitters is clearly superior to solution processing.


Figure 8: Device efficiency simulation of the fabricated OLEDs depicting the variation in EQE with varied PL quantum yield (vertical axis) and anisotropy factor (horizontal axis). The predicted EQEs are indicated with pink and blue rectangular marks for the DICzTRZ and ICzTRZ solution-processed OLEDs, respectively. The white star shows the predicted EQE for an evaporated ICzTRZ OLED with the orientation and PLQY taken from ref. [14]. All of the simulated EQEs agree fairly well with the experimental results.


Building upon our previously reported emitter, ICzTRZ, here we presented a dual emitter strategy consisting of two ICzTRZ moieties covalently linked together in the form of DICzTRZ. DFT calculations showed a much larger density of triplet states, which suggests that RISC will be faster in this compound compared to its parent. The twin design strategy leads to an enhancement in the molar extinction coefficient of the low-lying CT states, accompanied by a red-shift in the emission. The 20 wt % doped CzSi film of DICzTRZ emits in the blue at 488 nm and shows a photoluminescence quantum yield of 57.1%. The ΦPL of DICzTRZ is slightly lower than that of ICzTRZ (63% under N2 [14]), in line with its lower computed oscillator strength. DICzTRZ shows both prompt and delayed fluorescence, with a τp that remains largely unchanged from that of ICzTRZ, while its τd is significantly shorter. Unfortunately, the TDM of this material is not preferentially horizontally oriented in the solution-processed film, which is not unexpected in solution-processed films. The combination of its lower ΦPL and the vertical orientation of its TDM are the primary factors governing the relatively poorer device performance, with an EQEmax of 8.4%, compared to the vacuum-deposited OLED with ICzTRZ [14].

Supporting Information

The research data supporting this publication can be accessed at doi:10.17630/4a01d3e3-71bc-4ebb-9812-c4b838e13573.

Supporting Information File 1: Synthesis protocols, NMR spectra, supplementary photophysical measurements, computational data obtained from DFT and TD-DFT and electroluminescence data.
Format: PDF Size: 3.7 MB Download
Supporting Information File 2: xyz Coordinates corresponding to the ground state optimized geometry of DICzTRZ.
Format: XYZ Size: 14.8 KB Download


We thank EU Horizon 2020 Grant Agreement No. 812872 (TADFlife) for funding this project. Further support was obtained by the Helmholtz Association Program at the Karlsruhe Institute of Technology (KIT). The German Research Foundation (formally Deutsche Forschungsgemeinschaft DFG) in the framework of SFB1176 Cooperative Research Centre "Molecular Structuring of Soft Matter" (CRC1176, A4, B3, C2, C6) and the cluster 3D Matter Made To Order all funded under Germany’s Excellence Strategy -2082/1--390761711 are greatly acknowledged for financial contributions. We acknowledge support from the Engineering and Physical Sciences Research Council of the UK (grant EP/P010482/1), from the International Collaborative Research Program of Institute for Chemical Research, Kyoto University (grant # 2020-37 and 2021-37), and from JSPS KAKENHI Grant Number JP20H05840 (Grant-in-Aid for Transformative Research Areas, “Dynamic Exciton”). ZZ acknowledges the financial support from the China Scholarship Council (CSC, 201606890009) for his PhD studies. EZ-C is a Royal Society Leverhulme Trust Senior Research fellow (SRF\R1\201089).


  1. Liu, Y.; Li, C.; Ren, Z.; Yan, S.; Bryce, M. R. Nat. Rev. Mater. 2018, 3, 18020. doi:10.1038/natrevmats.2018.20
    Return to citation in text: [1]
  2. Wong, M. Y.; Zysman-Colman, E. Adv. Mater. (Weinheim, Ger.) 2017, 29, 1605444. doi:10.1002/adma.201605444
    Return to citation in text: [1]
  3. Baldo, M. A.; O'Brien, D. F.; You, Y.; Shoustikov, A.; Sibley, S.; Thompson, M. E.; Forrest, S. R. Nature 1998, 395, 151–154. doi:10.1038/25954
    Return to citation in text: [1]
  4. Milián-Medina, B.; Gierschner, J. Org. Electron. 2012, 13, 985–991. doi:10.1016/j.orgel.2012.02.010
    Return to citation in text: [1] [2]
  5. Cui, L.-S.; Nomura, H.; Geng, Y.; Kim, J. U.; Nakanotani, H.; Adachi, C. Angew. Chem., Int. Ed. 2017, 56, 1571–1575. doi:10.1002/anie.201609459
    Return to citation in text: [1]
  6. Rajamalli, P.; Chen, D.; Li, W.; Samuel, I. D. W.; Cordes, D. B.; Slawin, A. M. Z.; Zysman-Colman, E. J. Mater. Chem. C 2019, 7, 6664–6671. doi:10.1039/c9tc01498e
    Return to citation in text: [1]
  7. dos Santos, P. L.; Chen, D.; Rajamalli, P.; Matulaitis, T.; Cordes, D. B.; Slawin, A. M. Z.; Jacquemin, D.; Zysman-Colman, E.; Samuel, I. D. W. ACS Appl. Mater. Interfaces 2019, 11, 45171–45179. doi:10.1021/acsami.9b16952
    Return to citation in text: [1]
  8. Dias, F. B.; Penfold, T. J.; Monkman, A. P. Methods Appl. Fluoresc. 2017, 5, 012001. doi:10.1088/2050-6120/aa537e
    Return to citation in text: [1]
  9. El-Sayed, M. A. Acc. Chem. Res. 1968, 1, 8–16. doi:10.1021/ar50001a002
    Return to citation in text: [1]
  10. Schmidt, T. D.; Lampe, T.; Sylvinson M. R., D.; Djurovich, P. I.; Thompson, M. E.; Brütting, W. Phys. Rev. Appl. 2017, 8, 037001. doi:10.1103/physrevapplied.8.037001
    Return to citation in text: [1] [2]
  11. Xiang, S.; Lv, X.; Sun, S.; Zhang, Q.; Huang, Z.; Guo, R.; Gu, H.; Liu, S.; Wang, L. J. Mater. Chem. C 2018, 6, 5812–5820. doi:10.1039/c8tc01419a
    Return to citation in text: [1] [2] [3]
  12. Maeng, J. H.; Ahn, D. H.; Lee, H.; Jung, Y. H.; Karthik, D.; Lee, J. Y.; Kwon, J. H. Dyes Pigm. 2020, 180, 108485. doi:10.1016/j.dyepig.2020.108485
    Return to citation in text: [1] [2]
  13. Naqvi, B. A.; Schmid, M.; Crovini, E.; Sahay, P.; Naujoks, T.; Rodella, F.; Zhang, Z.; Strohriegl, P.; Bräse, S.; Zysman-Colman, E.; Brütting, W. Front. Chem. (Lausanne, Switz.) 2020, 8, 750. doi:10.3389/fchem.2020.00750
    Return to citation in text: [1] [2] [3]
  14. Zhang, Z.; Crovini, E.; dos Santos, P. L.; Naqvi, B. A.; Cordes, D. B.; Slawin, A. M. Z.; Sahay, P.; Brütting, W.; Samuel, I. D. W.; Bräse, S.; Zysman‐Colman, E. Adv. Opt. Mater. 2020, 8, 2001354. doi:10.1002/adom.202001354
    Return to citation in text: [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17]
  15. Park, H.-J.; Han, S. H.; Lee, J. Y. J. Mater. Chem. C 2017, 5, 12143–12150. doi:10.1039/c7tc03133e
    Return to citation in text: [1] [2] [3] [4] [5]
  16. Cho, Y. J.; Jeon, S. K.; Chin, B. D.; Yu, E.; Lee, J. Y. Angew. Chem., Int. Ed. 2015, 54, 5201–5204. doi:10.1002/anie.201412107
    Return to citation in text: [1] [2] [3]
  17. Kim, M.; Jeon, S. K.; Hwang, S.-H.; Lee, S.-S.; Yu, E.; Lee, J. Y. Chem. Commun. 2016, 52, 339–342. doi:10.1039/c5cc07999c
    Return to citation in text: [1] [2] [3]
  18. Wei, D.; Ni, F.; Wu, Z.; Zhu, Z.; Zou, Y.; Zheng, K.; Chen, Z.; Ma, D.; Yang, C. J. Mater. Chem. C 2018, 6, 11615–11621. doi:10.1039/c8tc02849d
    Return to citation in text: [1] [2]
  19. Chen, D.; Kusakabe, Y.; Ren, Y.; Sun, D.; Rajamalli, P.; Wada, Y.; Suzuki, K.; Kaji, H.; Zysman-Colman, E. J. Org. Chem. 2021, 86, 11531–11544. doi:10.1021/acs.joc.1c01101
    Return to citation in text: [1]
  20. Byeon, S. Y.; Kim, J.; Lee, D. R.; Han, S. H.; Forrest, S. R.; Lee, J. Y. Adv. Opt. Mater. 2018, 6, 1701340. doi:10.1002/adom.201701340
    Return to citation in text: [1]
  21. Sharma, N.; Spuling, E.; Mattern, C. M.; Li, W.; Fuhr, O.; Tsuchiya, Y.; Adachi, C.; Bräse, S.; Samuel, I. D. W.; Zysman-Colman, E. Chem. Sci. 2019, 10, 6689–6696. doi:10.1039/c9sc01821b
    Return to citation in text: [1]
  22. Kim, M.; Jeon, S. K.; Hwang, S.-H.; Lee, S.-S.; Yu, E.; Lee, J. Y. J. Phys. Chem. C 2016, 120, 2485–2493. doi:10.1021/acs.jpcc.5b09114
    Return to citation in text: [1]
  23. Senes, A.; Meskers, S. C. J.; Dijkstra, W. M.; van Franeker, J. J.; Altazin, S.; Wilson, J. S.; Janssen, R. A. J. J. Mater. Chem. C 2016, 4, 6302–6308. doi:10.1039/c5tc03481g
    Return to citation in text: [1]
  24. Senes, A.; Meskers, S. C. J.; Greiner, H.; Suzuki, K.; Kaji, H.; Adachi, C.; Wilson, J. S.; Janssen, R. A. J. J. Mater. Chem. C 2017, 5, 6555–6562. doi:10.1039/c7tc01568b
    Return to citation in text: [1]
  25. Gao, Y.; Hlil, A.; Wang, J.; Chen, K.; Hay, A. S. Macromolecules 2007, 40, 4744–4746. doi:10.1021/ma0702750
    Return to citation in text: [1]
  26. Zhang, Q.; Zhuang, H.; He, J.; Xia, S.; Li, H.; Li, N.; Xu, Q.; Lu, J. J. Mater. Chem. C 2015, 3, 6778–6785. doi:10.1039/c5tc00839e
    Return to citation in text: [1]
  27. Grimme, S. Chem. Phys. Lett. 1996, 259, 128–137. doi:10.1016/0009-2614(96)00722-1
    Return to citation in text: [1]
  28. Santos, P. L.; Ward, J. S.; Data, P.; Batsanov, A. S.; Bryce, M. R.; Dias, F. B.; Monkman, A. P. J. Mater. Chem. C 2016, 4, 3815–3824. doi:10.1039/c5tc03849a
    Return to citation in text: [1]
  29. Hosokai, T.; Matsuzaki, H.; Nakanotani, H.; Tokumaru, K.; Tsutsui, T.; Furube, A.; Nasu, K.; Nomura, H.; Yahiro, M.; Adachi, C. Sci. Adv. 2017, 3, e1603282. doi:10.1126/sciadv.1603282
    Return to citation in text: [1]
  30. Noda, H.; Nakanotani, H.; Adachi, C. Sci. Adv. 2018, 4, eaao6910. doi:10.1126/sciadv.aao6910
    Return to citation in text: [1]
  31. Samanta, P. K.; Kim, D.; Coropceanu, V.; Brédas, J.-L. J. Am. Chem. Soc. 2017, 139, 4042–4051. doi:10.1021/jacs.6b12124
    Return to citation in text: [1]
  32. Wada, Y.; Nakagawa, H.; Matsumoto, S.; Wakisaka, Y.; Kaji, H. ChemRxiv 2019, (ii), 1–21. doi:10.26434/chemrxiv.9745289.v1
    Return to citation in text: [1]
  33. Wada, Y.; Nakagawa, H.; Matsumoto, S.; Wakisaka, Y.; Kaji, H. Nat. Photonics 2020, 14, 643–649. doi:10.1038/s41566-020-0667-0
    Return to citation in text: [1]
  34. Cha, J.-R.; Lee, C. W.; Lee, J. Y.; Gong, M.-S. Dyes Pigm. 2016, 134, 562–568. doi:10.1016/j.dyepig.2016.08.023
    Return to citation in text: [1]
  35. Cho, Y. J.; Jeon, S. K.; Lee, S.-S.; Yu, E.; Lee, J. Y. Chem. Mater. 2016, 28, 5400–5405. doi:10.1021/acs.chemmater.6b01484
    Return to citation in text: [1]
  36. Wang, X.; Yan, Q.; Chu, P.; Luo, Y.; Zhang, Z.; Wu, S.; Wang, L.; Zhang, Q. J. Lumin. 2011, 131, 1719–1723. doi:10.1016/j.jlumin.2011.03.061
    Return to citation in text: [1]
  37. Crosby, G. A.; Demas, J. N. J. Phys. Chem. 1971, 75, 991–1024. doi:10.1021/j100678a001
    Return to citation in text: [1]
  38. Melhuish, W. H. J. Phys. Chem. 1961, 65, 229–235. doi:10.1021/j100820a009
    Return to citation in text: [1]
  39. Lampe, T.; Schmidt, T. D.; Jurow, M. J.; Djurovich, P. I.; Thompson, M. E.; Brütting, W. Chem. Mater. 2016, 28, 712–715. doi:10.1021/acs.chemmater.5b04607
    Return to citation in text: [1] [2]
  40. Ediger, M. D.; de Pablo, J.; Yu, L. Acc. Chem. Res. 2019, 52, 407–414. doi:10.1021/acs.accounts.8b00513
    Return to citation in text: [1]
  41. Kumar, M.; Pereira, L. ACS Omega 2020, 5, 2196–2204. doi:10.1021/acsomega.9b03253
    Return to citation in text: [1]
  42. Vecchi, P. A.; Padmaperuma, A. B.; Qiao, H.; Sapochak, L. S.; Burrows, P. E. Org. Lett. 2006, 8, 4211–4214. doi:10.1021/ol0614121
    Return to citation in text: [1]
Other Beilstein-Institut Open Science Activities