Tuning the emission and exciton utilization mechanisms of pyrazine-based multi-carbazole emitters and their use in organic light-emitting diodes

Abstract

Thermally activated delayed fluorescence (TADF) and hot excitons are two distinct exciton harvesting mechanisms that can lead to 100% internal quantum efficiency in organic light-emitting diodes (OLEDs). Herein, we show that with judicious molecular engineering, the resulting structurally similar compounds emit via distinct photophysical mechanisms, which has a direct consequence on the OLED efficiency. When the pyrazine core is substituted with four carbazoles, the molecule 4CzPyz shows TADF in doped PPT film, with Φ_PL of 75%, ΔE_ST of 0.23 eV and τ_d of 150 μs. The device based on 4CzPyz emits in the sky-blue (λ_EL = 486 nm) with an EQE_max of 24.1%. When one carbazole is replaced by an ortho-biphenyl, the ΔE_ST of 3CzBPz increases to 0.29 eV, the Φ_PL decreases to 56%, and the TADF character is largely suppressed in the PPT film. However, a RISC process between higher-lying triplet excited states and the S₁ state is hypothesized to be operational, supported by a combined photophysical and DFT study, to rationalize how the device with 3CzBPz shows an EQE_max of 9.6% (λ_EL = 464 nm), reflecting that greater than 86% of the excitons are converted into light in the OLED. When two ortho-biphenyl groups are connected to the pyrazine core, the ΔE_ST of 2CzBPz is further increased to 0.34 eV while the Φ_PL is reduced to 45% in the PPT film. The DFT and photophysical studies indicate that 2CzBPz should act as a traditional blue fluorescence emitter. The OLED devices with 2CzBPz bear this out and exhibit an EQE_max 3.2% at a λ_EL of 446 nm. These results show how subtle structural changes modulate the efficiency of the triplet exciton harvesting mechanisms and provide new design directions for highly efficient blue emitters for OLEDs.

---

Introduction

Reverse intersystem crossing (RISC), is a photophysical process that converts triplet excitons to singlet excitons.^1–3 RISC is central to the thermally activated delayed fluorescence (TADF) mechanism where the process is endothermic and converts non-emissive triplet excitons to emissive singlet excitons.^4–6 Organic light-emitting diodes (OLEDs) with compounds that emit TADF can achieve 100% internal quantum efficiency (IQE).^7,8 For TADF to be operational, the energy gap between the lowest-lying singlet (S₁) and triplet (T₁) excited states, ΔE_ST, must be sufficiently small, normally taken to be less than 200 meV.^9,10 The most common design rule for TADF emitters includes poorly electronically coupled electron donor (D) and acceptor (A) moieties in order to confine the highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) on the donor and acceptor units, respectively. The resulting small exchange integral induces a small ΔE_ST. Using this molecular design requires very electronically weak donor and acceptor groups in order to achieve high-energy emissive S₁ states for blue TADF emitters. Frequently, the limited choice of these groups leads to compounds that show too delocalized frontier molecular orbitals (FMO), leading to larger ΔE_ST and inefficient RISC.^11–13 Meanwhile, the presence of higher-energy triplet states has been suggested to assist in the harvesting of triplet exciton in the device.^14,15 One possible mechanism requires that the higher-energy triplet states T_n (n > 1) lie essentially degenerate with S₁ yet be much higher in energy than T₁, which renders RISC competitive with IC.^16–18 The nature of the excited states involved in this mechanism usually show mixed locally-excited (LE) and charge transfer (CT) character [so called hybridized local and charge-transfer (HLCT)] that consequently leads to enhanced spin–orbit coupling between S₁ and these higher-lying triplet states.^14,15 The nature of the HLCT state provides a balance between fast radiative transition rates manifested in high Φ_PL and sufficiently small energy gaps between S₁ and T_n (ΔE_S1Tn), facilitating RISC.^16,17

Compounds emitting via either a traditional TADF mechanism, involving RISC between T₁ and S₁ mediated by a small ΔE_ST, or emitting via a mechanism involving RISC from a T_n state to S₁ (n > 1) have been widely explored as emitters for OLEDs (Fig. 1). For TADF emitters based on a twisted D–A structure, the S₁ and T₁ states are usually of CT character, and according to El Sayed's rules, direct RISC from T₁ to S₁ of states with the same orbital symmetry is not allowed, which is reflected in the very low spin–orbit coupling (SOC) between these two states.^19–21 Previous research results have shown that intermediate triplet states between S₁ and T₁ can assist the RISC process, and as the orbital type of T_n usually differs from that of the S₁ state, strong SOC exists.^21–23 Thus, the RISC process is facilitated by spin-vibronic coupling between T₁ and T_n and enhanced SOC between T_n and S₁ when the excited states involved in RISC are energetically closely aligned.²² Our previous work on the pyrazine-based TADF emitter DTCz-Pz also showed that the presence of an intermediate T₂ state of HLCT character that is of different orbital type to S₁ (CT character) can provide an indirect route for RISC to occur despite the relatively large ΔE_ST of 0.27 eV in 2,8-bis(diphenyl-phosphoryl)-dibenzo[b,d]thiophene (PPT).¹² Makhseed et al. recently reported a TADF emitter 4CzPyz where four carbazoles were connected to a pyrazine core, and the multiple carbazoles can present slightly different electronic coupling with the central acceptor, thus providing a dense population of excited states.²⁴4CzPyz exhibited narrowed ΔE_ST of 0.23 eV and shorter delayed lifetime (τ_d) of 144 μs in 10 wt% bis[2-(diphenylphosphino)phenyl]ether oxide (DPEPO) film compared to corresponding values of 0.27 eV and 5.5 ms for DTCz-Pz in 7% PPT film.^12,24 However, the electroluminescence behavior of 4CzPyz was not explored and the increased molecular conjugation caused significantly red-shifted emission as 4CzPyz showed sky-blue emission with λ_PL of 502 nm, while for DTCz-Pz which only contains two tert-butylcarbazoles were connected to pyrazine, the emission is deep blue with λ_PL of 460 nm.^12,24

Fig. 1 Illustration of (a) TADF and (b) T_n state involved mechanisms in OLEDs, ISC and RISC are intersystem crossing and reverse intersystem crossing, respectively. PF, DF, IC, and RIC are prompt fluorescence, delayed fluorescence, internal conversion and reverse internal conversion, respectively.

ortho-Substituted biphenyl (BP) has been used to improve the RISC process by introducing a LE state in the two-step RISC mechanism without substantially increasing the conjugation in the molecule.²⁵ The compound 4mCzBN-BP emits from a CT state [E(S₁) = 2.80 eV] and has a T₁ state of HLCT character [E(T₁) = 2.69 eV] in toluene, both of which are energetically close to the triplet LE state of BP (2.95 eV). As a result, 4mCzBN-BP exhibited faster k_RISC of 2.28 × 10⁶ s⁻¹ in toluene, compared to 1.26 × 10⁶ s⁻¹ for the reference emitter 4mCzBN.²⁵ Inspired by this work, here we modified the reference emitter 4CzPyz by replacing one or two of the carbazole donors with ortho-substituted biphenyl groups and obtained two blue/deep blue emitters, 3CzBPz and 2CzBPz. The three compounds exhibit distinct photophysics and exciton harvesting mechanisms in OLEDs. The S₁ energies of 3CzBPz and 2CzBPz are destabilized to 2.94 eV and 3.00 eV, respectively, compared to 2.88 eV for 4CzPyz in 10 wt% PPT films, while the ΔE_ST values progressively increase from 0.23 eV (4CzPyz) to 0.29 eV (3CzBPz) and 0.34 eV (2CzBPz), respectively. The smaller ΔE_ST for 4CzPyz makes it possible to utilize triplet excitons in OLEDs via TADF; indeed, the device exhibited a maximum external quantum efficiency (EQE_max) of 24.1% and sky-blue emission with λ_EL of 486 nm. The close-to-resonant T₃/S₁ states in 3CzBPz allow the RISC process occurring via the higher-energy T_n channels. The OLED with this compound as the emitter shows pure blue emission with λ_EL of 464 nm and an EQE_max 9.6%. However, the too large ΔE_ST for 2CzBPz precludes either of these two exciton-harvesting mechanisms from being operational and the OLED based on 2CzBPz exhibits an EQE_max of only 3.2% but the device is deep blue with λ_EL of 446 nm.

Results and discussion

Synthesis

The synthesis of the three emitters is outlined in Scheme 1. The key intermediate, tetrachloropyrazine, (4ClPz) was synthesized in 40% yield from dioxopiperazine upon reaction with phosphorus pentachloride.²⁵4ClPz was reacted with excess carbazole under S_NAr conditions to obtain 4CzPyz in a yield of 60%, which is significantly improved compared to the 2.5% previously reported using a Suzuki–Miyaura cross-coupling strategy.²⁴ For the synthesis of 3CzBPz, 4ClPz was converted to the biphenyl-substituted intermediate (3ClBPz) via a Suzuki–Miyaura reaction, which was then reacted with carbazole to obtain 3CzBPz in a 70% yield.²⁶ For the synthesis of 2CzBPz, 4ClPz was reacted first with carbazole to obtain the dicarbazole intermediate 2Cz2ClPz, which was then directly reacted with 2-biphenylboronic acid via a Suzuki–Miyaura cross-coupling reaction to obtain 2Cz2BPz in a yield of 65%.²⁶ The three emitters were characterized by combination of ¹H and ¹³C NMR spectroscopy, X-ray diffraction (XRD), high-resolution mass spectrometry, melting point determination, high performance liquid chromatography (HPLC), and elemental analysis. The three emitters have melting points of over 380 °C and high degradation temperatures (T_d for 5 wt% weight loss), measured by thermogravimetric analysis (TGA), of 371, and 420 °C for 2CzBPz and 4CzPyz (Fig. S7, ESI‡), respectively. For 3CzBPz sublimation was, however, observed during the TGA, and the temperature responsible for 5 wt% weight loss is 407 °C.

Scheme 1 Synthesis route of 4CzPyz, 3CzBPz, and 2CzBPz.

Single crystals of 4CzPyz were grown via slow evaporation of a mixed solution of chloroform/methanol, while single crystals of 3CzBPz and 2CzBPz were obtained following gradient-temperature vacuum sublimation. The structures are illustrated in Fig. 2. The three compounds show similar degrees of twist of the carbazole rings away from the plane of the central pyrazine, 4CzPyz and 3CzBPz being the most similar. In 4CzPyz the carbazole groups are twisted slightly differently depending on which side of the pyrazine ring they are on. In the first independent molecule, one side shows dihedral angles between the pyrazine and the mean plane of the carbazole of 34.67° and 43.18°, the other of 49.68° and 54.20°. The second independent molecule shows a less extreme difference, one side showing angles of 42.91° and 43.45°, the other of 55.09° and 56.73°. For 3CzBPz, the carbazole at the ortho position to the biphenyl is twisted with a dihedral angle of 47.39°, while the dihedral angles for the carbazoles at the meta and para positions are larger at 56.01° and 52.71°, respectively. The dihedral angle between the carbazoles and the pyrazine in 2CzBPz are at the higher end of the range seen, at 55.06°, rather different to those of a previously reported green-emissive pyrazine-containing TADF emitter pDTCz-DPzS [38.20(8)°].⁷ As was the case for the previously reported structure of the DMF solvate of 4CzPyz,²⁷ the crystal structure of 4CzPyz shows several intermolecular C–H⋯π interactions with H⋯centroid distances between 2.53 and 2.67 Å and corresponding C⋯centroid distances between 3.425(9) and 3.593(10) Å (Fig. S8, ESI‡). The structure of 3CzBPz also showed some C–H⋯π interactions, with H⋯centroid distances of 2.69 Å and corresponding C⋯centroid distances of 3.6002(18) Å. This close packing mode is not observed in 2CzBPz.

Fig. 2 Thermal ellipsoid plots of the crystal structures of (a) and (b) the two independent molecules of 4CzPyz, (c) 3CzBPz, and (d) 2CzBPz. Ellipsoids are drawn at the 50% probability level, and minor components of disorder and hydrogens are omitted for clarity.

Theoretical calculations

Optimization of the ground state geometries of the three emitters was carried out by DFT calculation at the PBE0/6-31G(d,p) level of theory, starting from the X-ray crystal structures.^28,29 The Tamm–Dancoff approximation (TDA) to TD-DFT was applied to investigate the character of both the excited singlet and triplet states as well as to estimate ΔE_ST.²⁸ The spin–orbital coupling matrix element (SOCME) values between excited states were obtained using PySOC based on the optimized triplet state geometries.³⁰ The HOMO of 4CzPyz is distributed across the whole molecule while the LUMO is mainly localized on the pyrazine, whereas for 3CzBPz and 2CzBPz, the HOMOs are located on the carbazole moieties and the pyrazine, while LUMOs are delocalized on pyrazine and the biphenyl substituent. The HOMOs for the three emitters are predicted to be around −5.50 eV, while the LUMOs for 3CzBPz and 2CzBPz are destabilized to −1.62 eV and −1.60 eV, respectively, compared to −1.72 eV for 4CzPyz, which hints that the delocalization of the pyrazine weakens the electron-withdrawing character of the acceptor. The TDA-DFT calculations predict the three emitters to have similar, moderately large ΔE_ST values of around 0.40 eV. For 4CzPyz, the S₁ and T₁ are 3.09 eV and 2.56 eV, respectively, and the intermediate T₂ (2.99 eV) is energetically close to S₁ (Fig. 3). Moreover, the nature of T₂ is of HLCT character and different from the CT dominant character of S₁ and T₁, which leads to a much higher SOCME value () between T₂ and S₁ than 2 × 10⁻⁴ cm⁻² for direct SOC between T₁ and S₁ (Fig. S1, ESI‡). The S₁ and T₁ of 3CzBPz are destabilized, respectively, to 3.17 eV and 2.65 eV, both of which show dominant CT character, while the T₂ is 2.99 eV and T₃ is degenerated to S₁ at 3.19 eV, both possessing HLCT character. T₃ and S₁ exhibits the highest SOC with of 1.21 cm⁻² compared to 1.18 cm⁻² and 0.45 cm⁻² for and , respectively (Fig. S2, ESI‡). The high SOC and small energy gap (0.02 eV) between S₁ and T₃ indicate that RISC in 3CzBPz can occur between T₃ and S₁via high-energy triplet state mechanism.¹⁸ The S₁ and T₁ of 2CzBPz are 3.18 eV and 2.67 eV, respectively, both of which show dominant CT character. The T₂ and T₃ of 2CzBPz are 3.01 eV and 3.28 eV, and the energy offset between S₁/T₂ (0.17 eV) or S₁/T₃ (0.10 eV) seems to reduce RISC processes involving these states greatly, with the ΔE_ST also too large, this compound is predicted to be fluorescent.

Fig. 3 Theoretical modelling of the energies of the HOMO/LUMO and the S₁ and T₁ states of 4CzPyz, 3CzBPz, and 2CzBPz in the gas phase and the electron density distribution of the frontier molecular orbitals at PBE0/6-31G(d,p) level (isovalue = 0.02).

Electrochemistry

Cyclic voltammetry (CV) and differential pulse voltammetry (DPV) measurements of 4CzPyz, 3CzBPz, and 2CzBPz were carried out in DCM with tetrabutylammonium hexafluorophosphate (n-Bu₄NPF₆) as the supporting electrolyte to estimate the HOMO and LUMO energies (Fig. 4). The CV traces of 4CzPyz, 3CzBPz, and 2CzBPz show irreversible oxidation waves with E_pas of 1.16 V, 1.08 V and 1.07 V, respectively, which align with the corresponding oxidation potentials obtained from the DPV of E_ox = 1.12 V, 1.07 V, and 1.08 V, respectively. 4CzPyz possesses a pseudo-reversible reduction wave with E_pc of −1.80 V which aligns with the reduction potential E_red = −1.76 V obtained from the DPV, while no reduction waves are observed for 3CzBPz and 2CzBPz within the DCM solvent window. The corresponding HOMO levels of 4CzPyz, 3CzBPz and 2CzBPz inferred from the E_ox values in the DPVs are −5.46 eV, −5.41 eV and −5.42 eV, respectively, which are slightly stabilized than those predicted from the DFT calculations yet match the calculated trend. The HOMO value of 4CzPyz (−5.46 eV) is slightly destabilized compared to the reported value of −5.59 eV in DCM.²⁴ The LUMO level of 4CzPyz is −2.58 eV, which is moderately destabilized compared to the reported value of −2.83 eV in DCM,²⁴ while the LUMO levels for 3CzBPz and 2CzBPz are calculated to be −2.47 eV and −2.45 eV, respectively, where these are inferred from the HOMO energies and the optical gaps (E_g), which themselves were determined from the intersection of the normalized absorption and fluorescence spectra in toluene (2.94 eV and 2.97 eV, Fig. S5d, ESI‡). The LUMOs values of 4CzPyz, 3CzBPz and 2CzBPz are moderately more stabilized than the values obtained from DFT calculations, yet follow the calculated trend.

Fig. 4 CV (solid lines) and DPV (dashed lines) for 4CzPyz, 3CzBPz, and 2CzBPz in degassed DCM solution containing [nBu₄N]PF₆ as the supporting electrolyte and using Fc/Fc⁺ as an external standard (Fc/Fc⁺ = 0.46 eV versus SCE,³¹ scan rate = 100 mV s⁻¹).

Photophysics

The UV-vis absorption spectra of 4CzPyz, 3CzBPz, and 2CzBPz in toluene are shown in Fig. 5. The profiles match closely to the TDA-DFT calculated absorption spectra (Fig. S4, ESI‡). The TDA-DFT calculations predict that the S₁ of the three emitters should exhibit charge-transfer dominated hybrid transitions from HOMO to LUMO (Fig. S1–S3, ESI‡). 4CzPyz, 3CzBPz, and 2CzBPz all exhibit structured absorption around 450–380 nm. 4CzPyz showed the strongest hybrid CT absorption with molar absorptivity (ε) values of 2.4 × 10⁴ M⁻¹ cm⁻¹ (425 nm) and 2.3 × 10⁴ M⁻¹ cm⁻¹ (413 nm), while for 3CzBPz the absorption band is less intense and blue-shifted to 1.8 × 10⁴ M⁻¹ cm⁻¹ (408 nm) and 1.7 × 10⁴ M⁻¹ cm⁻¹ (398 nm). The absorption of 2CzBPz is further attenuated and blue-shifted, with ε of 0.9 × 10⁴ M⁻¹ cm⁻¹ (401 nm) and 1.1 × 10⁴ M⁻¹ cm⁻¹ (388 nm). The tendency of the ε values match the TDA-DFT results where the oscillator strengths (f) reduce from 0.37 for 4CzPyz to 0.22 and 0.05 for 3CzBPz and 2CzBPz, respectively. The three compounds exhibit structureless emission in toluene with λ_PL of 473 nm, 450 nm, and 440 nm for 4CzPyz, 3CzBPz, and 2CzBPz, respectively. The blue-shifted emission of 3CzBPz, and 2CzBPz are ascribed to the weaker character of the acceptors. The three emitters exhibit a moderately high Φ_PL in degassed toluene, of 71%, 65%, and 64% for 4CzPyz, 3CzBPz, and 2CzBPz, respectively, which are slightly reduced to 66%, 63%, and 61% after exposure to oxygen. The S₁ and T₁ values were extracted from the onsets of the fluorescence and phosphorescence spectra, respectively, in 2-methyltetrahydrofuran (2-MeTHF) glass at 77 K (Fig. 5(b)–(d)). At 77 K, the three emitters exhibit broad and featureless fluorescence as well as phosphorescence spectra, which suggest that both of these states possess mainly CT character. The S₁/T₁ states for 4CzPyz, 3CzBPz, and 2CzBPz are measured to be 2.94/2.67 eV, 3.06/2.72 eV, and 3.10/2.72 eV, respectively, which match the tendency of TDA-DFT calculation, and the corresponding ΔE_ST values are 0.27 eV, 0.36 eV, and 0.38 eV, respectively.

Fig. 5 (a) Absorption (solid) and emission (dashed) spectra of 4CzPyz, 3CzBPz, and 2CzBPz in toluene solution (λ_exc = 380 nm), prompt fluorescence (Fl) (1–100 ns) and phosphorescence (Ph) (1–10 ms) spectra of (a) 4CzPyz, (b) 3CzBPz and (c) 2CzBPz at 77 K in 2-MeTHF (λ_exc = 343 nm).

The time-resolved PL decays of these materials were measured in 10⁻⁵ M degassed toluene solution (Fig. 6(a)). 4CzPyz, 3CzBPz, and 2CzBPz exhibit prompt lifetime (τ_p) of 6.4 ns, 4.3 ns, and 6.8 ns, respectively; however, no delayed component was detected. The PL decay and oxygen insensitivity behavior of 4CzPyz, 3CzBPz and 2CzBPz in toluene indicate that in toluene these compounds behave as fluorophores only. As the TDA-DFT calculations indicated that a HLCT triplet exciton harvesting mechanism may be possible in 4CzPyz and 3CzBPz, we examined the effect of solvent polarity on the nature of the excited state (Fig. S5, ESI‡) by undertaking a Lippert–Mataga study for each of the three emitters. All materials showed emission that was insensitive to polarity changes in low-polarity solvents, such as cyclohexane, and showed broadened emission in high-polarity solvents such as acetonitrile and methanol. The behaviour is a hallmark of compounds with low-lying HLCT excited states. The Stokes shift versus the orientation polarizability (Δf) plots as well as the electric dipole moment (μ_e) of three emitters are presented in Fig. 6(b)–(d). Notably, the Lippert–Mataga plots of three emitters each show two regions where there is a linear relationship between solvent polarity and Stokes shift, indicative of compounds possessing excited states of mixed LE/CT character. The Lippert–Mataga plots reveal that the three emitters show increased CT character in high polarity solvents while in low polarity solvents the S₁ states possess more significant LE character. The smaller degree of red-shifting of the emission spectra and smaller μ_e in 2CzBPz suggest that there is greater LE character to the S₁ state in this compound than in 3CzBPz and 4CzPyz.

Fig. 6 (a) transient PL decay curves of three emitters in toluene (λ_exc = 379 nm), and solvatochromic Lippert–Mataga plots for (b) 4CzPyz, c() 3CzBPz, and 2CzBPz.

With a view to using these compounds as emitters in OLEDs, we then investigated the photophysics of the three emitters in solid-state matrices. The photoluminescence spectra of three emitters are insensitive to the medium, with λ_PL of 486 nm, 463 nm, and 446 nm for 4CzPyz, 3CzBPz and 2CzBPz (Fig. 7(a)), which are only modestly red-shifted compared to the corresponding λ_PL of 473 nm, 450 nm, and 440 nm in toluene, respectively. The highest Φ_PL value for 4CzPyz was achieved in 10 wt% doped PPT film at 75%, which is significantly higher than the 36% previously reported in the 10 wt% doped DPEPO film.²⁴ The Φ_PL values for 3CzBPz and 2CzBPz in PPT are 56%, and 38%, respectively, which are slightly lower than corresponding values in DPEPO (Fig. 7(b)). The tendency of the Φ_PL of the three emitters match the TDA-DFT calculated trend in oscillator strength.

Fig. 7 (a) Emission spectra and (b) Φ_PL of 4CzPyz, 3CzBPz, and 2CzBPz in 10 wt% doped hosts (λ_exc = 310 nm).

The three emitters in 10 wt% doped PPT films exhibit unstructured emission, with FWHM of 64 nm, 63 nm and 44 nm for 4CzPyz, 3CzBPz and 2CzBPz, respectively, and the phosphorescence is of dominant CT character with broader (FWHM ∼ 120 nm) and featureless emission (Fig. 7). The calculated ΔE_ST values of the three emitters determined from the onsets of the prompt fluorescence and phosphorescence spectra, are 0.23 eV (same value as previously reported in DPEPO),²⁴ 0.29 eV, and 0.34 eV for 4CzPyz, 3CzBPz and 2CzBPz (Fig. 8), respectively.

Fig. 8 Prompt fluorescence (PF) at 77 K (delay: 1 ns, gate: 100 ns), and phosphorescence spectra at 77 K (delay: 2 ms, gate: 4 ms) of (a) 4CzPyz, (b) 3CzBPz and (c) 2CzBPz in 10 wt% doped PPT film (λ_exc = 343 nm).

The room temperature time-resolved PL decays in 10 wt% doped PPT films are shown in Fig. 9. The prompt fluorescence decays of the three emitters show bi-exponential kinetics with τ_avg of 5.9, 5.1, and 7.6 ns for 4CzPyz, 3CzBPz and 2CzBPz (Fig. 9(a)), respectively. The doped films of 4CzPyz and 3CzBPz show delayed emission, with delayed lifetimes, τ_d (Fig. 9(b)) of 150 μs for 4CzPyz (the lifetime reported in the literature was 174 μs),²⁴ and 929 μs for 3CzBPz. The intensity of the delayed emission increases with increasing temperature (Fig. S6, ESI‡). The smaller ΔE_ST and the temperature dependence of the delayed component from 4CzPyz indicates that the compound in a 10 wt% doped film in PPT shows TADF. The weak TADF of the 3CzBPz doped film in PPT film can also be triggered with the assistance of higher-energy triplet excited states despite the larger ΔE_ST of 0.29 eV. However, there is no indication of TADF in 2CzBPz due to the too large ΔE_ST of 0.34 eV.

Fig. 9 Transient PL decay curves of 4CzPyz, 3CzBPz, and 2CzBPz in 10 wt% doped PPT films in (a) 100 ns time window and (b) 8 ms time window (λ_exc = 378 nm).

Electroluminescence properties

Next, we investigated the electroluminescence properties of these three emitters. Multilayer OLEDs were fabricated using these materials as dopants within the emissive layer, with the following device structures: ITO/HATCN (5 nm)/TAPC (40 nm)/TCTA (10 nm)/mCP (10 nm)/PPT: emitter (10 wt%) (30 nm)/PPT (10 nm)/TmPyPB (40 nm)/LiF (0.8 nm) Al (100 nm). Here, HATCN acts as the hole injection layer, TAPC and TCTA are used as hole transporting layers, and mCP is the exciton blocking layer. PPT was selected as the host material due to the good carrier transport ability and the high Φ_PL values of the three emitters in 10 wt% doped films, and TmPyPB is the electron-transporting material. The performance of the devices is summarized in Fig. 10. The devices based on the three emitters exhibit turn on voltage (V_on) at 3.0 V, and the maximum luminance (Lum_max) exceeded 10⁴ cd m⁻². The OLEDs based on 4CzPyz exhibit sky blue emission with λ_EL of 486 nm, and CIE coordinates of (0.20, 0.36). Thanks to the high Φ_PL and TADF character of 4CzPyz in PPT host (Fig. 7(b) and 9(b)), a maximum EQE (EQE_max) of 24.1% is achieved at 3 cd m⁻², which is much higher than our previously reported pyrazine-based TADF emitter DTCz-Pz (EQE_max: 11.6% at 2 cd m⁻²). The EQE reduced to 7.0% and 2.6% at 100 cd m⁻² and 1000 cd m⁻², respectively. For OLEDs with 3CzBPz, blue emission is achieved with λ_EL of 464 nm and CIE coordinates of (0.16, 0.22). The EQE_max for the device with 3CzBPz reached 9.6% at 1 cd m⁻² and this reduced to 3.1% and 2.0% at 100 cd m⁻² and 1000 cd m⁻², respectively. The severe efficiency roll-off of the devices with 4CzPyz and 3CzBPz are ascribed to the relatively larger ΔE_ST values and thus longer triplet exciton lifetimes, a correlation that has also been noted in previously reported pyrazine- or pyrimidine-based blue emitters (Table S3, ESI‡).^12,32–34

Fig. 10 Optimized devices based on 4CzPyz, 3CzBPz, and 2CzBPz: (a) current density–voltage–luminance characteristics, (b) EQE–luminance characteristics, (c) EL spectra, (d) energy level diagram of the device structure, (e) molecular structures of the materials used in the device.

The EQE is a product of four parameters and can be expressed as:

EQE = γΦ_Fη_S/Tη_out (1)

where γ represents the charge carrier balance of the device, Φ_PL is the photoluminescence quantum efficiency of the emitter, η_S/T accounts for the ratio of excitons that are able to decay radiatively, and η_out is the outcoupling efficiency. Considering the Φ_PL of 56% in 3CzBPz doped PPT film and presuming 20% of typical light outcoupling for OLEDs and 100% for hole–electron combination efficiency, the η_S/T of the 3CzBPz based OLEDs reaches 86%, and the slightly sub-linear relationship between luminance and current density of the device with 3CzBPz (Fig. S7, ESI‡) rules out triplet–triplet annihilation (TTA) as a dominant exciton harvesting process in the OLED.^3,35 The device results are consistent with a higher energy triplet state being involved in the exciton harvesting mechanism. The high EQE_max value also makes 3CzBPz amongst the most efficient HLCT emitters (Table S4, ESI‡). The device based on 2CzBPz exhibited deep blue emission with λ_EL of 446 nm and CIE coordinates of (0.15, 0.12). The EQE_max for the OLED with 2CzBPz is 3.2%. The η_S/T for 2CzBPz device is estimated to be 42%, indicating only a small portion of triplet excitons are utilized in OLEDs based 2CzBPz. The λ_EL of the devices with 4CzPyz, 3CzBPz, and 2CzBPz are progressively blue-shfited at 486 nm, 464 nm, and 446 nm, respectively, values that are almost identical to the corresponding λ_PL in the PPT film (Fig. 7(a) and 10(c)).

Conclusions

We investigated the excited state dynamics, photophysical properties, and exciton utilization mechanism in OLEDs of three pyrazine-based multi-carbazole containing emitters through a combined, detailed theoretical and experimental study. When four carbazoles are connected to the pyrazine core, the emitter 4CzPyz exhibits a moderate ΔE_ST of 0.23 eV, high Φ_PL of 74% and TADF character (τ_p: 5.9 ns, τ_d: 150 μs) in PPT film. The device based on 4CzPyz shows sky-blue emission (λ_EL = 486 nm) with EQE_max of 24.1% thanks to the highly efficient exciton utilization by TADF mechanism. When one carbazole is replaced by an ortho-biphenyl, the ΔE_ST of 3CzBPz is increased to 0.29 eV, Φ_PL decreases to 57%, and the TADF character is lost in PPT film. However, the photophysical properties and DFT calculations support that 3CzBPz can harvest triplet excitons via a higher-energy triplet state channel mechanism. This is reflected also in the OLED performance where the device shows an EQE_max of 9.6% at λ_EL of 464 nm; indeed, 86% of the excitons are estimated to be converted into light in the OLED based on 3CzBPz. When two ortho-biphenyl groups are connected to the pyrazine core, the ΔE_ST of 2CzBPz is further increased to 0.34 eV and Φ_PL is reduced to 45% in PPT film. The calculations and photophysics indicate 2CzBPz to be a deep blue fluorescence emitter, and the OLEDs based on 2CzBPz exhibits an EQE_max 3.2% with λ_EL of 446 nm. These results show the critical role of a suitably aligned LE triplet state in modulating the nature of the exciton harvesting mechanism, leading to dramatically different efficiencies in the OLEDs.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

The St Andrews team would also like to thank EPSRC (EP/P010482/1 and EP/L017008/1) for financial support. EZ-C is a Royal Society Leverhulme Trust Senior Research Fellow (SRF\R1\201089). D.C thanks the China Postdoctoral Science Foundation (No. 2022TQ0227).

References

1. C. M. Marian, Annu. Rev. Phys. Chem., 2020, 72, 617–640.
2. T. J. Penfold, E. Gindensperger, C. Daniel and C. M. Marian, Chem. Rev., 2018, 118, 6975–7025.
3. R. Ieuji, K. Goushi and C. Adachi, Nat. Commun., 2019, 10, 13044.
4. Y.-Z. Shi, H. Wu, K. Wang, J. Yu, X.-M. Ou and X.-H. Zhang, Chem. Sci., 2022, 13, 3625–3651.
5. J. Teng, Y. Wang, C. Chen, J. Teng and Y. Wang, J. Mater. Chem. C, 2020, 8, 11340–11353.
6. X. K. Chen, D. Kim and J. L. Brédas, Acc. Chem. Res., 2018, 51, 2215–2224.
7. P. L. dos Santos, D. Chen, P. Rajamalli, T. Matulaitis, D. B. Cordes, A. M. Z. Slawin, D. Jacquemin, E. Zysman-Colman and I. D. W. Samuel, ACS Appl. Mater. Interfaces, 2019, 11, 45171–45179.
8. D. Sun, E. Duda, X. Fan, R. Saxena, M. Zhang, S. Bagnich, X. Zhang, A. Köhler and E. Zysman-Colman, Adv. Mater., 2022, 34, 2110344.
9. P. Xiao, J. Huang, Y. Yu, J. Yuan, D. Luo, B. Liu and D. Liang, Appl. Sci., 2018, 8, 1–27.
10. F. B. Dias, K. N. Bourdakos, V. Jankus, K. C. Moss, K. T. Kamtekar, V. Bhalla, J. Santos, M. R. Bryce and A. P. Monkman, Adv. Mater., 2013, 25, 3707–3714.
11. Q. Zhang, J. Li, K. Shizu, S. Huang, S. Hirata, C. Adachi, P. C. Adachi, Q. Zhang, J. Li, K. Shizu, S. Hirata and H. Miyazaki, J. Am. Chem. Soc., 2012, 134, 14706–14709.
12. P. Rajamalli, D. Chen, S. M. Suresh, Y. Tsuchiya, C. Adachi and E. Zysman-Colman, Eur. J. Org. Chem., 2021, 2285–2293.
13. L. S. Cui, H. Nomura, Y. Geng, J. U. K. Kim, H. Nakanotani and C. Adachi, Angew. Chem., Int. Ed., 2017, 56, 1571–1575.
14. G. Sun, X. H. Wang, J. Li, B. T. Yang, Y. Gao and Y. Geng, Sci. Rep., 2021, 11, 17686–17688.
15. T. Liu, X. Chen, J. Zhao, W. Wei, Z. Mao, W. Wu, S. Jiao, Y. Liu, Z. Yang and Z. Chi, Chem. Sci., 2021, 12, 5171–5176.
16. D. Hu, L. Yao, B. Yang and Y. Ma, Philos. Trans. R. Soc., A, 2015, 373, 20140318.
17. Y. Xu, P. Xu, D. Hu and Y. Ma, Chem. Soc. Rev., 2021, 50, 1030–1069.
18. N. Sharma, M. Y. Wong, D. Hall, E. Spuling, F. Tenopala-Carmona, A. Privitera, G. Copley, D. B. Cordes, A. M. Z. Slawin, C. Murawski, M. C. Gather, D. Beljonne, Y. Olivier, I. D. W. Samuel and E. Zysman-Colman, J. Mater. Chem. C, 2020, 8, 3773–3783.
19. M. A. El-Sayed, J. Chem. Phys., 1963, 38, 2834–2838.
20. H. Noda, H. Nakanotani and C. Adachi, Sci. Adv., 2018, 4, eaao6910.
21. M. K. Etherington, J. Gibson, H. F. Higginbotham, T. J. Penfold and A. P. Monkman, Nat. Commun., 2016, 7, 13680.
22. H. Noda, X. Chen, H. Nakanotani, T. Hosokai, M. Miyajima, N. Notsuka, Y. Kashima, J. Brédas and C. Adachi, Nat. Mater., 2019, 18, 1084–1090.
23. J. U. Kim, I. S. Park, C. Chan, M. Tanaka, H. Nakanotani and C. Adachi, Nat. Commun., 2020, 11, 1765.
24. L. Salah, M. K. Etherington, A. Shuaib, A. Danos, A. A. Nazeer, B. Ghazal, A. Prlj, A. T. Turley, A. Mallick, P. R. McGonigal, B. F. E. Curchod, A. P. Monkman and S. Makhseed, J. Mater. Chem. C, 2021, 9, 189–198.
25. S. J. Woo, Y. H. Ha, Y. H. Kim and J. J. Kim, J. Mater. Chem. C, 2020, 8, 12075–12084.
26. T. Ishiyama, M. Murata and N. Miyaura, J. Org. Chem., 1995, 60, 7508–7510.
27. L. Salah, M. K. Etherington, A. Shuaib, A. Danos, A. A. Nazeer, B. Ghazal, A. Prlj, A. T. Turley, A. Mallick, P. R. McGonigal, B. F. E. Curchod, A. P. Monkman and S. Makhseed, J. Mater. Chem. C, 2021, 9, 189–198.
28. J. A. Pople, J. S. Binkley and R. Seeger, Int. J. Quantum Chem., 1976, 10, 1–19.
29. C. Adamo, J. Chem. Phys., 1999, 110, 6158–6170.
30. X. Gao, S. Bai, D. Fazzi, T. Niehaus, M. Barbatti and W. Thiel, J. Chem. Theory Comput., 2017, 13, 515–524.
31. N. G. Connelly and W. E. Geiger, Chem. Rev., 1996, 96, 877–910.
32. J. Liu, K. Zhou, D. Wang, C. Deng, K. Duan, Q. Ai and Q. Zhang, Front. Chem., 2019, 7, 1–9.
33. M. Cai, M. Auffray, D. Zhang, Y. Zhang, R. Nagata, Z. Lin, X. Tang, C.-Y. Chan, Y.-T. Lee, T. Huang, X. Song, Y. Tsuchiya, C. Adachi and L. Duan, J. Chem. Eng., 2021, 420, 127591.
34. T. Serevičius, J. Dodonova, R. Skaisgiris, D. Banevičius, K. Kazlauskas, S. Juršėnas and S. Tumkevičius, J. Mater. Chem. C, 2020, 8, 11192–11200.
35. M. K. Manna, S. Shokri, G. P. Wiederrecht, D. J. Gosztola and A. J. L. Ayitou, Chem. Commun., 2018, 54, 5809–5818.

---

Footnotes

† The research data supporting this publication can be accessed at https://doi.org/10.17630/b06557de-c785-4860-bb9e-2d299b2a3eec

‡ Electronic supplementary information (ESI) available. CCDC 2280783–2280785. For ESI and crystallographic data in CIF or other electronic format see DOI: https://doi.org/10.1039/d3tc02463f

---