The effect of bipolar charge transport of derivatives of 1-phenyl-1H-benzo[d]imidazole with horizontal molecular orientation on the performance of OLEDs based on thermally activated delayed fluorescence or phosphorescence

Abstract

The synthesis and properties of two derivatives of 1-phenyl-1H-benzo[d]imidazole with differing numbers of tert-butylcarbazole electron-donating moieties are reported. The compounds exhibit high thermal stability, with 5% weight loss temperatures exceeding 341 °C and glass transition temperatures of over 149 °C. They display moderate triplet energies of 2.63 and 2.66 eV. The synthesized compounds were employed as host materials in phosphorescence and TADF-based organic light-emitting diodes (OLEDs). An investigation of the angle-dependent emission intensity of light-emitting layers containing the phosphorescent emitter Ir(ppy)₂(acac) doped into the examined compounds revealed a notably high internal outcoupling efficiency in OLEDs, exceeding 30%. This efficiency is attributed to the significant horizontal molecular orientation factor, reaching up to 87%. Based on the characterization of the hosting properties of 1-phenyl-1H-benzo[d]imidazole derivatives, the most significant influence on device performance is attributed to their charge-transporting properties. An OLED with the phosphorescent emitter Ir(ppy)₂(acac) and a host material exhibiting bipolar charge transport demonstrated an external quantum efficiency of 13%. Additionally, the picric acid sensitivity of one of the compounds was examined. Triplet-facilitated emission was completely quenched upon the addition of a nitroaromatic explosive as a guest in a film.

---

1. Introduction

The technology of organic light emitting diodes (OLEDs) is undergoing rapid development owing to the advantages of color quality, response time and lower energy consumption of OLEDs when compared to competing devices in the market of display and illumination technology.¹ The singlet and triplet excitons are formed upon applying electricity in the ratio of 1 : 3 due to the spin statistics.^2,3 As a result, prompt fluorescent emitters have a limit of internal quantum efficiency (IQE) of 25%.⁴ The main path to increase the efficiency of OLEDs is utilization of the triplet excited states in electroluminescence (EL), potentially increasing the IQE value to 100%. In phosphorescence OLEDs (PhOLEDs), heavy metal complexes (such as iridium or platinum complexes) which facilitate spin–orbit coupling and in turn accelerate the radiative deactivation through phosphorescence are utilized. Despite the increase in performance of the devices, the use of heavy metal atoms brings certain problems, such as high production costs and environmental concerns.^5–7 PhOLEDs have been developed to date, with reported external quantum efficiency (EQE) values of over 30%.⁸

OLEDs exploiting thermally activated delayed fluorescence (TADF) emerged as a promising alternative to PhOLEDs.⁹ TADF is a delayed fluorescence of up-converted triplet excitons.¹⁰ Thermal motion of atoms facilitates the reverse intersystem crossing (RISC) making delayed fluorescence possible.¹¹ The energy splitting between the first singlet and triplet excited states has to be low for effective RISC.¹² Thus, the design strategy for a TADF molecule is based on the separation of highest occupied and lowest unoccupied molecular orbitals (HOMO and LUMO, respectively) on separate moieties.⁴ In many cases, the hybridization of locally exited (LE) and charge transfer (CT) states occurs forming new hybridized local and charge transfer (HLCT) states. The phenomenon of HLCT, as well as the hot-exciton transition, deals with the RISC at higher excited states rather than the first excited state. This leads to the emission of up-converted excitons even if the singlet–triplet energy splitting is not negligible.³ One of the most efficient and well-studied TADF emitters 1,2,3,5-tetrakis(carbazol-9-yl)-4,6-dicyanobenzene (4CzIPN)¹³ was reported to possess TADF properties operating through HLCT states.^14–16 Previously, our group reported on blue OLEDs with hosts exhibiting HLCT.¹⁷ Specifically, OLEDs with the di-tert-butylcarbazole-containing TADF emitter doped into the derivative of dimethoxycarbazolyl-disubstituted dibenzofuran as a host showed an EQE value of 5.8%.¹⁷ PhOLEDs containing imidazole derivatives with di-tert-butylcarbazole or carbazole moieties as HLCT-hosts exhibited EQE of up to 7.6%.¹⁸ Bipolar or electron-transporting derivatives of carbazole and benzimidazole exhibited good performance as hosts for TADF OLEDs and PhOLEDs.^19–22 Derivatives of carbazole exhibit good hole-transporting properties and excellent thermal stability as well as exciplex-forming properties and triplet emission which can be tuned by the purposeful molecular design.^23–29 Attachment of the tert-butyl group to the carbazole moiety of the HLCT-TADF emitter 4CzIPN led to the improvement of the morphology of the emitting layer (EML) and enhanced OLED characteristics.³⁰ An electron accepting benzimidazole moiety is a promising building block for the design of donor–acceptor–donor type emitters.^31,32 Imidazole-based emitters^33–36 including derivatives of benzimidazole^37,38 are characterized by hybridization of excited states due to the twisted geometries and conformational heterogeneity. Consequently, imidazole-based compounds have been successfully applied as hosts in PhOLEDs, with relatively high EQE values of over 26%, and low efficiency roll-offs at high luminescence (Table S1, ESI†).³⁹

The efficiencies of TADF-based and phosphorescence OLEDs can be improved by improving the internal outcoupling efficiency via molecular orientations of emitters in EMLs.^40–43 Many efficient TADF and phosphorescent emitters exhibit random molecular orientations in the EMLs.⁴⁴ Fortunately, their orientation can be improved using appropriate hosts.⁴⁵ Thus, charge injecting and charge-transporting properties, triplet channel energy losses, triplet harvesting abilities, and vertical or horizontal molecular orientations should be taken into account when new hosts for TADF and phosphorescence OLEDs are being developed.⁴⁶ However, it is difficult to obtain the best combinations of properties for a single host material. In addition, it is not clear which hosting properties have the strongest influence on the performance of OLEDs. We partly aimed to study this issue by developing compounds containing the same electron-accepting unit and the different number of the same electron-donating units.

Herein, we present the synthesis and properties of the derivatives of 1-phenyl-1H-benzo[d]imidazole and di-tert-butylcarbazole as hosts with TADF capabilities for phosphorescence and TADF-based OLEDs. Both synthesized compounds exhibit TADF. They were tested as emitters as well as hosts in PhOLEDs and TADF-based OLEDs. As the emitter, the compound with a single di-tert-butylcarbazole moiety (1tCzBzCN in Scheme 1) showed the best performance in host-free OLEDs. Additionally, as a host, compound 1tCzBzCN showed better performance in green phosphorescence OLEDs than compound 2tCzBzCN. The angle-dependent photoluminescence measurements revealed the horizontal orientation factors of 0.84 for the films containing 1tCzBzCN and 0.87 for the films containing 2tCzBzCN. This observation suggests an enhanced outcoupling factor for the emitting layers composed of green phosphorescence emitters doped in the studied compounds. Compound 1tCzBzCN exhibits bipolar charge carrier transport, while 2tCzBzCN exhibits a unipolar one. Compound 1tCzBzCN was utilized as a host for the fabrication of blue and green emitting PhOLEDs with external quantum efficiencies of 10.2 and 13%, respectively. In addition, in this study, we demonstrate the complete quenching of TADF emission of 2tCzBzCN in a solid state by picric acid. To the best of our knowledge, this is the first report on the optical sensing of nitroaromatic explosive compounds based on TADF quenching in the solid state.

Scheme 1 Synthesis of 1tCzBzCN and 2tCzBzCN.

2. Experimental section

2.1 Materials

N-Phenyl-1,2-phenylenediamine, 2,6-difluoro-4-formylbenzonitrile, and 3-fluoro-4-formylbenzonitrile were purchased from Fluorochem and used as received. Dimethylformamide (DMF), dimethyl sulfoxide (DMSO) and cesium carbonate (Cs₂CO₃) were purchased from Sigma-Aldrich and used as received. 3,6-Di-tert-butylcarbazole was synthesized according to the procedure reported earlier.⁴⁷ 2-Fluoro-4-benzonitril-1-phenyl-1H-benzo[d]imidazole (1, yield 70%) and 2,6-difluoro-4-benzonitril-1-phenyl-1H-benzo[d]imidazole (2, yield 47%) were synthesized from N-phenyl-1,2-phenylenediamine and the corresponding 4-formylbenzonitriles were synthesized according to the procedure described earlier.⁴⁸ After the initial workup, the target intermediates were recrystallized from hot methanol.

2-(3,6-Di-tert-butyl-9H-carbazol-9-yl)-4-((benzonitril)phenyl)-1-phenyl-1H-benzo[d]imidazole (1tCzBzCN). 2-Fluoro-4-benzonitril-1-phenyl-1H-benzo[d]imidazole (1) (1 g, 3.19 mmol) was added to the mixture of 3,6-di-tert-butylcarbazole (0.89 g, 3.19 mmol) and Cs₂CO₃ (2.1 g, 6.45 mmol) in DMSO (8 ml). The reaction mixture was stirred at 160 °C for 24 h under a nitrogen atmosphere. The reaction was monitored by TLC chromatography. Upon completion, the reaction mixture was cooled down to room temperature, the crude product was extracted with dichloromethane and dried with Na₂SO₄, and excess dichloromethane was removed via rotary evaporation. The crude product was purified using column chromatography on a silica gel. The mixture of tetrahydrofuran/hexane (1/4) was used as the eluent. The target compound was recrystallized from the same eluent to afford 1tCzBzCN (0.86 g, 46%) as white crystals. T_m = 287 °C (from DSC). ¹H NMR (400 MHz, DMSO) δ 8.24 (d, J = 8.1 Hz, 1H), 8.16 (d, J = 8.4 Hz, 2H), 8.07 (s, 2H), 7.88 (s, 1H), 7.77 (d, J = 8.0 Hz, 1H), 7.58 (s, 1H), 7.21 (t, J = 7.7 Hz, 1H), 7.06 (m, J = 14.3, 7.5 Hz, 2H), 6.93 (t, J = 7.4 Hz, 1H), 6.85 (t, J = 8.4 Hz, 2H), 6.56 (d, J = 8.7 Hz, 2H), 6.34 (d, J = 7.6 Hz, 2H), 1.32 (s, 18H). ¹³C NMR (101 MHz, DMSO) δ 166.39, 150.06, 142.79, 142.43, 137.99, 137.00, 136.23, 135.36, 134.22, 134.00, 130.61, 128.83, 127.87, 127.01, 126.66, 124.34, 123.37, 123.12, 122.51, 119.56, 115.82, 110.02, 108.78, 34.36, 31.73. MS (APCI+, 20 V), m/z: 572.29 [M + NH₄]⁺.

2,6-Bis(3,6-di-tert-butyl-9H-carbazol-9-yl)-4-((benzonitrile)phenyl)-1-phenyl-1H-benzo[d]imidazole (2tCzBzCN). 2,6-Difluoro-4-benzonitril-1-phenyl-1H-benzo[d]imidazole (2) (0.47 g, 1.42 mmol) was added to the mixture of 3,6-di-tert-butylcarbazole (0.77 g, 2.76 mmol) and Cs₂CO₃ (1.16 g, 3.56 mmol) in DMSO (8 ml). The reaction mixture was stirred at 160 °C for 24 h under a nitrogen atmosphere. The reaction was monitored by TLC. Upon completion, the reaction mixture was cooled down to room temperature, the crude product was extracted with dichloromethane and dried with Na₂SO₄, and excess dichloromethane was removed via rotary evaporation. The crude product was purified using column chromatography on a silica gel. The mixture tetrahydrofuran/hexane (1/3) was used as the eluent. The target compound was recrystallized from the same eluent to afford 2tCzBzCN (0.52 g, 37%) as yellow crystals. T_m = 412 °C (from DSC). ¹H NMR (400 MHz, CDCl₃) δ 8.11 (s, 4H), 7.94 (s, 2H), 7.83 (d, J = 7.9 Hz, 1H), 7.66–7.56 (m, 3H), 7.48 (d, J = 8.6 Hz, 4H), 7.41 (d, J = 6.9 Hz, 2H), 7.37–7.26 (m, 2H), 7.19 (d, J = 7.9 Hz, 1H), 7.10 (d, J = 8.6 Hz, 4H), 1.48 (s, 36H). ¹³C NMR (101 MHz, CDCl₃) δ 148.88, 144.12, 143.33, 142.90, 138.83, 137.75, 136.91, 136.30, 130.76, 129.43, 128.21, 127.47, 124.70, 124.25, 124.10, 123.81, 120.46, 116.64, 113.45, 111.93, 110.80, 109.25, 34.83, 32.01. MS (APCI+, 20 V), m/z: 850.53 [M + H]⁺.

3. Results and discussion

3.1 Synthetic procedures

Compounds 1tCzBzCN and 2tCzBzCN were synthesized in a simple two-step synthesis process (Scheme 1). In the first step, the corresponding fluoro-substituted 4-formylbenzonitriles were coupled with N-phenyl-1,2-phenylenediamine to form the 4-benzonitril-1-phenyl-1H-benzo[d]imidazole core and provide intermediate compounds. Following the first step, nucleophilic substitution reactions were carried out using Cs₂CO₃ as a base to introduce electron-donating 3,6-di-tert-butylcarbazole moieties and form the target compounds 1tCzBzCN and 2tCzBzCN.

3.2 Physical properties

Molecular optimization and orbitals. The electronic structures of 1tCzBzCN and 2tCzBzCN were optimized at their ground states at the B3LYP/6-31G** level. A notable difference in the molecular geometry of these compounds arises from the distinct dihedral angles between the benzoimidazole and benzonitrile moieties: 49° in 1tCzBzCN and 22° in 2tCzBzCN (Fig. 1). This variation is fundamentally attributed to the positional attachment of the benzoimidazole unit. In 1tCzBzCN, the ortho-substitution places the benzoimidazole moiety in close proximity to the bulky tert-butylcarbazolyl group. The steric repulsion between the tert-butyl group and the nearby benzoimidazole moiety and benzonitrile units forces the benzoimidazole unit to adopt a significantly twisted conformation (ca. 49°) to alleviate the spatial strain. Conversely, in 2tCzBzCN, the meta-substitution positions benzoimidazole further from the tert-butylcarbazole moiety, reducing steric interactions. This minimized strain enables the benzoimidazole moiety to adopt a more planar conformation relative to benzonitrile, resulting in a smaller twist (ca. 22°).

Fig. 1 Ground state optimized geometry and molecular charge distribution of 1tCzBzC and 2tCzBzC according to B3LYP/6-31G**.

The molecular orbital analysis provides key insights into the electronic distribution in 1tCzBzCN and 2tCzBzCN. In the HOMO, the wavefunction is localized on the benzonitrile and tert-butylcarbazole moieties in 1tCzBzCN, while in 2tCzBzCN, it extends over the benzonitrile and both tert-butylcarbazole moieties. This broader distribution in 2tCzBzCN reflects the stronger electron-donating environment provided by the two tert-butylcarbazole units, which contributes to its slightly higher HOMO energy (−5.30 eV) compared to that of 1tCzBzCN (−5.45 eV). In the LUMO, the wavefunction is predominantly localized on the benzonitrile and benzoimidazole moieties of both the compounds, highlighting the consistent electron-accepting nature of the benzonitrile unit. However, the LUMO energy of 2tCzBzCN (−2.11 eV) is slightly lower than that that of 1tCzBzCN (−1.89 eV). This stabilization arises from the enhanced π-conjugation between the benzonitrile and benzoimidazole units in 2tCzBzCN, facilitated by its smaller dihedral angle (22°), which allows for an improved orbital overlap.

Thermal properties. Thermogravimetric analysis (TGA) and differential scanning colorimetry (DSC) were utilized to study the thermal properties of the synthesized compounds. TGA measurements showed that both 1tCzBzCN and 2tCzBzCN exhibited excellent thermal stability. Their 5% mass-loss temperatures (T_−5%) were recorded at 341 and 426 °C, respectively (Fig. 2a and Table 1). However, the complete, single-step, loss of mass of the compounds observed during TGA measurements suggests that these values represent the temperatures corresponding to the beginning of sublimation rather than thermal degradation (Fig. 2a). Compounds 1tCzBzCN and 2tCzBzCN were obtained as purely crystalline compounds, and their melting temperatures (T_m) were observed at 287 and 412 °C (Fig. 2), respectively. The DSC analysis showed that both the compounds were capable of forming glasses. During the second heating scans, their glass transition temperatures (T_g) were recorded at 149 and 192 °C, respectively (Fig. 2a and c). For compound 2tCzBzCN, the crystallization signal was recorded at 277 °C during the second heating scan, suggesting that the glassy state is not stable. The solid amorphous state of 1tCzBzCN was more stable, as no crystallization signal was recorded in the second heating scan. The thermal characteristics of 1tCzBzCN and 2tCzBzCN are summarized in Table 1. Considerably higher values of T_m and T_g observed for 2tCzBzCN relative to those of 1tCzBzCN can be attributed to its higher molecular weight leading to stronger intermolecular interactions.

Fig. 2 TGA (a) and DSC (b) and (c) curves of 1tCzBzCN and 2tCzBzCN.

Table 1 Thermal and electrochemical characteristics of 1tCzBzCN and 2tCzBzCN

Compound	T −5%, °C	T m, °C	T g, °C	T cr, °C	E ^ox vs. Fconset, V	E ^red vs. Fconset, V	IPCV, eV	EACV, eV	E bg, eV
T −5% – 5% mass loss temperature; Tm – melting temperature; Tg – glass transition temperature; Tcr – crystallization temperature; E^oxonset – onset oxidation potential vs. Fc measured from CV; E^redonset – onset reduction potential vs. Fc measured from CV; IPCV – ionization potential, calculated from IPCV = E^ox vs. Fconset + 4.8; EACV – electron affinity, calculated from EACV = E^red vs. Fconset + 4.8; Ebg – band gap values calculated from Ebg = IPCV − EACV.
1tCzBzCN	341	287	149	—	0.79	−2.15	5.59	2.65	2.94
2tCzBzCN	426	412	192	277	0.94	−1.86	5.74	2.94	2.80

---

Electrochemical properties. The electrochemical properties of compounds 1tCzBzCN and 2tCzBzCN were studied utilizing cyclic voltammetry (CV). Ionization potential and electron affinity values were estimated from the oxidation and reduction onset potentials against ferrocene (E^{ox vs. Fc}_onset, E^{red vs. Fc}_onset), respectively. Both the synthesized compounds exhibited single oxidation peaks owing to the resident di-tert-butylcarbazole moieties in the molecules (Fig. 3). The ionization potential (IP_CV) values calculated from CV data were found to be 5.59 and 5.74 eV for compounds 1tCzBzCN and 2tCzBzCN, respectively. Compound 1tCzBzCN exhibited a lower electron affinity (EA_CV) of 2.65 eV than 2tCzBzCN (2.94 eV). Band gap energy values (E_bg) estimated using IP_CV and EA_CV were found to be 2.94 eV for 1tCzBzCN and 2.8 eV for 2tCzBzCN. The summary of data obtained from CV measurements is displayed in Table 1.

Fig. 3 CV curves of 2tCzBzCN and 1tCzBzCN.

Photophysical properties. Absorption spectroscopy and photoluminescence (PL) spectroscopy were used to study the photophysical properties of dilute solutions and solid films of compounds 1tCzBzCN and 2tCzBzCN. The photophysical characteristics are displayed in Table 2. The absorption bands observed at 290 and 294 nm for the solutions of compounds 1tCzBzCN and 2tCzBzCN, respectively, correspond to the local transitions of benzimidazole and di-tert-butylcarbazole fragments.⁴⁹ The absorption bands observed in the range of 300–350 nm are related to the n–π* transitions of di-tert-butylcarbazole fragments, and by π–π* transition of the benzimidazole fragment.⁵⁰ The low intensity absorption band or tail (in the case of 1tCzBzCN) observed in the range of 350–440 nm corresponds to a weak charge transfer (CT) between the di-tert-butylcarbazole fragment and the cyano-substituted phenyl ring.^48,51 The difference in PL spectra of the solutions of the compounds was due to the different number of electron-donating di-tert-butylcarbazole fragments as well as their position relative to the position of the cyano group. The absorption spectra of the films were found to be slightly bathochromically shifted in comparison to those of their solutions due to enhanced intermolecular interactions in the solid state.

Table 2 Major photophysical characteristics of the dilute toluene solutions and the films of compounds 1tCzBzCN and 2tCzBzCN

Compounds	PLQYtol^a	PLQYfilm^b	PLQYdoped^c	S1, eV^d	T1, eV^d	ΔEST, eV^d	τ 1, ns^a^e	τ 2, ms^b^f
a Toluene solutions at room temperature in air. b Neat films at room temperature in air. c Films of the molecular mixtures with Ir(ppy)2(acac) (10 wt%) at room temperature in air. d Films at 77 K temperature under an argon atmosphere. e Lifetime of prompt fluorescence at room temperature, in air. f Lifetime of delayed fluorescence at room temperature, in a vacuum.
1tCzBzCN	0.26	0.20	0.63	2.99	2.66	0.33	4.22, 10.11	0.58
2tCzBzCN	0.09	0.39	0.48	2.68	2.63	0.05	1.85, 4.03	0.46

---

PL spectra of the dilute solutions of compounds 1tCzBzCN and 2tCzBzCN in solvents of different polarities manifested the positive solvatochromic effect highlighting the CT nature of emission (Fig. 4a). The addition of the second di-tert-butylcarbazole donor moiety resulted in a redshift of PL spectra of 2tCzBzCN when compared to 1tCzBzCN, likely due to the extension of π-conjugation. Unlike 2tCzBzCN, the solution of 1tCzBzCN exhibited a redshift of 5 nm when the non-polar toluene (0.36 D) was changed to a moderately polar THF (1.75 D). An even longer bathochromic shift of 27 nm was observed when THF was changed to a highly polar DMF (3.82 D) (Fig. 4a). The increasing redshift with the increase of polarity of the solvent leads to the switch from LE character of emission to CT as for previously reported HLCT emitters.^52–55 The HLCT nature of the compounds is further substantiated by quantum chemical calculations, as elaborated later in this section. The removal of oxygen led to the enhancement of the PL only for the film of 2tCzBzCN with a 3.3-fold increase in intensity and no additional peaks (Fig. S3, ESI†). No PL intensity increase was observed for the film of 1tCzBzCN. The increase of PL intensity is attributed to the utilization of triplet excitons in the emission as the triplet energy is quenched in the presence of oxygen.² Collisional interactions with molecular oxygen could cause the drop of PLQY values for the films as the values were estimated for the samples under ambient air conditions. Deoxygenated toluene solutions of the compounds exhibited only prompt fluorescence with lifetimes of up to 10 ns (Fig. S1, ESI†). PL decay curves of the films of the compounds recorded in a vacuum were characterized by the appearance of the long-lived components of the emission with lifetimes up to 0.5 ms (Table 2 and Fig. S2, ESI†). The energy values of the first singlet S₁ and triplet T₁ excited states were estimated from the onsets of fluorescence and phosphorescence (delay of 1 ms) bands, respectively, recorded under an inert atmosphere at 77 K (Fig. 4b). These values were taken for the calculation of the singlet–triplet energy splitting (ΔE_ST) (Table 2). The thermal activation of delayed fluorescence was confirmed as the TADF component with sub-ms lifetimes increased upon heating (Fig. 4c). The lifetime derived from the PL decay is slightly lower for 2tCzBzCN (0.46 ms) with ΔE_ST of 0.05 eV than that of 1tCzBzCN (0.58 ms) with ΔE_ST of 0.33 eV. This observation shows that more efficient TADF has a shorter lifetime (Table 2).^29,56 There is experimental evidence of the HLCT nature of TADF of 1tCzBzCN.^14,16 The relatively high ΔE_ST together with the undoubtfully TADF nature of emission suggests a hot exciton upconversion which occurs at higher energy levels than S₁ and T₁. The films of both compounds show no change in PL spectral behavior in the temperature range of 77 to 300 K. This indicates that the singlet excitons and the triplet excitons upconverted via RISC are deactivated radiatively from practically the same excited states with energy levels close to the triplet excited states from which phosphorescence occurs at low temperatures (Fig. S4, ESI†). The emission band with a vibronic substructure related to the phosphorescence disappears upon heating.

Fig. 4 Absorption and PL spectra of the films and dilute toluene, THF and DMF solutions (a); PL and phosphorescence spectra, recorded at 77 K of the films (b); PL decay curves, recorded at different temperatures of the films (c) of compounds 1tCzBzCN and 2tCzBzCN. PL decay curves of the films of the mixtures of picric acid (5 wt%) doped in 2tCzBzCN recorded at room temperature in a vacuum. Inset: PL spectra of the films (d). PL spectra (e), PL decay curves (f), recorded at room temperature of the films of the molecular mixtures of Ir(ppy)₂acac (10 wt%) and compounds 1tCzBzCN and 2tCzBzCN.

Time-dependent density functional theory (TD-DFT) calculations were performed using the LC-ω*PBE functional and the 6-31G** basis set to get insightful aspects of excited state properties. The range separation parameter “ω” is tuned for the gaseous medium. The tuned ω value is 0.178 Bohr⁻¹ for 1tCzBzCN and 0.155 Bohr⁻¹ for 2tCzBzCN. The theoretical vertical excited-state energies were adjusted by applying a redshift of 0.4 eV for 1tCzBzCN and 0.2 eV for 2tCzBzCN (as summarized in Table S2, ESI†) to achieve closer alignment with experimental results. This adjustment effectively accounts for the influence of the molecular environment, providing a more accurate depiction of the excited-state dynamics. The TD-DFT calculations unveiled the presence of four triplet states (T₁–T₄) lying below the S₁ in 1tCzBzCN, highlighting the potential for hot exciton RISC due to multiple intermediate triplet levels. To unravel the nature of these excitations, the natural transition orbitals (NTOs) were generated (Fig. S5, ESI†). The NTOs of S₁ exhibit a combination of localization on the benzonitrile moiety and CT from the tert-butylcarbazole moiety to the benzoimidazole fragment. This observation is consistent with the HLCT nature of the S₁ state. The T₁ and T₃ states share the similar excitation pattern, showing localization on benzonitrile and benzoimidazole units with CT from the tert-butylcarbazole moiety. In contrast, T₂ is predominantly localized on the tert-butylcarbazole unit, while T₄ shows a mixed localization on the benzonitrile and tert-butylcarbazole moieties. Despite the highest spin–orbit coupling (SOC) value between S₁ and T₁ (0.70 cm⁻¹, Fig. S6, ESI†), the significant ΔE_(S1–T1) gap (0.4 eV) renders the S₁–T₁ pathway thermodynamically unfavourable for efficient RISC. However, T₄ emerges as a viable mediator in the hot exciton RISC. T₄ exhibits the smallest energy gap with S₁ (0.03 eV) and the second-highest SOC value (0.24 cm⁻¹) for the S₁–T₄ pathway. Additionally, the T₄–T₃ energy gap (0.1 eV) is larger than the S₁–T₄ gap, reducing the likelihood of internal conversion and favouring RISC as the dominant process.

The NTOs for the S₁ state of 2tCzBzCN reveal a characteristic HLCT nature, with electronic density localized on the benzonitrile moiety and charge transfer from the tert-butylcarbazole to the acceptor units. Similarly, the T₁ state shows predominant localization on the benzonitrile moiety, minor contributions from the benzoimidazole unit, and charge transfer from the tert-butylcarbazole fragment to the acceptor moieties. In the T₂ state, the NTOs depict a mixture of localization on benzonitrile and benzoimidazole units with charge transfer from one of the tert-butylcarbazole moieties. This variation in orbital character for both T₁ and T₂, compared to S₁, enables both states to act as potential pathways for RISC. Among these, the S₁–T₂ pathway is characterized by a small energy gap but exhibits a low SOC value, indicating that this pathway is relatively weak for RISC. In contrast, the S₁–T₁ pathway, with a moderate theoretical energy gap of 0.15 eV, demonstrates the significantly higher SOC value (0.90 cm⁻¹), marking it as the dominant and highly efficient pathway for RISC. The pronounced SOC value for the S₁–T₁ pathway translates to a faster RISC rate in 2tCzBzCN, facilitating efficient conversion of triplet excitons to singlet states. These dynamics contribute to the shorter PL lifetime of 2tCzBzCN, compared to that of its counterpart (1tCzBzCN).

Additionally, the sensitivity of TADF of compound 2tCzBzCN to the presence of nitroaromatic explosive compounds was examined. Expectedly, the film of picric acid was found to be non-emissive. The PL spectra and PL decay curves of the film of the molecular mixture of picric acid (5 wt%) and 2tCzBzCN are shown in Fig. 4d. For comparison, the PL spectrum of the film of neat 2tCzBzCN is shown. The peak of the PL spectrum of the molecular mixture of 2tCzBzCN and picric acid appeared at 578 nm. It was considerably redshifted with respect to the peak of the PL spectrum of the neat layer of 2tCzBzCN which was observed at 500 nm. The bathochromic shift is explained by intermolecular interactions of the two compounds. Electronic excitation energy transfer occurs between the host and the guest effectively quenching TADF. Consequently, the long-lived component of TADF with a lifetime of up to 0.34 ms was eliminated in the PL decay curve of the molecular mixture of 2tCzBzCN and picric acid. Only prompt fluorescence was observed when picric acid was added (5 wt%). At a higher concentration of picric acid of 25 wt%, the host–guest system became non-emissive demonstrating that picric acid totally quenched prompt fluorescence as well as TADF. To gain insights into the interaction mechanism between 2tCzBzCN and picric acid, quantum chemical calculations were conducted at the LC-ωPBE/6-31G** theoretical level. The optimized geometry of the complex of 2tCzBzCN and picric acid is shown in Fig. S7 (ESI†). It highlights the spatial arrangement of the interacting species. The molecular orbitals of the complex show that the HOMO is primarily localized on the electron-rich tert-butylcarbazole moieties of 2tCzBzCN, while the LUMO is concentrated on the highly electron-deficient trinitrophenol (picric acid). This clear transfer of charge density from 2tCzBzCN to picric acid signifies strong non-covalent interactions. This charge transfer mechanism plays a pivotal role in modulating the exciton dynamics, providing a theoretical basis for the observed quenching of emission of 2tCzBzCN by picric acid. The quantum chemical calculations further confirm a substantial interaction energy of −8.03 kcal mol⁻¹, indicating that the formation of the complex is both thermodynamically favourable and spontaneous.

The solid films of emitters 1tCzBzCN and 2tCzBzCN at room temperature in air demonstrated relatively low photoluminescence quantum yields of 0.2 and 0.39. To pursue efficient OLED performance, it is necessary to add efficient emitters to enhance the formation of excitons. Before device fabrication, the energy transfer between the host and the dopant was analyzed using PL spectra and PL decay curves of vacuum-deposited layers of 1tCzBzCN and 2tCzBzCN, and their molecular mixtures with 10 wt% bis[2-(2-pyridinyl-N)phenyl-C](acetylacetonato)iridium(III) (Ir(ppy)₂acac) (S₁ ∼ 2.47 eV and T₁ ∼ 2.45 eV). The PL spectra and PL decay curves are shown in Fig. 4e and f and Fig. S2 (ESI†). The whole PL spectra of the molecular mixtures are attributed to the emission of Ir(ppy)₂acac, illustrating the efficient energy transfer from the hosts (1tCzBzCN or 2tCzBzCN) to the guest (Ir(ppy)₂acac). The yields are 0.63 for the mixture containing 1tCzBzCN and 0.48 for the mixture containing 2tCzBzCN (Fig. S8, ESI†). TPL decay curves of the layers of 1tCzBzCN and 2tCzBzCN shown in Fig. S2 (ESI†) exhibit long lived components with the lifetimes of over 4 ms, which evidence the possibility of ISC or RISC. After mixing the synthesized compounds with Ir(ppy)₂acac, long lived components disappeared in the PL decay curves (Fig. 4f). They revealed emission lifetimes of less than 10 μs. This observation leads to a conclusion that Ir(ppy)₂acac facilitates exciton emission, corresponding to the increased PLQY. The PL decay curve of the molecular mixture of 1tCzBzCN and Ir(ppy)₂acac is characterized by the single exponential function, indicating efficient energy transfer from the host to the dopant, leading to the enhancement of PLQY up to 0.63. This may result from very fast and efficient Förster energy transfer (FRET) and Dexter energy transfer (DET) between the host and the guest. On the other hand, for the adequate representation of the PL decay curve of the molecular mixture of 2tCzBzCN and Ir(ppy)₂acac, the double exponential fit was required. This observation can be attributed to the ISC process, implying a relatively slow or inefficient FRET between the host and the dopant. This resulted in a lower PLQY of 0.48.

Charge-transporting properties. The charge-transporting properties of vacuum-deposited films of 1tCzBzCN and 2tCzBzCN were investigated by conducting time-of-flight (TOF) measurements (Fig. 5 and Fig. S9, ESI†). The typical two-electrode structure ITO/film/Al of TOF samples with one optically transparent electrode indium–tin oxide (ITO) was used. The thicknesses (d) of the films were measured using a profilometer Profilm3D (Fig. S10, ESI†). The TOF samples were excited by a 350 nm laser beam through the ITO side. The positive or negative voltages were selectively applied to ITO, simultaneously recording the TOF photocurrent transients for holes or electrons, respectively (Fig. 5a–c and Fig. S9, ESI†). From the TOF current transients plotted in log–log scales, the transit times (t_tr) for holes and electrons were estimated for 1tCzBzCN (Fig. 5a and b). For the films of 2tCzBzCNt_tr values only for holes were obtained (Fig. 5c and Fig. S9, ESI†). Thus, 1tCzBzCN exhibited bipolar charge carrier transport, whereas 2tCzBzCN showed only hole-transporting properties. Charge carrier mobility dependences on the electric field were adequately expressed using a Poole–Frenkel function μ_e,h = μ^{(βe,h·E1/2)}_0e,h. Here, μ_e,h are hole and electron mobilities at different electric fields (E), μ_0e,h are hole and electron mobilities at E of equal to zero, β_e,h is the Poole–Frenkel factor for holes or electrons (Fig. 5d). The samples of compounds 1tCzBzCN and 2tCzBzCN showed very close μ_0h values of 7.5 × 10⁻¹⁰ and 9.1 × 10⁻¹⁰ cm² V⁻¹ s⁻¹, respectively. The slightly different β_h values of 0.0086 and 0.0074 cm^1/2 V^−1/2 were observed for 1tCzBzCN and 2tCzBzCN, respectively, resulting in the considerable difference of μ_h values (2.34 × 10⁻⁶ and 8.75 × 10⁻⁷ cm² V⁻¹ s⁻¹ at E of 8.7 × 10⁵ V cm⁻¹). Such a difference could be attributed to the twice the number of tert-butyl groups in 2tCzBzCN. Apparently, tert-butyl groups may hinder HOMO–HOMO overlapping of neighbouring molecules of 2tCzBzCN in vacuum-deposited films thus reducing its hole mobility in comparison to that of 1tCzBzCN. In addition, the higher number of tert-butyl groups in 2tCzBzCN may prevent LUMO–LUMO overlapping of its neighbouring molecules. Our presumption is supported by previous reports on the different HOMO–LUMO separations of donor–acceptor type compounds caused by only slight modification of the substituents on the donor moieties.^24,57 In addition, molecular packing of the molecules of 1tCzBzCN and 2tCzBzCN in their films should be different, leading to the different HOMO–HOMO and LUMO–LUMO overlapping of neighbouring molecules. This results in the different charge-transporting properties of the films of 1tCzBzCN and 2tCzBzCN. As a result, 2tCzBzCN did not exhibit electron transport. The sample 1tCzBzCN showed electron transporting properties with μ_0e of 4.6 × 10⁻⁸ cm² V⁻¹ s⁻¹, β_e of 0.0047 cm^1/2 V^−1/2 and μ_e of 3.58 × 10⁻⁶ cm² V⁻¹ s⁻¹ at E of 8.7 × 10⁵ V cm⁻¹ (Fig. 5d).

Fig. 5 TOF transients of vacuum-deposited films of compounds 1tCzBzCN (a) and (b) and 2tCzBzCN (c) under applied positive voltages for holes and negative voltages for electrons at an optically transparent electrode ITO and the corresponding charge carrier mobilities (d) as the Poole–Frenkel function of electric fields. Inset: A fragment of data of thickness measurements of compound 1tCzBzCN using a profilometer Profilm3D.

3.3 Performance of organic light-emitting diodes

The series of OLEDs with the structure ITO/TAPC (50 nm)/mCP (10 nm)/EML (30 nm)/DPPS (55 nm)/LiF/Al were fabricated. ITO and Al were anode and cathode electrodes, respectively. The layers of 1,1-bis[(di-4-tolylamino)phenyl]cyclohexane (TAPC) and diphenyl-bis[4-(pyridin-3-yl)phenyl]silan (DPPS) were selected as hole and electron transport layers, respectively. N,N-Dicarbazolyl-3,5-benzene (mCP) was utilized for the deposition of the exciton blocking layer. LiF was utilized as the electron injection layer. As the light-emitting layer (EML), the neat layers of 1tCzBzCN (devices A), 2tCzBzCN (devices C) or the layers of 1tCzBzCN (devices B, E and F) and 2tCzBzCN (device D) co-deposited with the different guests were used. Green phosphorescent emitter Ir(ppy)₂(acac) (10 wt%) was used as a guest for the fabrication of OLEDs B and D. Sky-blue phosphorescent emitter bis[2-(4,6-difluorophenyl)pyridinato-C₂,N] (picolinato)iridium(III) (FIrpic) (9 wt%) was used as a guest in device E. The green TADF emitter 1,2,3,5-tetrakis(carbazol-9-yl)-4,6-dicyanobenzene (4CzIPN) was used as a guest in device F. The heterostructure of OLEDs and the molecular structures of the functional compounds are presented in Fig. 6a and b.

Fig. 6 The general structure of the fabricated OLEDs (a). The chemical structures of the functional compounds (b). EL spectra (c), current density–luminance–voltage (d) and CE–PE–EQE–luminance plots (e).

The EL spectra of devices A and C employing EMLs of neat compounds 1tCzBzCN and 2tCzBzCN correspond well with the PL spectra of the films of the respective compounds. The higher PLQY was observed for the film of 2tCzBzCN than for the film of 1tCzBzCN (Table 3). OLED A with the EML of 1tCzBzCN demonstrated more than twice as high external quantum efficiency compared to that of the 2tCzBzCN-based device (C). Thus, usage of 1tCzBzCN containing one electron-donating substituent allowed reaching the considerably higher external quantum efficiency of OLEDs compared to that of the device with 2tCzBzCN containing two di-tert-butylcarbazole moieties. This is due to the charge imbalance caused by 2tCzBzCN which also explains significantly higher turn-on and lit-on voltages of OLED C when compared to those of OLED A (Fig. 5d and 6d). Together with the PL data of guest–host systems, this observation points to 1tCzBzCN being a more efficient host than 2tCzBzCN. Among OLEDs based on Ir(ppy)₂(acac), the device with 1tCzBzCN has superior values of current efficiency (CE), power efficiency (PE) and EQE relative to OLEDs with 2tCzBzCN. For the estimation of the performance of 1tCzBzCN as a host, OLEDs E and F utilizing FIrpic and 4CzIPN as emitters, were fabricated and characterized. OLEDs with EMLs based on guest–host systems showed EL spectra representing the emission of the guests without the EL contribution of the host.

Table 3 EL characteristics of OLEDs

Device	EML	V on , V	CE^b, cd A^−1	PE^b, lm W^−1	EQE^b, %	λ, nm	1931 CIEx,y
a Turn-on voltage at 1 mA cm^−2 and the lit-on voltage 1 cd m^−2. b Maximum value/at 100 cd m^−2/at 1000 cd m^−2.
A	1tCzBzCN	4.0/3.0	5.6/2.1/—	5.9/3.2/—	2.0/1.3/—	472	(0.19, 0.31)
B	1tCzBzCN – Ir(ppy)2(acac)	4.3/2.6	54.3/53.8/48.1	57/46.3/32.1	13.0/12.4/11.47	522	(0.31, 0.64)
C	2tCzBzCN	7.6/6.2	14.5/8.23/—	7.0/3.3/—	0.9/0.5/—	496	(0.22, 0.48)
D	2tCzBzCN – Ir(ppy)2(acac)	4.9/4.0	19.4/18.8/13.8	13.6/12.7/7.3	4.4/4.3/3.2	526	(0.35, 0.62)
E	1tCzBzCN – FIrpic	4.0/2.8	25.6/25.8/21.4	26.8/22.5/13.4	10.2/10.0/7.9	472	(0.15, 0.32)
F	1tCzBzCN – 4CzIPN	5.2/3.1	35.6/34.1/26.7	30.7/24.7/13.7	10.4/10.2/7.88	514	(0.28, 0.58)

---

Time-resolved electroluminescence (TREL) spectra of devices A and C based on EMLs of the neat compounds have characteristic spikes due to imperfect charge transport at recombination sites. The inset of Fig. 7 a is an enlarged view of the spikes, which reflect a great deal of trap states inside the layer of 2tCzBzCN which impede charge carrier transport, prevent charge carrier recombination, and degrade the device performance. While device B has a long lifetime of emission, the TREL signal is shortened in comparison to that of OLED D, highlighting more efficient energy transfer. FIrpic and 1tCzBzCN have similar wavelengths of the EL peaks which are conventionally viewed as unfavorable for efficient host–guest interactions. Nevertheless, the long radiative lifetime of emission of OLED E (Fig. 7b) makes it possible to facilitate FRET even with a small spectral overlap of emission spectrum of the host and the absorption band of the guest due to vibronic coupling.⁵⁸ Thus, OLED F based on green emitting 4CzIPN, and E based on sky-blue emitting FIrpic showed close values of maximum EQE of ca. 10%. For further investigation of the output coupling effect of devices B and D, the angle-dependent PL measurements were performed. They allow studying the optical transition dipole of the systems of Ir(ppy)₂(acac) doped in 1tCzBzCN and 2tCzBzCNvia a laser (excitation at 325 nm) and collecting p-polarized fluorescence intensities at different emission angles as shown in Fig. 7. Two ideal cases, such as the molecular arrangement in a perfect horizontal orientation factor (Θ of 1) and an isotropic dipole orientation factor (Θ of 0.66), were also simulated to be the baseline in comparison to the experimental angle-dependent PL profiles. After fitting the two main peaks of the experimental angle-dependent PL spectra, high Θ values of 0.84 and 0.87 were obtained for OLEDs B and D, respectively. This indicates that the EMLs of devices B and D provide potentially high outcoupling efficiency of more than 30%. Device B showed a maximum EQE of 13%. This value correlates with the PLQY of the film of the corresponding guest–host system.

Fig. 7 TREL signals of OLEDs A, C, and D (a) and B, E, and F (b) Inset: an enlarged view of the spikes. Simulated (Θ of 1, 0.66), experimental, and fitting (Θ = 0.92) angle-dependent PL intensity profiles of the films of the molecular mixtures of Ir(ppy)₂acac (10 wt%) and compounds 1tCzBzCN (c) and 2tCzBzCN (d).

4. Conclusions

Two derivatives with 1-phenyl-1H-benzo[d]imidazole and di-tert-butylcarbazole groups were synthesized and analyzed for their potential as hosts for OLEDs with TADF capabilities. TADF of 2tCzBzCN was completely quenched by the addition of picric acid due to electronic excitation energy transfer. This demonstrates the great potential of TADF for optical sensing of nitroaromatic explosives. TADF of the synthesized emitters originates from the hybridization of local excited and charge transfer states. The solid films of the compounds show moderate values of photoluminescence quantum yields of up to 39%. The derivative of 1-phenyl-1H-benzo[d]imidazole with the single tert-butylcarbazolyl group performs better in organic light emitting diodes than the derivative with two tert-butylcarbazolyl groups. The effective electronic excitation energy transfer from the host to the guest is achieved due to vibronic coupling. The angle-dependent photoluminescence measurements performed for the films of the molecular mixtures of the compounds with the green phosphorescent emitter revealed the horizontal orientation factors of 0.84 and 0.87, suggesting enhanced out-coupling factors for the emitting layers. The performance of the compounds in OLEDs is primarily influenced by their charge-transport characteristics. The derivative of phenyl-1H-benzo[d]imidazole with the single tert-butylcarbazolyl group with hole and electron mobilities of 2.34 × 10⁻⁶ and 3.58 × 10⁻⁶ cm² V⁻¹ s⁻¹, respectively, at an electric field of 8.7 × 10⁵ V cm⁻¹ exhibited the best performance. OLEDs with this compound as a host showed an external quantum efficiency of 13%. The results of this study show that the search for hosts with high and balanced hole and electron mobilities, together with high internal outcoupling efficiency and triplet harvesting ability can result in the development of OLEDs with high external quantum efficiency.

Author contributions

Conceptualization: Simas Macionis, Jurate Simokaitiene, and Dalius Gudeika; methodology and investigation: Simas Macionis, Oleksandr Bezvikonnyi, Ehsan Ullah Rashid, Jiun-Haw Lee, Zi-Wen Su, Chia-Hsun Chen, Mariia Stanitska, and Oleksandr Navozenko; validation and data curation: Rita Butkute, Ruta Budreckiene, and Dmytro Volyniuk; formal analysis and writing of original draft: Simas Macionis, Jurate Simokaitiene, Oleksandr Bezvikonnyi, Dmytro Volyniuk, and Ehsan Ullah Rashid; writing – review and editing: Tien-Lung Chiu and Juozas. V. Grazulevicius; supervision and funding acquisition: Oleksandr Navozenko and Juozas. V. Grazulevicius.

Data availability

The data supporting this article have been included as part of the ESI.†

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This project has received funding from the Research Council of Lithuania (LMTLT), project “Exploitation of the solid-state enhanced long-lived emission of organic emitters for the detection of nitroaromatic explosive compounds”, agreement no. S-LU-24-6. J. H. Lee acknowledges the support and funding from the NSTC, grants 113-2221-E-002-102-MY3 and 112-2221-E-002-216-MY3. T. L. Chiu acknowledges the support and funding from the NSTC, grants 113-2622-E-155-002 and 112-2221-E-155-028-MY3.

References

1. E. Tankelevičiūtė, I. D. W. Samuel and E. Zysman-Colman, J. Phys. Chem. Lett., 2024, 15, 1034–1047.
2. T. Serevičius, R. Skaisgiris, G. Kreiza, J. Dodonova, K. Kazlauskas, E. Orentas, S. Tumkevičius and S. Juršenas, J. Phys. Chem. A, 2021, 125, 1637–1641.
3. S. W. Park, D. Kim and Y. M. Rhee, Int. J. Mol. Sci., 2023, 24, 12362.
4. Y. Im, M. Kim, Y. J. Cho, J. A. Seo, K. S. Yook and J. Y. Lee, Chem. Mater., 2017, 29, 1946–1963.
5. H. Yersin and W. J. Finkenzeller, Triplet Emitters for Organic Light-Emitting Diodes: Basic Properties, John Wiley & Sons, Ltd, 2008.
6. H. Yersin, Triplet Emitters for OLED Applications. Mechanisms of Exciton Trapping and Control of Emission Properties, Springer, Berlin, Heidelberg, 2012.
7. A. Jana, V. G. Sree, Q. Ba, S. C. Cho, S. U. Lee, S. Cho, Y. Jo, A. Meena, H. Kim and H. Im, J. Mater. Chem. C, 2021, 9, 11314–11323.
8. J. Jayabharathi, V. Thanikachalam and S. Thilagavathy, Coord. Chem. Rev., 2023, 483, 215100.
9. J. Cameron, A. Klimash, E. J. Hussien, F. Hacivelioğlu and P. J. Skabara, Sustainable Strategies in Organic Electronics, 2022, pp. 229–280.
10. A. Endo, K. Sato, K. Yoshimura, T. Kai, A. Kawada, H. Miyazaki and C. Adachi, Appl. Phys. Lett., 2011, 98, 083302.
11. B. Zhou, D. Yan, B. Zhou and D. Yan, Adv. Funct. Mater., 2019, 29, 1807599.
12. M. Chapran, P. Pander, M. Vasylieva, G. Wiosna-Salyga, J. Ulanski, F. B. Dias and P. Data, ACS Appl. Mater. Interfaces, 2019, 11, 13460–13471.
13. H. Uoyama, K. Goushi, K. Shizu, H. Nomura and C. Adachi, Nature, 2012, 492, 234–238.
14. H. Nakanotani, Y. Tsuchiya and C. Adachi, Chem. Lett., 2021, 50, 938–948.
15. T. Hosokai, H. Matsuzaki, H. Nakanotani, K. Tokumaru, T. Tsutsui, A. Furube, K. Nasu, H. Nomura, M. Yahiro and C. Adachi, Sci. Adv., 2017, 3, e1603282.
16. H. Noda, X. K. Chen, H. Nakanotani, T. Hosokai, M. Miyajima, N. Notsuka, Y. Kashima, J. L. Brédas and C. Adachi, Nat. Mater., 2019, 18, 1084–1090.
17. O. Bezvikonnyi, D. Gudeika, D. Volyniuk, A. Bucinskas and J. V. Grazulevicius, Mater. Sci. Eng., B, 2021, 273, 115441.
18. O. Bezvikonnyi, R. S. Bernard, V. Andruleviciene, D. Volyniuk, R. Keruckiene, K. Vaiciulaityte, L. Labanauskas and J. V. Grazulevicius, Materials, 2022, 15, 8495.
19. Z. J. Gao, T. H. Yeh, J. J. Xu, C. C. Lee, A. Chowdhury, B. C. Wang, S. W. Liu and C. H. Chen, ACS Omega, 2020, 5, 10553–10561.
20. J. J. Huang, Y. H. Hung, P. L. Ting, Y. N. Tsai, H. J. Gao, T. L. Chiu, J. H. Lee, C. L. Chen, P. T. Chou and M. K. Leung, Org. Lett., 2016, 18, 672–675.
21. S. Y. Chang, G. T. Lin, Y. C. Cheng, J. J. Huang, C. L. Chang, C. F. Lin, J. H. Lee, T. L. Chiu and M. K. Leung, ACS Appl. Mater. Interfaces, 2018, 10, 42723–42732.
22. Y. M. Chen, W. Y. Hung, H. W. You, A. Chaskar, H. C. Ting, H. F. Chen, K. T. Wong and Y. H. Liu, J. Mater. Chem., 2011, 21, 14971–14978.
23. C. Li, C. Duan, C. Han, H. Xu, C. Li, C. Duan, C. Han and H. Xu, Adv. Mater., 2018, 30, 1804228.
24. C.-Y. Chan, L.-S. Cui, J. Uk Kim, H. Nakanotani, C. Adachi, C. Chan, L. Cui, J. U. Kim, H. Nakanotani and C. Adachi, Adv. Funct. Mater., 2018, 28, 1706023.
25. S. M. Bonesi, M. Mesaros, F. M. Cabrerizo, M. A. Ponce, G. M. Bilmes and R. Erra-Balsells, Chem. Phys. Lett., 2007, 446, 49–55.
26. H. Ikebe, K. Nakao, E. Hisamura, M. Furukori, Y. Nakayama, T. Hosokai, M. Yang, G. Liu, T. Yasuda and K. Albrecht, Aggregate, 2024, 5, e405.
27. P. Ledwon, P. Zassowski, T. Jarosz, M. Lapkowski, P. Wagner, V. Cherpak and P. Stakhira, J. Mater. Chem. C, 2016, 4, 2219–2227.
28. A. Kadashchuk, Y. Skryshevski, A. Vakhnin, S. Toliautas, J. Sulskus, R. Augulis, V. Gulbinas, S. Nespurek, J. Genoe and L. Valkunas, J. Phys. Chem. C, 2014, 118, 22923–22934.
29. Y. A. Skryshevski and A. Y. Vakhnin, Ukr. J. Phys., 2019, 64, 406.
30. Y. Joo Cho, K. Soo Yook, J. Yeob Lee, Y. J. Cho, K. S. Yook and J. Y. Lee, Adv. Mater., 2014, 26, 6642–6646.
31. E. G. Cansu-Ergun, Polym. Rev., 2018, 58, 42–62.
32. C. Tardío, J. Álvarez Conde, A. M. Rodríguez, P. Prieto, A. de la Hoz, J. Cabanillas-González and I. Torres-Moya, Molecules, 2023, 28, 4631.
33. J. Jayabharathi, G. Goperundevi, V. Thanikachalam and S. Panimozhi, ACS Omega, 2019, 4, 15030–15042.
34. W. C. Chen, Z. L. Zhu and C. S. Lee, Adv. Opt. Mater., 2018, 6, 1800258.
35. J. Jayabharathi, J. Anudeebhana, V. Thanikachalam and S. Sivaraj, RSC Adv., 2020, 10, 8866–8879.
36. J. Tagare and S. Vaidyanathan, J. Mater. Chem. C, 2018, 6, 10138–10173.
37. Y. Liu, T. Tao, H. C. Hu, H. Li and X. Ouyang, Dyes Pigm., 2021, 188, 109191.
38. Y. H. Gao, C. Chen, Q. Tang, B. Su, G. Zhang, B. X. Bo and W. L. Jiang, J. Mater. Sci.: Mater. Electron., 2017, 28, 7204–7211.
39. C. Liu, T. Li, M. Sun, M. Xie, Y. Zhou, W. Feng, Q. Sun, S.-T. Zhang, S. Xue, W. Yang, C. Liu, T. Li, M. Sun, M. Xie, Y. Zhou, W. Feng, Q. Sun, S. Xue, W. Yang and S.-T. Zhang, Adv. Funct. Mater., 2023, 33, 2215066.
40. T. Hu, Y. Guo, G. Han and Y. Yi, J. Phys. Chem. C, 2024, 128, 3148–3154.
41. C. Mayr and W. Brütting, Chem. Mater., 2015, 27, 2759–2762.
42. S. Sohn, M. J. Kim, S. Jung, T. J. Shin, H. K. Lee and Y. H. Kim, Org. Electron., 2015, 24, 234–240.
43. D. Yokoyama, J. Mater. Chem., 2011, 21, 19187–19202.
44. C. K. Moon, K. H. Kim, J. W. Lee and J. J. Kim, Chem. Mater., 2015, 27, 2767–2769.
45. T. Marcato and C. J. Shih, Helv. Chim. Acta, 2019, 102, e1900048.
46. T. Chatterjee and K. T. Wong, Adv. Opt. Mater., 2019, 7, 1800565.
47. Y. Liu, M. Nishiura, Y. Wang and Z. Hou, J. Am. Chem. Soc., 2006, 128, 5592–5593.
48. S. Macionis, D. Gudeika, D. Volyniuk, O. Bezvikonnyi, V. Andruleviciene, J. Haw Lee, B.-A. Fan, C.-H. Chen, B.-Y. Lin, T.-L. Chiu and J. V. Grazulevicius, Adv. Opt. Mater., 2021, 9, 2002227.
49. S. H. Cheng, W. Y. Hung, M. H. Cheng, H. F. Chen, A. Chaskar, G. H. Lee, S. H. Chou and K. T. Wong, J. Mater. Chem. C, 2014, 2, 8554–8563.
50. S. Y. Chang, G. T. Lin, Y. C. Cheng, J. J. Huang, C. L. Chang, C. F. Lin, J. H. Lee, T. L. Chiu and M. K. Leung, ACS Appl. Mater. Interfaces, 2018, 10, 42723–42732.
51. Y. M. Huang, T. Y. Chen, D. G. Chen, H. C. Liang, C. H. Wu, M. M. Lee, T. L. Chiu, J. H. Lee, Y. C. Chiu, P. T. Chou and M. K. Leung, J. Mater. Chem. C, 2021, 9, 2381–2391.
52. A. Kundu, S. Karthikeyan, Y. Sagara, D. Moon and S. P. Anthony, ACS Omega, 2019, 4, 5147–5154.
53. J. Jayabharathi, R. Ramya, V. Thanikachalam, P. Jeeva and E. Sarojpurani, RSC Adv., 2019, 9, 2948–2966.
54. J. Jayabharathi, V. Thanikachalam, R. Ramya and S. Panimozhi, RSC Adv., 2019, 9, 33693–33709.
55. S. Arunlimsawat, P. Funchien, P. Chasing, A. Saenubol, T. Sudyoadsuk and V. Promarak, Beilstein J. Org. Chem., 2023, 19, 1664–1676.
56. T. Serevičius, R. Skaisgiris, J. Dodonova, L. Jagintavičius, D. Banevičius, K. Kazlauskas, S. Tumkevičius and S. Juršėnas, ACS Appl. Mater. Interfaces, 2020, 12, 10727–10736.
57. M. Mamada, H. Katagiri, C. Y. Chan, Y. T. Lee, K. Goushi, H. Nakanotani, T. Hatakeyama and C. Adachi, Adv. Funct. Mater., 2022, 32, 2204352.
58. K. Stavrou, L. G. Franca, A. Danos and A. P. Monkman, Nat. Photonics, 2024, 18, 554–561.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4tc04802d

---