Backbone modulation of thermally activated delayed fluorescence polymers for efficient orange-red emission in solution-processed OLEDs

Abstract

Thermally activated delayed fluorescence (TADF) enables near-unity internal quantum efficiency by harvesting triplet excitons via reverse intersystem crossing, but designing orange-red TADF polymers with high radiative rates and suppressed non-radiative loss remains challenging. Herein, design of orange-red TADF polymers via backbone engineering that tunes excited-state energies and energy-transfer pathways is reported. Carbazole and dibenzofuran (DBF) units are incorporated into the polymer backbones with various loadings of a naphthalimide–dimethylacridine unit. Altering DBF linkage (3,7- vs. 2,8-) modulates the photophysical properties of the polymer, enhancing energy transfer and exciton migration while reducing non-radiative decay. The polymer pNAI-DBF3705 exhibits a photoluminescence quantum yield of 78% and an accelerated reverse intersystem crossing rate of 8.76 × 10⁵ s⁻¹. Solution-processed OLEDs based on pNAI-DBF3705 deliver a maximum external quantum efficiency of 10% and improved operational stability at high driving voltages. These results highlight backbone engineering as an effective strategy to optimize excited-state dynamics in TADF polymers for high-performance, solution-processable orange-red OLEDs.

---

Introduction

Organic light-emitting diodes (OLEDs) utilizing thermally activated delayed fluorescence (TADF) have emerged as a cutting-edge technology in display and lighting applications due to their ability to achieve near 100% internal quantum efficiency without relying on rare or expensive heavy metals.^1–4 Unlike conventional fluorescent OLEDs, which only utilize singlet excitons, or phosphorescent OLEDs that depend on heavy-metal complexes to harvest triplet excitons, TADF-based OLEDs exploit both singlet and triplet excitons by leveraging the reverse intersystem crossing (RISC) process. This process efficiently converts non-emissive triplet excitons into emissive singlet excitons, significantly enhancing the overall electroluminescence efficiency of the device.^5–8 The fundamental principle behind TADF is minimizing the energy gap (ΔE_ST) between the lowest singlet and triplet states, enabling efficient thermal up-conversion of triplet excitons via RISC. This leads to higher external quantum efficiencies (EQEs), while eliminating the need for expensive and environmentally concerning heavy-metal-based phosphorescent materials.^9–11

TADF based OLEDs are actively being developed to optimize efficiency, stability, and cost-effectiveness.^12,13 Among these, TADF polymers, particularly for solution-processed polymer light-emitting diodes (PLEDs), have garnered significant attention due to their potential for large-scale, low-cost manufacturing.^11,14,15 Conjugated TADF polymers, in particular, exhibit enhanced charge transport properties due to their extended π-conjugation along the polymeric backbones, making them highly suitable for high-performance OLEDs. However, the design and synthesis of efficient orange-red TADF polymers remain challenging.^16–18 In addition, inefficient energy transfer from the polymer backbone to red-emitting TADF chromophores can cause exciton loss as backbone emission in higher energy regions.^19–21

Backbone engineering is therefore a crucial strategy to tune photophysical properties and improve TADF polymer performance.^11,18 Incorporating donor–acceptor (D–A) motifs, selecting appropriate co-hosts, and modulating connectivity can enhance charge transport, exciton diffusion, and energy-transfer efficiency.^22–24 Carbazole derivatives are widely used as hole-transporting hosts because of their high triplet energies and favorable hole mobility.²⁵ However, their inefficient energy transfer to red-emitting TADF units often necessitates the incorporation of a secondary host to improve exciton management. Dibenzofuran (DBF) has emerged as an effective co-host material due to its high singlet and triplet energy levels, efficient energy transfer properties, and strong hole transport capabilities.^11,26 Integrating DBF into the polymeric backbone enhances energy transfer efficiency, minimizes exciton loss, and improves overall OLED performance. DBF facilitates a more effective energy transfer pathway, ensuring optimal exciton migration to the emissive sites. Additionally, the rigid molecular structure of DBF contributes to reduced non-radiative transitions, thereby improving the photoluminescence quantum yield (PLQY). Its ability to facilitate balanced charge injection and transport further enhances device efficiency.^27–32 This dual role of DBF as both a co-host and a charge transport facilitator establishes it as a key component in the development of high-performance orange-red TADF polymers.¹¹

In this study, we have designed a novel orange-red conjugated polymer backbone incorporating both carbazole and DBF units (Fig. 1). To introduce TADF properties, we have strategically incorporated high-efficiency naphthalimide-dimethylacridine (NAI-DMAc) moieties, where dimethylacridine acts as the electron donor and is integrated directly into the polymeric chains. To fine-tune the optoelectronic properties, we synthesized polymers with varying concentrations of TADF units. Furthermore, two distinct polymer series were meticulously developed to regulate the excited-state characteristics by controlling the connection sites between DBF and TADF segments.³³ All synthesized polymers exhibit orange-red emission centered around 615 nm with distinct TADF behavior. Notably, the positioning of DBF units within the polymeric backbone significantly influences the excited-state dynamics and photophysical properties. The pNAI-DBF37 series, in which DBF is incorporated at the 3,7-positions, exhibits a higher PLQY and an accelerated reverse intersystem crossing rate (k_RISC) compared to those with DBF linked at the 2,8-positions. Among these, the pNAI-DBF3705, containing 5% TADF units, achieves an exceptional PLQY of 78% and a k_RISC of 8.56 × 10⁵ s⁻¹. OLEDs fabricated using pNAI-DBF3705 demonstrate a maximum EQE_max of 10% while maintaining superior operational stability at high voltages. These results highlight that the combination of DBF and carbazole in the polymer backbone is highly effective for optimizing orange-red TADF performance.

Fig. 1 The design and synthetic pathways for the two series of TADF polymers.

Results and discussion

Molecular design and synthesis

Two series of TADF polymers were synthesized via Suzuki polymerization by incorporating carbazole (Cz) and dibenzofuran (DBF) units into the polymer backbone. Due to its higher energy level compared to carbazole, DBF acts as an exciton transporter within the polymer backbone, as shown in Fig. S1, facilitating energy transfer to the carbazole units. To thoroughly understand how specific connectivity influences polymer properties, DBF units were systematically introduced at distinct positions (3,7 and 2,8) within polymeric chains. Two distinct polymer series were synthesized with bisborate-DBF incorporated explicitly at either the 3,7-positions (designated as pNAI-DBF3702/3705/3710) or the 2,8-positions (pNAI-DBF2802/2805/2810). For each series, three different molar ratios (2%, 5%, and 10%) of TADF-emitting units (NAI-DMAc) were integrated to systematically optimize the photophysical properties.¹⁸ The synthetic pathways for the monomers are depicted in Scheme S1, and their chemical structures were verified through ¹H NMR, ¹³C NMR, and mass spectrometry, as shown in Fig. S2–S16. The relative molar content of TADF units in the polymers was determined using ¹H NMR spectra by analyzing the integral ratios of hydrogen atoms from DMAc moieties and alkyl carbazole units, Fig. S17–S18. The measured TADF unit proportions were consistent with the intended feed ratios across all synthesized polymers.

The molecular weight distribution and polydispersity indices (PDIs) were analyzed using gel permeation chromatography (GPC). The weight-average molecular weights (M_w) of the polymers ranged from 7.7 kDa to 34.8 kDa, with PDIs varying between 1.2 and 2.0 (Fig. S19 and Table S1). Thermal properties were evaluated via differential scanning calorimetry (DSC) and thermogravimetric analysis (TGA). The observed glass transition temperatures (T_g) ranging from 99 °C to 163 °C indicate excellent structural rigidity and thermal stability (Fig. S20), which are crucial attributes for stable OLED operation under practical device conditions. Notably, increasing TADF unit content substantially increases T_g, further underscoring the structural stability benefits provided by TADF moieties. The decomposition temperatures exceeded 400 °C (Fig. S21), confirming excellent thermal stability, which is essential for practical optoelectronic applications. The incorporation of long alkyl chains into the carbazole units notably enhanced solubility in common organic solvents, significantly facilitating solution-processing techniques and enabling efficient film formation for OLED device fabrication.

Density functional theory calculations

Density functional theory (DFT) calculations were conducted to gain deeper insights into how the polymer backbone structure influences the photophysical behavior of TADF polymers incorporating dibenzofuran (DBF). For computational simplicity, representative molecular fragments consisting of one TADF unit, two DBF units, and two extended carbazole segments were modeled. The frontier molecular orbitals (FMOs), specifically the highest occupied molecular orbitals (HOMOs) and lowest unoccupied molecular orbitals (LUMOs), were computed using Gaussian software at the B3LYP/6-31G(d,p) level of theory.

Analysis of the computational results revealed distinct characteristics of orbital distribution. The LUMOs are consistently localized on the electron-accepting naphthalimide (NAI) cores within the TADF units, exhibiting LUMO energies of approximately −2.55 eV for the pNAI-DBF37 series and −2.57 eV for the pNAI-DBF28 series, as shown in Fig. 2a. Conversely, significant differences were observed in HOMO distributions due to variations in DBF connectivity within the polymer backbone. Polymers with DBF units linked specifically at the 3,7-positions demonstrate an extended and more uniform HOMO distribution along the polymeric backbone, enabling a broader electron delocalization. In contrast, polymers containing DBF units connected at the 2,8-positions exhibit twisted conformations, significantly limiting HOMO extension onto the DBF units, as shown in Fig. 2a. Notably, oxygen atoms in the DBF units contribute significantly to the HOMOs in the pNAI-DBF28 series but not in the pNAI-DBF37 series. This difference led to a slightly smaller energy gap (E_g) for the pNAI-DBF37 series due to the broader HOMO extension. Moreover, the incorporation of DBF into the polymer backbone not only optimizes the HOMO–LUMO distribution but also enhances the oscillation strength of the TADF emitters. This improvement in oscillation strength directly translates into enhanced light-emitting properties, facilitating more efficient exciton utilization and better performance in OLED devices. The enhanced electron delocalization resulting from the rigid structure of DBF further contributes to the increase in PLQY and RISC efficiency. The isolated distributions of the frontier molecular orbitals (FMOs) result in small ΔE_ST values of 6 meV for the pNAI-DBF37 series and 7 meV for the pNAI-DBF28 series, promoting efficient RISC. These findings underscore the significant role of DBF in improving the electronic structure and optical properties of TADF polymers, leading to enhanced performance in light-emitting applications.

Fig. 2 (a) Frontier molecular orbital analyses of two representative polymer fragments, displaying the spatial distributions of their HOMO and LUMO, along with their respective energy levels. (b) Electrostatic potential mapped onto isosurfaces derived from the total self-consistent field electron density for these polymeric fragments.

Electrostatic potential analysis revealed continuous electron-rich regions along the backbones of the pNAI-DBF37 series, promoting efficient carrier and exciton migration.³⁴ In contrast, the twisted backbones of the pNAI-DBF28 series lead to more dispersed electron-rich regions, suggesting that the pNAI-DBF37 series may exhibit superior charge transport properties, as shown in Fig. 2b. These findings underscore the importance of backbone design in optimizing the photophysical and charge transport properties of DBF-based TADF polymers.

Electrochemical properties

The electrochemical properties of the six polymers were investigated by cyclic voltammetry (CV) in degassed anhydrous acetonitrile. As shown in Fig. S22, all polymers display quasi-reversible oxidation and reduction processes. In line with frontier molecular orbital (FMO) simulations, the LUMO levels remain nearly constant at approximately –3.50 eV, dominated by reduction processes of the naphthalimide units. Conversely, the oxidation is mainly associated with the electron-donating dimethylacridine moieties and the polymer backbone. Within each polymer series, decreasing the TADF unit content progressively lowers the HOMO level and widens the energy gap (E_g). For polymers with identical TADF loadings, the pNAI-DBF37 series exhibits consistently higher HOMO levels and smaller E_g values than the corresponding pNAI-DBF28 series, in good agreement with FMO predictions. These results indicate that backbone connectivity modulates HOMO distribution and thereby governs the electrochemical characteristics of the polymers.

Natural transition orbital calculation

The excited-state properties of TADF polymers were examined using computational analysis based on natural transition orbitals (NTOs). Time-dependent density functional theory (TD-DFT) calculations were conducted to evaluate the excited-state behavior of representative polymeric fragments, with NTO analysis performed using Multiwfn software.³⁵ To quantify spin–orbit coupling matrix elements (SOCMEs) between singlet and triplet states, calculations were carried out using the ORCA 4.1.1 package at the B3LYP/G-TZVP level.³⁶ A comprehensive overview of excited-state energy diagrams for two representative polymer structures is provided in Fig. S23, while Table S2 presents the computed SOCMEs for transitions between the S₁ and T_n states.

As illustrated in Fig. 3, the spatial distributions of hole and electron densities in the excited states are depicted in blue and pink, respectively. The overlap integral 〈Ψ_hΨ_e〉, which measures the spatial correlation between hole and electron wavefunctions, serves as an indicator of localized versus charge-transfer excitation. Both the S₁ (singlet charge-transfer, ¹CT) and T₁ (triplet charge-transfer, ³CT) states of these polymer series exhibit pronounced CT characteristics, while the T₂ states demonstrate localized triplet (³LE) behavior. FMO analysis reveals that while the electron density distributions are largely similar across both polymer series, differences in hole density patterns arise due to varying oxygen atom contributions. Specifically, in the pNAI-DBF37 series, the hole density region is largely devoid of oxygen participation, whereas in the pNAI-DBF28 series, oxygen atoms significantly contribute to the hole orbitals. This structural variation results in a slightly increased localization index for the ³CT state in pNAI-DBF37 (0.118) compared to pNAI-DBF28 (0.116).

Fig. 3 The illustration of the energy-level diagrams for excited states of two polymer series and the natural transition orbital analysis for representative segments of these polymers and spin–orbit coupling matrix elements for (a) the pNAI-DBF37 series and (b) the pNAI-DBF28 series.

A key finding is the reduced energy gap (ΔE_1CT-3LE) between the ¹CT and the nearly degenerate ³LE states in the pNAI-DBF37 series. According to the second-order vibronic coupling mechanism, this smaller energy separation enhances the efficiency of RISC by facilitating spin–flip transitions mediated via the ³LE state.³⁷ These results underscore the critical role of DBF substitution patterns, specifically the 3,7-position, in governing excited-state dynamics, with the pNAI-DBF37 series exhibiting superior photophysical properties. Further analysis employing Fermi's golden rule emphasizes the critical role of SOC in determining k_RISC.³⁸ Specifically, the pNAI-DBF37 series exhibits stronger SOC interactions between ¹CT and ³CT states (〈¹CT|Ĥ_SOC|³CT〉), significantly enhancing k_RISC compared to the pNAI-DBF28 series (〈¹CT|Ĥ_SOC|³CT〉). This improved coupling in the pNAI-DBF37 polymer arises from optimized molecular orbital distributions, facilitating efficient spin–flip transitions and effective charge transfer processes.

Although the SOC value between the ¹CT and ³LE, represented as (〈¹CT|Ĥ_SOC|³LE〉), is slightly lower for the pNAI-DBF37 series compared to pNAI-DB28, the smaller ΔE_ST between the ¹CT and ³CT states in pNAI-DBF37 (6 meV) relative to pNAI-DB28 (7 meV) significantly enhances the RISC efficiency. According to second-order vibronic coupling mechanisms, smaller ΔE_ST values strongly favor efficient RISC processes.¹⁷ Therefore, the combined effects of reduced ΔE_ST and enhanced SOC between ¹CT and ³CT states create an effective dual-channel RISC mechanism, substantially boosting the overall photophysical performance of the pNAI-DBF37 polymer. This clearly demonstrates that strategic backbone engineering, focusing on precise tuning of excited-state energies and SOC characteristics, is essential for optimizing the performance of TADF polymers.

Photophysical properties

A comprehensive photophysical analysis was conducted to elucidate the influence of polymer backbone structures on their optical properties. As illustrated in Fig. 4a, the UV-visible absorption spectra of the synthesized polymer series in a pure film exhibit a characteristic bimodal absorption in the 200–400 nm range, attributed to π–π* transitions. The distinct connection sites of DBF units play a crucial role in modulating the HOMO distribution along the polymeric backbone, directly affecting absorption characteristics. The pNAI-DBF37 series, exhibiting a broader HOMO delocalization, undergoes a more pronounced redshift in absorption compared to the pNAI-DBF28 series, consistent with CV measurements. A weak absorption band appearing between 400 and 600 nm is ascribed to charge transfer transitions associated with TADF units. In terms of photoluminescence (PL) behavior, all polymers display an intense emission centered around 620 nm in the solid state. Notably, at an equivalent TADF unit fraction, the emission peak of pNAI-DBF3702 is redshifted by approximately 5 nm relative to pNAI-DBF2802 due to enhanced HOMO delocalization, corroborating FMO simulations and CV results. A progressive redshift in PL spectra is observed with an increasing TADF unit proportion, indicative of a decreasing E_g, aligning with electrochemical findings. However, the extent of this redshift differs between polymer series: the pNAI-DBF37 series display a modest shift of 12 nm from pNAI-DBF3702 to pNAI-DBF3710, whereas the pNAI-DBF28 series undergo a more substantial shift of 21 nm from pNAI-DBF2802 to pNAI-DBF2810. These disparities likely arise from variations in polymeric aggregation states governed by DBF attachment positions. The extended backbone and larger spatial occupation of pNAI-DBF37 polymers reduce intermolecular interactions, thereby limiting spectral shifts, while the more compact structure of the pNAI-DBF28 series enhances aggregation-induced effects, amplifying the redshift.

Fig. 4 Ultraviolet-visible absorption spectra (dashed lines) and photoluminescence spectra (solid lines) of pure polymer films for (a) the pNAI-DBF37 series and (b) the pNAI-DBF28 series. (c) Photoluminescence quantum yields of polymer films (10 wt%) doped into the mCP-CN host matrix (90 wt%), measured in a vacuum. These values were obtained by adjusting the PLQYs measured in air based on the integrated intensity ratio between vacuum and air emission spectra. Transient photoluminescence decay profiles of doped polymer films under vacuum conditions for (d) the pNAI-DBF37 series and (e) the pNAI-DBF28 series. (f) Temperature-dependent transient photoluminescence decay curves of the doped pNAI-DBF3705 films.

The photoluminescence quantum yields (PLQYs) of polymer-doped films (10 wt%) were systematically investigated using 9H-carbazole-3-carbonitrile (mCP-CN) as the host matrix. The host material mCP-CN was selected based on the superior PLQY values of the doped films, as shown in Fig. S24. The PLQYs measured under vacuum conditions were extrapolated from their air-measured counterparts using the integration ratio of steady-state spectra in both environments. As illustrated in Fig. 4c and Fig. S25, all polymer films exhibit an enhanced steady-state emission intensity under vacuum compared to air-exposed conditions, suggesting the active involvement of triplet excitons in the photoluminescence process. This enhancement is attributed to the well-known oxygen quenching effect on triplet excitons in air, a characteristic behavior of TADF materials. Notably, at an equivalent TADF unit fraction, the pNAI-DBF37 series exhibit superior PLQYs, with pNAI-DBF3705 achieving the highest value of 78% (Fig. 4c). This enhancement is likely due to the elongated and spatially expanded backbone structure of the pNAI-DBF37 series, which effectively suppresses aggregation-induced quenching of triplet excitons. Additionally, a more uniform HOMO distribution facilitates efficient energy transfer from the polymer backbone to the TADF units. Within each polymer series, the highest PLQY is observed for materials containing 5% TADF units. In contrast, pNAI-DBF2802 and pNAI-DBF3702 exhibit lower PLQYs, primarily due to inefficient energy transfer between the polymer backbone and the TADF units, leading to the formation of exciton traps. Conversely, polymers with a higher TADF unit concentration, such as pNAI-DBF3710 and pNAI-DBF2810, exhibit diminished PLQYs, likely due to exciton quenching caused by aggregation at elevated exciton densities. The PLQY trend reflects a balance between energy transfer and concentration quenching. The optimum ratio (5%) achieves efficient exciton funneling without aggregation-induced quenching, yielding the highest PLQY (78%) with accelerated k_RISC. These findings highlight the intricate interplay between the polymer backbone structure, exciton dynamics, and aggregation effects in determining the overall photophysical efficiency of TADF-based conjugated polymers.

The fluorescence decay dynamics further elucidate the impact of DBF attachment positions on the photophysical behavior of the polymers, as illustrated in Fig. 4d and e. The correlation between the TADF unit concentration and polymeric lifetimes is consistent with the observed PLQY trends. Polymers with higher TADF content, such as pNAI-DBF3710 and pNAI-DBF2810, exhibit shorter delayed fluorescence lifetimes of approximately 1.3–1.6 μs and reduced delayed fluorescence ratios of 53.1 and 61.3, primarily due to exciton quenching at elevated TADF unit concentrations. Conversely, while pNAI2802 and pNAI3702 possess fluorescence lifetimes and delayed fluorescence ratios comparable to pNAI2805 and pNAI3705, their lower radiative transition rates suggest the presence of exciton traps, leading to reduced luminescence efficiency. Notably, pNAI-DBF2805 and pNAI-DBF3705 achieve a well-balanced interplay between the prompt fluorescence decay rate (k_p) and the reverse intersystem crossing rate (k_RISC) within their respective series, effectively minimizing non-radiative losses from the singlet excited state and optimizing overall exciton utilization. Furthermore, the k_RISC of pNAI-DBF3705 is significantly higher at 8.76 × 10⁵ s⁻¹, attributed to its shorter fluorescence lifetime, likely benefiting from its backbone architecture. This suggests that the strategic incorporation of DBF units and precise control over their attachment positions via backbone engineering can effectively enhance RISC, thereby improving the performance of TADF polymers.

To further verify the TADF characteristics, fluorescence decay measurements of doped polymer films were conducted at different temperatures in a vacuum, as presented in Fig. 4f and Fig. S26, and Table S3. Taking pNAI-DBF3705 as a representative example, its k_RISC decreases from 8.76 × 10⁵ s⁻¹ at 300 K to 1.87 × 10⁵ s⁻¹ at 77 K. The observed positive temperature dependence of all polymers reinforces the intrinsically endothermic nature of the RISC process, providing clear evidence of their TADF properties. These results underscore the critical role of structural engineering in tuning the exciton dynamics of conjugated polymers for enhanced optoelectronic performance. All photophysical data are presented in Table 1.

Table 1 Summary of the photophysical properties of all polymers

	λAbs^a [nm]	λPL^b [nm]	PLQY^c [%]	τ p [ns]	τ d [μs]	φ d [%]	k p [10^6 s^−1]	k d [10^5 s^−1]	k RISC [10^5 s^−1]	k ^sr [10^5 s^−1]	k ^snr [10^5s^−1]
a The peak wavelength observed in the ultraviolet-visible absorption spectra of the pure polymer films. b The peak wavelength in the photoluminescence spectra of pure polymer films. c The photoluminescence quantum yields of the doped films in a vacuum, calculated using the PLQYs of the films in air and the ratio of the steady-state emission intensities of the corresponding films in a vacuum to those in air. d The prompt fluorescence lifetime component measured in a vacuum. e The delayed fluorescence lifetime component measured in a vacuum. f The delayed fluorescence component, determined by fitting the transition decay curves in a vacuum. g The decay rate for prompt fluorescence. h The decay rate for delayed fluorescence. i The rate constants for reverse intersystem crossing. j The rate constant for singlet exciton radiative recombination. k The nonradiative rate constants for the singlet excited state.
pNAI-DBF3710	238/363	624	65	27.1	1.32	53.1	11.3	2.62	5.58	34.4	18.5
pNAI-DBF3705	238/363	615	78	28.6	1.96	68.8	8.51	2.74	8.76	20.8	5.9
pNAI-DBF3702	238/363	612	55	27.8	2.31	72.2	5.49	1.72	6.19	8.39	6.9
pNAI-DBF2810	252/302	628	54	26.7	1.60	61.3	7.84	2.07	5.35	16.4	13.9
pNAI-DBF2805	252/302	620	60	28.6	2.18	74.0	5.47	2.04	7.82	8.56	5.7
pNAI-DBF2802	252/302	613	45	26.4	2.48	75.0	4.26	1.36	5.46	4.79	5.9

---

Device properties

Prior to fabricating solution-processed OLEDs, the film-forming properties of six different polymers were analyzed using atomic force microscopy (AFM). The thin films were prepared via spin-coating under the same conditions employed for device fabrication. As depicted in Fig. S27, uniform and smooth polymeric films were successfully obtained when 10 wt% of the polymers were incorporated into the mCP-CN host matrix. Within a scanned area of 5 × 5 μm², the doped films exhibit low surface roughness (R_a), ranging from 0.282 to 0.350 nm. The architecture of the solution-processed OLEDs is illustrated in Fig. 5a, comprising an indium tin oxide (ITO) anode, a 30 nm poly(3,4-ethylenedioxythiophene):poly(styrenesulfonate) (PEDOT:PSS) hole-injecting and hole-transporting layer, a 40 nm emissive layer containing the doped emitters, a 45 nm 1,3,5-tri[(3-pyridyl)-phen-3-yl]benzene (TmPyPB) electron-transporting layer, a 0.9 nm lithium fluoride (LiF) electron-injection layer, and a 130 nm aluminum (Al) cathode. The inclusion of 10 wt% emitters within the mCP-CN host was aimed at optimizing the balance between hole and electron injection and transport. Fig. 5a presents the energy level alignment of each OLED layer, with polymer energy levels determined via CV, while values for other materials were sourced from existing literature. All EL data are presented in Table 2.

Fig. 5 (a) Architecture of the solution-processed device and the corresponding energy level alignment. (b) Electroluminescence spectra measured at an approximate luminance of 10 cd m⁻². (c) Current density–voltage–luminance (J–V–L) characteristics. (d) Relationship between EQE and luminance.

Table 2 Electroluminescence properties of solution-processed OLED based polymers

	V on [V]	L max [cd m^−2]	ELpeak^c [nm]	CEmax^d [cd A^−1]	PEmax^e [lm W^−1]	EQEmax,100,500^f [%]	CIE^g [x,y]
a Turn-on voltage measured at a luminance of 1 cd m^−2. b Peak luminance achieved by the device. c Electroluminescence peak wavelength recorded at a luminance of 10 cd m^−2. d Maximum current efficiency. e Highest power efficiency observed. f Maximum external quantum efficiency, along with values measured at 100 cd m^−2 and 500 cd m^−2. g Chromaticity coordinates based on the Commission Internationale de l'Éclairage (CIE) system.
pNAI-DBF3710	4.8	1392	622	9.73	6.37	6.8, 3.3, 2.9	0.59, 0.38
pNAI-DBF3705	5.0	2127	616	12.90	7.50	10.0, 7.2, 4.7	0.59, 0.39
pNAI-DBF3702	4.8	1186	608	7.94	5.19	5.0, 3.1, 2.4	0.55, 0.38
pNAI-DBF2810	5.4	1393	616	7.40	3.90	5.1, 3.2, 2.6	0.59, 0.39
pNAI-DBF2805	5.4	1454	612	10.10	5.82	6.1, 3.5, 2.9	0.56, 0.39
pNAI-DBF2802	5.0	599.6	604	7.54	4.74	4.2, 2.1, 1.5	0.52, 0.38

---

As demonstrated in Fig. 5b and Fig. S28a, the EL spectra of the TADF polymers exhibit emission peaks within the 604–622 nm range, corresponding to the orange-red region on the Commission Internationale de L’Eclairage (CIE) color coordinates. The pNAI-DBF37 series displays redshifted emissions compared to the pNAI-DBF28 series, aligning with the molecular simulation and photophysical characterization results. Notably, due to the lower concentration of TADF units, pNAI-DBF3702 and pNAI-DBF2802 exhibit emission peaks at 608 nm and 604 nm, respectively. In contrast, the emission peaks shift to 622 nm and 616 nm for pNAI-DBF3710 and pNAI-DBF2810. Similarly, the polymers containing 5% TADF units, pNAI-DBF3705 and pNAI-DBF2805, show emission peaks at 616 nm and 612 nm, respectively.

For pNAI-DBF3702 and pNAI-DBF2802, the lower proportion of TADF units introduces more structural defects, which may contribute to noticeable electroluminescence instability at high operating voltages. As shown in Fig. 5c and Fig. S28b, this instability becomes evident in the J–V–L curves, which display distinct turning points when the applied voltage exceeds 6 V. Additionally, the maximum luminance for these polymers is relatively low, reaching just above 1100 cd m⁻². A significant efficiency drop is observed at high current densities, leading to the lowest EQE_max among all of the investigated polymers.

The turn-on voltage for both pNAI-DBF3710 and pNAI-DBF3702 is 4.8 V, while that for pNAI-DBF3705 is 5.0 V and those for the pNAI-DBF28 series are slightly higher at 5.4 V. This difference may be attributed to the more continuous electrostatic potential distributions in the pNAI-DBF37 series, which facilitate more efficient energy transfer from the polymeric backbone to the TADF units. Consequently, at the same proportion of TADF unit concentrations, the device performance of the pNAI-DBF37 series surpasses that of the pNAI-DBF28 series. Furthermore, as illustrated in Fig. 5d and Fig. S28c, the efficiency roll-off in the pNAI-DBF28 series is more pronounced than that in the pNAI-DBF37 series. This discrepancy is likely due to variations in the polymeric chain structure and aggregation behavior, which are influenced by the distinct attachment positions of the DBF units within the polymer backbone. At higher operating voltages, the polymer backbones of pNAI-DBF3710 and pNAI-DBF3705 maintain robust exciton transport capabilities, contributing to improved overall device stability and performance.

As expected, and in agreement with simulation calculations and photophysical characterization, pNAI-DBF3705 demonstrates the highest device performance among the six polymers due to its optimized molecular architecture and doping concentration. This polymer achieves a maximum luminance (L_max) of 2127 cd m⁻², a peak current efficiency (CE_max) of 12.90 cd A⁻¹, a maximum power efficiency (PE_max) of 7.50 lm W⁻¹, and an EQE_max of 10.0%. The superior performance of pNAI-DBF3705 can be attributed to the suppression of concentration-induced quenching. This study demonstrates that optimizing the polymeric backbone structure and host material selection can significantly enhance the EQE and mitigate efficiency roll-off in orange-red TADF polymeric OLEDs. Among the investigated polymers, pNAI-DBF3705 exhibits the highest performance, maintaining exceptional stability and efficiency across different brightness levels. Its superior characteristics make it a promising candidate for next-generation high-efficiency orange-red TADF polymeric OLED applications.

Conclusions

In conclusion, this study successfully demonstrates the strategic design and synthesis of novel orange-red TADF polymers through backbone engineering by incorporating DBF and carbazole units in conjugation with efficient NAI-DMAc based TADF emitters. Comprehensive computational, electrochemical, and photophysical analyses have shown that suitable modification in the polymer backbone, particularly the site-specific attachments of DBF units, profoundly influences the molecular orbital distribution, spin–orbit coupling, and energy transfer properties. The optimized polymer pNAI-DBF3705 exhibits an exceptionally high PLQY of 78% and an accelerated k_RISC of 8.76 × 10⁵ s⁻¹, resulting in significantly improved OLED performance with an EQE_max of 10% and enhanced operational stability. These results highlight the potential of precise molecular engineering in TADF polymer design, providing valuable guidelines for developing highly efficient and stable, and solution-processable orange-red emitters for OLED applications.

Author contributions

L. H., S. Y. and Z. R. initiated and supervised the project. M. U. synthesized and characterized the TADF polymers. M. U. and L. H. performed photophysical and electrochemical measurements of the TADF emitters. L. H. and Y. L. completed the preparation and characterization of OLED devices. Z. C. and M. U. performed the computational calculation. M. U., Z. R. and L. H. wrote the manuscript. All authors discussed the progress of the research and reviewed the manuscript.

Conflicts of interest

There are no conflicts to declare.

Data availability

The data supporting this article, including test methods, synthesis procedures and supplementary photophysical data have been included as part of the supplementary information (SI). See DOI: https://doi.org/10.1039/d5tc03166d.

Acknowledgements

The financial support from the National Natural Science Foundation of China (nos. 52273164, and 22405027), the Shandong Provincial Natural Science Foundation (ZR2022ZD37) and the Natural Science Foundation of the Jiangsu Higher Education Institutions of China (24KJB150003) is gratefully acknowledged.

Notes and references

1. Y. X. Hu, J. Miao, T. Hua, Z. Huang, Y. Qi, Y. Zou, Y. Qiu, H. Xia, H. Liu, X. Cao and C. Yang, Nat. Photonics, 2022, 16, 803–810.
2. T. Yang, J. Liang, Y. Cui, Z. Li, X. Peng, S.-J. Su, Y. Wang and C. Li, Adv. Opt. Mater., 2023, 11, 2201191.
3. Y. Liu, C. Li, Z. Ren, S. Yan and M. R. Bryce, Nat. Rev. Mater., 2018, 3, 18020.
4. Z. Zhang, T. Ding, Q. Zhou, Y. Sun, M. Qu, Z. Zeng, Y. Ju, L. Li, K. Wang and F. Chi, Renewable Sustainable Energy Rev., 2021, 148, 111263.
5. M. I. Alam, M. R. Nagar, J.-H. Jou and S. Vaidyanathan, Chem. Mater., 2024, 36, 8649–8662.
6. J. Liu, Z. Zhao, Q. Li, L. Hua, H. Zhao, C. Yu, W. Cao and Z. Ren, ACS Appl. Mater. Interfaces, 2023, 15, 51474–51482.
7. S. Diesing, L. Zhang, E. Zysman-Colman and I. D. W. Samuel, Nature, 2024, 627, 747–753.
8. K. Shi, Y. Xie, L. Hua, S. Li, Z. Yang, Y. Yin, Z. Wang, S. Ying, Y. Liu, Z. Ren and S. Yan, ACS Mater. Lett., 2024, 6, 1491–1503.
9. Z. Huang, B. Lei, D. Yang, D. Ma, Z. Bin and J. You, Angew. Chem., Int. Ed., 2022, 61, e202213157.
10. Y. Liu, X. Tong, X. Chen, Y. Wang, S. Ying, Z. Ren and S. Yan, ACS Appl. Mater. Interfaces, 2021, 13, 8997–9005.
11. M. U. Khan, L. Hua, Y. Liu, H. Zhao, Y. Guo, Y. Wang, S. Yan and Z. Ren, J. Mater. Chem. C, 2025, 13, 4523–4532.
12. Y. X. Zhang, Y. Y. Zheng, C. S. Shi, J. Y. Zhang, T. Wang, N. Sun and Q. Wang, J. Mater. Chem. C, 2022, 10, 2034–2041.
13. P. Li, S.-L. Lai, Z. Chen, W. K. Tang, M.-Y. Leung, M. Ng, W.-K. Kwok, M.-Y. Chan and V. W.-W. Yam, Chem. Sci., 2024, 15, 12606–12615.
14. X. Yin, Y. He, X. Wang, Z. Wu, E. Pang, J. Xu and J. A. Wang, Front. Chem., 2020, 8, 725.
15. T. Wang, Y. Zou, Z. Huang, N. Li, J. Miao and C. Yang, Angew. Chem., Int. Ed., 2022, 61, e202211172.
16. Y. Guo, L. Hua, H. Zhao, Y. Liu, S. Yan and Z. Ren, J. Phys. Chem. C, 2024, 128, 11924–11932.
17. Y. Guo, Z. Zhao, L. Hua, Y. Liu, B. Xu, Y. Zhang, S. Yan and Z. Ren, ACS Appl. Mater. Interfaces, 2023, 16, 1225–1233.
18. L. Hua, Y. Liu, H. Zhao, S. Chen, Y. Zhang, S. Yan and Z. Ren, Adv. Funct. Mater., 2023, 33, 2303384.
19. Z. Zhao, X. Tong, Y. Liu, R. Wan, H. Li, S. Yan and Z. Ren, ACS Appl. Mater. Interfaces, 2023, 15, 4257–4266.
20. T. Wang, K. Li, B. Yao, Y. Chen, H. Zhan, Z. Xie, G. Xie, X. Yi and Y. Cheng, Adv. Funct. Mater., 2020, 30, 2002493.
21. Y. Yang, L. Zhao, S. Wang, J. Ding and L. Wang, Macromolecules, 2018, 51, 9933–9942.
22. X. Tong, Z. Zhao, L. Hua, Y. Zhang, B. Xu, Y. Liu, S. Yan and Z. Ren, Angew. Chem., Int. Ed., 2023, 33, 2305324.
23. Y. Yang, K. Li, C. Wang, H. Zhan and Y. Cheng, Chem. Asian J., 2019, 14, 574–581.
24. T. Wang, B. Yao, K. Li, Y. Chen, H. Zhan, X. Yi, Z. Xie and Y. Cheng, Adv. Opt. Mater., 2021, 9, 2001981.
25. P. Khammultri, P. Chasing, C. Chitpakdee, S. Namuangruk, T. Sudyoadsuk and V. Promarak, RSC Adv., 2021, 11, 24794–24806.
26. X. Lan, J. Zeng, J. Chen, T. Yang, X. Dong, B. Z. Tang and Z. Zhao, Angew. Chem., Int. Ed., 2025, 64, e202414488.
27. T. Wang, Z. Huang, H. Zhang, J. Miao and C. Yang, Adv. Funct. Mater., 2024, 34, 2408119.
28. C. Adachi, M. A. Baldo, S. R. Forrest, S. Lamansky, M. E. Thompson and R. C. Kwong, Appl. Phys. Lett., 2001, 78, 1622–1624.
29. M. Wang, Z. Fu, R. Cheng, J. Du, T. Wu, Z. Bin, D. Wu, Y. Yang and J. Lan, Chem. Commun., 2023, 59, 5126–5129.
30. L. Wan, Z. Cheng, X. Ma, Y. Jiang, Z. Yan, Y. Yan, Z. Su, F. Liu and P. Lu, Chem. Eur. J., 2025, 31, e202404653.
31. W.-W. Tao, K. Wang, J.-X. Chen, Y.-Z. Shi, W. Liu, C.-J. Zheng, Y.-Q. Li, J. Yu, X.-M. Ou and X.-H. Zhang, J. Mater. Chem. C, 2019, 7, 4475–4483.
32. D. R. Lee, S.-H. Hwang, S. K. Jeon, C. W. Lee and J. Y. Lee, Chem. Commun., 2015, 51, 8105–8107.
33. L. Hua, H. Wu, S. Yan, W. Zhu, Z. Ren and Y. Wang, Chem. – Eur. J., 2025, 31, e202501391.
34. M. Bian, D. Zhang, Y. Wang, Y.-H. Chung, Y. Liu, H. Ting, L. Duan, Z. Chen, Z. Bian, Z. Liu and L. Xiao, Adv. Funct. Mater., 2018, 28, 1800429.
35. T. Lu and F. Chen, J. Comput. Chem., 2012, 33, 580–592.
36. F. Neese, Wiley Interdiscip. Rev.: Comput. Mol. Sci., 2012, 2, 73–78.
37. M. K. Etherington, J. Gibson, H. F. Higginbotham, T. J. Penfold and A. P. Monkman, Nat. Commun., 2016, 7, 13680.
38. Z. Zhao, Y. Liu, L. Hua, S. Yan and Z. Ren, Adv. Funct. Mater., 2022, 32, 2200018.

---