High‑Efficiency Trilayer Phosphorescent OLEDs Without Electrode Modification Layers: Design, Mechanism, and Performance

Abstract

Functional layers are widely used to fine‑tune carrier injection and balance in OLEDs, but each added layer increases fabrication time and cost. Leveraging the dual functionality of certain materials can eliminate this need. Using impedance spectroscopy and transient electroluminescence, we show that di‑[4-(N,N-ditolyl‑amino)-phenyl] cyclohexane (TAPC) and 4,7‑diphenyl‑1,10‑phenanthroline (Bphen) act simultaneously as carrier‑injection and transport layers. The resulting trilayer OLEDs match the efficiency of conventional multilayer devices while simplifying the architecture. Detailed analysis of recombination and quenching mechanisms reveals that TAPC blocks electrons, Bphen suppresses hole accumulation, and balanced carrier transport reduces efficiency roll‑off.

Background

Organic light‑emitting devices (OLEDs) are pivotal for solid‑state lighting and high‑resolution displays. Conventional designs introduce several functional layers—anode modification layers (AMLs), cathode modification layers (CMLs), hole‑blocking layers (HBLs), and electron‑blocking layers (EBLs)—to optimize injection and confinement. However, each extra layer adds deposition steps, prolongs production, and increases cost, hindering industrial scalability. Advances in material chemistry now allow single compounds to fulfill multiple roles. For example, deoxyribonucleic acid‑cetyltrimethylammonium complexes serve as both high‑mobility hole transporters and electron‑blocking layers, while 4,4′,4″‑tris(carbazol‑9‑yl)-triphenylamine (TCTA) can act as both hole‑transport layer (HTL) and host in the emitting layer (EML). Such multifunctionality offers a path to simpler, cost‑effective OLEDs, yet reports on phosphorescent white OLEDs (PHWOLEDs) with minimal architectures remain scarce.

Impedance spectroscopy (IS) and transient electroluminescence (EL) have emerged as powerful diagnostics for OLED operation. The first peak in capacitance–voltage (C‑V) curves often aligns with the device turn‑on voltage, and the peak inflection is highly sensitive to carrier accumulation and injection barriers. Transient EL, driven by short voltage pulses, probes carrier dynamics and can reveal recombination pathways and delay times, essential for optimizing device performance. Here we combine IS and transient EL to demonstrate that TAPC and Bphen can serve as dual‑function layers, enabling efficient trilayer PHOLEDs without AMLs or CMLs. We further dissect the recombination and exciton‑quenching mechanisms governing device performance.

Methods/Experimental

Device Fabrication

The small‑molecule materials TAPC, Bphen, TmPyPB, and CBP were sourced from Luminescence Technology Corporation; dopants Ir(ppy)₃, Ir(MDQ)₂(acac), and FIrpic were obtained from Xi’an p‑OLED. All compounds were used as received. Glass substrates with patterned indium tin oxide (ITO) stripes were cleaned sequentially with Decon 90, deionized water, oven drying, and a 5‑minute plasma clean. PEDOT:PSS was spin‑coated to ~40 nm thickness and annealed at 120 °C for 10 min. Subsequent layers and the Mg:Ag (15:1) cathode were deposited by thermal evaporation under ~5 × 10⁻⁴ Pa, at 1–2 Å s⁻¹, monitored by a quartz crystal oscillator. Device active areas were 10 mm² defined by a shadow mask. Four distinct trilayer architectures were studied, differing in the HTL (TAPC) and ETL (Bphen) positioning, with or without additional thin interlayers (MoO₃, LiF, PEDOT:PSS).

Characterizations

Current density–voltage–luminance (J‑V‑L) curves and EL spectra were recorded in ambient air at room temperature using a Goniophotometric Measurement System (GP‑500). Transient voltage decay was measured with a high‑speed diode (1N4531), a Rigol DG5102 waveform generator, and a Rigol DS4054 oscilloscope, with signal averaging. Transient EL employed 1 ms voltage pulses from the DG5102, triggered the collection of EL via an avalanche photodiode (C30902) and a time‑correlated single‑photon counting system. Capacitance–voltage data were acquired with a TH2829C impedance analyzer (100 mV amplitude, 1 kHz), sweeping 0–10 V.

Results and Discussion

Efficient OLEDs Simplified Without AML

To eliminate the AML, TAPC was chosen as the HTL due to its HOMO aligning with ITO’s work function. Three devices were compared: (i) ITO/TAPC/CBP:10 wt % Ir(ppy)₃/TmPyPB/LiF/Mg:Ag (device D₁), (ii) ITO/MoO₃/TAPC/CBP:… (D₂), and (iii) ITO/PEDOT:PSS/TAPC/… (D₃). C‑V curves for all devices exhibit a first peak at ~3 V, indicating unchanged turn‑on voltage without an AML (Fig. 1a). The J‑V characteristics (Fig. 1b) show three distinct regimes: (I) leakage/Ohmic, (II) trap‑controlled volume current, and (III) partially filled‑trap conduction. D₁ demonstrates the highest low‑bias current density, attributed to a robust ITO/TAPC dipole that enhances hole injection. Although D₁ also shows the steepest J‑V slope in regime III (m = 11 vs. 7–8 for D₂/D₃), the higher trapping density is likely a consequence of TAPC’s morphology without a wetting layer.

Transient voltage decay under 5 V bias (Fig. 2b) reveals two decay components: a fast τ₁ (~100 µs) and a slower τ₂ (ms range). The rapid fall is governed by hole injection and transport; D₁ decays fastest, confirming superior hole injection through ITO/TAPC alone. The τ₂ component, influenced by internal resistance, shows minimal variation across devices. Transient EL delay times (t_d) extracted from 9 V pulses (Fig. 2d) are 0.32 µs (D₁), 1.05 µs (D₂), and 0.48 µs (D₃), underscoring that D₁ maintains excellent hole injection without an AML.

Efficient OLEDs Simplified Without CML

We further simplified the ETL by leveraging the self‑doping effect of Bphen with Mg:Ag cathodes. Devices S₁–S₃ used ITO/TAPC/CBP:10 wt % Ir(ppy)₃/ y/Mg:Ag, where y = TmPyPB, Bphen, or LiF. S₃ (Bphen) shows identical turn‑on voltage (3 V) and similar J‑V curves to S₁, confirming that Bphen alone provides robust electron injection. Transient EL onset times (t_d) are 0.32 µs (S₁), 1.05 µs (S₂), and 0.48 µs (S₃), again demonstrating that Bphen can replace a conventional CML.

Performance Comparison Between Simple Trilayer and Multilayer OLEDs

Three green PHOLEDs were fabricated: device 1 (with MoO₃ and LiF), device 2 (with LiF only), and device 3 (trilayer TAPC/CBP:10 wt % Ir(ppy)₃/Bphen). J‑V‑L curves (Fig. 3b) show device 3 has slightly lower current density and luminance, yet its turn‑on voltage remains 3 V, indicating that removing AML/CML does not impair carrier injection. Remarkably, device 3 exhibits the lowest efficiency roll‑off (Fig. 3c). We modeled exciton‑quenching using the following rate equations: (4)–(7) (see equations). The model captures the experimental EQE trends and demonstrates that triplet–triplet annihilation (TTA) and triplet–polaron annihilation (TPA) dominate at higher current densities. Device 3’s balanced carrier transport suppresses hole accumulation at the EML/ETL interface, reducing TPA and thereby improving high‑brightness efficiency.

Analysis of Exciton Recombination in Monochrome PHOLEDs

Low phosphorescent dopant concentrations create traps for charge carriers, leading to two recombination pathways: Langevin recombination (bimolecular) and trap‑assisted recombination. To probe these mechanisms, we fabricated devices with CBP doped 10 wt % Ir(ppy)₃ (green), 5 wt % Ir(MDQ)₂(acac) (red), and 15 wt % FIrpic (blue). Transient EL under 1 ms pulses (Fig. 5b,c) shows that the green and red devices’ rise times lengthen with increasing reverse bias, indicating that reverse bias releases trapped carriers. This behavior is absent in the blue device, suggesting negligible trap‑assisted recombination.

Capacitance–voltage measurements (Fig. 6a) reveal two pronounced peaks for green and red devices, with the first peak matching the turn‑on voltage. The subsequent peak decline near 3–3.5 V corresponds to trap‑assisted recombination of captured holes with electrons. In contrast, the blue device displays a single peak, confirming that only Langevin recombination occurs. The mathematical model (Eqs. 4–7) predicts higher polaron densities than triplet densities at low current densities, supporting the presence of TPA in the red device. TTA is mitigated by high Langevin recombination rates that reduce triplet density.

To further suppress quenching, we evaluated different host materials (CBP, TCTA, 26DCzPPy, TPBi). CBP’s balanced charge mobility and high triplet energy effectively limit both TTA and TPA, making it the optimal host for the studied devices.

Single‑Layer White OLEDs

We also fabricated a trilayer WOLED: ITO/TAPC/CBP:FIrpic:Ir(MDQ)₂(acac)/Bphen/Mg:Ag. The device turns on below 3 V and reaches a current efficiency of 21 cd A⁻¹. Normalized EL spectra (Fig. 7c) show a slight reduction of the red component at higher bias, attributable to trap‑assisted recombination of the red dopant. At 5 840 cd m⁻², the CIE coordinates (0.39, 0.39) correspond to warm‑white emission.

Conclusions

We demonstrate that TAPC and Bphen can each perform dual roles as injection and transport layers, enabling efficient phosphorescent OLEDs with a simple trilayer architecture (TAPC/EML/Bphen) that rivals conventional multilayer devices. Impedance spectroscopy and transient EL confirm the coexistence of Langevin and trap‑assisted recombination in green and red devices. By modeling TTA and TPA quenching, we rationalize the observed efficiency roll‑off behavior. Our approach delivers high‑efficiency, low‑roll‑off white OLEDs while reducing fabrication complexity and cost, advancing the practical deployment of OLED technologies.

Abbreviations

26DCzPPy:

2,6‑Bis(3‑(carbazol‑9‑yl)‑3‑phenylpyridine)

AML:

Anode modification layer

Bphen:

4,7‑Diphenyl‑1,10‑phenanthroline

C:

Capacitance

CBP:

4,4′‑N,N′‑Dicarbazole‑biphenyl

CE‑L‑EQE:

Current efficiency‑luminance‑external quantum efficiency

CML:

Cathode modification layer

C‑V:

Capacitance–voltage

C‑V‑L:

Capacitance–voltage–luminance

EBL:

Electron‑blocking layer

EL:

Electroluminescence

EML:

Emitting layer

EQE:

External quantum efficiency

ETL:

Electron‑transporting layer

FIrpic:

Bis[(4,6‑difluorophenyl)-pyridinato‑N,C²′]-picolinato Ir(III)

HBL:

Hole‑blocking layer

HOMO:

Highest occupied molecular orbital

HTL:

Hole‑transporting layer

IQE:

Internal quantum efficiency

Ir(MDQ)₂(acac):

Iridium(III) bis‑(2‑methyldibenzo‑[f,h]‑quinoxaline) acetylacetonate

Ir(ppy)₃:

Tris(2‑phenylpyridine) iridium

IS:

Impedance spectroscopy

ITO:

Indium tin oxide

J‑V:

Current density–voltage

J‑V‑L:

Current density–voltage–luminance

LUMO:

Lowest unoccupied molecular orbital

OLEDs:

Organic light‑emitting devices

PEDOT:PSS:

Poly(3,4‑ethylenedioxythiophene)-poly(styrene sulfonate)

PHWOLEDs:

Phosphorescent white OLEDs

TAPC:

Di‑[4-(N,N-ditolyl‑amino)-phenyl] cyclohexane

TCTA:

4,4′,4″‑Tris(carbazol‑9‑yl)-triphenylamine

TmPyPB:

1,3,5‑Tri(m‑pyrid‑3‑yl‑phenyl) benzene

TPA:

Triplet‑polaron annihilation

TPBi:

2,2′‑[2″‑1,3,5‑benzinetriyl]-tris(1‑phenyl‑1‑H‑benzimidazole)

TTA:

Triplet‑triplet annihilation