Interfacial dipole-engineered fiber optoelectronic devices with enhanced efficiency and stability

Abstract

Organic light-emitting diodes (OLEDs) have been developed to enhance device lifetime, efficiency, and operational stability. However, the widely used hole injection layer (HIL) material poly(3,4-ethylenedioxythiophene):poly(styrene sulfonate) (PEDOT:PSS) exhibits limitations such as high work function and acidity, which degrade device performance. This study introduces a [2-(9H-carbazol-9-yl)ethyl]phosphonic acid (2PACz) self-assembled monolayer (SAM) as an alternative to PEDOT:PSS. 2PACz-based OLEDs achieved lower turn-on voltages and higher external quantum efficiencies (EQEs) compared with PEDOT:PSS-based devices. The maximum EQE of green and red fiber organic light emitting diodes (FOLEDs) were 10.71% and 8.97%, respectively, representing 16.9% and 12.9% improvements compared with those of reference devices using PEDOT:PSS as the HIL. Furthermore, compared with TiO₂ fiber-shaped dye-sensitized solar cells (FS-DSSCs), the incorporation of TiO₂/2PACz increased the power conversion efficiency (PCE) from 5.67% to 6.53%, corresponding to an improvement of approximately 17%. Notably, the TiO₂/2PACz-based fiber-shaped gas sensors (FS-GSs) also exhibited enhanced gas sensing characteristics, including increased response and sensitivity, highlighting the multifunctionality and broad applicability of this interfacial engineering strategy across diverse optoelectronic platforms.

Graphical abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s40580-026-00549-x.

Introduction

Organic light-emitting diodes (OLEDs) serve as key components in modern display and lighting technologies owing to their vivid image quality and high efficiency [1, 2]. In OLED structures, the hole injection layer (HIL) is essential for facilitating hole injection and reducing energy barriers at the anode interface. Since the early development of OLEDs, poly(3,4-ethylenedioxythiophene) (PEDOT):poly(styrene sulfonate)(PSS) has been widely adopted as an HIL material owing to its water solubility with solution-based processability. This compatibility enables its application to both flexible [3, 4] and rigid substrates, supporting diverse applications. Furthermore, PEDOT:PSS features low surface roughness, thereby enhancing charge injection efficiency and achieving high conductivity. Its energy level, approximately 5.0 eV, closely aligns with that of indium tin oxide (ITO), enabling effective hole injection. These characteristics have established PEDOT:PSS as a widely utilized material in optoelectronic devices, including OLEDs [5–7], organic solar cells [8, 9], and perovskite solar cells [10, 11]. However, PEDOT:PSS presents several limitations. The acidic nature of PEDOT:PSS causes degradation at the ITO interface and facilitates the diffusion of indium into the PEDOT:PSS layer. This etching effect disrupts interface stability and lowers charge injection efficiency, ultimately reducing device performance [12]. The high work function of PEDOT:PSS creates a hole injection barrier at the ITO interface, leading to an increased turn-on voltage (V_on) [13]. Furthermore, the hygroscopic PSS group absorbs moisture, which impacts surrounding layers and results in device instability [14, 15]. These limitations are not specific to OLEDs but extend to other optoelectronic devices as well [16].

To address these limitations, several strategies have been explored, including the use of neutral PEDOT:PSS [17], metal-doped PEDOT:PSS [18], sandwich architectures wherein PEDOT:PSS is layered between protective materials [19], and replacement with alternative organic or inorganic materials [20, 21]. This study investigates the potential of self-assembled monolayers (SAMs) as a viable alternative to conventional HILs. SAMs form ultra-thin molecular films that strongly bond to metal oxide surfaces, effectively reducing interface resistance, enhancing charge injection efficiency, and minimizing material degradation [22]. Unlike PEDOT:PSS, SAM-based HILs address acidity and stability issues, offering enhanced performance and durability in OLEDs [23, 24].

This study focused on utilizing [2-(9H-carbazol-9-yl)ethyl]phosphonic acid (2PACz) as the HIL in fiber-type optoelectronic devices. The 2PACz molecules form an oriented self-assembled monolayer, where the anchoring phosphonic acid group ensures strong surface binding [25], while the carbazole moiety induces a net dipole moment, leading to an upward shift of the vacuum level and an increase in the effective work function. This interfacial energetic modulation reduces the hole injection barrier and facilitates efficient charge injection, as evidenced by the reduced turn-on voltage and enhanced current density [26]. Moreover, the improved injection characteristics promote balanced carrier transport, thereby suppressing charge accumulation and minimizing non-radiative recombination losses, which is consistent with the enhanced optoelectronic device efficiency and stability [27–30].

In fiber-shaped OLEDs (FOLEDs) comprising 2PACz, the V_on for green and red devices decreased from 4.02 to 3.55 V and from 4.3 V to 4.07 V, respectively, compared with those of reference devices using PEDOT:PSS as the HIL layer. Furthermore, the maximum external quantum efficiency (EQE_max) for green devices increased from 8.23% to 10.71% and from 7.34% to 8.97%, respectively. In planar OLEDs, the V_on for green and red devices decreased from 5.53 to 4.07 V and from 5.34 to 4.36 V, respectively, whereas the EQE_max increased from 4.67% to 5.46% and from 7.98% to 9.01%. Optoelectronic devices utilizing 2PACz exhibit improved power conversion efficiency (PCE) compared with previous studies [31, 32], which is attributed to the suppression of charge recombination and the promotion of charge transport efficiency. The optimized fiber-shaped dye-sensitized solar cells (FS-DSSCs) exhibits a PCE of up to 6.53%, representing an improvement of 15% compared with that of a reference device.

Experimental section

Materials for FOLEDs

ITO-patterned 1 in. × 1 in. glass substrates were used for planar OLEDs, while polyethylene terephthalate (PET)/ITO substrates were used for FOLEDs. The substrates were cleaned using acetone and isopropyl alcohol (IPA), both obtained from Samchun Chemical. Ethanol (Samchun Chemical) served as the solvent for the HIL material 2PACz (> 98.0%; Tokyo Chemical Industry Co., Ltd.), which was prepared as a 0.5 mM solution. The emission layer (EML) was composed of a host blend of tris(4-carbazoyl-9-ylphenyl)amine (TCTA) and 1,3,5-tris(1-phenyl-1H-benzimidazol-2-yl)benzene (TPBi) doped with tris[2-(p-tolyl)pyridine]iridium(III) (Ir(mppy)₃) and DG-Red-2 as the green and red emitter, respectively. Ir(mppy)₃ was sourced from LumTec and DG-Red-2 was prepared following methods described in previous studies. Chlorobenzene (CB), used as the solvent for EML preparation, and lithium fluoride (LiF), used as the electron injection layer (EIL), were both obtained from Sigma-Aldrich. Aluminum, which served as the cathode material, was sourced from Taewon Science.

Fabrication of the FOLEDs

The flexible PET substrate coated with ITO (PET/ITO) underwent a cleaning process via sequential ultrasonic treatment in deionized water, acetone, and IPA. Subsequently, the substrate was treated with UV-ozone for 10 min. The IPA solution was mixed with PEDOT:PSS (AI 4083) in a volume ration of 1:1 for use at the HIL. The mixture was vortexed and ultrasonicated for 10 min, and then filtered using PVDF. The HIL solution was spin-coated onto the PET/ITO substrate at 4,000 rpm for 30 s, resulting in a film thickness of 40 nm after drying at 100 °C for 10 min. For the EML precursor, a mixture of 48.5% TCTA:48.5% TPBi:3% Ir(mppy)₃ in CB with a concentration of 10 mg mL⁻¹ was prepared. The EML precursor was spin-coated onto the PEDOT:PSS film at a speed of 1,000 rpm for 40 s, resulting in a film thickness of 40 nm after drying at 80 °C for 5 min. Next, the ETL, EIL, and cathode layer were sequentially thermally evaporated onto the EML inside a vacuum chamber at a pressure of less than 6 × 10⁻⁶ Torr. The ETL consisted of TPBi with a thickness of 40 nm, the EIL was LiF with a thickness of 1 nm, and the cathode was aluminum with a thickness of 100 nm.

Materials for FS-DSSCs

Titanium wire (99.7%, Ф250 μm; Sigma-Aldrich) was utilized as the working electrode and platinum wire (99.9%, Ф125 μm; Wooillmetal) was used as the counter electrode in solid-state FDSSCs (SS-FDSSCs). TiO₂ paste (18NR-T; Greatcellsolar) was diluted with anhydrous ethanol 99.8% (1:1 w/w) to prepare a colloid TiO₂ solution and stirred at room temperature for 24 h using a magnetic bar. N719 dye (Greatcellsolar) was dissolved in anhydrous ethanol 99.8% at a concentration of 0.5 mM. The iodine-based electrolyte (I⁻/I₃⁻), which is a commercial electrolyte (High Stability Electrolyte; EL-HSE, and High Performance Electrolyte; EL-HPE) was purchased from Greatcellsolar. 4AT (> 98.0%) was purchased from Tokyo Chemical Industry Co., Ltd. 4AT derivatives were prepared using methanesulfonyl chloride (> 99.7%) and p-toluenesulfonyl chloride (≥ 98%), both purchased from Sigma-Aldrich.

Preparation of the Li-TFSI film

Bis(trifluoromethane) sulfonamide lithium salt (Li-TFSI) was dissolved in a propylene carbonate solution at 3 wt%. During the casting process, this solution was carefully poured into a suitable container with optical clear resin to prevent the formation of bubbles. The Li-TFSI mixture was then subjected to a drying process and cured using UV light.

Fabrication of FS-DSSCs

After cutting the Ti wire (60 mm), it was washed for 10 min in an ultrasonic bath using deionized water, acetone, and IPA. The cleaned Ti wire was dried for 10 min and then subjected to ozone treatment for 10 min. Subsequently, it was electrically heated at a current of 2.0 A for 20 s in air, during which oxygen in the air was oxidized on the surface of the ozone-treated Ti wire to generate a compact TiO₂ (c-TiO₂) layer. For the ETL, the TiO₂ colloid solution was diluted in anhydrous ethanol at 1:1 (w/w) and stirred for 12 h. The Ti wire on which the c-TiO₂ was formed was subjected to dip-coating in the diluted TiO₂ solution at room temperature and annealing at 120 °C for 3 min. This process was repeated four times to control the thickness of the mesoporous-TiO₂ (mp-TiO₂) layer. The mp-TiO₂ layer formed on the Ti/c-TiO₂ wire was crystallized by electric heating by applying a current of 1.2–2.0 A for 5 min. The Ti/c-TiO₂/mp-TiO₂ wire was immersed in the N719 dye solution for 20 h. A Li-TFSI film, which is a solid-state electrolyte, was wound around the Ti/c-TiO₂/mp-TiO₂/N719 dye wire and then spirally wound with Pt wire (counter electrode). The fabricated device was immersed in the iodine-based redox couple electrolyte that was excellently penetrated and absorbed by the Li-TFSI film, and then dried at room temperature to maintain a solid-state device. As a final step, the completed FS-DSSCs were inserted into a transparent Teflon tube to seal both ends with an optical adhesive that was cured with UV light. The experimental methods and processes are detailed in our previous reports [33, 34].

Fabrication of FS-GSs

After cutting the Ti wire (60 mm), it was washed for 10 min in an ultrasonic bath using deionized water, acetone, and IPA. The cleaned Ti wire was dried for 10 min and then subjected to ozone treatment for 10 min. Subsequently, it was electrically heated at a current of 2.0 A for 20 s in air, during which oxygen in the air was oxidized on the surface of the ozone-treated Ti wire to generate a compact TiO₂ (c-TiO₂) layer. For the ETL, the TiO₂ colloid solution was diluted in anhydrous ethanol at 1:1 (w/w) and stirred for 12 h. The Ti wire on which the c-TiO₂ was formed was subjected to dip-coating in the diluted TiO₂ solution at room temperature and annealing at 120 °C for 3 min. This process was repeated two times to control the thickness of the mesoporous-TiO₂ (mp-TiO₂) layer. The mp-TiO₂ layer formed on the Ti/c-TiO₂ wire was crystallized by electric heating by applying a current of 1.2–2.0 A for 5 min.

Characterization

Atomic force microscopy (AFM) measurements were carried out using an environment control AFM system (AFM 5300E, HITACHI) with an Rh-coated Si tip (SI-DF3-R). A Si probe with a stiffness of 0.97 N/m and a resonance frequency of 24 kHz was used in the non-contact mode AFM with a scan area of 1 μm × 1 μm and a pixel resolution exceeding 512 × 256 at room temperature and ambient pressure. The surface morphology was analyzed by calculating the average surface smoothness of the ITO/PEDOT:PSS and ITO/2PACz surfaces. The transmittance was determined employing UV–Visible-NIR spectroscopy (Cary 5000, Agilent Technologies). The current density–voltage–luminance (J–V–L) characteristics, CIE 1931 chromaticity diagram, EQE, current efficiency (CE), and power efficiency (PE) of the FOLEDs were measured using an in-house setup by integrating the McScience M3000 OLED Parameter test system. The LT₅₀ measurements were conducted on non-encapsulated FOLEDs under ambient conditions. Ultraviolet photoelectron spectroscopy (UPS) was performed using a gas discharge UV source housed at ESCALAB MKII (He I radiation; hν = 21.2 eV). The take-off angle was set to 45° relative to the sample surface. During UPS, the sample was biased at − 3.0 V relative to the ground potential to observe low-energy secondary electron cutoff.

The UV photoelectron emission was detected with a concentric hemispherical analyzer with a pass-energy of 50 eV. Photoluminescence (PL) spectra were recorded using a Hamamatsu Quantaurus-QY spectrometer by exciting the samples at 365 nm. The PL lifetime was measured using a Hamamatsu Quantaurus-Tau fluorescence lifetime measurement system, where pulsed laser excitation at 375 nm provided emission decay data for exponential fitting. The J–V properties of the SS-FDSSCs were measured using a solar simulator (Newport 450 W 3 A solar system, Oriel) at AM 1.5, with illumination as a light source. The calibrated light intensity was set to 1 Sun (100 mW cm⁻²) using a standard silicon cell. The effective area of the device was defined as the projected area transmitted by the mask, which is equal to the diameter of the photoanode multiplied by its length (1 cm). The incident photon-to-current conversion efficiency (IPCE) for EQE spectra was measured using an incident photon-to-current conversion efficiency measurement system (QuantX 300, Oriel) with an Oriel Cornerstone™ 130 1/8 m monochromator operated in AC mode. The impedance response was measured from 1 Hz to 1 MHz with an oscillation amplitude of 15 mV in the dark (Bio-Logic VMP-3). The experimental data were simulated using the commercial Z-view software to estimate the values of each component of the corresponding equivalent circuits.

The gas-sensing performance of the functionalized TiO₂/2PACz-based FS-GSs was evaluated using a customized gas-sensing system. The setup consisted of a test chamber (VPX-G400, WIT Co., Ltd., Korea) equipped with inlet and outlet valves to regulate the gas flow. Target gases such as H₂ and high-purity air were introduced into the chamber through a mass flow controller (MFC, GMC1200, ATOVAC Co., Ltd., Korea) for precise flow-rate control. The sensor device was mounted on a stage holder inside the chamber and electrically connected to a Source Meter unit (Model 2450, Keithley Instruments, USA) interfaced with a computer for real-time current monitoring. Sensing tests were performed by exposing the fiber-shaped device to H₂ diluted in air while maintaining the chamber temperature at 30 °C and applying a bias voltage of 1 V. To evaluate gas selectivity, the sensor was subsequently exposed to various analytes, including H₂, CO, CO₂, NH₃, and NH₄, each supplied from 100-ppm gas cylinders. Repeatability was confirmed through multiple measurement cycles conducted under identical H₂ concentrations and experimental conditions, and the current responses were recorded for each cycle. The gas response (R) was calculated using the following relation:

where I_t and I_air represent the instantaneous current in target gas and pure air, respectively. The response time was defined as the period required for the sensor current to reach 90% of its maximum variation upon gas exposure, while the recovery time corresponded to the duration needed to achieve 90% of the return to baseline during desorption.

Results and discussion

Figure 1a illustrates the device architecture of the green FOLEDs. The devices were fabricated on ITO-coated substrates, serving as the anode. A conventional PEDOT:PSS HIL) was used as a reference, while a 2PACz SAM was introduced as an alternative interfacial layer to evaluate its effect on charge injection and device performance. As shown in Fig. 1b, 2PACz forms a densely packed SAM on the ITO surface via its phosphonic acid anchoring group, inducing an interfacial dipole that enhances the anode work function and improves energy level alignment for hole injection. Figure 1c shows the current density–voltage–luminance (J–V–L) characteristics of the green FOLEDs using PEDOT:PSS and 2PACz as the HIL. The device incorporating 2PACz exhibits a pronounced lower V_on compared with that of the reference device with PEDOT:PSS, indicating more efficient hole injection at the anode/organic interface. At a given voltage, the 2PACz-based device exhibits a higher current density and luminance than the PEDOT:PSS-based device. This behavior is attributed to the reduced hole injection barrier and improved energy level alignment provided by the interfacial dipole of the 2PACz SAM, which leads to more efficient charge injection and improved charge balance. Figure 1d shows EQE as a function of current density for green FOLEDs employing PEDOT:PSS and 2PACz as the HIL. The device incorporating 2PACz consistently exhibits a higher EQE over the entire measured current density range compared with the PEDOT:PSS-based device. Figure 1e shows the CE and PE as a function of current density for green FOLEDs employing PEDOT:PSS and 2PACz as the HIL. The device incorporating 2PACz exhibits consistently higher CE and PE across the entire measured current density range compared with the PEDOT:PSS-based device. Figure 1f shows the normalized electroluminescence (EL) spectra of the green FOLEDs employing PEDOT:PSS and 2PACz as the HIL, as well as the corresponding CIE color coordinates. Both devices exhibit nearly identical EL spectra with a peak emission centered at approximately the same wavelength, indicating that the introduction of 2PACz does not alter the intrinsic emission characteristics of the emissive layer.

Fig. 1.

Characteristics of FOLEDs with 2PACz. a Schematic diagram of the green FOLED structure, a luminescence photograph, and the ESP map of 2PACz calculated by DFT. b Energy level diagram of FOLED. c J–V–L characteristics with the PEDOT:PSS and 2PACz-based devices. d EQE and e CE and PE as functions of current density. f EL spectra; the inset shows the CIE 1931 chromaticity diagramd

The role of 2PACz in the FOLED was investigated via density functional theory (DFT) simulations. All calculations were performed using the NWChem 7.2 package at the B3LYP/cc-PVTZ level of theory. The calculated HOMO level of 2PACz is − 5.76 eV, which is consistent with the experimental result shown in Fig. 1b. Compared with PEDOT:PSS (5.2 eV), the HOMO level of 2PACz is more closely aligned with that of TCTA (5.8 eV), thereby reducing V_on. Furthermore, the enhanced CE achieved by 2PACz is supported by its calculated dipole moment. The dipole moment of 2PACz is 1.9 D, which is of the same order of magnitude as those of related materials such as MeO-2PACz and MeO-PhPACz [35]. The electrostatic potential (ESP) map of 2PACz (Fig. 1a), further illustrates the formation of an interfacial dipole layer when 2PACz is anchored on the ITO surface.

Figure 2 shows the electroluminescent performance of red FOLEDs employing either PEDOT:PSS or 2PACz as the HIL. Compared with green FOLEDs, red FOLEDs typically suffer from less favorable energy level alignment owing to the mismatch HOMO energy levels of red phosphorescent emitters, which often results in increased hole injection barriers and efficiency roll-off. The J–V–L characteristics (Fig. 2c) revealed that the 2PACz-based red FOLED exhibits a lower V_on and enhanced luminance at a given voltage compared with the PEDOT:PSS-based device. Figure 2d shows that the 2PACz-based red FOLED maintains a higher EQE across the entire current density range with suppressed efficiency roll-off. This behavior suggests improved charge balance and reduced non-radiative recombination, which are essential for achieving stable performance in long-wavelength OLEDs. The 2PACz-based FOLEDs achieved a maximum EQE of 8.97%, representing a substantial 28% improvement compared with that (6.39%) of the PEDOT:PSS-based device The 2PACz-based FOLEDs attained a high CE of 7.42 cd/A and PE of 3.18 lm/W (Fig. 2e). Compared with the PEDOT:PSS-based device, the normalized EL spectra and CIE color coordinates (Fig. 2f) remain unchanged with the introduction of 2PACz, indicating preserved emission characteristics and color purity of the red emitter. These key parameters of the PEDOT:PSS- and 2PACz-based green and red FOLEDs are summarized in Table 1. Furthermore, the EL performance of green and red glass-based OLEDs is summarized in Figs. S1 and S2, respectively, and Table S1. The variation in efficiency observed among the glass-based OLEDs likely arises from differences in interfacial morphology, film formation, and subtle variations in processing conditions, which can influence charge injection and recombination dynamics. Nevertheless, the overall trend remains consistent, with the 2PACz-modified devices exhibiting systematically improved performance compared to the reference devices.

Fig. 2.

Characteristics of FOLEDs comprising 2PACz. a Schematic of red FOLED structure and a luminescence photograph. b Energy level diagram of FOLED. c J–V–L characteristics with the PEDOT:PSS- and 2PACz-based devices d EQE and e CE and PE as functions of current density. f EL spectra; the inset shows the CIE 1931 chromaticity diagram

Table 1.

Performance of FOLEDs (10 devices)

Color	HIL	Von (V)	Lmax (cd m−2)	EQEmax (%)	CEmax (cd A−1)	PEmax (Im W−1)	CIE (x, y)
Green	PEDOT:PSS	4.02 ± 0.1	12,957	8.23 ± 1.5	28.63 ± 3.6	9.44 ± 3.8	(0.30 ± 0.1, 0.62 ± 0.1)
Green	2PACz	3.55 ± 0.1	23,591	10.71 ± 1.1	37.44 ± 3.4	16.38 ± 3.6	(0.31 ± 0.1, 0.62 ± 0.1)
Red	PEDOT:PSS	4.30 ± 0.1	2496	6.39 ± 1.3	5.34 ± 1.4	2.58 ± 1.1	(0.66 ± 0.1, 0.34 ± 0.1)
Red	2PACz	4.07 ± 0.1	6297	8.97 ± 1.1	7.42 ± 1.1	3.18 ± 1.0	(0.66 ± 0.1, 0.34 ± 0.1)

The UPS data, as shown in Fig. S3, provide insights into the electronic properties of the ITO/PEDOT:PSS and ITO/2PACz structures. The work function was calculated using the following equation:

where E_F is the Fermi level binding energy, E_SC is the secondary electron cutoff, and hν is the photon energy (21.22 eV in this experiment). The measured E_F values for PEDOT:PSS and 2PACz were 1.5 and 1.8 eV, respectively, while the corresponding E_SC values were 17.5 and 17.3 eV. Based on these values, the calculated work function for the ITO/PEDOT:PSS and ITO/2PACz structures were 5.2 and 5.7 eV, respectively. Figure S4 shows the conductive atomic force microscopy (C-AFM) current maps of ITO surfaces modified with (a) PEDOT:PSS and (b) 2PACz, measured over an area of 5 μm × 5 μm under the same bias conditions. The PEDOT:PSS-coated ITO exhibits a relatively low and uniform current distribution with an average surface current (Sₐ) of approximately 0.36 nA, indicating limited local charge transport pathways at the anode interface. By contrast, the ITO surface modified with 2PACz displays a significantly enhanced and spatially heterogeneous current distribution, with a substantially increased average surface current of approximately 29 nA. The presence of highly conductive regions suggests the formation of efficient local charge transport channels induced by the densely packed 2PACz SAM.

Figure 3a shows the AFM images of ITO surfaces modified with PEDOT:PSS and 2PACz. The PEDOT:PSS-coated ITO exhibits a relatively smooth surface with a root-mean-square (RMS) roughness of 0.91 nm. By contrast, the ITO surface treated with 2PACz shows an increased RMS roughness of 1.95 nm, in conjunction with a more granular surface morphology. The increased surface roughness observed for the 2PACz-modified ITO can be attributed to the formation of a densely packed SAM anchored via phosphonic acid groups. Despite the increase in roughness, the 2PACz layer remains ultrathin and uniform, without introducing morphological defects that could deteriorate device performance. Figure 3b shows the optical transmittance spectra of bare ITO, ITO/PEDOT:PSS, and ITO/2PACz in the wavelength range of 400–800 nm. All samples exhibit high optical transparency across the visible region, which is essential for efficient light outcoupling in OLED devices. Compared with bare ITO, the introduction of PEDOT:PSS leads to a slight reduction in transmittance, particularly in the shorter wavelength region, due to absorption by the polymer layer. By contrast, the ITO/2PACz sample maintains a transmittance comparable to that of bare ITO over the entire visible range, indicating negligible optical loss induced by the ultrathin 2PACz SAM. Figure 3c and d shows the PLQY of green and red emitting films, respectively, as a function of aging time for the PEDOT:PSS- and 2PACz-modified films. The PLQY of the PEDOT:PSS films deteriorates more rapidly with aging, indicating progressive degradation of the emissive layer under ambient conditions. However, the 2PACz-modified films exhibit a significantly slower PLQY decay and maintains a higher PLQY over the entire aging period. This improved photoluminescence stability suggests that the 2PACz interfacial layer effectively suppressed non-radiative recombination pathways and mitigates interfacial degradation processes. The operational lifetime of the FOLEDs was evaluated at a constant current density to assess the stability of the PEDOT:PSS and 2PACz films, as shown in Fig. 3e. The LT₅₀ values at an initial luminance of 100 cd/m² were 32 and 63 min for the PEDOT:PSS and 2PACz FOLEDs, respectively. The extended operational lifetime clearly indicates improved FOLEDs stability achieved via interfacial engineering with 2PACz as the HIL. Therefore, the enhanced stability can be attributed to the chemically robust and non-acidic nature of 2PACz, which suppresses interfacial degradation processes and reduces non-radiative loss pathways.

Fig. 3.

a AFM images of the PEDOT:PSS and 2PACz films. b Transmittance of the PEDOT:PSS and 2PACz films. PLQY of c green and d red PEDOT:PSS and 2PACz films. e Normalized luminance decay of FOLEDs

Figure 4a illustrates the architecture of the FS-DSSCs investigated in this study. The coaxial configuration of an FS-DSSCs comprises a Ti wire core, an mp-TiO₂/2PACz composite photoanode, an N719 dye, a Li-TFSI film solid electrolyte, and a Pt counter electrode. As shown in Fig. 4b and Table 2, the FS-DSSCs comprising 2PACz exhibited the highest PCE of 6.53% (J_SC = 13.3 mA cm⁻²; V_OC = 0.68 V, FF = 72.6%). The conventional FS-DSSCs showed a PCE of 5.67% (J_SC = 12.1 mA cm⁻²; V_OC = 0.66 V, FF = 70.2%). Therefore, the incorporation of 2PACz resulted in higher J_SC and FF compared with those of pristine TiO₂, thereby resulting in an enhanced PCE. This enhancement is attributed to the improved interfacial charge transfer at the TiO₂/N719 dye interface induced by the 2PACz interfacial layer.

Fig. 4.

a Schematic of an SS-FDSSC. b J–V curves for the TiO₂ and TiO₂/2PACz devices and c corresponding Nyquist plots. The inset shows the equivalent circuit. d EQE spectra and integrated current density. e Photocurrent density as a function of internal voltage. f J_SC for the TiO₂ and TiO₂/2PACz devices as a function of the light intensity on a double-logarithmic scale

Table 2.

Performance of SS-FS-DSSCs (10 devices)

Device	Voc (V)	Jsc (mA/cm2)	FF (%)	PCE (%)	Integrated Jsc (mA/cm2)
Reference	0.66 ± 0.1	12.1 ± 1.1	70.2 ± 1.2	5.67 ± 0.5	11.8
2PACz	0.68 ± 0.1	13.3 ± 0.8	72.6 ± 1.1	6.53 ± 0.5	13.0

Electrochemical impedance spectroscopy (EIS) was employed to elucidate the interfacial charge-transfer characteristics of the pristine TiO₂ and TiO₂/2PACz FS-DSSCs. Measurements were performed under dark conditions at a bias of 0.7 V over the frequency range from 0.1 Hz to 1 MHz. The results are summarized in Table 3 and the corresponding Nyquist plots are shown in Fig. 4c. The Nyquist plots were fitted using an equivalent circuit model, R_S – (R₁║CPE₁)–(R₂║CPE₂), comprising series resistance (R_S), two charge-transfer resistances (R_CE and R_WE), and their respective constant-phase elements (CPE₁ and CPE₂). The high-frequency arc corresponded to charge transfer at the Pt│I⁻/I₃⁻ interface (R_CE), while the larger low-frequency semicircle reflected the charge-transport process at the photoanode│electrolyte interface (R_WE) [36]. For the mp-TiO₂-based FS-DSSC, the R₁ and R₂ values were 16.9 and 92.8 Ω, respectively; by contrast, for the TiO₂/2PACz-based FS-DSSC, the values were 19.8 and 98.9 Ω. Furthermore, the electron lifetimes were calculated from R₂ and CPE₂ [37]. The electron lifetime values of the TiO₂ and TiO₂/2PACz devices were 2.1 × 10⁻¹ and 2.60 × 10⁻¹ s, respectively. These results suggest that higher R₂ values, which are indicative of slower charge-transfer rates at the photoanode/electrolyte interface, may imply a longer electron lifetime, indicating materials with potentially enhanced electrical properties. Longer electron lifetimes, associated with slower recombination rates, can positively impact electrical performance and device efficiency. Consequently, the significant increase in J_SC and improved FF contributed to the enhanced PCE of the TiO₂/2PACz-based FS-DSSCs. The EQE spectra of the TiO₂ and TiO₂/2PACz FS-DSSCs are shown in Fig. 4d. The TiO₂/2PACz FS-DSSC exhibited a higher EQE response compared with that of the TiO₂ device, which is consistent with the J_SC values derived from the J–V curves. The integrated J_SC values, calculated by integrating the EQE spectra, for bare TiO₂ and TiO₂/2PACz were 11.8 and 13.0 mA cm⁻², respectively. The J–V curves and EQEs consistently indicate more efficient charge extraction and reduced charge recombination in the TiO₂/2PACz FS-DSSCs compared with those of the TiO₂ device.

Table 3.

EIS parameters of FS-DSSCs with TiO₂ and 2PACz

Device	Rs(Ω)	R1(Ω)	CPE1(F)	R2(Ω)	CPE2(F)
TiO2	16.9	92.8	1.25 × 10−6	164	1.29 × 10−3
TiO2 + 2PACz	19.8	98.9	1.34 × 10−6	190	1.41 × 10−3

The photoinduced charge extraction and recombination dynamics of the TiO₂/2PACz with FS-DSSCs were further investigated by analyzing the photocurrent density (J_ph) as a function of the internal voltage (V_int), as shown in Fig. 4e. The net photocurrent was obtained as J_ph = J_light − J_dark, allowing isolation of the intrinsic balance of charge generation, recombination, and transport under illumination. The internal volage was defined as V_int = V₀ – V_app, where V₀ is the potential at which J_ph = 0 and V_app represents the applied voltage. The obtained J_ph exhibited a linear increase with voltage at lower V_int levels; however, it saturated at higher V_int, indicating a voltage sufficient to sweep away all carriers to the electrode. Consequently, the saturated photocurrent (J_ph,sat) is constrained by the number of absorbed photons, and is expressed as follows:

where q is the elementary charge (1.602 × 10⁻¹⁹ C) and t_A is the thickness of the active layer (30 μm in conventional DSSCs, representing the thickness of the mp-TiO₂ layer where the dye is adsorbed). G_MAX denotes the maximum photoexcited carrier generation rate per unit volume [38]. The J_ph,sat values for mp-TiO₂ and TiO₂/2PACz were 9.12, and 10.25 mA cm⁻², respectively, corresponding to G_max values of approximately 1.9 × 10¹⁹ and 2.1 × 10¹⁹ cm⁻³ s⁻¹. Consequently, TiO₂/2PACz enhances the effective photon absorption and subsequent carrier generation within the TiO₂ framework.

The light intensity dependence of J_SC was examined to further evaluate the impact of TiO₂/2PACz on the photoelectrochemical behavior of the FS-DSSCs. As shown in Fig. 4f, J_SC exhibited an approximately linear augmentation with incident light intensity, following the power law J_SC ∝ L^α, where α represents an exponential factor indicating the number of photons required to generate electrons in an external circuit, with α typically < 1 [39]. The data were analyzed on a double-logarithmic scale and fitted to the power law, yielding α values of 0.94 ± 0.01 and 1.01 ± 0.02 for the TiO₂ and TiO₂/2PACz devices, respectively. Additional long-term stability data are provided in Fig. S5, which shows that the TiO₂/2PACz -modified devices maintain a high fraction of their initial efficiency under ambient storage conditions. These trends suggest that the incorporation of 2PACz may provide modest improvements in mechanical robustness, which is likely due to the strengthened composite structure and enhanced interfacial adhesion within the mp-TiO₂ layer.

Figure 5a illustrates the hydrogen gas sensing mechanism of the TiO₂/2PACz fiber sensor. In air, ionized oxygen species (O²⁻) are formed on the TiO₂ surface by electron trapping. Upon H₂ exposure, hydrogen reacts with the adsorbed oxygen ions to form H₂O, releasing electrons back to the TiO₂ conduction band. The resulting reduction of the depletion layer leads to a decrease in sensor resistance. As shown in Fig. 5b, the TiO₂/2PACz sensor exhibits a rapid response time of 6 s upon H₂ exposure, which is attributed to the efficient electron release resulting from the reaction between hydrogen and ionized oxygen species on the TiO₂ surface (Fig. 5a). The chemically anchored 2PACz layer facilitates interfacial charge transfer, leading to an enhanced and faster sensing response compared with bare TiO₂. After H₂ removal, the sensor gradually recovers with a recovery time of 122 s attributed to re-adsorption of oxygen species onto the TiO₂ surface. Figure 5c shows that the TiO₂/2PACz sensor exhibits a pronounced selectivity toward hydrogen, with a response exceeding 60%, while negligible responses are observed for NH₃, CO, CO₂, and CH₄. This high selectivity originates from the preferential diffusion of small H₂ molecules through the 2PACz layer and their efficient reaction with ionized oxygen species on the TiO₂ surface, resulting in rapid electron release and a strong sensing signal. As shown in Fig. 5d, the TiO₂/2PACz sensor exhibits highly reproducible responses during repeated H₂ exposure and removal cycles. The stable cycling behavior originates from the reversible surface reaction between hydrogen and ionized oxygen species on TiO₂, in conjunction with the chemically anchored 2PACz layer that ensures interfacial stability and suppresses performance degradation.

Fig. 5.

H₂ gas-sensing characteristics of a TiO₂/2PACz fiber sensor. a Schematic of the fiber-shaped gas. b Response and recovery curve. c Selectivity against various gases (all at 100 ppm). d Repeatability at 100 ppm H₂

Conclusion

This study demonstrated that integrating functionalized 2PACz into a mp-TiO₂ photoanode provides an effective approach to enhancing the performance of optoelectronic devices. Green 2PACz FOLEDs exhibited an EQE of 10.71% and a CE of 37.44 cd A⁻¹, while the red device achieved an EQE of 8.97% and a CE of 7.42 cd A⁻¹. The combination of 2PACz and TiO₂, which provided an excellent charge extraction characteristic and effective photon absorption, facilitated achieving a PCE of up to 6.53%. Moreover, incorporation of 2PACz into a fiber gas sensor enabled a strong and repeatable H₂ sensing response, whereas pristine TiO₂ fibers exhibited negligible activity, highlighting a promising extension of these materials toward multimodal fiber electronics. Overall, this study demonstrates that the structural simplicity, chemical robustness, and functional universality of 2PACz make it a promising building block for next-generation small-scale, flexible, and multifunctional device platforms.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Jae Ho Kim: conceptualization, validation, writing-original draft, Youngjin Kim: conceptualization, formal analysis, Yonghun Kim: validation, Jung-Dae Kwon: validation, Jinhee Heo: formal analysis, Byung-Sik Moon: conceptualization, Bo Seung Jang: formal analysis, Dahyun Lee: formal analysis, Won June Kim: conceptualization, formal analysis, writing-original draft, Jin-Woo Oh: conceptualization, validation, writing-original draft, Hyung Woo Lee: conceptualization, validation, writing-original draft, Myungkwan Song: conceptualization, validation, methodology, writing—review & editing.

J. H. Kim and Y. Kim contributed equally to this work.

Funding

This research was supported by the Fundamental Research Program PNKA460 of the Korea Institute of Materials Science (KIMS), a National Research Council of Science & Technology (NST) grant by Korea government (MSIT) (Grant No. GTL24041-000), and a National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (RS-2024-00406152).

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no conflict of interest.