Highly conductive polymer electrodes for polymer light-emitting diodes

The study addresses the limitation of indium tin oxide (ITO) as a rigid transparent anode in OLEDs, which constrains device flexibility. The research question is whether a highly conductive polymer electrode—specifically polybenzodifuranedione (PBFDO)—can replace ITO to enable flexible, solution-processed transparent electrodes without sacrificing device performance. The motivation stems from the need for electrodes that combine high conductivity, optical transparency, smoothness, and mechanical flexibility. However, commercial PBFDO solutions exhibit high viscosity and a high freezing point due to strong intermolecular hydrogen bonding and DMSO crystallization, resulting in poor film quality by spin coating. The study proposes introducing n-butanol to modulate hydrogen bonding, lower viscosity and freezing point, improve solution processability, and thereby fabricate high-quality PBFDO films for efficient OLEDs, including flexible devices.

Flexible transparent electrodes have been pursued using conductive nanomaterials such as silver nanowires, carbon nanotubes, and graphene. Silver nanowire-based electrodes provide high conductivity and transparency and have been integrated with elastomers, but interfacial compatibility with polymer matrices can limit long-term stability. Conductive polymers (e.g., polyaniline, polythiophene, PEDOT:PSS) are attractive due to intrinsic flexibility but often suffer from lower conductivity; methods like acid treatment, plasticizers, and topological supramolecular networks have improved PEDOT:PSS, achieving high conductivity and stretchability for intrinsically stretchable LEDs. Recently, the n-type self-doped conducting polymer PBFDO has emerged as a promising ITO alternative due to higher conductivity than PEDOT:PSS and a suitable work function. Yet, commercial PBFDO solutions in DMSO show high viscosity and precipitation from strong inter- and intramolecular hydrogen bonding and a high DMSO freezing point, hindering spin-coated film quality. This work targets these solution-process limitations to enable PBFDO as a practical flexible anode.

Materials: PBFDO conducting solution and PFSO10 (blue polymer) were sourced from Dongguan Volt-Amp Optoelectronics Technology Co., Ltd. PEDOT:PSS (PH 4083) and Super Yellow (SY) came from Heraeus and Sigma Aldrich, respectively. Glass substrates (Luoyang Guluo) and PEN substrates (Teijin) were used. Solution modification: The modified PBFDO solution was prepared by mixing n-butanol with PBFDO mother liquor at a 1:2 volume ratio. The rationale was to form hydrogen bonds between n-butanol and both DMSO and PBFDO, disrupting PBFDO–PBFDO H-bonds to lower viscosity and interacting with DMSO to depress the freezing point. Film fabrication: The modified solution was filtered through a hydrophilic PTFE membrane, spin-coated onto cleaned glass (or PEN), and baked at 120 °C for 10 min. Multiple spin coatings yielded PBFDO thicknesses from 15 to 65 nm. Prior to coating PEDOT:PSS, PBFDO films were rinsed with ethanol and heated at 120 °C for 15 min. Device fabrication: Standard green/yellow devices: Glass (or PEN)/PBFDO/PEDOT:PSS/SY/CsF (1 nm)/Al (100 nm). The SY solution (chlorobenzene, 6 mg mL−1) was spin-coated to ~70 nm and baked at 100 °C for 10 min before cathode deposition. Devices without PEDOT:PSS: Glass/PBFDO/SY/CsF/Al and Glass/ITO/SY/CsF/Al. Blue devices: Glass/PBFDO/PEDOT:PSS/PFSO10/CsF/Al and ITO reference using the same stack. Characterization: FTIR assessed hydrogen bonding in films and solutions; DSC measured freezing points; temperature-dependent viscosity was measured by rheometry; UV–Vis measured transmittance; ellipsometry measured refractive index; AFM assessed surface roughness and surface potential; SKPM (using HOPG as reference, 4.6 eV) and UPS measured work function; XRD probed crystallinity; four-probe measured sheet resistance and conductivity; device J–V–L and EL spectra were recorded with a CS200 meter and USB2000+ spectrometer. Performance comparisons were made to ITO electrodes with varied sheet resistances (15, 45, 110, 450 Ω sq−1).

- Hydrogen-bond modulation: FTIR showed hydrogen-bond signatures in PBFDO film (3200–3550 cm−1) and unmodified solution (3431 cm−1). Adding n-butanol broadened and downshifted the hydrogen-bond peak (to ~3392 cm−1), consistent with n-butanol engaging in H-bonding with DMSO (peak near 3380 cm−1) and PBFDO carbonyls (near 3360 cm−1). - Freezing point and viscosity improvement: DSC showed the freezing point decreased from −1 °C (unmodified) to −27 °C (modified). Temperature–viscosity curves indicated the modified solution had ~4 mPa·s lower viscosity than the unmodified solution at the same temperature, improving filtration and spin-coating quality. Droplet-spread tests showed improved wetting/coverage on glass for the modified solution. - Optical properties: PBFDO films were largely insoluble in DMSO once formed (absorbance unchanged after rinse). Spin-coated films with 15–65 nm thickness showed high transmittance: ~91% (15 nm), ~87% (25 nm), and 76–80% (35–45 nm) across 380–780 nm. The refractive index of PBFDO was 1.23–1.40 across the visible (lower than glass at ~1.51), which avoids total internal reflection at the glass/electrode interface, aiding outcoupling. - Morphology and energy levels: AFM showed smooth films with RMS roughness ~2.06 nm. SKPM (referenced to HOPG) gave an average surface potential translating to a PBFDO work function of 5.32 eV; UPS measured 5.33 eV, consistent between methods. XRD showed a prominent (100) peak with lamellar spacing of 10.89 Å, indicating regular molecular stacking conducive to charge transport. - Electrical properties: Sheet resistance decreased with thickness: 763 Ω sq−1 (15 nm), 410.2 Ω sq−1 (25 nm), down to 128 Ω sq−1 (65 nm). Conductivity increased from 790 S cm−1 (15 nm) to 1271 S cm−1 (65 nm). - Device performance (SY emitters with PEDOT:PSS): All devices had Vth = 2.6 V, indicating efficient carrier injection. The PBFDO (25 nm) device achieved LEmax = 12.8 cd A−1, comparable to ITO with similar sheet resistance (12.7 cd A−1). Maximum luminance (Lmax) was lower for PBFDO devices than ITO due to lower conductivity. EL spectra exhibited microcavity-related narrowing/broadening with thickness. - Flexible device on PEN: With PBFDO (25 nm), the flexible device showed Vth = 2.6 V, LEmax = 7.7 cd A−1, Lmax = 4078 cd m−2; the lower efficiency was attributed to PEN’s lower transmittance (~82%) and outcoupling. - Devices without PEDOT:PSS: Glass/PBFDO/SY/CsF/Al showed Vth = 2.6 V and LEmax = 5.1 cd A−1, while the ITO counterpart exhibited a higher Vth = 4.4 V and much lower LEmax = 0.8 cd A−1 due to a large hole-injection barrier from ITO (4.8 eV) to SY (HOMO 5.6 eV). PBFDO’s WF (~5.3 eV) facilitated hole injection. - Stability: EL spectra were stable from 4–20 V. Under an initial luminance of 1000 cd m−2, PBFDO-based devices decayed to ~670 cd m−2 after 75 h, whereas ITO-based devices reached ~670 cd m−2 in ~4 h and further degraded to ~630 cd m−2 after ~43 h, indicating superior operational stability for PBFDO-based devices. - Blue OLEDs (PFSO10): PBFDO-based blue devices exhibited bright blue emission with Vth ~3.0 V and LEmax ~2.5 cd A−1, comparable to ITO references.

Introducing n-butanol into the PBFDO/DMSO solution effectively modulates intermolecular hydrogen bonding. This reduces viscosity and depresses the freezing point, enabling filtration and uniform spin coating to form smooth, conductive, and highly transparent PBFDO films. The films exhibit a work function around 5.3 eV and regular molecular stacking, which together with PEDOT:PSS alignment supports efficient hole injection and transport. The low refractive index of PBFDO relative to glass mitigates total internal reflection at the electrode/substrate interface, improving light extraction. Consequently, OLEDs with PBFDO electrodes deliver threshold voltages identical to ITO-based devices and comparable peak luminous efficiencies, despite PBFDO’s somewhat lower conductivity limiting maximum luminance. The PBFDO electrode also enables simplified device structures without PEDOT:PSS due to favorable hole-injection energetics, outperforming ITO in that configuration. Flexible devices on PEN confirm that PBFDO functions as a practical flexible transparent anode, though substrate optical properties influence efficiency. Stability tests indicate PBFDO-based devices have comparable or superior operational stability to ITO-based controls under the tested conditions.

This work demonstrates that solution-processable PBFDO, when modified with n-butanol to tune hydrogen bonding, can be spin-coated into high-quality, smooth, conductive, and highly transparent films suitable as flexible transparent anodes for OLEDs. The PBFDO electrodes provide appropriate energy-level alignment (WF ~5.3 eV) and a low refractive index that aids outcoupling, enabling OLEDs with threshold voltages of 2.6 V and peak efficiencies up to 12.8 cd A−1, comparable to ITO-based devices. Flexible OLEDs on PEN with PBFDO achieved LEmax of 7.7 cd A−1, and devices without PEDOT:PSS benefited from PBFDO’s favorable hole injection, outperforming ITO in that simplified stack. Stability assessments showed robust spectral and luminance retention. These results position PBFDO as a promising replacement for ITO in flexible optoelectronics. Future work could focus on further increasing PBFDO film conductivity while maintaining transparency, optimizing optical outcoupling in flexible substrates, and expanding compatibility with a broader range of emissive polymers and device architectures.

- The maximum luminance (Lmax) of PBFDO-based devices is lower than that of ITO-based devices, attributable to the comparatively lower conductivity of PBFDO films. - Flexible devices on PEN exhibit reduced efficiency, linked to the lower transmittance (~82%) and limited light extraction of the PEN substrate. - Measured PBFDO work function (~5.32–5.33 eV) differs from some literature values, potentially due to substrate, thickness, and surface adsorption effects. - While operational stability is promising, broader lifetime testing (e.g., under encapsulation variations, environmental stress, or mechanical cycling) was not detailed in the presented results.