Ambipolar blend-based organic electrochemical transistors and inverters

The study addresses the challenge of realizing ambipolar organic electrochemical transistors (OECTs) that can conduct both holes and electrons to simplify CMOS-like circuit fabrication and enable dual ionic sensing (cations and anions) in bioelectronics. Existing OMIECs are predominantly p-type, with few n-type options and even fewer truly ambipolar materials; designing new ambipolar OMIECs requires complex synthesis and precise control of energy levels and morphology. Drawing inspiration from ambipolar behavior in OFETs and from bulk-heterojunction strategies in OPVs, the authors hypothesize that blending separate p-type and n-type mixed conductors into a single active layer can yield ambipolar OECT operation without new material synthesis. Their goal is to demonstrate ambipolar OECT devices and inverters using a blend of a p-type glycolated polythiophene and an n-type fullerene derivative, achieving balanced performance, stability, and high gain while preserving independent ionic/electronic pathways for each component.

Prior work shows that ambipolar OFETs can reduce processing complexity and enable polymorphic logic, while ambipolar OECTs can shrink device footprint and increase circuit density. Ambipolarity in bioelectronics enables detection of both cations and anions in one device. Strategies to induce ambipolarity through side-chain modification or backbone engineering have been reported, but typically degrade one polarity’s performance and have not delivered stable, high-performing ambipolar OECTs. The OMIEC materials landscape remains heavily p-type, with recent progress in n-type polymers and small molecules. Bulk-heterojunction concepts from OPVs and ambipolar OFETs suggest that mixing p- and n-type components can yield complementary transport. However, achieving ambipolarity in OECTs requires materials that support ionic transport and avoiding detrimental charge transfer at the p:n interface. The authors position their approach against this backdrop, leveraging a misaligned HOMO/LUMO to minimize interfacial charge transfer and enable additive, independently tunable behavior.

• Materials selection: p-type polymer poly(2-(3,3'-bis(2-(2-(2-methoxyethoxy)ethoxy)ethoxy)-[2,2'-bithiophen]-5-yl)thieno[3,2-b]thio phene), p(g2T-TT), and n-type fullerene derivative PrC60MA (C60,N,N,N-trimethyl-1-(2,3,4-tris(2-(2-methoxyethoxy)ethoxy)phenyl)methanaminium monoadduct). The pair was chosen for high mixed conduction, ion-accessible side chains, and substantial HOMO/LUMO misalignment to suppress interfacial charge transfer.
• Device concept: Blend-based ambipolar OECTs using a bulk-heterojunction morphology so each component transports its respective carriers while enabling ionic access.
• Blend optimization: Considering p(g2T-TT) has ~15× higher µC* than PrC60MA, fullerene-rich blends were explored (99:1, 95:5, 90:10 w:w). Balanced n- and p-type figures-of-merit (µC*) were achieved at 95:5 (PrC60MA:p(g2T-TT)).
• Film preparation: Materials dissolved in chloroform; typical concentrations: p(g2T-TT) 5 mg/mL, PrC60MA 10 or 30 mg/mL; blends prepared by mixing pristine solutions. Spin coating at 1000 rpm 60 s then 3000 rpm 10 s to yield ~60–70 nm films (thicker ~175 nm films for VPI). Thermal anneal 120 °C for 20 min. All processing under inert atmosphere (glovebox). Thickness measured by profilometry.
• OECT fabrication and measurement: Three-terminal geometry with Au source/drain and Ag/AgCl pellet gate; aqueous 0.1 M KCl electrolyte. Channel dimensions typical: L = 30 µm, W = 1000 µm, d = 60–70 nm. Output and transfer characteristics measured for both polarities; transconductance g_m computed using g_m = (Wd/L)·(µC*)/(V_th − V_ol). Stability assessed under alternating polarity pulsing (ON ~3 s, interval ~3 s) for 100 cycles. Transient response (T_ON) extracted.
• Spectroelectrochemistry (SEC) and cyclic voltammetry (CV): Films on FTO used as working electrodes; 0.1 M KCl electrolyte; Ag/AgCl reference, Pt counter. SEC spectra recorded over +0.3 to −0.9 V; blend spectra compared to “rule of mixtures” prediction.
• Electrochemical quartz crystal microbalance (EQCM): Gold-coated QCM sensors coated with films; 0.1 M KCl; CV at 10 mV/s; mass changes extracted using the Sauerbrey equation; analysis focused on cation (K+) uptake/release by PrC60MA and anion (Cl−) uptake/release by p(g2T-TT).
• Structural characterization: Out-of-plane XRD (Cu Kα) and GIWAXS to assess crystallinity, lamellar stacking, and coherence length (Scherrer analysis) of pristine and blend films. Complementary morphology visualization via vapor-phase infiltration (VPI) of diethylzinc/water to form ZnO contrast followed by cross-sectional HRSEM with BSE detector to reveal phase distribution and network connectivity.
• Inverter fabrication and testing: Two identical ambipolar OECTs connected in series (one to ground, one to V_DD) with common gate. Voltage transfer characteristics (VTC) recorded for V_DD = ±0.7, ±0.8, ±0.9 V; gains obtained from VTC slope; cycling stability over 20 square-wave cycles.

• Ambipolar OECT performance from a simple binary blend:
  • Optimized blend composition: 95:5 (w:w) PrC60MA:p(g2T-TT), balancing µC* for both polarities.
  • Device characteristics (L = 30 µm, W = 1000 µm, d = 60–70 nm, 0.1 M KCl): well-balanced output/transfer curves with low hysteresis.
  • Table 2 (blend 95:5):
    • n-type: g_m,max = 3.0 ± 0.6 S cm−3; µC* = 11.8 ± 1.4 F cm−2 V−1 s−1; V_th = 0.649 ± 0.018 V.
    • p-type: g_m,max = 4.8 ± 0.2 S cm−3; µC* = 22.8 ± 0.9 F cm−2 V−1 s−1; V_th = −0.090 ± 0.003 V.
  • Compared to pristine materials (Table 1, this work): PrC60MA (n): V_th = 0.620 V; g_m,max = 6.1 S cm−1; µC* = 21.7 F cm−2 V−1 s−1. p(g2T-TT) (p): V_th = 0.001 V; g_m,max = 97.6 S cm−1; µC* = 324.3 F cm−2 V−1 s−1.
• Stability and dynamics:
  • Alternating polarity cycling (100 cycles): no significant current degradation; ON/OFF ratios maintained >10^3.
  • Transient response: p-type T_ON = 466 ± 56 ms; n-type T_ON = 20 ± 3 ms (≈25× faster for n-type), consistent with lower polymer content limiting anion doping pathways.
• Electrochemical/optical independence and additivity:
  • SEC of blend matches rule-of-mixtures prediction; distinct, independent spectral signatures of p(g2T-TT) polaron formation (≈830 nm) and PrC60MA radical anion extending to NIR; slight red shift and increased A0-0/A0-1 ratio indicates enhanced J-aggregation of polymer in blend.
  • CV reveals distinct n-doping, p-doping, and neutral regions; EQCM shows mass gain on both polarities within OECT voltage window (≈100 ng/cm² for p-type vs ≈700 ng/cm² for n-type), confirming anion and cation exchange in the same film.
• Microstructure and morphology:
  • GIWAXS/XRD: blend remains crystalline with oriented PrC60MA lamellae; lamellar d-spacing essentially unchanged, but PrC60MA out-of-plane coherence length reduced by ~20% (from 41.8 ± 0.8 nm to 33.6 ± 0.8 nm), correlating with reduced n-type mobility and µC* in blend.
  • VPI-HRSEM: ZnO contrast indicates uniform bulk-heterojunction-like interpenetrating network rather than a bilayer, with ZnO infiltration across the blend and selective staining of polymer phase.
• Inverter performance (two identical ambipolar OECTs):
  • VTCs in both quadrants centered around Vin ≈ +0.3 V with small hysteresis (ΔVin ≈ 0.05 V).
  • High gains (slope of VTC): forward/back peaks of 62/35 at V_DD = 0.7 V; 74/48 at 0.8 V; 82/56 at 0.9 V; and 33/49 at −0.7 V; 62/49 at −0.8 V; 56/51 at −0.9 V.
  • Inverter stability: robust over 20 cycles for both positive and negative V_DD.
• Design principle validated: Significant HOMO/LUMO misalignment prevents detrimental charge transfer; p- and n-type figures of merit can be tuned via composition, yielding balanced ambipolar operation using readily available materials.

The findings validate a simple and versatile bulk-heterojunction strategy to realize ambipolar OECTs and complementary-like circuits without synthesizing new ambipolar OMIECs. By deliberately misaligning the HOMO of the p-type polymer and LUMO of the n-type fullerene, the two components operate largely independently, enabling additive electrochemical and spectroscopic responses and decoupled optimization. The optimized 95:5 blend delivers balanced µC*, high transconductance densities for both polarities, ON/OFF ratios >10^3, and fast n-type switching, while maintaining stability over extensive alternating polarity cycling. Structural data link a modest reduction in PrC60MA crystalline coherence length to decreased n-type performance in blends, while VPI-HRSEM confirms an interpenetrating network that supports mixed conduction for both ions and electrons. Inverters built from identical ambipolar devices show high gain and symmetric operation in both voltage quadrants, highlighting the potential for simplified circuit fabrication. Nonetheless, ambipolar devices’ inability to fully switch off introduces Z-shaped VTCs, higher static power consumption, and lower static noise margins compared to ideal CMOS. Overall, the approach advances ambipolar OECT technology by enabling independent material tuning, offering pathways to enhance performance and tailor operating voltages for bioelectronic sensing that simultaneously detects cations and anions.

This work introduces and demonstrates a blend-based, bulk-heterojunction paradigm for ambipolar OECTs and logic inverters, using a p(g2T-TT):PrC60MA system. The devices exhibit balanced ambipolar transport, high transconductance, robust cycling stability with ON/OFF >10^3, and high inverter gains with minimal hysteresis. Electrochemical and optical studies confirm independent, additive behavior of the two components, and microstructural analyses reveal a uniform interpenetrating morphology that supports mixed conduction. Compared to synthesizing single ambipolar materials, the blend approach simplifies materials selection and allows independent optimization of p- and n-type performance, opening a broad design space across the growing OMIEC library. Future research should focus on: (i) optimizing morphology and crystallinity to increase µC* for both polarities; (ii) tuning energy levels and thresholds to center operation around 0 V; (iii) improving p-type response speed by increasing accessible polymer pathways; (iv) minimizing static power consumption and improving noise margins in ambipolar circuits; and (v) leveraging blend tunability for advanced, reconfigurable bioelectronic sensors capable of dual cation/anion detection.

• Performance trade-offs: The blend’s µC* values, while balanced, remain below state-of-the-art p-type OECTs (reported up to ~500 F cm−2 V−1 s−1), and n-type performance, though closer, still trails top performers.
• Switching dynamics: p-type transient response is significantly slower than n-type, likely due to low polymer content limiting anion-accessible pathways.
• Off-state limitations: Ambipolar devices cannot be fully turned off, leading to Z-shaped VTCs, higher static power consumption, and reduced static noise margins in inverters compared to CMOS.
• Doping difficulty and threshold shifts: Blending shifts thresholds further from the OFF voltage, indicating more difficult doping for both polarities.
• Potential electrochemical side reactions: PrC60MA exhibits oxygen reduction reaction (ORR) contributions due to a relatively shallow LUMO, which can complicate electrochemical responses.
• Morphological impact: Reduced PrC60MA coherence length (~20%) in blends decreases n-type mobility/µC*.
• Operating point: The ambipolar operating voltages are centered around ~+0.3 V rather than 0 V, which may complicate direct CMOS-like biasing in some applications.