Photocatalytic doping of organic semiconductors

The study addresses the challenge of doping organic semiconductors (OSCs) efficiently using mild, widely available dopants. Conventional molecular dopants are often strong oxidants or reductants, reactive, and unstable—especially n-dopants—leading to by-products, cost, and limited scalability. Weak oxidants/reductants typically require thermal or radiative activation or metal-catalysed bond cleavage and there is no single dopant able to perform both p- and n-doping. The authors hypothesize that photoredox catalysts (PCs), serving as electron shuttles under light, can mediate either oxidation (p-doping) using air (O2) or reduction (n-doping) using weak amines, enabling efficient, room-temperature, solution-based doping of diverse OSCs while preserving microstructure. They further ask whether a single photocatalytic platform can achieve both p- and n-doping, even simultaneously, with minimal consumption of chemicals and high device-relevant conductivities.

Prior work established chemical doping as central to enhancing performance in OSC devices (LEDs, photovoltaics, thermoelectrics, FETs, electrochemical devices). Many effective strategies rely on strong oxidants/reductants via ground-state electron transfer, but such dopants can be unstable, produce by-products, be expensive, and especially problematic for n-doping stability. Activation of weak dopants typically employs thermal or radiation processes or metal-catalysed bond cleavage. Photoredox catalysts are widely used in organic synthesis for selective redox reactions using weak sacrificial oxidants/reductants, suggesting potential applicability to OSC doping. However, no prior approach enabled both p- and n-doping via the same catalytic system, nor broadly used air as the p-dopant under mild conditions while maintaining OSC microstructure and delivering high conductivities.

General concept and materials: The photocatalytic doping platform comprises a photocatalyst (primarily acridinium derivatives), a weak oxidant or reductant, an organic salt for charge compensation, and π-conjugated OSC polymers (p- or n-type). Acridinium-based PCs used: Acr-Me+ (10-methylacridinium), Mes-Acr-Me+, and Mes-Acr-Ph+; comparisons included perylene diimide (PDI-CC7) and eosin Y. Redox-inert counterions: LiTFSI or [EMIM][TFSI] stabilize charges. Weak oxidants: dioxygen (air) or diphenyl disulfide ((PhS)2). Weak reductant: triethylamine (Et3N/Et2N). Photocatalytic p-doping procedure (case study PBTTT): PBTTT thin films were prepared (spin-coated) and immersed in a PC solution: 0.01 M Acr-Me+ and 0.1 M LiTFSI in BuOAc:CH3CN (3:1). Films were irradiated under 455 nm blue light (50 mW cm−2) in air up to 12 min. After irradiation, films were removed, washed with clean solvent (BuOAc:CH3CN = 3:1), and dried in N2. The PC solution was recovered and reused. Controls included no PC, PC in dark, and no light. Characterization included UV–vis–NIR absorption, XPS/UPS, GIWAXS, and conductivity (2- and 4-probe). Mechanistic studies: Transient absorption spectroscopy (TAS), photoluminescence (PL), and absorption studies analyzed excited-state dynamics and photoinduced electron transfer (PET). Optically transparent amines (Et2N/Et3N) were used to generate reduced PC species in situ and monitor reduction/regeneration cycles with O2 or (PhS)2. DFT (M06/6-311G**, aug-cc-pVTZ with SMD solvent model) computed Gibbs free-energy profiles for catalytic cycles. Generality tests: A range of polymers with differing ionization potentials (IPs) and side chains were studied: P(g42T-T), P(g42T-TT), P3HT, PBTTT, gDPP-g2T, and PTQ1. Multiple PCs (Acr-Me+, Mes-Acr-Me+, Mes-Acr-Ph+, PDI, eosin Y) were evaluated under identical conditions (air as oxidant for p-doping). Conductivity was measured versus irradiation time and wavelength; microstructure was probed by GIWAXS. Photocatalytic n-doping: n-type BBL films were covered with a PC solution (0.01 M Mes-Acr-Me+ in BuOAc:CH3CN = 3:1) containing 0.1 M [EMIM][TFSI] and 0.1 M Et2N (or Et3N). Processes were conducted in N2 glovebox; irradiation used 455 nm and 390 nm LEDs (each 50 mW cm−2). After irradiation, films were washed and dried under N2. Spectroscopy and conductivity assessed doping. Simultaneous p- and n-doping: Physically separated p-type P(g,2T-T) and n-type BBL films were immersed in the same Mes-Acr-Me+ solution with [EMIM][TFSI] (no added amine), irradiated to photoactivate PC as a redox shuttle transferring electrons from P(g,2T-T) to BBL. Films were then washed and dried. Only the organic salt was consumed to maintain charge neutrality. A planar thermoelectric generator was fabricated by drop-casting P(g,2T-T) and BBL on PET and simultaneously photocatalytically doped, then tested for current/voltage/power output under thermal gradients. Key experimental parameters: Typical p-doping solution: BuOAc:CH3CN = 3:1, [LiTFSI] = 0.1 M, [PC] = 0.01 M; in air for O2 oxidant or in N2 for (PhS)2 (0.1 M). n-doping: [EMIM][TFSI] = 0.1 M, Et3N/Et2N = 0.1 M, [PC] = 0.01 M, in N2. Light: 455 nm (and 390 nm for n-doping), 50 mW cm−2; irradiation durations up to ~12 min (p-doping) or minutes for n-doping. Conductivity measured with appropriate 2-/4-probe geometries depending on magnitude.

- Photocatalytic p-doping requires both PC and light: O2 alone or PC in the dark does not dope PBTTT; blue-light irradiation in presence of Acr-Me+ yields strong polaronic absorption and large conductivity increases. - PBTTT case: Conductivity increased from ~10−5 S cm−1 (undoped) to >700 S cm−1 after 10 min irradiation; stability maintained over 30 days. UPS shows work function increases (e.g., from ~4.04 eV undoped to up to ~4.72 eV with longer irradiation). XPS detects TFSI (F 1s, O 1s) only upon illumination, indicating counterion ingress; GIWAXS shows reduced π–π stacking distance and increased lamellar spacing with preserved chain orientation, consistent with side-chain localization of TFSI and oxidation of PBTTT. - Bulk doping: Films from 16–60 nm reached similar maximum conductivity and polaronic intensity; thinner films dope faster. Doping level scales with light dose and depends on wavelength; PC solution is reusable. - Mechanism: TAS/PL indicate PET dominated by PC excited state; electrons transfer from OSC to PC*. Reduced PC forms in presence of amine and is re-oxidized by O2 or (PhS)2, regenerating PC and closing the catalytic cycle. A distinct 658 nm photoinduced absorption signifies the excited reduced PC. DFT shows both reduced and excited reduced PC can be oxidized by O2 (for p-doping), with the excited reduced state favored; analogous favorability for BBL reduction by reduced/excited reduced PC for n-doping. - Generality: Multiple OSCs (P(g42T-T), P(g42T-TT), gDPP-g2T, P3HT, PBTTT, PTQ1) can be photocatalytically p-doped by acridinium PCs; even high-IP PTQ1 (~5.3 eV) shows slight doping. Estimated PC excited-state EAs (EA*) of 5.9–6.9 eV enable oxidation of polymers with IPs 4.3–5.3 eV. Conductivity enhancements correlate inversely with the energy barrier (IPOSC − EA*PC). - High conductivities: P(g42T-T) reached up to ~3,000 S cm−1 (air as oxidant). Eosin Y under blue (455 nm) or green (525 nm) light gave similar conductivities. Photocatalytic doping produced P(g42T-T) films with highest crystallinity among methods tested. - Photocatalytic n-doping (BBL): Without PC/light, Et2N barely dopes BBL. With PC and light, BBL shows strong negative polaron bands (400, 720, 865 nm) and conductivity increases by >5 orders of magnitude, from <10−5 S cm−1 to >1 S cm−1 within ~2 min. - Simultaneous p-/n-doping: With Mes-Acr-Me+ as shuttle, electrons transfer from P(g,2T-T) (p-type) to BBL (n-type) when films are immersed together (physically separated) and irradiated; P(g,2T-T) reaches typical ~200 S cm−1 while BBL reaches ~0.1 S cm−1; only [EMIM][TFSI] is consumed to maintain charge neutrality. Doping observed only with PC present. - Device demonstration: A planar thermoelectric generator on 25-µm PET with simultaneously photocatalytically doped P(g,2T-T) and BBL legs delivered power outputs comparable to state-of-the-art modules doped by conventional (separate) methods; devices showed stable, reversible responses across load resistances and temperature gradients.

The work demonstrates that photoredox catalysis can mediate both oxidation and reduction of OSCs under mild, solution-processable, room-temperature conditions using widely available weak dopants (air O2, amines). By using PCs as recyclable redox shuttles activated by light, the approach overcomes the limitations of strong, reactive dopants, minimizes by-products, and decouples doping strength from ground-state thermodynamics via excited-state redox potentials. The method preserves OSC microstructure while enabling bulk doping, as evidenced by GIWAXS and spectroscopy, and achieves high electrical conductivities (up to 3,000 S cm−1 for p-type and >1 S cm−1 for n-type in minutes). The generality across multiple polymers and PCs, the tunability of doping level via light dose, and the ability to perform simultaneous p- and n-doping with minimal chemical consumption directly address longstanding challenges in OSC doping. The successful thermoelectric module further underscores technological relevance for scalable, ambient-compatible manufacturing of organic electronics.

This study introduces a general, efficient photocatalytic doping strategy for organic semiconductors that operates at room temperature and uses recyclable, air-stable photocatalysts with weak, widely available dopants. The method enables p-doping with O2, n-doping with amines, and even simultaneous p-/n-doping, delivering high conductivities while preserving microstructure and allowing precise control via light dose. It also facilitates the insertion of redox-inert counterions into undoped films without detrimental structural changes and demonstrates device-level utility in thermoelectric generators. Future work could optimize photocatalyst structures and spectral response, expand OSC material scope (including higher-IP/EAs), refine counterion selection and transport, integrate with diverse device architectures, explore continuous and patterned photodoping for manufacturing, and further elucidate kinetics and interfacial energetics to maximize efficiency and stability.

- Simultaneous p-/n-doping can be thermodynamically limited by alignment of polymer electrochemical potentials, potentially capping achievable doping levels. - n-doping and simultaneous doping required inert (N2) conditions and dual-wavelength irradiation (390/455 nm), which may complicate processing compared with p-doping in air. - The approach consumes TFSI-based counterions (LiTFSI or [EMIM][TFSI]) and relies on solvent exposure; ion ingress and solvent compatibility must be managed for specific device stacks. - Doping efficiency is sensitive to the energetic offset between polymer IP/EA and PC excited-state redox potentials; very high-IP polymers (e.g., PTQ1) exhibit only slight doping. - Light-dose dependence necessitates careful irradiation control for uniform, scalable processing; very thick films may require longer exposure despite evidence of bulk doping in the studied thickness range.