Poly(arylene-ethynylene)s in Synthetic Heterogeneous Visible Light Photocatalysis: Structure–Reactivity Principles and Opportunities in Flow

Abstract

Visible light photocatalysis has reshaped synthetic organic chemistry by enabling mild and selective radical processes. However, its widespread adoption is still constrained by the reliance on homogeneous photocatalysts, typically precious-metal complexes or organic dyes, which complicate recovery, recyclability, scalability, and integration into telescoped continuous flow processes. Heterogeneous photocatalysts (hPCs) offer a sustainable alternative, and among emerging candidates, π-conjugated polymers (π-CPs) have recently attracted attention as tunable, metal-free semiconductors capable of efficient visible light harvesting. This perspective highlights poly­(arylene-ethynylene)­s (PAEs), a distinctive subclass of π-CPs whose modular, conjugated backbones enable control over morphology, band structure, and excited-state behavior. Historically developed for optoelectronic applications, PAEs are now gaining traction as photocatalysts in synthetic organic chemistry, including in continuous flow reactors. Yet, the absence of systematic structure–reactivity relationships limits rational catalyst design and restricts broader deployment. Here, we examine emerging trends that link PAEs structure to photocatalytic function and articulate guiding principles for designing next-generation polymeric photocatalysts. By integrating concepts from materials science and synthetic methodology, this perspective aims to establish PAEs as robust, recyclable, and versatile platforms for sustainable photocatalysis.

1. Introduction

Over the past few decades, visible light photocatalysis has become a cornerstone of modern organic synthesis, enabling mild and selective processes that unlock radical reactivity, challenging bond constructions, and previously inaccessible reaction pathways. While homogeneous molecular photocatalysts (PCs) have driven many of these advances, their broader implementation is constrained by limited catalyst recovery, photostability, and compatibility with sustainable continuous or multistep processing. These limitations have spurred increasing interest in heterogeneous organic semiconductor photocatalysts (hPCs), which combine molecular-level tunability with the inherent recyclability of solid-state materials, including straightforward integration in continuous flow systems.

Among emerging hPC platforms, conjugated polymers (π-CPs) have attracted attention due to their modular synthesis, broad visible light absorption, and structural robustness under photocatalytic conditions. Within this class, poly­(arylene-ethynylene)­s (PAEs)constructed from aromatic units linked by ethynylene spacersoccupy a distinctive position. PAEs are readily accessible from commercially available building blocks via well-established cross-coupling strategies, enabling their tuning for a wide range of photocatalytic applications. Although their semiconducting properties have been recognized for more than two decades, PAEs have historically been applied in molecular electronics and photonic materials. Only recently, driven by the resurgence of visible light photocatalysis, PAEs have emerged as heterogeneous photocatalysts for organic transformations. Key to their reactivity is the efficient π-electron delocalization upon irradiation, which generates long-lived excited states capable of productive single electron and energy transfer events (SET and EnT, respectively, Scheme ). ,

1. Conceptual Evolution and Positioning of PAEs in Photocatalysis.

Despite this promise, the development of PAE-based photocatalysts has largely proceeded through empirical exploration with limited articulation of general structure–reactivity relationships. This has hindered rational catalyst design and constrained their broader adoption in synthetic organic chemistry. We contend, however, that the accessibility and structural regularity of PAEs make them particularly well suited as model systems for probing structure–reactivity relationships in heterogeneous photocatalysis.

In parallel, while continuous flow photocatalysis has been widely recognized as a strategy to overcome photon- and mass-transfer limitations inherent to batch photochemical reactions, the use of PAEs in flow systems remains comparatively underdeveloped.

In this perspective, we analyze recent advances in PAE-based heterogeneous visible light photocatalysis with an emphasis on identifying unifying structure–reactivity principles and highlighting opportunities enabled by continuous flow operation. Our goal is to establish a conceptual and forward-looking framework for the rational design and implementation of PAEsand organic semiconductor photocatalysts more broadlyin synthetic photocatalysis, spanning material design, reaction development, and flow-based processing. Finally, we discuss future directions of research and identify key advances needed to fully exploit the potential of PAEs as heterogeneous photocatalysts in synthetic organic chemistry.

2. Background on PAEs Photocatalysis

2.1. Effect of Light and Formalism

To illustrate the defining structural and photophysical characteristics of PAEs, poly­(para-phenylene ethynylene) (1)the simplest member of this family, composed exclusively of benzene ringsis considered as a model system (Scheme a). In its ground-state, 1 is a rigid, rod-like polymer that adopts a conformation of successively twisted phenylacetylene repeat units. It is insoluble in common organic solvents and displays strong visible light absorption, with an absorption tail extending beyond 800 nm. Despite high thermal stability (>300 °C), polymer 1 exhibits a very low rotational barrier (<1 kcal·mol^–1), enabling facile backbone planarization under mild activation. Accordingly, visible light irradiation promotes rotation of π-orbitals between adjacent units, generating a planarized, conjugated backbone and triggering π-electron delocalization over approximately 3–10 repeat units. The resulting excited-state (1*) is commonly described as a cumulene-like (or quinoid) structure displaying separated positive and negative charges, where reactions can take place. As we will see later, the charge separation and migration strongly influence the excited-state lifetimes of PAEs and therefore their reactivity. As for the simplest PAE 1, the triplet excited-state lifetime (τ) reaches ca. 200 μs, sufficiently long to sustain productive photocatalytic processes.

2. Background on PAEs Photocatalysis.

In general, the excited-state electronic structure of PAEs is commonly described using simplified energy diagrams based on the semiconductor band-theory formalism (Scheme a, lower left). In the ground-state (i.e., the twisted backbone), electrons occupy the valence band (VB); upon photoexcitation with energy exceeding the band gap (E _G), electrons are promoted to the conduction band (CB), generating excitons (or electron–hole pairs, i.e., excited-states). As noted above, the extended conjugation of PAEs facilitates π-electron delocalization, enabling spatial separation and migration of the electron–hole pair (i.e., the quinoid form). Because the charge transport is dominated by photoexcited electrons that move toward the material surface, PAEs are classified as n-type semiconductors.

2.2. Reactivity of Excited PAEs

Photoexcited PAEs accumulate electrons at their surface and are intrinsically predisposed toward single-electron reduction chemistry. This electronic bias rationalizes the widespread use of molecular oxygen as a reaction partner in PAE-based heterogeneous photocatalysis, where O₂ functions as the excited-electron quencher. As a result, aerobic oxidations, including the conversion of sulfides to sulfoxides, amines to imines, and arylboronic acids to phenols, constitute the most extensively reported transformations (Scheme b). These reactions may proceed through complementary mechanistic pathways, encompassing SET processes that generate superoxide (O₂ ^•–) and EnT pathways involving singlet oxygen (¹O₂). The catalytic cycle depicted in Scheme b for amine oxidation is likewise applicable to sulfide and arylboronic acid oxidations, although the latter requires a sacrificial tertiary amine (R₃N).

3. Structure–Reactivity Relationship of PAEs

Understanding how structure governs photocatalytic activity is essential for moving beyond the empirical optimization of PAEs toward rational catalyst design. In these materials, subtle variations at both the molecular level and the macromolecular level can profoundly influence light absorption, charge separation, and interfacial reactivity. Drawing on selected literature, the following section examines how these structural parameters collectively modulate the photophysical and catalytic behavior of solid-state PAEs, with the goal of establishing clear structure-activity relationships. Specifically, we address the following questions: What role does the CC linkage play in governing the reactivity of the PAEs? How do substituents on the aromatic rings modulate the PAE reactivity? How do macromolecular characteristicssuch as surface area, porosity, and particle sizeimpact the overall photocatalytic activity of PAEs (Figure )?

1.

Structure–reactivity relationship study of PAEs at the molecular and macromolecular scales.

3.1. Importance of the CC Triple Bond

The CC triple bond is an inherent structural feature of PAEs, but its specific role in determining their reactivity is crucial. A study led by Lan and Bai addressed this aspect by comparing the photocatalytic activity of triazine-derived PAEs incorporating different linkers: single (C–C, 2), double (CC, 3), and triple bonds (CC, 4), in benchmark oxidation reactions of thioanisole 5 and benzylamine 7. Across all cases, CC triple bond-containing PAE 4 consistently afforded the highest yields of the oxidized products99% of sulfoxide 6 and imine 8when used as a hPC. Comprehensive analytical studies were performed on materials 2–4, revealing distinct differences in their optical properties. In the absorption spectra, PAE 4 exhibited a markedly enhanced absorption profile in the visible region (^absλ_tail (4) > 800 nm) relative to 3 (^absλ_tail (3) = 750 nm) and 2 (^absλ_tail (2) = 680 nm). These spectral differences correlated with the band gaps, which decreased progressively from 2.05 eV for 2 to 1.92 eV for 3, and from 1.86 eV for 4. The measurement of redox potentials demonstrated that all materials could reduce dioxygen (E _red[O₂/O₂ ^•–] = −0.59 V vs SCE) via SET at the CB (excited electrons), with hPC 4 displaying the highest reductive power (E[4e-] = −1.30 V vs SCE). However, the redox potentials at the VB (holes) were thermodynamically insufficient to promote single electron oxidations on thioanisole (E _ox[PhSMe/PhS^•+Me] = +1.59 V vs SCE), discrediting the photoredox pathway proposed by the authors. Instead, we concluded that an EnT pathway via ¹O₂ might be the main mechanism. Finally, photoluminescence and time-resolved fluorescence spectroscopy revealed that 4 possessed the longest singlet excited-state lifetime (34 ns), slightly exceeding those of 3 (30 ns) and 2 (23 ns). Although these lifetimes cannot be directly correlated with the predominant EnT pathway, they nonetheless reflect enhanced exciton stability in 4, accounting for its better catalytic activity (Scheme ).

3. Importance of the CC Triple Bond.

Taken together, these results demonstrate that incorporation of π-linkers into the polymer backbone markedly enhances the photocatalytic efficiency in oxidative transformations. This improvement stems from the increased conjugation, as revealed by spectroscopic analyses, which broadens absorption into the visible region and enables more efficient light harvesting. Although the differences between CC and CC π-linkers are relatively subtle here, the absorption profiles and band gap measurements indicate that CC linkers provide improved absorption characteristics thanks to their linear geometry, which reduces steric hindrance, strengthens the orbital overlap, and enhances conjugation. This results in more effective electron delocalization upon irradiation and prolonged excited-state lifetimes, ultimately giving better photocatalytic efficiency. The CC motif also increases the single electron reducing power of the polymers, as reflected in their redox potentials. Collectively, these attributes confer improved catalytic performance to PAEs, consistent with additional studies (not shown here) that underscored the pivotal role of alkyne linkers in conjugated polymers (Practical Highlight 1).

3.2. Position of the CC Triple Bond

High conjugation is a key driver of the light harvesting capabilities of PAEs. A critical design element in conjugated PAE systems is the positioning of the CC triple bonds on the aromatic core. While most reported PAEs incorporate meta-trisubstituted aromatic linkers (e.g., 1,3,5-triethynylbenzene), resulting in nonfully conjugated backbones, our group has demonstrated that fully conjugated linkers can significantly enhance photocatalytic performances (Scheme ).

4. Influence of the CC Triple Bond Position.

To evaluate this effect, three poly­(phenylene-ethynylene)­s (PPEs) were synthesized: the fully conjugated linear PPE (1), the 1,3-branched (9), and the 1,3,5-trisubstituted one (10). In the model benzylamine oxidation reaction, the expected imine product 8 was obtained in comparable yields using 1 and 9 (67% and 66%, respectively) and in slightly lower yield with 10 (60%). Notably, under continuous flow conditions, photocatalyst 1 delivered a quantitative yield (100%), whereas 9 and 10 afforded only 72%. Photophysical studies revealed a lower band gap for 1 (2.60 eV) and 9 (2.54 eV) relative to 10 (2.79 eV), consistent with their absorption profiles. Interestingly, material 10 exhibited a higher oxidative power (E[10h ⁺ ] = +1.86 V vs SCE) compared with 9 (E[9h ⁺ ] = +1.76 V vs SCE) and 1 (E[1h ⁺ ] = +1.67 V vs SCE). Also, PAEs 1 and 10 displayed a higher reductive power (−0.93 V vs SCE) in comparison with 9 (E[9e ^– ] = −0.78 V vs SCE). Additionally, the changes in aromatic substitutions afforded very different surface areas for the three materials (calculated with the Brunauer–Emmett–Teller method, BET). As such, 1,4- and 1,3-disubstituted materials 1 and 9 displayed no porosity (S _BET < 50 m² g^–1), while 1,3,5-trisubstituted polymer 10 was highly porous (S _BET(10) = 1031 m² g^–1). Mechanistic investigations indicated that both photoredox and energy transfer pathways were operative under these conditions.

Therefore, these investigations demonstrated that linear 1,4-diethynylbenzene precursors are better suited for heterogeneous photocatalysis in general, thanks to the better conjugation of the polymer. It also highlighted that modifying the position of the CC triple bond on the aromatic rings had a non-negligible influence on the redox potentials and that higher redox potentials and higher porosity do not necessarily mean higher reactivity. Additionally, the differences in reactivity between batch and flow systems revealed that traditional batch conditions can be sometimes limited for distinguishing subtle differences in photocatalysts’ performance. In contrast, flow systems, due to their enhanced surface-area-to-volume ratio and more homogeneous irradiation, enable accelerated reaction kinetics and more accurate comparison of catalyst efficiencies. In this case, the reaction time decreased from 19 h in batch to 9 min in flow, allowing for clear differentiation among the three photocatalysts, which appeared nearly equivalent under batch conditions (Practical Highlight 2).

3.3. Influence of Planarity

Conjugated polymers require planar architectures to enable efficient charge transport. Such planarity may arise either from rotational relaxation upon excitation or from an intrinsically planar ground-state structure. A study conducted by Zhang, Ruan, Qian, and Rueping highlighted this effect by incorporating various 9-fluorenone derivatives with different branching patterns into pyrene-derived PAEs (C2-branched 11; C3-branched 12, Scheme a). Interestingly, branching patterns strongly influenced the ground-state planarity of the polymers. DFT calculations showed that polymer 11 was fully planar, whereas polymer 12 adopted a curved geometry. Also, 11 displayed a slightly lower band gap (1.78 eV) than 12 (1.87 eV), rationalized from minor differences in their absorption profiles. Submitting 11 and 12 in the SET-mediated photocatalytic aerobic oxidation of N-aryl tetrahydroisoquinolines (such as 13) demonstrated the superior catalytic activity of 11, which afforded a 96% yield of product 14 (instead of 26% for 12). In addition to its improved planarity, the measurement of redox potentials also provided insights into the better reactivity of 11. While hole redox potential measurements (1.05 and 1.30 V vs SCE for 11 and 12, respectively) demonstrated that both materials could thermodynamically oxidize N-aryl tetrahydroisoquinolines (E _ox = +0.89 V vs SCE), the CB potentials (−0.73 V vs SCE for 11 and −0.57 V vs SCE for 12) highlighted that 11 was thermodynamically capable to reduce dioxygen, whereas 12 lies near the reduction threshold (E _red[O₂/O₂ ^•–] = −0.59 V vs SCE). As such, this difference likely accounts for the lower activity of 12 in this transformation.

5. Influence of Planarity of PAE and Reactivity.

Thus, this study showed that slight molecular changes around the CC triple bond can have a strong impact on reactivity due to planarity effects. In particular, variations in the planarity exert a direct influence on the redox properties of these materials. As observed earlier, more planar architectures exhibit enhanced reducing power, whereas increased torsion correlates with greater oxidative capability. Because PAEs operate as n-type semiconductors, the reductive strength of the CB is more consequential than the oxidative ability of the VB, accounting for the superior catalytic activity of 11 in SET processes.

In the same vein, Zhao, Geng, and Tan recently demonstrated that modifying the backbone of pyrene-containing PAEs using linkers with varying degrees of ring strain or rigidity could effectively alter the mode of reactivity. For this, a series of pyrene-containing PAEs interconnected with different linkers1,4-diethynylbenzene (15), 4,4′-diethynylbiphenyl (16), and 3,4′-diethynylbiphenyl (17)were synthesized (Scheme b).

DFT calculations revealed that while compound 15 was fully planar, compounds 16 and 17 exhibited torsion angles of 37° and 43°, respectively. The absorption profiles as well as the band gap were almost similar for all three materials, enabling the study of the influence of the torsion angle as a sole changing parameter. For this, the three materials were first engaged in the photocatalytic oxidation of phenylboronic acid 18 to phenol 19, known to occur through a SET mechanism under these conditions (triethylamine being used as a sacrificial single electron reductant). As anticipated, planar catalyst 15 delivered the highest yield of phenol 19 (99%), whereas 16 and 17 provided only moderate yields (43% and 36%, respectively). Interestingly, when applied to the oxidative [4 + 2]-photocycloaddition between N-phenylmaleimide (20) and N,N-dimethylaniline (21), known to proceed via an EnT mechanism here, the trend was reversed: photocatalyst 16 delivered the best performance, affording cycloadduct 22 in 99% yield.

The redox potentials of materials 15–17 were determined, revealing that all three hPCs are thermodynamically capable of oxidizing triethylamine (E _ox[Et₃N/Et₃N^+•] = +0.90 V vs SCE) and reducing dioxygen. Consistent with prior observations, the planar derivatives exhibited greater single electron reducing power relative to that of the twisted analogues, but this sole interpretation is insufficient to rationalize the pronounced divergence observed in their photochemical reactivity. To probe the origin of these differences, the singlet exciton binding energiesnamely the energy required to separate the electron–hole pair into free charge carriersof 15, 16, and 17 were measured to be 30, 45, and 88 meV, respectively, demonstrating that the planar system 15 achieved more efficient charge separation than either 16 or 17. Fluorescence quenching experiments in the presence of O₂ revealed a substantial reduction in the excited-state lifetime of 15, whereas the lifetimes of 16 and 17 remained largely unchanged. Taken together, these data supported the conclusion that 15 underwent productive photochemistry primarily from its singlet excited-state via a SET pathway, facilitated by its superior charge-separation capacity, a rare case in semiconductor photocatalysis. In contrast, the twisted derivatives exhibited longer triplet-state lifetimes (13, 23, and 40 μs for 15, 16, and 17, respectively), consistent with a more efficient EnT engagement. DFT calculations corroborated this interpretation by indicating that the twisted architectures possessed an increased propensity for intersystem crossing (ISC), thereby accounting for their enhanced performance in the [4 + 2]-photocycloaddition.

Collectively, these studies underscored the critical influence of the aromatic ring planarity on the photophysical and photochemical behavior of PAEs. Enhanced planarity strengthens π conjugation, increases visible light absorption, and promotes charge separation. Undeniably, planar systems exhibit a higher single electron reducing power but diminished oxidizing ability relative to their twisted counterparts. Because PAEs behave as n-semiconductors, the reductive capability of their CB dominates the oxidative strength of their VB in photoredox pathways. Moreover, distortions can facilitate ISC and preferentially channel reactivity through triplet-state EnT mechanisms. Overall, these findings further highlight that the excited-state behavior of PAEs can be systematically tuned through modulation of the planarity of the aromatic substituents (Scheme , Practical Highlight 3).

3.4. Electronic Effects in PAEs

A critical aspect to consider in the design of PAEs is the electronic influence of their substituents. As discussed above, charge separation in PAEs is a key determinant of the photocatalytic reactivity. While much of the focus has been placed on enhancing conjugation to promote charge separation, this is not the only strategy available. The incorporation of electron-donating or electron-withdrawing monomers can generate donor–acceptor architectures with tailored electronic properties that facilitate exciton dissociation. So far, a few studies investigated this aspect through the prism of organic synthesis. Unfortunately, these studies only display the influence of electron density on resulting redox potentials, neglecting the kinetic parameters, which might prevent the full understanding of the structure–reactivity relationship in this section.

In 2015, the group of Zhang investigated this aspect by examining the reactivity of single-atom-modified PAEs incorporating benzochalcogendiazole acceptors bearing oxygen (23), sulfur (24), or selenium (25) atoms (Scheme a). UV–vis spectroscopy revealed that 23 and 24 exhibited similar absorption profiles, with maxima around 800 nm, while 25 extended into the near-infrared region (>1000 nm). This trend was reflected in their band gaps: 2.70 and 2.77 eV for 23 and 24, respectively, and a narrower gap of 2.41 eV for 25. Redox potential measurements indicated that the oxygen-containing 23 had the highest single electron reducing power (−1.22 V vs SCE), followed by 24 (−1.08 V vs SCE) and 25 (−1.06 V vs SCE), consistent with the increasing electron-accepting nature of the chalcogen atoms. Intriguingly, 24 exhibited the highest oxidative power (+1.69 V vs SCE), but the reason for this remains unclear. The authors conducted EPR spectroscopy, which showed a more intense signal for 24, suggesting a longer-lived exciton, although no direct lifetime measurements were reported. Then, they tested those materials in the benchmark benzylamine 7 oxidation reaction occurring through both SET and EnT mechanisms and found that sulfur-containing 24 outperformed its analogues, delivering a 74% conversion of amine 7 (no yield was provided). Although some photophysical parameters remain unmeasured, the superior performance of 24 may result from a more persistent excited state in the presence of a sulfur atom.

6. Donor–Acceptor Effects in PAEs.

In a similar vein, the same group investigated the donor–acceptor effects by systematically varying the ratio of donor to acceptor in PAEs made of triethynylbenzene as donor units and phenyl or benzobisthiadiazole (BBT) as acceptor units. As such, they prepared different PAEs containing 0% (26), 10% (27), 50% (28), and 100% (29) of BBT relative to the phenyl substituentBBT being considered as a stronger acceptor compared to phenyl (Scheme b). As expected, increasing the proportion of BBT in the polymer backbone led to larger red-shifts and smaller band gaps. This trend was consistent with a stronger donor–acceptor character: when the acceptor content increased, the energy gap between the valence and conduction bands narrowed. The redox properties followed the same logic. Polymers richer in BBT showed lower oxidative and reductive potentials (in absolute terms), again reflecting the influence of the donor–acceptor effect on the energy levels. These polymers were then tested in the oxidative [4 + 2]-photocycloaddition between N-phenylmaleimide (20) and N,N-dimethylaniline (21), which was found to occur through concomitant SET and EnT pathways. Interestingly, the material containing 10% BBT (27) provided the highest reactivity, affording cycloadduct 22 in 82% yield. Increasing the acceptor content further proved detrimental: polymers 28 and 29 gave only 43% and 29% yield, respectively, and the BBT-free material (26) delivered only traces of the product. Although the exact origin of this reactivity pattern is not yet fully understood, the authors suggested that polymer 27 strikes the best compromise between band gap and redox potential, features that are both critical for effective visible light photocatalysis.

Altogether, these studies underscore the critical role of donor–acceptor tuning in modulating the photophysical and photocatalytic properties of PAEs. While increasing acceptor content generally lowers the band gaps and redox potentials, excessive acceptor incorporation can lead to insufficient oxidative and reducing power, ultimately diminishing reactivity. Thus, achieving high photocatalytic performance requires a fine balance between light absorption, exciton dissociation, and redox potentialsideally within a band gap below 3 eV to ensure visible light activation (Practical Highlight 4).

3.5. Surface Areas, Porosity, and Particle Sizes in PAEs

Because heterogeneous photocatalysis relies on interfacial reactivity between a solid-state photocatalyst and a liquid reaction medium, macromolecular features, particularly surface-related phenomena, are critical for predicting the reactivity of PAEs. High porosity in PAEs is often assumed to enhance reactivity by increasing the surface contact with reagents, and many studies have therefore focused on designing highly porous conjugated polymers with large specific surface areas. However, as shown earlier, a greater porosity does not necessarily translate into improved reactivity. In heterogeneous synthetic photocatalysis, performance depends also on pore accessibility since pores must be large enough to allow reagent diffusion. A study reported by Patra and co-workers further supports this observation (Scheme a). They synthesized a series of porous PAEs (30–32) incorporating a boron dipyrromethene (BODIPY) core with varying substituents (R) to vary the surface area and pores size. The resulting materials displayed thereby a broad range of surface areas and pore volumes: 30 (R = n-octyl, S _BET = 322 m² g^–1, pore diameter = 3.2 nm), 31 (R = 4-tert-butylphenyl, S _BET = 571 m².g^–1, 1.4 nm), and 32 (R = phenyl, S _BET = 1010 m² g^–1, 0.7 nm). Despite these significant differences, all three photocatalysts displayed nearly identical reactivity in the oxidation of thioanisole 5, with conversion rates ranging from 91% to 96% (no yields were given), supporting the notion that porosity is not a reliable indicator of reactivity.

7. Influence of the Surface Area, Pore, and Particle Size on PAE’s Reactivity.

As noted above, PAEs generate electron–hole pairs upon irradiation, followed by charge migration to the surface, where chemical transformations occur. However, charge delocalization in PAEs spans only ∼10 monomer unitscorresponding to ca. 0.5 nm when PPE 1 is considered as a model. Maximizing photocatalytic efficiency therefore requires maximizing the fraction of polymer chains located within this active surface region. Conventional PAE syntheses, however, do not permit such precise control over polymer chain length. To address this limitation, Vilela and co-workers prepared nanometric PAE layers by performing silica-nanoparticle-templated Sonogashira polymerizations of 1,3,5-triethynylbenzene (33) and dibromobenzothiadiazole (34). This strategy produced silica nanoparticles uniformly coated with a thin PAE shell. Subsequent removal of the silica core using aqueous ammonium hydrogen difluoride yielded hollow spherical PAE nanoparticles with diameters of 10–30 nm. Varying the amount of silica template (mg mL^–1 relative to the polymerization medium) afforded distinct hollow spheres (35–38). Surface analyses revealed that higher template loadings (higher silica concentration) produced materials with larger surface areas and pore volumes, while the pores diameters remained essentially constant at ∼1.4 nm (Scheme b). The photocatalysts were assessed in the singlet oxygen-mediated [4 + 2]-photocycloaddition of α-terpinene (39) to ascaridole (40) under flow. Polymer 38, which possessed the largest surface area and pore volume, delivered the highest performance with 96% conversion of 39, and the conversion decreased with decreasing surface area. Although the correlation was not strictly proportional, these results demonstrated that thin-layer PAEs with higher-surface areas and pore volumes outperformed nontemplated PAEs with lower-surface areas such as 35.

Thus, the effect of the surface area on PAE reactivity must be considered with caution. While high porosity is often associated with improved reactivity in heterogeneous photocatalysis, this relationship is not unequivocal. The pore size and volume, in conjunction with the particle dimensions, must be carefully optimized to facilitate the efficient mass transport of reactants to the catalytic sites. In this line, because charge transport in PAEs is typically confined to approximately ten monomer units, nanometric-layer architectures maximize the proportion of accessible active sites, thereby enhancing reactivity (Practical Highlight 5).

4. PAEs as a Photocatalyst for Continuous Flow Synthesis of Organic Compounds

The combination of photocatalysis and continuous flow processing provides a highly efficient environment for light-driven chemistry, and the advantages become even more pronounced when solid-state photocatalysts are employed. Flow reactors provide well-defined geometries that ensure uniform and efficient irradiation, thereby mitigating photon attenuation and the uneven light distribution that often limits batch photochemical processes. In addition, the superior heat and mass transfer characteristics of flow minimize temperature and concentration gradients, suppressing side reactions and maintaining homogeneous reaction conditions. These features collectively contribute to more reliable kinetics and enhance reproducibility in photocatalytic transformations. Moreover, the small reaction volumes and continuous throughput afford intrinsic safety benefits, particularly when dealing with high-energy intermediates or reactive radical species commonly accessed under photocatalytic conditions. Embedding solid-state photocatalysts into this controlled flow environment further strengthens these advantages: the heterogeneous catalyst bed provides a stable and reproducible photon–catalyst interface, avoids issues of dissolution or leaching, and enables clean product streams without the need for catalyst separation. Solid-state materials typically exhibit high photostability under extended illumination, allowing long operational lifetimes and facilitating straightforward scale translation. Together, the synergy between flow photochemistry and heterogeneous catalytic architectures delivers a robust, safe, and sustainable platform for advancing the practical scope of photocatalytic synthesis.

Despite the increasing interest in photoactive polymers for photocatalysis, applications of PAE-based materials under continuous flow conditions remain remarkably limited, with only a handful of examples reported to date. The first demonstration by Vilela and co-workers, mentioned above, employed the hollow-sphere PAE nanoparticle 38 as a heterogeneous photosensitizer within a flow photoreactor. In their study, the polymeric catalyst was dispersed with the substrate (α-terpinene 39, 0.1 M solution), mixed with a continuous supply of oxygen (2 equiv), and passed through an irradiated microstructured channel, enabling efficient generation of singlet oxygen and achieving high substrate conversion (96%) with short residence times (38 s, Scheme a). Although comparisons with a molecular photosensitizer (TPP: tetraphenylporphyrin) revealed higher productivity for the homogeneous analogue (1 mmol min^–1 instead of 0.1 mmol min^–1 with PAE 38) attributed to its intrinsically greater singlet oxygen quantum yield, the PAE material offered compelling practical advantages. In particular, it could be recovered by simple filtration, exhibited excellent resistance to photodegradation and oxidative conditions, and maintained high selectivity over multiple reuse cycles with only modest performance decline. Further comparison with a small-molecule analogue corresponding to the repeating unit of the polymer confirmed that the chromophore segment within the PAE structure was responsible for the observed photoactivity, while also revealing that the polymeric form displayed significantly greater photostability.

8. Applications of PAEs as Photocatalysts under Continuous Flow Settings.

The second example was reported in 2017 by Zhang and co-workers, who unveiled the photocatalytic activity of PAE-coated glass fibers integrated into a fixed-bed continuous flow photoreactor (Scheme b). Commercial glass fibers (∼10 μm diameter) were used as the support, and a polybenzothiadiazole coating (structurally analogous to 38) was formed directly on the surface by the Sonogashira cross-coupling polymerization of 1,3,5-triethynylbenzene (33) and 4,7-dibromo-2,1,3-benzothiadiazole (34). This procedure yielded a uniform coating of PAE (∼80 nm), providing a surface area of 136 m² g^–1 and a pore diameter of 1.5 nm.

The resulting photocatalyst 41 exhibited strong visible light absorption (^absλ_tail > 700 nm), a band gap of 2.48 eV, and redox potentials of −1.06 V vs SCE (CB) and +1.42 V vs SCE (VB). Photocatalyst 41 was first applied to the photoreductive dehalogenation of α-bromoacetophenones (42) under white light irradiation. For this transformation, 200 mg of 41 was packed into a glass column reactor. Recirculation of the substrate solution at 0.5 mL min^–1 afforded the corresponding acetophenones (43) in excellent yields after 2 h (91–95%). The hPC was then evaluated in the enantioselective α-alkylation of aliphatic aldehydes under similar conditions. The coupling reaction between aldehydes (44) with bromoalkanes (45) in the presence of the MacMillan imidazolidinone catalyst (46) and 1,6-lutidine proceeded to conversion up to 99% and 95% ee upon continuous recirculation between 5 and 15 h. In addition to favorable redox potentials to reduce alkyl bromides (E _red[42] ∼ −0.49 V vs SCE), the nanoscale thickness of the catalyst layer enhanced light penetration, which was confirmed by control experiments employing a monolithic catalyst that showed diminished activity. The system also demonstrated excellent mechanical and chemical robustness: five consecutive runs showed no significant loss in the photocatalytic performance, supporting the stability of the glass-fiber coating approach. Finally, this work represents the first application of PAE in a catalytic asymmetric reaction under flow conditions.

The third and most recent example was reported by our group in 2025. As mentioned earlier, this work demonstrated the photocatalytic activity of the simplest PAE, poly­(para-phenylene-ethynylene) 1, for the aerobic oxidation of amines under both batch and continuous flow conditions (Scheme c). For the flow setup, photocatalyst 1 (50 mg) was premixed with commercial glass beads (5.5 g). Electrostatic interactions between 1 and the beads ensured efficient incorporation into the packed-bed reactor and preserved the bed homogeneity over extended operation. Amines (0.1 M, 0.1 mL min^–1) were delivered together with a stream of oxygen (1.5 mL min^–1) into a photoreactor that was irradiated with white light. A residence time of only 9 min was sufficient to convert a broad range of amines, including more than 25 substrates and several natural products (such as compound 47 arising from (±)-salsolidine), into the corresponding imines in good to excellent yields (45–100%). The high oxidation potential of 1 (+1.67 V vs SCE) enabled the smooth oxidation of challenging substrates such as n-hexylamine (E _ox = +1.48 V vs SCE, forming compound 48). Also, the flow mode consistently outperformed batch reactions, likely due to improved mixing of the triphasic reaction medium and superior irradiation of the catalytic bed. Certain substrates, such as n-hexylamine and tetrahydroisoquinoline, were successfully oxidized into compounds 48 and 49 only under flow conditions, underscoring the decisive role of reactor design in enabling transformations inaccessible in batch. Importantly, the single-pass flow system sustained uninterrupted operation for over 24 h with no detectable loss of catalytic activity, highlighting both the stability of the heterogeneous photocatalyst and the inherent capacity of the flow setup to enable automatic catalyst reutilization.

The continuous production of imines enabled a telescoped two-step sequence to access α-acylaminoamidesprivileged peptidomimetic scaffoldsvia an Ugi three-component reaction with isonitriles and carboxylic acids. Because the Ugi step required high imine concentrations for acceptable yields, the photocatalytic module was intensified by connecting two packed-bed reactors in series (18 min total residence time), allowing the use of more concentrated amine solutions (0.5 M) and generating imines at 0.24 M. The crude imine stream was then pumped into a heated coil reactor at 50 °C (50 min residence time) together with cyclohexyl isocyanide 50 and a carboxylic acid, affording α-acylaminoamides 51–53 in moderate to good overall yields (43–74%). Of particular interest, compound 52isolated as a mixture of two diastereomershas been reported to exhibit antibiotic activity against E. coli. Despite moderate overall yields, productivity metrics highlighted the clear advantage of flow, with 52 produced at 158 mg h^–1 under continuous conditions compared to only 2.7 mg h^–1 using the analogous batch protocol.

Altogether, these early examples demonstrate that the integration of solid-state photocatalysts into continuous flow systems provides a powerful and practical foundation for modern photocatalysis. The precise control over photon delivery, mass transport, and thermal management inherent to flow systems complements the robustness and longevity of heterogeneous materials, enabling highly reproducible, scalable, and safe photochemical processes. As illustrated by recent advances, these combined technologies not only enhance the efficiency of established transformations but also unlock reactivity profiles that remain inaccessible in batch.

5. Conclusions and Future Outlook

Over the past few years, the photochemistry of PAE materials has opened new opportunities for sustainable synthetic photocatalysis. By surveying key literature precedents, we have outlined the features that render PAEs effective in organic synthesis, with particular attention given to the molecular and macromolecular attributes that enhance their photocatalytic performance. A dedicated emphasis on continuous flow systems further illustrates how this enabling technology maximizes the efficiency, stability, and synthetic reach of PAE-based photocatalysts.

Looking forward, several directions are expected to shape the future development of PAE-based photocatalysis:

• From “black-box” to mechanistic predictability.

A central limitation of current PAE photocatalysis is its largely empirical nature. Future progress will require rigorous integration of thermodynamic descriptors with kinetic analysis, supported by quantitative benchmarks such as emission quantum yields and time-resolved spectroscopic measurements. Crucially, these photophysical parameters must be correlated with macromolecular featuresincluding surface area, porosity, and particle sizeto enable accurate prediction of PAE behavior in organic synthesis and to establish a truly design-driven framework. Collectively, overcoming the prevailing “black-box” approach will require interdisciplinary efforts that integrate synthetic polymer chemistry, advanced spectroscopy, materials characterization, and reaction engineering to translate structure–property relationships into a predictive photocatalyst design.

• Rational molecular and macromolecular design.

Advances in the PAE performance are expected to arise from deliberate molecular engineering aimed at improving the thermodynamic and kinetic profiles. In particular, incorporation of donor–acceptor motifs, heteroatom-doped backbones, or electronically polarized linkers could expand accessible redox windows beyond current limits (≈± 2 V vs SCE), enabling more demanding oxidative and reductive transformations and moving beyond predominantly oxygen-mediated chemistry.

• Reassessing photophysical models: molecular vs semiconductor behavior.

Although PAEs are often described using the semiconductor band theory, the finite extent of charge delocalization in many systems raises the possibility that molecular and semiconductor behaviors may coexist. In such cases, classical excited-state quenching pathways may compete with or even dominate over bandlike charge transport. Recognizing and exploiting this dual or multimodal behavioralready suggested in related materials such as metal–organic frameworks (MOFs)could fundamentally reshape design principles for PAE photocatalysts.

• Enabling asymmetric and cooperative photocatalysis.

An especially compelling opportunity lies in the incorporation of chiral motifs or ligands within PAE frameworks to enable asymmetric photocatalysis under heterogeneous conditions. By embedding chiral environments directly into the polymer backbone or side chains, PAEs could support cooperative photoinduced activation and enantioselective substrate control within a single solid material. Such an approach would address a long-standing challenge in photocatalysis as truly asymmetric transformations remain exceedingly rare in heterogeneous systems. If realized, chiral PAEs could bridge homogeneous asymmetric photocatalysis and recyclable solid-state catalysis, opening fundamental new avenues for enantioselective synthesis.

• Toward predictive, multiscale, and data-driven catalyst design.

The convergence of density functional theory, advanced spectroscopy, and data-driven modeling offers a powerful opportunity to move PAE photocatalysis from empirical optimization toward a genuinely predictive catalyst design. Beyond idealized oligomer models, future efforts should integrate ground-state DFT and time-dependent DFT with mesoscale descriptions to capture the effects of backbone conformation, finite conjugation length, and interchain interactions on excited-state energetics. When combined with experimental benchmarks, such multiscale approaches could enable quantitative prediction of redox potentials, triplet energies, charge localization, and interfacial electron-transfer pathways. In parallel, the development of structure–photophysics–reactivity maps derived from curated experimental data sets and machine learning tools may establish generalizable design rules for organic semiconductor photocatalysts and guide the targeted synthesis of next-generation PAEs.

• Leveraging reactor engineering as a reactivity tool.

The compatibility of PAEs with continuous flow operation offers more than scalability; it provides a platform for discovering fundamentally new reactivity. By precisely controlling photon flux, residence time, and mass transport, flow systems can access reaction pathways and selectivity that are unattainable in batches, offering deeper mechanistic insight. Integrating advanced catalyst design with innovative reactor engineering promises not only to expand the synthetic scope of PAEs but also to establish heterogeneous flow photochemistry as a transformative approach for sustainable and precise chemical synthesis.

Glossary

Abbreviations

BBT

benzobis­(thiadiazole)

BET

Brunauer–Emmett–Teller analysis

BODIPY

boron dipyrromethene

CB

conduction band

COF

covalent-organic framework

DFT

density functional theory

E[e−]

redox potential of the excited electron

E[h+]

redox potential of the hole

ee

enantiomeric excess

E _G

band gap

EnT

energy transfer

hPC

heterogeneous photocatalyst

ISC

intersystem crossing

MOF

metal–organic framework

PAE

poly­(arylene-ethynylene)

PC

photocatalyst

π-CP

π-conjugated polymer

SCE

saturated calomel electrode

SET

single-electron transfer

TPP

tetraphenyl porphyrin

VB

valence band

The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. CRediT: Basile Weyl data curation, formal analysis; Benjamin Laroche conceptualization, data curation, formal analysis, funding acquisition, writing - original draft, writing - review & editing.

The authors declare no competing financial interest.

Published as part of JACS Au special issue “Continuous Flow Chemistry”.