Ultra-low loading porphyrin-incorporated conjugated polymer dots as photocatalysts for aerobic oxidation of sulfides in water

Abstract

The development of metal-free, low-loading catalysts, together with the use of water as a benign solvent is desirable for advancing facile and efficient catalysis of organic transformations in an environmentally friendly manner. The direct oxidation of sulfides to sulfoxides is a key process in organic synthesis and pharmaceuticals. This study introduces metal-free porphyrin-integrated poly[9,9′-dioctylfluorenyl-2,7-diyl)-co-(1,4-benzo-thiadiazole)] polymer dots (PFBT-TPP Pdots) as photocatalysts for photo-aerobic sulfide oxidation in water. PFBT-TPP Pdots were synthesized by the Suzuki–Miyaura cross-coupling process. DFT calculations confirmed the conjugation and delocalization characteristics of TPP incorporated into PFBT. The conjugated polymers were converted into Pdots using coprecipitation with poly(styrene-co-maleic anhydride). Pdots with a diameter of around 30 nm demonstrated exceptional dispersion in water. PFBT-TPP Pdots efficiently photocatalyzed 0.1 M of thioanisole or other sulfide derivatives, attaining remarkable conversion efficiency (80–100%) and selectivity (88–100%) at a low catalytic loading (50 μg of conjugated polymer) in water, under 1 atm O₂, at ambient temperature, and under 3 h-LED illumination (20 W, λ = 456 nm). The photocatalytic process was scaled to produce gram quantities of the isolated product with high selectivity and no signs of over-oxidation. This demonstrates a sustained and practical application. The computational analysis and experimental findings illustrated that PFBT-TPP Pdots effectively generated reactive oxygen species during photocatalysis, leading to enhanced catalytic efficiency compared to PFBT Pdots and their metal-containing counterparts. This study highlights the potential of Pdots for efficient, selective photocatalysis in aqueous environments, overcoming common solubility challenges for substrates and products.

---

1 Introduction

Direct oxidation of sulfides is one of the crucial chemical processes on which scientists have long focused. The generation of versatile sulfoxides, which serve as important precursors in organic synthesis^1,2 and the pharmaceutical industry,^3,4 is a manifestation of its significance. Such a process is also applicable to oxidative desulfurization for fuel treatment^5,6 and pollution reduction.⁷ The conventional method for oxidizing sulfides on a large scale involves using severe conditions and harmful oxidants such as peroxides, peracids, and inorganic oxidants at elevated reaction temperatures.⁸ The paradigm transition toward greener production has been realized in recent times. The utilization of photon energy in the synthesis of fine chemicals and commodities has significantly advanced the possibility of developing sustainable chemical processes, including sulfide oxidation.^9,10 Much attention has been devoted to the photo-oxidation of sulfides with O₂ under environmentally benign conditions. Reactive oxygen species (ROS) such as singlet oxygen (¹O₂) and/or superoxide radicals (˙O₂⁻), are oxidants generated by energy transfer (ET) or single electron transfer (SET) from the activated photosensitizer (PS) or photocatalyst (PC) to O₂, respectively.

The utilization of various categories of organic and inorganic compounds such as PSs or PCs has garnered interest. Organic dyes, for example, porphyrins,^11,12 Rose Bengal,¹³ BODIPY,¹⁴ methylene blue,¹⁵ riboflavin,¹⁶ 9,10-dicyanoanthracene,¹⁷ and thioxanthones¹⁸ were used as PSs or PCs in photo-aerobic oxidation of sulfides. Their drawbacks are the degradation of organic dye-based PSs during photocatalysis and their difficult-to-tune photophysical properties. Over-oxidation and C–H oxidation have occurred concurrently in certain instances. The use of inorganic catalytic materials, Bi₄O₅Br₂ and Pt/BiVO₄, as PCs was witnessed.^19,20 However, sluggish reactivity and poor selectivity were detected. It was also discovered that the combined organic–inorganic composite dye-Ti-MCM-41,²¹ dye-TiO₂,^22,23 CdS/g-C₃N₄,²⁴ and metal organic frameworks²⁵ functioned as efficient PCs. Nonetheless, additional redox mediators, including trimethylamine, 2,2,6,6-tetramethylpiperidine-1-oxyl (TEMPO), and H₂O₂, are necessary for these catalytic systems. Coordination metal complexes were also found to function as active photocatalysts for sulfide conversion.^26,27 However, their cost and contamination of metal residue in the products were proved to be disadvantageous. Another category of heterogeneous photocatalysts burgeoning in this field is organic polymeric materials, graphitic carbon nitride (g-C₃N₄)^28,29 and conjugated polymers.^30,31 These materials possess adjustable electronic properties and synthetic variation. It is crucially noteworthy that sulfide oxidation by homogeneous and heterogeneous photocatalysts is commonly conducted in organic solvents. The oxidation reaction in water, a solvent known for its cost-effectiveness and environmental benignity, is hindered by the restricted solubility of photocatalysts and sulfide substrates. This limitation leads to challenges in mass transfer.^32,33 Therefore, the precedent literature featuring photo-aerobic sulfide oxidation in water is scarce.

Polymer dots (Pdots) are an emerging class of conjugated organic polymer nanoparticles characterized by a diameter that is smaller than 100 nm. Pdots contain a folded conjugated polymer that facilitates intense π–π stacking interactions between inter- and intra-chains, resulting in charge localization within the particles via the creation of excimers and exciplexes.³⁴ A polymeric surfactant or amphiphilic polymer forms an exterior layer, enabling Pdots to disperse in water. Their distinctive optical, electronic, luminescence, and physical properties, and water solubility lead to initial practical utilities in biological applications, including phototherapeutics, cell imaging, and biosensing.^35–37 In addition, Pdots have also been implemented in catalysis, particularly in the processes involving multiple charge transfers and aqueous media. There has been a focus on utilizing Pdots with various chemical architectures, as well as their combination with nanoparticles, for photo-driven hydrogen evolution over the past few years.^38–41 Furthermore, Tian et al. only recently demonstrated the intriguing application of Pdots in photocatalytic CO₂-to-CO conversion in aqueous solutions, achieving an appreciable rate with 100% selectivity.⁴²

In this study, we investigate the potential of Pdots based on poly[9,9′-dioctylfluorenyl-2,7-diyl)-co-(1,4-benzo-thiadiazole)] (PFBT) and its incorporated porphyrin (PFBT-TPP) and metalated porphyrin (PFBT-(M)TPP) derivatives, as photocatalysts for aerobic oxidation of sulfide in aqueous media. Within our scope of study, PFBT with 1 mol% TPP incorporated (PFBT-1%TPP) displayed the best activity toward thioanisole oxidation, achieving a superior sulfoxide production of 1.33 mol g⁻¹ h⁻¹ and over 99% selectivity. A broad range of sulfides were fully converted into the corresponding sulfoxides with similar rates to thioanisole. Our photocatalytic system can be scaled up to gram-scale production. The photocatalytic examination revealed that both ¹O₂ and ˙O₂⁻ are oxidants in this case. The experimental findings were supported by DFT calculations. To the best of our knowledge, this is the first-time illustration of Pdots successfully functioning as a photocatalyst in sulfoxidation in water, with no cocatalysts or sacrificial electron donors needed.

2 Experimental

All chemicals and reagents, including benzaldehyde, bromobenzaldehyde, pyrrole, BF₃·OEt, 2,3-dichloro-5,6-dicyanobenzoquinone (DDQ), 9,9-diocylfluoro-2,7-dibobonic acid (F-BOH), Pd(PPh₃)₄, 4,7-dibromo-2,1,3-benzothiadiazole (BT-Br₂), and Poly(styrene-co-maleic anhydride) were purchased from Sigma-Aldrich, and all sulfide reagents were purchased from Tokyo Chemical Industry (TCI). Commercially available chemicals were used without further purification unless stated otherwise. All solvents (analytical grade) were purchased from commercial sources and were dried on molecular sieves and stored under a nitrogen atmosphere before use. Ultrapure water was obtained from a Merck Milli-Q® Reference type 1 Type Water Purification System.

Synthetic protocols and detailed characterization of 5-phenyldipyrromethane, 5,15-bis(4-bromophenyl)-10,20-diphenylporphyrin (TPP-Br₂), ZnTPP-Br₂, and CoTPP-Br₂ are provided in the ESI.†

2.1 Synthesis of conjugated polymers

F-BOH (0.214 g, 1.0 eq.), BT-Br₂, TPP-Br₂, and Pd(PPh₃)₄ (0.026 g, 0.05 eq.) were added to a pressurized tube. The amounts of BT-Br₂ and TPP-Br₂ were calculated according to the specific mole percentage fraction regarding the amount of F-BOH, which was always fixed at 50%. The tube was purged with N₂. A degassed 2 M K₂CO₃ solution (4 mL) and THF (16 mL) were transferred to the tube via a canula. The reaction mixture was heated at 85 °C under stirring for 48 h. After that, bromobenzene was added as an end-capping agent and the reaction was prolonged for another 12 h. The reaction was quenched with water, then extracted with chloroform and purified by precipitation into methanol. The obtained solid was filtered and further washed with DMF, followed by MeOH to get rid of the remaining starting materials and low-molecular-weight polymers before drying in a vacuum oven at 60 °C. Metalloporphyrin-based conjugated polymers were synthesized using the same procedure in which CoTPP-Br₂ or ZnTPP-Br₂ was added instead of TPP-Br₂.

2.2 Preparation of Pdots

The solution of the conjugated polymer and the solution of poly(styrene-co-maleic anhydride) (PSMA) in THF were prepared at a concentration of 1.0 mg mL⁻¹. To 5 mL of THF 250.0 μL of conjugated polymer solution and 187.5 μL of PSMA solution were added, and the mixture was quicky injected into 10 mL of Milli-Q water under sonication. THF was then removed by purging the solution mixture with a stream of N₂ under continuous sonication at 25 °C for 30 min to obtain a clear solution of Pdots.

2.3 Photocatalytic reactions

In a 10 mL glass tube equipped with a stir bar, 2 mL of a Pdot solution was added and sealed with a septum. The solution was purged with O₂ gas for 10 min, and then thioanisole (or other sulfides) (0.2 mmol) was added using a syringe. The reaction tube was equipped with an O₂ balloon. The reaction was illuminated using a home-made photoreactor (blue LED strip 20 W). The stirring was kept constant, and air flow was applied to dissipate heat from the light source throughout the photoreaction. Once the reaction time elapsed, the reaction mixture was extracted with ethyl acetate (3.5 mL × 3). The organic phase was collected and dried over Na₂SO₄. The product analysis was carried out using a Varian CP-3800 Gas Chromatograph equipped with a flame ionization detector (FID). A HP-5 capillary column (0.32 mm × 30 m) was used for the separation of reaction mixtures. The temperature program was set as 80 °C, 80–150 °C (10 °C min⁻¹) and 150–280 °C (20 °C min⁻¹). Quantitative analysis was carried out by the internal standard method using dodecane as an internal standard. The sulfide conversion and product selectivity were calculated with the equations shown below.

2.4 Computational method

All theoretical calculations were performed using TURBOMOLE version 7.5.0.⁴³ Density Functional Theory (DFT)⁴⁴ calculations were carried out using Becke's three-parameter hybrid exchange functional combined with the Lee–Yang–Parr correlation functional (B3LYP),⁴⁵ in combination with split valence and polarization functions on the heavy atom (def-SV(P))⁴⁶ basis set to optimize the ground state structures. A closed-shell restricted formalism was used to optimize the singlet ground state of PFBT-1%TPP and PFBT-1%ZnTPP complexes, whereas an unrestricted formalis was necessary to optimize the doublet ground state of the PFBT-1%CoTPP complex due to its unpaired electron. While a restricted formalism forces electrons to pair in occupied molecular orbitals in the conventional spin-up–spin-down manner, an unrestricted formalism results in α (spin down) and β (spin up) sets of singly occupied orbitals. Therefore, in the PFBT-1%CoTPP complex, the singly occupied molecular orbital (SOMO) becomes the α-SOMO, and each doubly occupied orbital is computed as two (one α and one β) SOMOs, such that the highest occupied molecular orbital (HOMO) is the α-SOMO-1 and β-SOMO. Calculations of vertical excitation energies were carried out using the optimized structures using the time dependent DFT (TDDFT) method at the same basis set. These calculations yielded energy levels of ground and excited states, oscillator strengths, and primary configurations of the low-lying electronically excited states. All TDDFT calculations were based on the optimized ground-state geometries of the respective models.

3 Results and discussion

3.1 Design, synthesis and characterization of porphyrin-based conjugated polymers

The molecular designs are based on donor–acceptor type conjugated copolymers containing fluorene and benzothiadiazole units (PFBT). To further improve the photocatalytic properties, tetraphenylporphyrin (TPP) and their metalated derivatives ((M)TPP), including zinc tetra-phenylporphyrin (ZnTPP) and cobalt tetraphenylphorphyrin (CoTPP), were incorporated into the main chains of PFBT, resulting in a series of PFBT-TPP and PFBT-(M)TPP polymers (Fig. 1a). We hypothesized that, during photocatalysis, the TPP units will serve as an additional or co-photosensitizer, empowering the generation of photocarriers and the charge transfer along the polymer main chains, thereby enhancing the efficiency of photocatalyzed chemical conversion. These conjugated polymers were synthesized by Suzuki–Miyamura cross-coupling between the diboronic acid-functionalized 9,9′-dioctylfluorene monomer and a mixture of 4,7-dibromo-2,1,3-benzothiadiazole and dibromoporphyrin monomers with Pd(PPh₃)₄ as a catalyst. In earlier studies, we attempted to increase the mole fraction of TPP units in PFBT-TPP up to 30%. Nonetheless, the resultant products were mainly oligomers, short-chain polymers, and exhibited alterations in solubility in THF. Consequently, PFBT polymers including 1 mol% and 5 mol% TPP units, together with 1 mol% of (M)TPP, demonstrating commendable photocatalytic activity (vide infra), were synthesized and will be the subject of the subsequent discussion. The purified polymers were acquired with yields ranging from 20–40%.

Fig. 1 Design, synthesis and characterization of porphyrin-based conjugated polymers. (a) Molecular structures of the polymers. (b) Absorption spectra of the polymers in THF. (c) Emission spectra of the polymers in THF (excitation wavelength = 460 nm). (d) Frontier molecular orbitals (HOMO and LUMO) of PFBT-1%TPP, PFBT-1%CoTPP, and PFBT-1%ZnTPP square planar complexes, and the PFBT chain.

We first characterized the chemical structures of the parent PFBT polymer and porphyrin-containing PFBT-TPP polymers. ¹H-NMR spectroscopy confirmed the presence of TPP units in the PFBT-TPP polymers (Fig. S1†). Due to low loading of TPP units, IR spectra of PFBT-TPP showed almost no difference to that of PFBT (Fig. S2†). The mole fractions of TPP units incorporated in PFBT-TPP polymers were 0.74 and 4.3 mol% for PFBT-1%TPP and PFBT-5%TPP, respectively (Fig. S1†), as calculated by using the ¹H-NMR integral ratio between the methyl protons of dioctyl substituents in the 9,9′-dioctylfluorenyl-2,7-diyl unit against N–H protons of TPP rings. For metalloporphyrin-containing polymers, the actual contents of incorporated metals were 44% (PFBT-1%CoTPP) and 77% (PFBT-1%ZnTPP) as analyzed by inductively coupled plasma optical emission spectroscopy. These values were slightly lower than the feeding TPP value of 1 mol%. Furthermore, the number average molecular weight (M_n) and polydispersity index (PDI) of all polymers were analyzed by gel permeation chromatography (Table S1†). Adding higher mole fractions of TPP units led to shorter polymer chains (M_n = 47.2 kg mol⁻¹ for PFBT, 38.8 kg mol⁻¹ for PFBT-1%TPP and 15.3 kg mol⁻¹ for PFBT-5%TPP), while the molecular weight distribution became narrower (i.e. lower PDI values) with the increasing mole fraction of TPP units. The incorporation of 1 mol% of (M)TPP units also reduced the M_n of the polymers as compared to that of the non-metalated PFBT-1%TPP polymer. These trends likely stem from the decreasing solubility of growing polymer chains as more TPP units are added, thus hindering the formation of longer TPP-containing polymer chains in the synthetic medium.

Photophysical properties of PFBT, PFBT-TPP and PFBT-(M)TPP polymers were compared in their molecularly dissolved state. The solution of PFBT polymer in THF displayed the main absorption peak at 455 nm (Fig. 1b). For TPP- and (M)TPP containing polymers, new absorption peaks, including the intense Soret band at 418 nm and a group of weak Q bands between 500 and 650 nm, could be observed, thus indicating the presence of TPP- and (M)TPP units in the conjugated polymers (Fig. 1b and S3†). For photoluminescence characterization, all polymers showed emission bands at 534 nm, which originated from the conjugated PFBT backbone (Fig. 1c).^42,47 Interestingly, the emission intensity became lower when incorporating TPP or (M)TPP units in the main chains (Fig. 1c and S4†). Such emission quenching strongly suggests the energy transfer between PFBT chains and TPP or (M)TPP units.⁴⁸ Therefore, we can conclude that the porphyrin derivatives were successfully incorporated into the main chains of PFBT polymers. To understand the influence of TPP units on the electronic structures of the copolymers, the energy levels of frontier molecular orbitals of PFBT-1%TPP were computationally calculated (Fig. S5†). It can be seen that the HOMOs and LUMOs of PFBT-1%TPP are distributed over both PFBT and TPP units, signifying that the TPP unit is well conjugated into the PFBT main chain, and a certain degree of delocalization is anticipated. The computational analysis of PFBT demonstrated a smaller bandgap when a larger conjugated system was constructed (Fig. S6†). and PFBT (n = 4) was calculated to have a bandgap of 2.70 eV, a LUMO level at −2.60 eV and a HOMO level at −5.31 eV. Once (M)TPP was integrated to PFBT, a slight change in bandgap values was observed in the range of 2.60–2.70 eV, while the LUMO and HOMO levels of PFBT-TPP and PFBT-(M)TPP were changed in energy to ca. −2.46 eV and −5.06 to −5.16 eV, respectively (Fig. 1c).

3.2 Preparation and characterization of Pdots

Pdots were prepared by coprecipitation of PFBT-based polymers and PSMA. In brief, the conjugated polymer and PSMA were dissolved in THF. The mixture was then quickly poured into water under sonication. Evaporating THF yielded Pdots, in which PFBT-based polymers formed nanoparticles stabilized by the PSMA surfactant, as a solution in water (Fig. 2a). Transmission electron microscopy revealed that these nanoparticles are spherical with an average size of 33–37 nm (Fig. 2b and S7†). These results corroborated well with their hydrodynamic diameter of approximately 36–53 nm as analyzed by dynamic light scattering (Fig. S8†). All Pdot samples also displayed a narrow size distribution.

Fig. 2 Preparation and characterization of Pdots from porphyrin-based conjugated polymers. (a) Preparation process of Pdots. (b) Transmission electron micrographs of PFBT-1%TPP Pdots. The scale bar is 50 nm. (c) Absorption spectra of Pdots in water. (d) Emission spectra of Pdots in water (excitation wavelength = 460 nm).

Pdots exhibited different photophysical properties when compared to the loose polymer solution in THF. In particular, the absorption spectrum of Pdots was slightly broadened and red-shifted (Fig. 2c and S9†), indicating the interchain interactions and the aggregation of polymer chains in Pdots.^42,49–52 The formation of Pdots also affects the photoluminescence of the polymers (Fig. 2d and S10†); the emission was quenched and also red-shifted when compared to that in THF. The decrease in fluorescence intensity could be attributed to the interchain interactions of conjugated polymer aggregates inside the Pdot core, leading to a prolonged excited-state lifetime as well as more non-radiative relaxation and migration of electrons to the quenching sites.^42,53 An attempt to measure the accurate fluorescence lifetimes to probe these effects was unsuccessful due to the instrumental limitation for sub nanosecond analysis. In addition to the emission of the PFBT backbone, the characteristic fluorescence bands of TPP units could be clearly observed at 655 nm and 720 nm.⁵⁴(M)TPP-based Pdots displayed similar absorption and emission profiles to the benchmark polymers. However, the TPP fluorescence was quenched and blue-shifted in (M)TPP-based Pdots (e.g., 600 nm and 650 nm observed in PFBT-1%ZnTPP Pdots).⁵⁵ Some physical properties of Pdots are summarized in Table S2†.

3.3 Photocatalytic aerobic oxidation of sulfide derivatives using Pdots

We initially examined the visible light-mediated (blue-LED) aerobic oxidation of thioanisole (0.2 mmol, 0.1 M) as the model substrate utilizing PFBT Pdots (87.5 μg) as the photocatalyst, under a static O₂ balloon, at ambient temperature and pressure, all conducted in water. The products from the reaction were extracted using ethyl acetate and quantified by GC-FID (Fig. S11†). In the absence of O₂, light, or PFBT-Pdots, negligible amounts of oxidized products were detected suggesting that these components are indispensable for the oxidation of thioanisole in the given system. Thioanisole conversion and methyl phenyl sulfoxide (MPS) production were observed in the photocatalytic reaction.

A time profile of the photocatalytic process was obtained over a 24-h duration (Fig. 3a). The conversion of thioanisole progressively increased, achieving >98% conversion after 10 h, with >95% selectivity. Moreover, there was no evidence of sulfone formation or other degradation products when the photocatalytic activity was tested for an additional duration of 14 h. This may be attributed to the formation of hydrogen bonds between water and thioanisole, thereby preventing over-oxidation to the corresponding sulfone.^56,57

Fig. 3 Photocatalytic oxidation of sulfide using Pdots. (a) Time profile of sulfide conversion and sulfoxide production catalyzed by PFBT Pdots. (b) Time profile of sulfide conversion catalyzed by PFBT Pdots and PFBT-TPP Pdots. (c) Time profile of sulfide conversion catalyzed by PFBT Pdots, PFBT-TPP Pdots and PFBT-(M)TPP Pdots. All experiments were performed in the presence of 0.2 mmol sulfide, 2 mL of Pdot aqueous solution, and an O₂ balloon, under photo-irradiation with a blue LED (λ_max = 456 nm).

Recognizing the potential of PFBT Pdots in photocatalytic aerobic oxidation of sulfide, we sought to modify the structure of the conjugated PFBT polymer by incorporating TPP units into the polymer chain. We anticipated that porphyrin would serve as a promising candidate for enhancing the generation of ROS, as demonstrated in prior literature studies.^58,59PFBT-1%TPP Pdots was evaluated as a photocatalyst in the oxidation of sulfides under similar circumstances. The photocatalytic efficiency was significantly enhanced, achieving 100% conversion of thioanisole to MPS within 3 h with >99% selectivity (Fig. 3b).

Under a limited oxygen concentration atmosphere (i.e. air zero), thioanisole conversion was compromised and decreased to 25%. Furthermore, it was noted that the photocatalytic activity remained unchanged despite the increase in the amount of TPP units integrated into the conjugated polymer chain, as PFBT-5%TPP Pdots served as a photocatalyst. Thus, only PFBT-1%TPP Pdots will be subjected to further photocatalytic investigation and discoursed in this work.

We then assessed the enhanced catalytic efficacy of the metal-free photocatalyst, PFBT-1%TPP Pdots, in comparison to PFBT-1%ZnTPP Pdots and PFBT-1%CoTPP Pdots (Fig. 3c).

The photocatalyzed conversion of thioanisole to MPS using PFBT-1%ZnTPP Pdots was much slower than that using PFBT-1%TPP Pdots; nonetheless, the conversion steadily increased with time and a full conversion was achieved within 10 h, akin to the process catalyzed by PFBT Pdots. In contrast to the others, the photocatalysis by PFBT-1%CoTPP Pdots was sluggish. An extended induction time was noticed throughout the initial 8 h, with around 25% sulfide conversion recorded at 12 h. This comparative experiment marked the superior photocatalytic performance of the metal-free conjugated polymer dots for sulfide oxidation under the studied conditions.

Having established the optimal reaction conditions and the best performing photocatalyst, the scope of sulfide substrates was expanded to cover electron-donating and electron-withdrawing derivatives of thioanisole, dialkyl thioether, and diphenyl thioether. The photo-oxidation of various sulfides (0.2 mmol, 0.1 M) employing PFBT-1%TPP Pdots as the photocatalyst under identical conditions was conducted with 3 h irradiation. Table 1 demonstrates that photocatalysis with PFBT-1%TPP Pdots consistently achieved good yield and selectivity across many substrates, indicating broad substrate compatibility. Nonetheless, the conversion efficiency of these substrates may be influenced by their electronic effects. The electron-withdrawing aryl-substituted sulfide (Entry 3, Table 1) attained only 81% conversion and 96% selectivity. This observation is frequently reported in other studies in this field.^60,61 Dibutyl thioether underwent photocatalysis, resulting in 100% conversion with 88% selectivity of dibutyl sulfoxide and 12% of dibutyl sulfone (Entry 4, Table 1). Dibutyl thioether was the only substrate that underwent over-oxidation to yield sulfone under these catalytic conditions. This is anticipated to be attributed to the most electron-rich sulfur atom of dibutyl sulfide among all tested substrates. Lastly, the oxidation of diphenyl sulfide is commonly recognized as more challenging than that of analogous dialkyl or alkyl aryl thioethers due to the reduced electron density on the sulfur atom. The oxidation of diphenyl sulfide with singlet oxygen is considerably less active in comparison to the oxidation of other sulfides.^11,62 The comprehensive photo-oxidation of diphenyl sulfide to the corresponding sulfoxide (100% conversion with ≥99 selectivity, Entry 5 Table 1), suggests that our catalytic system may use both ¹O₂ and ˙O₂⁻ as ROS (vide infra).

Table 1 Photooxidation of various disulfides using PFBT-1%TPP Pdots^a

Entry	Substrate	Product	Conv. (%)^b	Sel (%)^b
a All experiments were performed in the presence of 0.2 mmol sulfide, 2 mL of PFBT-1%TPP Pdot aqueous solution (50 mg of PFBT-1%TPP), and an O2 balloon, under 3 h photo-irradiation with a blue LED (λmax = 456 nm). b Determined by GC-FID using dodecane as an internal standard. Three repeat experiments were carried out and the underlying error estimate is ±3%.
1			100	≥99
2			100	95
3			81	96
4			100	88 (12% sulfone)
5			100	≥99

---

The endeavours to increase the extent of thioanisole photooxidation were evident. The gram-scale reaction was executed successfully with 10 mmol thioanisole (1.2 g) and 100 mL of PFBT-1%TPP Pdot aqueous solution under 3 h of blue LED illumination (Fig. S12†). The resulting product, MPS, was obtained in a 95% isolated yield with high NMR purity (Fig. S13†). Crystallization in cold methanol was utilised to obtain additional purification. This effective scale-up reaction shed light on the practical use of the green and sustainable photocatalytic sulfide oxidation. It is essential to highlight that our research showcases effective sulfide photo-oxidation at milligram to gram scales in pure water, a phenomenon that is rather uncommon due to the insolubility of sulfides in water and limited mass transfer. Our photocatalysis results were incorporated into a small compilation of studies on light-mediated aerobic oxidation of sulfides in aqueous or water-based binary systems for thorough and comparative analysis (Table S3†).

3.4 Mechanistic investigation

Solid-state diffuse reflectance UV-vis spectroscopy and the Tauc plots (Fig. S14†) were utilized to determine the band gaps, which were found to be as follows: 2.37 eV for PFBT, 2.29 eV for PFBT-1%TPP and PFBT-1%ZnTPP, and 2.30 eV for PFBT-1%CoTPP. The positive slopes of the Mott–Schottky plots confirmed the n-type characteristics of the conjugated polymers (Fig. S15†). The flat-band potentials (E_FB) were estimated to be −1.90 V vs. NHE for PFBT, −1.52 V vs. NHE for PFBT-1%TPP, −1.27 V vs. NHE for PFBT-1%CoTPP, and −0.91 V vs. NHE for PFBT-1%ZnTPP, which suggests that ˙O₂⁻ species can be generated from O₂ in the presence of Pdots containing such conjugated polymers (E(³O₂/˙O₂⁻) = −0.33 V vs. NHE).⁶³ Generally porphyrin and metalloporphyrin moieties have long been known to be in their triplet excited state upon visible light irradiation. A singlet oxygen, ¹O₂, can be generated through energy transfer from a triplet excited porphyrin to O₂.⁶⁴ To further the understanding of the photocatalytic process, controlled studies and scavenger-additive tests were conducted (Fig. 4). In the absence of light or oxygen, photocatalytic reactions did not occur. The use of para-benzoquinone (p-BQ) as an ˙O₂⁻ scavenger in the PFBT Pdots-photocatalyzed thioanisole-to-MPS reaction under standard conditions with 3 h of illumination led to a decrease in sulfide conversion from 100% to 47% while maintaining over 99% selectivity. A slight drop in sulfide conversion (from 100% to 91%) was found upon the addition of an equivalent of 1,4-diazabicyclo[2.2.2]octane (DABCO) as an ¹O₂ scavenger in the process. Substantially reduced sulfide conversions were seen with the addition of p-BQ (11% conversion, >99% selectivity) and DABCO (55% conversion, >99% selectivity) in the PFBT-1%TPP Pdot-catalyzed photo-oxidation of thioanisole to MPS. Although distinguishing between the two photo-oxidation pathways involving ˙O₂⁻ and ¹O₂ is non-trivial, the results suggested that the ROS generated from the light activation of PFBT Pdot are mostly based on ˙O₂⁻ through the SET mechanism. As anticipated, the integration of porphyrin units in PFBT-1%TPP Pdot significantly enhanced the amount of ¹O₂ produced via the ET process. Overall, the elevated formation of ROS was observed in the PFBT-1%TPP Pdot-photocatalyzed process.

Fig. 4 Controlled and scavenger-additive experiments. All experiments were performed in the presence of 0.2 mmol sulfide, 2 mL of Pdot aqueous solution, and an O₂ balloon, under 3 h photo-irradiation with a blue LED (λ_max = 456 nm). The additive scavenger was added at a 0.1 : 1 mole ratio to the substrate. The photocatalytic reactions were catalyzed by (a) PFBT Pdots and (b) PFBT-1%TPP Pdots.

Equipped with this information, we were able to suggest plausible processes for photo-oxidation of sulfide (Fig. 5). Upon the illumination of PFBT-1%TPP Pdot, electrons in the HOMO are promoted to the LUMO, generating free electrons and positively charged holes. This consequently facilitates two concurrent routes for ROS generation via ET and SET processes. For the ET pathway, the generated ¹O₂ reacts with electron-rich sulfide producing a diradical peroxysulfoxide intermediate (A). This species then reacts with another sulfide giving an unstable hypervalent intermediate (B), which eventually transforms into the corresponding sulfoxide.⁶⁵ Simultaneously, the positively charged hole oxidizes sulfide forming a radical cation (C), which then reacts with ˙O₂⁻, leading to the peroxide cation intermediate (D) in aqueous solution. The intermediate interacts with an additional sulfide, subsequently undergoing deprotonation to produce the equivalent sulfoxide. The process, including proton transfer stages, suggests that a protic medium such as water may significantly facilitate the synthesis of sulfoxide.⁶⁶

Fig. 5 Proposed mechanism for the photo-oxidation of sulfide to sulfoxide via electron transfer (ET) and single electron transfer (SET) reactions.

3.5 Computational analysis

To investigate the influence of porphyrin-based and metal-porphyrin doped conductive polymers on singlet oxygen (¹O₂) generation, we constructed three distinct models including PFBT-1%TPP, PFBT-1%ZnTPP, and PFBT-1%CoTPP, representing configurations in the absence and presence of a metal center (Fig. 6 and S16†). Both PFBT-1%TPP and PFBT-1%ZnTPP complexes exhibited closed-shell systems. In the PFBT-1%ZnTPP complex, the Zn metal center adopted a square planar coordination geometry with a Zn²⁺ oxidation state, comprising a fully occupied 3d-orbital, [Ar] 3d¹⁰ electron configuration. However, an open-shell system was observed for the PFBT-1%CoTPP square planar complex. In this case, the Co metal center took on a Co²⁺ oxidation state with an [Ar] 3d⁷ electron configuration. According to the crystal field theory for square planar complex splitting, the presence of one unpaired electron in the d_xy orbital results in a doublet state for this open-shell system.

When considering the excitation states and corresponding energy levels of PFBT-1%TPP as a closed-shell system, we found that the primary singlet excitation was S₀ → S₅ (Fig. 6). This excitation process involved the HOMO-2 → LUMO + 1 transition, corresponding to an intramolecular charge transfer (ICT) within the polymer chain. For the emission process, the main fluorescence peaks of PFBT-1%TPP observed at 534 nm and 655 nm (Fig. 1c and 2d) originated from relaxation pathways involving LUMO + 3 → HOMO (S₂ → S₀) and LUMO → HOMO (S₁ → S₀) transitions, thus indicating a strong contribution from the TPP unit. In the context of the triplet excited state, the relative energy level of the S₁ state closely aligned with that of the T₅ state, allowing a population of intersystem crossing (ISC) toward the triplet state. This ISC was driven by the energetic difference between excited singlet and triplet states with a intersystem crossing energy gap (ΔE_ISC) of 0.15 eV. Subsequently, energy transfer via processes like triplet–triplet annihilation or triplet−triplet energy transfer led to singlet oxygen (¹O₂) generation.

Fig. 6 Optimized structures of porphyrin-based conjugated polymers. Selected frontier molecular orbitals involved in excitations and relaxations of (a) PFBT-1%TPP and (b) PFBT-1%CoTPP square planar complexes. Energies are in eV.

An introduction of metal centers in PFBT-(M)TPP polymers could affect the generation of ¹O₂ to a certain degree. In PFBT-1%ZnTPP, the Zn metal center led to a slight alteration in the primary singlet excitation, transitioning to S₀ → S₇ (HOMO-2 → LUMO + 1, Fig. S16†). Given the minor adjustments in relative energy levels of S₁ and T₅ states (ΔE_ISC = 0.27 eV), processes like triplet–triplet annihilation or triplet−triplet energy transfer could also potentially facilitate the generation of ¹O₂.⁶⁷ Interestingly, upon introducing Co metal in the PFBT-1%CoTPP complex, we observed that the dominant doublet excitation is the D₀ → D₃₂ transition, corresponding to a balanced mixture of β-SOMO-2 → β-LUMO and α-SOMO-2 → α-LUMO type excitations in the doublet orbitals. These transitions involved the ICT within the polymer chain and pyridinic N moiety, while other transitions were negligible. Upon relaxation from D₁₃ to D₀, the excited state of PFBT-1%CoTPP underwent radiative decay from β-LUMO + 2 to β-SOMO-1, leading to the emission at 534 nm. This transition was also associated with the ICT within the polymer chain. As shown in Fig. 6b, there was no driving force present to trigger quartet formation through ISC, either in the ground-state geometry or in a relaxed Q₁ state. Consequently, there was no energy pathway for quartet-triplet annihilation to induce or excite the transition from triplet ground state oxygen (³O₂) to singlet oxygen (¹O₂) in the excited state.⁶⁸ Furthermore, it is noteworthy that the theoretical calculations indicated that both absorption and emission of PFBT-(M)TPP polymers predominantly originated from the PFBT backbone, further corroborating the experimental data, which showed no distinct TPP-based spectral features.

Taken together, computational analysis aligned with the experimental results in terms of the maximum absorption wavelengths (PFBT-1%TPP: calculated 448 nm and experimental 450 nm; PFBT-1%CoTPP: calculated 448 nm and experimental 450 nm), the primary contributions to the emission spectra, and the photo-aerobic phenomena. This underpins the superior photo-aerobic efficacy bestowed by the metal free PFBT-1%TPP to its metal-containing counterparts.

4 Conclusions

This study demonstrated, for the first time, the feasibility of using metal-free porphyrin-based conjugated polymer dots (PFBT-1%TPP Pdots) as efficient photocatalysts for the aerobic oxidation of sulfides in water. PFBT-1%TPP Pdots, exhibiting exceptional dispersion in water, facilitated the efficient photooxidation of thioanisole and other sulfide derivatives, achieving a remarkable conversion efficiency of 80–100% and high selectivity of 88–100% when using low catalytic loading under 3 h of LED illumination. The photocatalytic process was scaled to produce gram quantities of the isolated product with high selectivity and no signs of over-oxidation. This demonstrates a sustained and practical application. The computational analysis and experimental findings illustrated that PFBT-1%TPP Pdots effectively generated ROS (both ¹O₂ and ˙O₂⁻) during light activation, leading to enhanced photocatalytic efficiency compared to PFBT Pdots and their metal-containing counterparts. We anticipate that our approach will offer a sustainable alternative to traditional methods for the direct oxidation of sulfides to sulfoxides, which commonly rely on harsh conditions and toxic reagents. Future research directions could explore the application of these photocatalysts for other organic transformations in aqueous media.

Data availability

The data supporting this article have been included as part of the ESI.†

Author contributions

Wissuta Boonta: investigation, methodology, writing – origin draft. Ratanakorn Teerasaranyanon: investigation, methodology, data curation. Chattarika Sukpattanacharoen: computational calculations, writing. Sanhawat Rumporee: investigation, data curation. Phurinat Lorwongkamol: preliminary investigation. Nattapong Paiboonvorachat: data curation, analysis. Pongphak Chidchob: data curation, writing – review and editing. Suwit Suthirakun: computational calculations, data curation. Junjuda Unruangsri: conceptualization, funding acquisition, supervision, writing – review and editing.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

The authors gratefully acknowledge the financial support from the Photocatalysts for Clean Environment and Energy Research Unit, Chulalongkorn University and the International Science Partnerships Fund (ISPF, award number 1203758747), British Council. We would also like to thank NSTDA Supercomputer Center (ThaiSC) and Institute of Science, Suranaree University of Technology for the computational resources. W. B. and R. T. were supported by the Development and Promotion of Science and Technology Talents Project (DPST) and Ratchadapisek Somphot Fund for Postdoctoral Fellowship, Chulalongkorn University, respectively.

Notes and references

1. A. P. Pulis and D. J. Procter, Angew. Chem., Int. Ed., 2016, 55, 9842–9860.
2. E. Wojaczynska and J. Wojaczynski, Chem. Rev., 2020, 120, 4578–4611.
3. S. Nishiguchi, T. Izumi, T. Kouno, J. Sukegawa, L. Ilies and E. Nakamura, ACS Catal., 2018, 8, 9738–9743.
4. P. Kalaba, M. Ilic, N. Y. Aher, V. Dragacevic, M. Wieder, M. Zehl, J. Wackerlig, S. Beyl, S. B. Sartori, K. Ebner, A. Roller, N. Lukic, T. Beryozkina, E. R. P. Gonzalez, P. Neill, J. A. Khan, V. Bakulev, J. J. Leban, S. Hering, C. Pifl, N. Singewald, J. Lubec, E. Urban, H. H. Sitte, T. Langer and G. Lubec, J. Med. Chem., 2020, 63, 391–417.
5. S. Otsuki, T. Nonaka, N. Takashima, W. Qian, A. Ishihara, T. Imai and T. Kabe, Energy Fuels, 2000, 14, 1232–1239.
6. F. Figueras, J. Catal., 2004, 226, 25–31.
7. M. L. Marin, L. Santos-Juanes, A. Arques, A. M. Amat and M. A. Miranda, Chem. Rev., 2012, 112, 1710–1750.
8. S. Caron, R. W. Dugger, S. G. Ruggeri, J. A. Ragan and D. H. Ripin, Chem. Rev., 2006, 106, 2943–2989.
9. X. Yang and D. Wang, ACS Appl. Energy Mater., 2018, 1, 6657–6693.
10. E. Skolia, P. L. Gkizis and C. G. Kokotos, ChemPlusChem, 2022, 87, e202200008.
11. A. G. Mojarrad and S. Zakavi, Catal. Sci. Technol., 2018, 8, 768–781.
12. S. M. Bonesi, M. Fagnoni, S. Monti and A. Albini, Tetrahedron, 2006, 62, 10716–10723.
13. X. Gu, X. Li, Y. Chai, Q. Yang, P. Li and Y. Yao, Green Chem., 2013, 15, 357–361.
14. H. Wang, G. W. Wagner, A. X. Lu, D. L. Nguyen, J. H. Buchanan, P. M. McNutt and C. J. Karwacki, ACS Appl. Mater. Interfaces, 2018, 10, 18771–18777.
15. D. Madhavan and K. Pitchumani, Tetrahedron, 2001, 57, 8391–8394.
16. J. Dad'ová, E. Svobodová, M. Sikorski, B. König and R. Cibulka, ChemCatChem, 2012, 4, 620–623.
17. S. M. Bonesi, I. Manet, M. Freccero, M. Fagnoni and A. Albini, Chem.–Eur. J., 2006, 12, 4844–4857.
18. X. Li, Y. Wang, L. Yang, Z. Zhang and X. Xie, Tetrahedron, 2022, 110, 132708.
19. W. Zhao, C. Yang, J. Huang, X. Jin, Y. Deng, L. Wang, F. Su, H. Xie, P. K. Wong and L. Ye, Green Chem., 2020, 22, 4884–4889.
20. B. Zhang, J. Li, B. Zhang, R. Chong, R. Li, B. Yuan, S.-M. Lu and C. Li, J. Catal., 2015, 332, 95–100.
21. F. Huang, H. Hao, W. Sheng, X. Dong and X. Lang, J. Colloid Interface Sci., 2023, 630, 921–930.
22. H. Hao, X. Li and X. Lang, Appl. Catal. B: Environ., 2019, 259, 118038.
23. X. Lang, J. Zhao and X. Chen, Angew. Chem., Int. Ed., 2016, 55, 4697–4700.
24. Y. Xu, Z.-C. Fu, S. Cao, Y. Chen and W.-F. Fu, Catal. Sci. Technol., 2017, 7, 587–595.
25. H. Wei, Z. Guo, X. Liang, P. Chen, H. Liu and H. Xing, ACS Appl. Mater. Interfaces, 2019, 11, 3016–3023.
26. W. P. To, Y. Liu, T. C. Lau and C. M. Che, Chem.–Eur. J., 2013, 19, 5654–5664.
27. A. Casado-Sanchez, R. Gomez-Ballesteros, F. Tato, F. J. Soriano, G. Pascual-Coca, S. Cabrera and J. Aleman, Chem. Commun., 2016, 52, 9137–9140.
28. J. Li, Y. Chen, X. Yang, S. Gao and R. Cao, J. Catal., 2020, 381, 579–589.
29. P. Zhang, Y. Wang, H. Li and M. Antonietti, Green Chem., 2012, 14, 1904–1908.
30. X. Li, X. Ma, F. Zhang, X. Dong and X. Lang, Appl. Catal., B, 2021, 298, 120514.
31. J. Li, Z. An, J. Sun, C. Tan, D. Gao, Y. Tan and Y. Jiang, ACS Appl. Mater. Interfaces, 2020, 12, 35475–35481.
32. B. L. Miller, T. D. Williams and C. Schöneich, J. Am. Chem. Soc., 1996, 118, 11014–11025.
33. C. Xing, J. Deng, R. Tan, M. Gao, P. Hao, D. Yin and D. Yin, Catal. Sci. Technol., 2017, 7, 5944–5952.
34. Y. Hong, J. W. Lam and B. Z. Tang, Chem. Commun., 2009, 4332–4353.
35. D. Chen, I. C. Wu, Z. Liu, Y. Tang, H. Chen, J. Yu, C. Wu and D. T. Chiu, Chem. Sci., 2017, 8, 3390–3398.
36. L. Guo, J. Ge and P. Wang, Photochem. Photobiol., 2018, 94, 916–934.
37. X. Men and Z. Yuan, ACS Appl. Polym. Mater., 2020, 2, 4319–4330.
38. L. Wang, R. Fernandez-Teran, L. Zhang, D. L. Fernandes, L. Tian, H. Chen and H. Tian, Angew. Chem., Int. Ed., 2016, 55, 12306–12310.
39. A. Liu, L. Gedda, M. Axelsson, M. Pavliuk, K. Edwards, L. Hammarstrom and H. Tian, J. Am. Chem. Soc., 2021, 143, 2875–2885.
40. P.-J. Tseng, C.-L. Chang, Y.-H. Chan, L.-Y. Ting, P.-Y. Chen, C.-H. Liao, M.-L. Tsai and H.-H. Chou, ACS Catal., 2018, 8, 7766–7772.
41. M. H. Elsayed, M. Abdellah, Y. H. Hung, J. Jayakumar, L. Y. Ting, A. M. Elewa, C. L. Chang, W. C. Lin, K. L. Wang, M. Abdel-Hafiez, H. W. Hung, M. Horie and H. H. Chou, ACS Appl. Mater. Interfaces, 2021, 13, 56554–56565.
42. B. Cai, M. Axelsson, S. Zhan, M. V. Pavliuk, S. Wang, J. Li and H. Tian, Angew. Chem., Int. Ed., 2023, 62, e202312276.
43. R. Ahlrichs, M. Bär, M. Häser, H. Horn and C. Kölmel, Chem. Phys. Lett., 1989, 162, 165–169.
44. F. Furche and R. Ahlrichs, J. Chem. Phys., 2002, 117, 7433–7447.
45. A. D. Becke, Phys. Rev. A, 1988, 38, 3098–3100.
46. F. Weigend and R. Ahlrichs, Phys. Chem. Chem. Phys., 2005, 7, 3297–3305.
47. Y. He, S. Gong, R. Hattori and J. Kanicki, Appl. Phys. Lett., 1999, 74, 2265–2267.
48. C. Wu, Y. Zheng, C. Szymanski and J. McNeill, J. Phys. Chem. C, 2008, 112, 1772–1781.
49. J. Pecher and S. Mecking, Chem. Rev., 2010, 110, 6260–6279.
50. C. Szymanski, C. Wu, J. Hooper, M. A. Salazar, A. Perdomo, A. Dukes and J. McNeill, J. Phys. Chem. B, 2005, 109, 8543–8546.
51. C. Wu, C. Szymanski and J. McNeill, Langmuir, 2006, 22, 2956–2960.
52. C. Wu, B. Bull, C. Szymanski, K. Christensen and J. McNeill, ACS Nano, 2008, 2, 2415–2423.
53. T.-Q. Nguyen, I. B. Martini, J. Liu and B. J. Schwartz, J. Phys. Chem. B, 2000, 104, 237–255.
54. M. Uttamlal and A. Sheila Holmes-Smith, Chem. Phys. Lett., 2008, 454, 223–228.
55. Z. Valicsek and O. Horváth, Microchem. J., 2013, 107, 47–62.
56. E. Skolia, P. L. Gkizis, N. F. Nikitas and C. G. Kokotos, Green Chem., 2022, 24, 4108–4118.
57. B. Zhao, G. B. Hammond and B. Xu, Tetrahedron Lett., 2021, 82, 153376.
58. R. Chen, J.-L. Shi, Y. Ma, G. Lin, X. Lang and C. Wang, Angew. Chem., Int. Ed., 2019, 58, 6430–6434.
59. J. Jiang, Z. Liang, X. Xiong, X. Zhou and H. Ji, ChemCatChem, 2020, 12, 3523–3529.
60. Y. Gao, H. Xu, S. Zhang, Y. Zhang, C. Tang and W. Fan, Org. Biomol. Chem., 2019, 17, 7144–7149.
61. Q. Fan, L. Zhu, X. Li, H. Ren, G. Wu, H. Zhu and W. Sun, Green Chem., 2021, 23, 7945–7949.
62. C.-L. Gu, C. S. Foote and M. L. Kacher, J. Am. Chem. Soc., 2002, 103, 5949–5951.
63. W. Sheng, H. Hao, F. Huang, F. Zhang and X. Lang, Chem. Eng. J., 2022, 430, 133071.
64. Y. Z. Chen, Z. U. Wang, H. Wang, J. Lu, S. H. Yu and H. L. Jiang, J. Am. Chem. Soc., 2017, 139, 2035–2044.
65. Q. Song, W. Li, F. Shan, X. Peng, L. Wang, Z. Wang and X.-Q. Yu, Nano Lett., 2024, 24, 13895–13902.
66. L. Ding, Y. Zhao, Y. Li, C. Ge, L. Lv, H. Ji, W. Song, C. Chen and J. Zhao, J. Catal., 2023, 426, 328–335.
67. H. Wang, S. Jiang, S. Chen, D. Li, X. Zhang, W. Shao, X. Sun, J. Xie, Z. Zhao, Q. Zhang, Y. Tian and Y. Xie, Adv. Mater., 2016, 28, 6940–6945.
68. C. M. Wehrmann, M. Imran, C. Pointer, L. A. Fredin, E. R. Young and M. S. Chen, Chem. Sci., 2020, 11, 10212–10219.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d5se00146c

‡ These authors contributed equally.

---