Swimming upstream – photocatalytic depolymerization of lignosulfonate in seawater

Abstract

The valorization of lignin, a major component of lignocellulosic biomass, is essential for reducing reliance on fossil-based resources. Lignosulfonates, water-soluble lignin derivatives produced in large volumes by the pulp and paper industry, remain underutilized due to lignin's structural complexity and the harsh conditions typically required for its depolymerization. Here, we present a mild and sustainable photocatalytic method for depolymerizing sodium lignosulfonate (LS) using sodium anthraquinone-2-sulfonate (AQ2S), a non-toxic, water-soluble, and commercially available photocatalyst. The reaction proceeds under UV LED irradiation at room temperature and ambient pressure in saline aqueous media, including untreated Baltic Sea seawater, without the need for organic solvents, acids, bases, or hazardous additives. Salinity is crucial for stabilizing AQ2S and maintaining its activity during prolonged irradiation. Computational methods at the density functional theory (DFT) level provide insight into the reaction mechanism and catalyst behavior, showing that in the absence of salt, AQ2S is deactivated via hydroxylation reactions. Conversely, in saline media it remains catalytically active in mediating LS depolymerization. The process is successful while yielding lower-molecular-weight lignin fragments, as confirmed by SEC, UV-vis spectroscopy, NMR, and FTIR, with evidence of decreased aromaticity and preserved structural motifs. This study demonstrates the utility of seawater as a green reaction medium and introduces an eco-friendly, mechanistically informed strategy for lignin conversion.

---

Bruno V M Rodrigues

Dr Bruno Manzolli Rodrigues earned his B.Sc. (2009) and Ph.D. (2014) in Chemistry from the University of São Paulo, Brazil. He served as a Professor at Universidade Brasil (2016–2022) and as a Senior Researcher at Stockholm University (2020–2022). In 2022, he joined the University of Wuppertal, Germany, as a Senior Research Associate. Author of over 70 peer-reviewed articles, and recipient of the 2024 John C. Warner Early Career Researcher Prize, he has secured multiple third-party grants. His research centers on catalysis for biomass valorization, renewable feedstocks, and sustainable chemistry, integrating both fundamental investigations and applied solutions for green technologies.

---

Green foundation

1. This work introduces a novel, sustainable approach to valorize sodium lignosulfonate using a commercial, non-toxic photocatalyst under UV LED irradiation. The process requires no harmful additives, organic solvents, or special solubilizing agents, thereby minimizing both environmental impact and energy use.

2. The depolymerization of lignosulfonate could be performed even in pure, untreated seawater, a truly abundant and renewable solvent. This eliminates the need for purified freshwater or additional reagents, which are typically required in similar systems. The photocatalyst remains active and can be reusable under UV LED light, a low-energy and scalable irradiation source. The reaction yields lower-molecular-weight fragments, representing a successful valorization pathway for a biomass waste.

3. Future research should explore in situ product use without extraction and adapt the process to solar light or continuous flow to enhance scalability and reduce energy input.

Introduction

Valorization of biomass-derived wastes is of great importance in tackling the problems arising from the world's dependence on fossil-based resources. Lignosulfonates (Fig. 1a), as part of lignocellulosic residues generated by the pulp and paper industry in large quantities, are particularly interesting due to their water solubility, abundance and chemical versatility.^1,2 Derived from the sulfite pulping process, lignosulfonates make up 90% of commercial lignin.³ Despite their potential, they are often underutilized and typically used for low-value applications, such as dispersants or are combusted for energy recovery.^3,4

Fig. 1 (a) A proposed chemical model structure for lignosulfonate according to Pukánszky et al.;¹⁹ sulfonate groups are present in the aliphatic side chains; (b) chemical structure of sodium anthraquinone-2-sulfonate (AQ2S).

Lignin's complex, heterogeneous, and chemically highly stable structure poses significant challenges for efficient and selective conversion into value-added compounds,⁵ such as monomers, dimers, and oligomers.^6–9 Traditional depolymerization methods, such as pyrolysis, acid/base catalyzed reactions, and chemical oxidation, often require high pressures and temperatures or involve the use of toxic reagents or organic solvents, which undermine both process sustainability and scalability.^10–12 In this context, one promising strategy for lignin conversion is the utilization of renewable solar energy for photochemical activation.

Recent advances in photocatalytic lignin depolymerization have led to the exploration of a wide range of heterogeneous and homogeneous photocatalysts. Among the heterogeneous catalysts, metal oxides, especially TiO₂ or Cu-based ones,^13–18 have been widely investigated for this purpose, both in their pure form and as composites doped with various other metals.

Additionally, metal sulfides,^20–22 metal–organic frameworks (MOFs),^23,24 or carbon-based catalysts, such as polymeric graphitic carbon nitride (g-C₃N₄),^25–27 have demonstrated photocatalytic activity toward lignin depolymerization. While heterogeneous systems offer the advantage of easy catalyst recovery and the possibility for reuse, structure–property relationships with respect to reaction mechanisms require sophisticated spectroscopic methods to provide an atomic-level understanding.^28,29 Another drawback is parasitic light absorption and poor contact between the insoluble catalyst and the lignin substrate.¹⁶

Homogeneous photocatalysts or photoinitiators³⁰ address some of these challenges by enabling molecular-level interactions between the catalyst or the reagent and dissolved lignin.³⁰ This improved contact enhances catalyst activity and enables possible electron transfer routes. However, many systems rely on expensive iridium-based catalysts or complex organic molecules requiring harmful organic solvents.^31–33 Anthraquinone-based catalysts, on the other hand, such as 2-bromoanthraquinone, have recently emerged as effective photocatalysts for lignin degradation under light irradiation.³⁴ Their favorable redox properties and ability to mediate hydrogen transfer and oxidative processes, as well as to produce reactive oxygen species (ROS) while being commercially available and cheap, make them favorable candidates. However, many of these systems typically require the use of organic solvents to ensure sufficient solubility of both the catalyst and the lignin substrate. Notably, 2-bromoanthraquinone has often been employed in conjunction with ethylene dibromide (EDB) as a solvent. While EDB can provide the necessary solubility and reaction environment, it is a highly toxic, persistent and environmentally hazardous compound, which significantly undermines the overall sustainability of such systems.^34,35 As a result, the development of solvent systems that maintain photocatalytic efficiency while minimizing ecological and health impacts, remains a critical challenge.

Commercially available anthraquinone-sulfonates, such as sodium anthraquinone-2-sulfonate (AQ2S, Fig. 1b), offer distinct advantages. These compounds are non-toxic, inexpensive, and, above all, highly water-soluble, eliminating the need for organic solvents.³⁶ Their compatibility with aqueous systems makes them particularly well-suited for the depolymerization of lignosulfonates, which are water-soluble, thus ensuring optimal contact and efficient photocatalytic activity. Although a known limitation of anthraquinone-sulfonates is their deactivation via photoreactions with water under irradiation, the presence of salts can significantly enhance their lifetime and reactivity.^37–39

In this study, we aim to demonstrate that AQ2S can retain its photocatalytic activity over extended periods of UV irradiation in saline media. Given that both sodium lignosulfonate and AQ2S are highly soluble in aqueous salt solution, we explore the use of seawater, the most abundant and underexploited solvent on Earth, as the reaction medium, which may significantly reduce the environmental footprint.⁴⁰ We propose a photocatalytic method for the depolymerization of sodium lignosulfonate under mild conditions, using UV LED irradiation at ambient pressure and room temperature. The system employs AQ2S, a commercially available, water-soluble, and non-toxic⁴¹ photocatalyst, functioning in saline media, without the need for additional acids or bases, organic solvents or hazardous additives (Scheme 1). Notably, it remains effective even in unfiltered and untreated seawater, further improving process sustainability by eliminating water purification steps. The effectiveness of this approach is evaluated using various analytical techniques, including UV-vis spectroscopy, NMR, SEC and FTIR. Overall, this study introduces a robust and environmentally friendly approach to lignin valorization that leverages a seawater-based reaction system.

Scheme 1 Schematic overview of the main findings presented in this manuscript. The scheme illustrates the UV-induced depolymerization of sodium lignosulfonate using anthraquinone-2-sulfonate (AQ2S) as catalyst in the presence of NaCl in aqueous solution. Reactions containing both NaCl and AQ2S—or seawater as a substitute for NaCl—show a significantly greater degree of depolymerization compared to controls lacking either component.

Results and discussion

To determine an appropriate light source for photocatalysis, a UV-vis absorption spectrum of AQ2S was recorded, revealing the expected peaks at 200 nm and 255 nm, corresponding to aromatic π → π* transitions, as well as the benzoic band at 286 nm and n → π* transitions at 333 nm (Fig. 2).⁴² Given these strong absorption features in the UV region, UV LED irradiation was deemed suitable.

Fig. 2 UV-vis absorption spectra of AQ2S in water (solid blue line) and its degradation product in pure water after 4 days of UV irradiation (dashed blue line). Characteristic electronic transitions of AQ2S are labelled at their corresponding absorption peaks.

Even though AQ2S is a known catalyst for various transformations in synthetic chemistry, its use in aqueous solutions is generally limited by its tendency to undergo hydroxylation in water, resulting in a loss of reactivity.⁴³ To assess the stability of AQ2S in deionized water under UV LED irradiation, a visible color change from colorless to a bright orange solution is observed even after a short period of exposure (Fig. S1b). This color change is accompanied by alterations in both the UV-vis absorption spectrum and the NMR spectrum, compared to that of the unaltered compound (Fig. 2 and Fig. S2). Since the proposed hydroxylation mechanism of AQ2S in water does not specify a regioselective site of hydroxylation,³⁶ the broad and nonspecific nature of the NMR spectrum (Fig. S2) is most likely due to the formation of multiple hydroxylated derivatives. Evidence for the formation of hydroxylated derivatives was further supported by FTIR spectra of pure AQ2S and of the orange product obtained after irradiation (Fig. S10). While the spectrum of pure AQ2S did not exhibit a broad absorption band characteristic of hydroxy groups, such a feature was clearly present in the irradiated sample. In contrast, the catalyst in aqueous NaCl solution, used to simulate seawater, does not show any considerable color change. Considering that the photocatalytic activity of anthraquinone sulfonates proceeds via the formation of hydrogen peroxide in oxygenated aqueous media,^44,45 H₂O₂ production was employed as a proxy to evaluate the catalyst's activity. This was also considered to confirm its suitability for the degradation of LS, since H₂O₂ in alkaline solution has long been reported in literature to depolymerize lignin under irradiation.⁴⁶ Testing the irradiated solutions with commercial, semiquantitative peroxide test stripes revealed that the bright orange AQ2S solution in deionized water showed no detectable amounts of H₂O₂ after four days, whereas the solution containing NaCl consistently produced measurable amounts even after the same irradiation period (Fig. S1a). This was interpreted as evidence that AQ2S retains its photocatalytic activity for longer in simulated seawater than in pure water.

Based on these findings, photochemical depolymerization reactions were carried out using LS (2 g L⁻¹) and AQ2S (5 wt%) under UV LED irradiation in 0.5 M aqueous NaCl solution. Control reactions were also performed: (i) AQ2S and LS in deionized water, (ii) LS in NaCl solution without AQ2S and (iii) AQ2S and LS in NaCl solution without irradiation. Technical lignins such as lignosulfonates contain a variety of chromophores, providing them with a naturally dark color that brightens as depolymerization progresses.⁴⁷ Using the color change as first indicator of depolymerization, a clear trend was observed over a 10-day irradiation period (Fig. 3). Initially appearing as a dark brown solution, the sample containing AQ2S and LS in 0.5 M aqueous NaCl solution got progressively lighter over time, eventually turning completely colorless and transparent. This clearly indicated substantial breakdown of chromophoric structures within LS, suggesting a shortening of macromolecular chains. In contrast, both control samples either lacking AQ2S or using deionized water instead of NaCl solution, showed minimal to no visible color change, apart from a change to a more orange solution in the presence of AQ2S without salt, indicating the degradation of the catalyst to the deactivated hydroxylated version. That both solutions did not reach a colorless state, indicates that neither the NaCl solution alone, nor AQ2S in the absence of salt is sufficient to induce significant depolymerization under the applied conditions. The necessity of UV light irradiation was also confirmed, as the AQ2S-LS sample in salt water showed no observable color change in the dark. These findings align with earlier observations, suggesting that AQ2S is less stable in deionized water, possibly limiting its catalytic lifetime and hindering effective depolymerization.

Fig. 3 Photographs of beakers containing reaction mixtures under UV LED irradiation over 10 days. The top row shows LS in 0.5 M aqueous NaCl (without AQ2S), the middle row shows AQ2S and LS in deionized water, and the bottom row shows AQ2S and LS in 0.5 M aqueous NaCl. Images were taken after 0, 3, 5, 7, and 10 days of irradiation; the corresponding time points are indicated above each column.

To gain a more precise understanding of the depolymerization process, UV-vis absorption spectroscopy was employed. This technique monitors the decline of LS's characteristic absorption features over time, thereby reflecting the loss of chromophoric structures. While typically used to quantify lignin concentration in solution,⁴⁸ UV-vis can also be used to monitor lignin degradation.⁴⁹ In the absence of AQ2S, the spectrum of LS irradiated in 0.5 M NaCl (Fig. 4c) showed only a small decrease in absorption between 300–400 nm after 3 days, with no further change thereafter. The characteristic peak at 275 nm, remained largely unaffected, suggesting negligible structural changes and confirming that depolymerization does not occur without the catalyst. The reference reaction without irradiation exhibited no significant changes in absorption spectra after 10 days in the dark (Fig. S5). As expected, based on the observed color change, the spectra for LS and AQ2S in deionized water (Fig. 4b) showed diminished catalyst stability with time of irradiation. After 3 days, the AQ2S peak at 255 nm had already significantly declined and was no longer clearly visible. Concurrently, the decline of the prominent LS peak at 275 nm slowed markedly: starting from an absorption of 0.23 at day 0, it decreased to 0.18 by day 3, then gradually only to 0.14 by day 10. This points to the partial retention of chromophoric groups and only partial depolymerization. In contrast, combining both AQ2S and NaCl produced a much more pronounced degradation of LS (Fig. 4a). The catalyst peak at 255 nm remained clearly visible for at least 7 days, and was still detectable after 10 days, suggesting sustained catalyst stability. Meanwhile, the 275 nm peak, corresponding to lignin chromophores, underwent a significant decline from 0.25 at day 0, to 0.22 (day 3), 0.20 (day 5), then sharply to 0.14 (day 7) and ultimately to just 0.02 by day 10. This rapid decline after day 5 marks the apparent onset of accelerated depolymerization. The combination of low absorption and the colorless appearance of the final solution strongly suggests an extensive breakdown of chromophoric and aromatic structures. When normalizing the peak at 275 nm at 0 days to 1 and plotting the relative changes over time (Fig. 4d), the differences between conditions become even clearer. While LS in salt water alone shows negligible decline, and only slightly more evident decrease for LS and AQ2S in deionized water, the reaction containing LS and AQ2S in salt water seems to have an onset after 5 days with a subsequent stark decline, confirming effective depolymerization.

Fig. 4 (a) UV–vis absorption spectra of sodium lignosulfonate (LS) and AQ2S in 0.5 M aqueous NaCl solution over 0–10 days of UV LED irradiation. (b) UV–vis absorption spectra of LS and AQ2S in deionized water over the same irradiation period. (c) UV–vis absorption spectra of LS in 0.5 M aqueous NaCl solution without AQ2S under identical conditions. (d) Changes in absorbance at 275 nm over time, with maximum absorbance normalized to 1. Data are shown for LS in 0.5 M NaCl (orange triangles), LS and AQ2S in deionized water (green circles), and LS and AQ2S in 0.5 M NaCl (pink squares). Dashed lines are included for visual guidance.

Even though the UV-vis spectra for AQ2S and LS in simulated seawater point to a strong decrease in aromatic content, additional analytical techniques confirm that significant portions of the lignin structure remain intact. FTIR spectra recorded from the salt residues obtained after removing the solvent from the crude mixture, reveal the retention of key functional groups throughout the whole irradiation period (Fig. 5). A broad absorption band at around 3261 cm⁻¹, attributed to O–H stretching vibrations,⁵⁰ is consistently observed in all spectra, indicating the persistence of hydroxyl functionalities. Similarly, distinct peaks near 2930 cm⁻¹, generally assigned to C–H stretching in methylene (CH₂) or methoxy (OCH₃) groups in lignin, remain evident across all measured samples.^50,51 A prominent absorption at around 1589 cm⁻¹, corresponding to carbonyl groups in sodium lignosulfonate¹ as well as aromatic skeletal vibrations,⁵² gradually decreases in intensity with prolonged UV LED exposure. However, this band is still visible even after 10 days of continuous irradiation, suggesting partial retention of these structural motifs. Notably, the S=O symmetric stretching vibration at 1048 cm⁻¹, characteristic of sulfonate groups in lignosulfonates,⁵³ also persists throughout the irradiation period. In contrast, the intensities of peaks at 1507 cm⁻¹ and 1457 cm⁻¹, attributed to aromatic skeletal vibrations⁵² and aromatic ring stretching,⁵⁰ respectively, decrease markedly over time. This trend indicates a progressive reduction in aromatic content, consistent with depolymerization and structural modification of the lignosulfonate backbone.

Fig. 5 FTIR spectra recorded for solids extracted from LS and AQ2S irradiated for 10 days in 0.5 M aqueous NaCl solution with UV LED light. Solids were obtained by removing the solvent under reduced pressure after the respective time of irradiation. Shown are spectra for 3, 5, 7 and 10 days of irradiation compared to 0 days.

Complementary information can be observed via NMR spectroscopy. Crude ¹H NMR spectra show that many signals present in the untreated mixture remain visible in the irradiated samples, particularly in the aliphatic region (Fig. S4). Shifts in chemical shifts and relative intensities are likely due to similar functional groups but in modified chemical environments, resulting from chain scission, decreasing macromolecular size during depolymerization.

To further assess the decrease in chain length of LS, the progression of depolymerization was monitored via size exclusion chromatography (SEC, Table 1), revealing clear temporal trends in molar mass distribution and dispersity. A key strength of this study lies in a rigorous baseline definition, with both untreated (LS) and 0-day control (AQ2S + LS – 0 day) samples confirming the structural stability of LS in the reaction medium absent of light. Their very similar molar mass metrics, across peak molar mass (M_p), number-average (M_n), weight-average (M_w), and dispersity (Đ), establish that neither dissolution nor ionic strength from NaCl induces structural perturbation, affirming that subsequent molecular changes stem solely from photocatalytic activation.

Table 1 Results of size exclusion chromatography (SEC) for lignosulfonate (LS) and depolymerized lignosulfonate samples. Depolymerization was achieved by irradiating aqueous LS solutions with UV light for varying durations (0 to 10 days) in the presence of AQ2S as a photocatalyst

Sample	M p/Da	M n/Da	M w/Da	Đ
LS	1738	451	5473	12
AQ2S + LS – 0 day	1846	411	4886	12
AQ2S + LS – 3 days	1928	386	4851	13
AQ2S + LS – 5 days	1928	378	4114	11
AQ2S + LS – 7days	339	281	1711	6
AQ2S + LS – 10 days	349	193	488	3

---

Early irradiation (up to day 5) resulted in only modest shifts in molar mass, pointing out that initial interaction with reactive species or photocatalyst does not immediately translate into extensive bond cleavage, which suits the observations made via UV-vis measurements. This induction period may reflect a kinetic lag phase, possibly due to the need for progressive photoactivation or accumulation of reactive intermediates. During this phase, the subtle decrease in M_n and M_w may correspond to peripheral side-chain scission or oxidation events insufficient to induce backbone fragmentation.

A pivotal transition is observed by day 7, at which point SEC data exhibit a pronounced drop in all molar mass parameters, alongside a sharp decline in Đ. Such a shift strongly indicates that depolymerization reaches a critical threshold where the photocatalytic process becomes markedly more efficient. The emergence of a more narrowly distributed product pool suggests that selective cleavage mechanisms predominate at this stage, favoring reproducible breakdown into low-molecular-weight fragments. By day 10, the depolymerization is fully manifested, with M_w reduced by nearly an order of magnitude and dispersity values approaching 3, which is an indicative of a considerably more uniform and lower-mass oligomeric product. The sustained decrease in M_p and M_n over time further underscores the progressive, non-random nature of bond scission, likely driven by photocatalyst-mediated activation of specific interunit linkages within the lignin macromolecule. Taken together, these results delineate a two-phase kinetic profile: an initial latency period marked by limited reactivity, followed by an accelerated depolymerization phase characterized by enhanced efficiency and product uniformity. Such behavior aligns with emerging understandings of lignin photocatalysis, where structural complexity and radical scavenging capacity can initially impede reactivity until key thresholds of activation are crossed.

Following a liquid–liquid extraction of the depolymerized solution using ethyl acetate, the resulting solids were dissolved in deuterated DMSO for NMR analysis, revealing clear signals of AQ2S in the spectrum (Fig. S3). Compared to measurements of the raw material, no spectral shifts or additional signals are observed in the relevant region after 10 days of irradiation, further confirming the catalyst's stability under the applied conditions. Applying ESI-HRMS analysis, the catalyst stability could be further proven as the respective signal could be determined before and after the full irradiation period (Fig. S7) as expected from the NMR spectra. Additionally, the recyclability of the catalyst was assessed. Upon achieving a colorless solution from the standard reaction mixture (100 mg LS, 5 wt% AQ2S, 0.5 M aqueous NaCl) after 10 days of irradiation, a second batch of 100 mg LS was added without supplementing additional AQ2S (Fig. S6). Remarkably, after 10 additional days of irradiation, a colorless solution was obtained again, demonstrating that AQ2S retained its photocatalytic activity even after 20 full days of irradiation in aqueous solution. When a third batch of 100 mg was added, a color change to a light yellow was observed, although complete discoloration was not achieved. Nevertheless, this highlights the scalability and robustness of the system, as the reaction can accommodate increased LS loading and prolonged irradiation periods without increasing the AQ2S concentration. A liquid–liquid extraction with ethyl acetate after 30 days of irradiation confirmed the stability of the catalyst, as signals attributable to AQ2S were still detected (Fig. S12).

Having established the effectiveness of 0.5 M aqueous NaCl solution as reaction a medium, the next step was to explore the use of natural seawater as a truly sustainable and readily available reaction medium. Due to the slightly lower solubility of LS in unfiltered, untreated seawater, a reduced concentration of 1.2 g L⁻¹ of LS was used with 5 wt% AQ2S. Since earlier SEC analysis confirmed that a combination of a colorless solution and decreased UV-vis absorption correlates with increased depolymerization, both visual observation and UV-vis spectroscopy were used to monitor the reaction (Fig. 6a and b). Notably, the depolymerization reaction in natural seawater followed a trend nearly identical to that observed in 0.5 M NaCl solution. During the initial phase (0–3 days), only minor changes in color and absorption at 275 nm were observed. Between days 3 and 5, the reaction rate appeared to increase, and by day 7, the solution was nearly colorless and transparent with only residual absorption. Importantly, the AQ2S-associated peak at 255 nm remained clearly visible up to day 7, indicating that the catalyst remains stable even in unfiltered, untreated seawater. The ability of the system to operate with seawater instead of purified fresh water offers a significant advantage compared to conventional methods. Seawater accounts for about 97% of the Earth's water, while of the remaining 3% that is fresh water, approximately 2.5% is unavailable for human use, being locked up in glaciers, ice caps, or the atmosphere.⁵⁴ Given the abundance of seawater and scarcity of fresh water, along with the additional energy demands associated with distillation to supply large volumes of purified water for industrial processes, the use of seawater makes the process significantly greener and enhances its potential for up-scale.⁵⁵

Fig. 6 (a) Photographs of solutions containing sodium lignosulfonate (LS) and AQ2S in seawater, taken after 0, 3, 5, 7, and 10 days of UV LED irradiation. The gradual loss of color over time serves as an indirect indicator of depolymerization. (b) UV–vis absorption spectra of the same solutions, showing a decrease in absorbance over time. A characteristic peak associated with AQ2S is marked with a dashed grey line.

Computational study

The computational study aimed to evaluate the catalytic performance of AQ2S and its deactivation upon hydroxylation. The irradiation frequency used in this work (see Experimental section) is consistent with both the homolytic bond breaking of some lignin model sites (with the formation of H˙)⁵⁷ and the activation of AQ2S from its singlet ground state to the lowest excited triplet state.⁴² Generally, anthraquinone sulfonate derivatives act as radical traps, leading to the release of H˙ and the formation of H₂O₂ after reaction with O₂, followed by the recovery of the active form of the catalyst.^45,57 Moreover, the formation of hydroxylated derivatives of AQ2S was found to impair the catalyst's activity; however, this limitation can be mitigated in the presence of NaCl.

In this context, for a rational approach to the complexity of the system under investigation, we split the system in a series of three sets of molecular models: (i) the catalytic activity of AQ2S, which should act as a radical trap, a step propaedeutic to the formation of H₂O₂; (ii) the formation of the hydroxylated forms to explain the loss of the AQ2S's catalytic activity; (iii) the role of NaCl in hampering the hydroxylation of AQ2S, hence allowing the lignin depolymerization process.

Catalytic behavior of AQ2S. The first step focused on elucidating the catalytic role of AQ2S. We previously stated the necessity of using UV light irradiation for the reaction to occur, even in the presence of AQ2S. The catalyst was investigated in its neutral from, hence also considering the Na⁺ counterion, as reported by Bedini et al.⁴² As stated earlier, the irradiation at the working frequency used in this study could generate both H˙ radicals from lignin and the excited triplet state of AQ2S (³AQ2S); thus, we considered the occurrence of both species in the system.

In Fig. 7, we evaluated the activity of the AQ2S in its ground (¹AQ2S) and triplet (³AQ2S) states, respectively. The mechanisms in Fig. 7 show that both ¹AQ2S and ³AQ2S are efficient in trapping H˙, with path a* being more favorable than path a, due to its radical, barrierless hydrogenation process. Mechanism a* is in very good agreement with the proposed mechanism for the similar 9,10-anthraquinone-2,6-disulphonate disodium salt.^45,57 It is important to note that path a* implies an acid–base step, with the deprotonation of ²INT2, which should lead to a steadily decreasing pH, even if partially buffered by the hydrolysis reaction of 1HPX⁻. This step could account for the lowering of the pH observed during the reaction. Indeed, the measured starting pH ≈ 10 at the beginning of the reaction, in the AQ2S-LS-NaCl system, decreased to around 7 after 10 days of irradiation. A small drawback in path a* could be due to the slight endergonic process (ΔG° = +7.47 kcal mol⁻¹) involving the conversion of ²AQ2S⁻ to ¹AQ2Svia a charge transfer process involving ³O₂ to form O₂˙⁻; however, this value represents a non-prohibitive Gibbs free energy barrier.

Fig. 7 Likely mechanism paths of ¹AQ2S and ³AQ2S. Gibbs free energies of activation (red values) and formation (black values) are given in kcal mol⁻¹. Dashed bonds in the transition state (TS) indicate forming (blue) and breaking (red) bonds. The spin state for each model is given as a superscript on the left of the label. INT = intermediate; TS = transition state; ¹HPX⁻ = hydrogen peroxide anion. Gray box on the top left indicates the measured starting and ending pH of the reaction up to 10 d and the pK_a of H₂O₂ in water solution, respectively. The Energy Transfer (ET) has been calculated according to Brédas et al. and Tacchi et al.^45,56

According to the energetics in Fig. 7, the relaxation of ³AQ2S to ¹AQ2S through path b appears less favorable than path a*, due to the high Gibbs free energy activation, for which the relaxation of ¹O₂ can not occur through the formation of H₂O₂via radical hydrogenation at the mild conditions of the reaction. Thus, the relaxation to ¹AQ2S preferably occurs through the ²AQ2S⁻ intermediate.

Finally, the second hydrogenation of AQ2S (path c) is outcompeted by the deprotonation process in path a*. The molecular model mechanisms reported here support the effectiveness of AQ2S as a catalyst.

AQ2S hydroxylation. It is well known that the triplet form of anthraquinone and its derivatives promotes the formation of their hydroxylated forms in aqueous solution, and that experimental conditions greatly affect the reaction yield, with mono- and/or di-hydroxylated species potentially forming depending on factors such as anthraquinone, auxiliary educts, oxygen concentrations, and pH.^42,58–61 The leading hypotheses, inferred from experimental evidence, suggest a role for ³AQ2S in promoting both the formation of HO˙ and the final ¹AQ2OH hydroxylated anthraquinone species, along with the production of HOO˙/O₂˙⁻ species that ultimately lead to the formation of H₂O₂. On these grounds, the AQ2S hydroxylation is experimentally assessed mainly by a radical mechanism involving O₂.⁴² In this work, we wanted to focus on the potential role of the OH⁻ ion in the hydroxylation of ³AQ2S, by considering the starting pH ≈ 10, as well as the occurrence of H˙ produced by the photoactivation of LS. Moreover, for the sake of comparison, we also considered the likely hydroxylation process involving ¹AQ2S. Results are reported in Fig. 8. From a general perspective, both ¹AQ2S and ³AQ2S readily form the relevant hydroxylated species. ¹AQ2SOH (i.e., 2-hydroxy-8,10-anthraquinone) is known to be an orange-brown powder. Fig. 8 shows that this species undergoes deprotonation and radical hydrogenation, forming final dianion products that are inactive as radical traps.

Fig. 8 Hydroxylation mechanism for ¹AQ2S and (pink path). Gibbs free energies of activation (red values) and formation (black values) are given in kcal mol⁻¹. Dashed bonds in the transition state (TS) indicate forming (blue) and breaking (red) bonds, respectively. The spin state for each model is given as a superscript on the left of the label. INT = intermediate; TS = transition state; P_n = (n = 1, 2, 3) final Product.

For ³AQ2S, both the addition of H˙ and OH⁻ (Fig. 7 and 8) proceed without energy barriers, making the two processes competitive. In fact, for the LS-AQ2S system (without NaCl), the formation of an orange-colored solution supports the formation of hydroxylated species. In this system, the starting pH ≈ 10 decreased to 8.7 after 10 days of irradiation. These findings indicate a limited consumption of OH⁻, likely involved both in the catalytic process (Fig. 7) and in hydroxylation reactions (Fig. 8), leading to the rapid formation of the deactivated catalyst. As previously noted, the reaction depicted in Fig. 7 concludes at a final pH of approximately 7. Given that the deprotonation of ²INT2 is a key step, this pH likely represents a limiting threshold beyond which further reaction progression via the deprotonation step is inhibited (Fig. 7).

Finally, for both ¹AQ2S and ³AQ2S, alternative hydrogenation pathways, e.g., those leading to 1-hydroxy-9,10-anthraquinone sulfonate, should be considered. Indeed, experimental evidences consistently indicates the formation of a mixture of hydroxylated forms.^58–61 This implies that the several mechanism proposed are characterized by very close energetic paths. Therefore, we focused on the formation of ¹AQ2OH. This compound is a very likely product,^59,60 and represents one of the off-cycle paths that inhibit the catalyst's activity.

Effect of NaCl on the reaction. The final step of the investigation aimed to elucidate the role of the salt in preserving catalyst activity, presumably by hindering the hydroxylation process. Accordingly, we examined the influence of salt concentration on the inhibition of OH⁻ approaching the substrate, primarily through purely electrostatic interactions.

To investigate this effect, we developed simplified models in which only one Na⁺ or Cl⁻ ion, and one, two, or three NaCl units were placed near the approaching OH⁻. These models revealed that, starting from two NaCl units, OH⁻ is effectively prevented from approaching the catalyst. This inhibitory effect becomes more pronounced with increasing numbers of NaCl units, likely due to electrostatic interactions between Na⁺ and OH⁻. The observed decline in catalyst activity, despite the presence of Na⁺ from LS, further supports the importance of salt concentration. At the experimental LS concentration of approximately 3.7 × 10⁻³ M (see Experimental section), the resulting Na⁺ concentration is expected to be relatively low. In contrast, the addition of NaCl elevates the solution concentration to about 0.5 M (see Experimental section), greatly exceeding the sodium ion content introduced by LS alone.

Fig. 9 presents selected final structures obtained from both transition state and geometry optimization calculations based on ¹TS5 in Fig. 7. In these structures, OH⁻ remains distant from the reactive site in the presence of the NaCl, illustrating the influence of the salt on the system.

Fig. 9 Effect of NaCl on the approaching OH⁻ for ¹TS5 and ¹INT3. C–OH distances for ¹TS5, ¹TS5;2NaCl, and ¹TS5;3NaCl are 2.25, 3.76, and 3.85 Å, respectively. Same behaviour holds both for the triplet states of AQ2S. white = H; grey = C; red = O; teal = Na; green = Cl; pale brown = S. The spin state for each model is given as a superscript on the left of the label.

An additional consideration involves the potential scavenging effect of Cl⁻ ions towards OH˙ radicals, which may be generated through the action of ³AQ2S,^43,58,59 and can lead to the hydroxylation of AQ2S via a radical mechanism. According to Atinault et al.,⁶² the rate constant for the formation of ClOH˙⁻ is k = 6.1 × 10⁹ (L mol⁻¹ s⁻¹) in neutral, concentrated NaCl aqueous solutions. Although this value pertains to specific conditions, it highlights a very rapid reaction, supporting the idea that Cl⁻ could effectively scavenge OH˙ radicals and inhibit radical-mediated hydroxylation of ³AQ2S, particularly under the high NaCl concentrations used in this study.

Experimental

Materials

Sodium lignosulfonate (Grade 2, GAC-NaLS-2) was obtained from GREEN AGROCHEM-LIGNIN. Anthraquinone-2-sulfonic acid sodium salt monohydrate (97%) was purchased from Sigma-Aldrich; sodium chloride (99.5%) and anhydrous sodium sulfate (99%) were purchased from Grüssing GmbH; ethyl acetate (≥99.5%) was purchased from Fisher Scientific. All reagents were used as received without further purification. Quantofix® Peroxide 25 test stripes were purchased from Macherey-Nagel. As light source, an UV LED lamp LEDtube 060-T8-10-UV-365 (10 W, 60 cm) by METOLIGHT was used. Seawater was collected from the coast of the Baltic Sea near Maasholm Bad, Germany at 54 70009° N, 10 00 106° E on November 2, 2024 and used without filtration or further purification.

Depolymerization reaction

For the photochemical depolymerization experiments, sodium lignosulfonate (100 mg) and AQ2S (5 mg) were weighed and transferred to a 50 mL glass beaker equipped with a magnetic stirring bar. Deionized water (50 mL) was added and the mixture was stirred at room temperature until complete dissolution. Subsequently, sodium chloride (1.46 g) was added and stirring continued until complete solubilization. The beaker was then closed with plastic foil to minimize evaporation of the solvent, placed on a stirring plate, and irradiated from above with UV LED light with a distance of 3 cm between solution and LEDs for up to 10 days. As reference reaction without irradiation, the beakers were covered in aluminum foil additionally to being closed with plastic foil, and stored in a dark place for 10 days. For reactions in seawater, LS (60 mg) and AQ2S (3 mg) were dissolved in 50 mL unfiltered and untreated seawater. The beakers were then again closed with plastic foil and placed under a lamp.

For FTIR and selected NMR measurements, the reaction was stopped after the respective number of days, the solvent was removed under reduced pressure and the resulting solid samples were dried in an oven at 60 °C overnight. For MS measurements and additional NMR measurements, a liquid–liquid extraction was performed: the pH of the reaction mixture was adjusted to 4–5 using 1 M HCl, followed by the addition of NaCl (10 g) and stirring until complete dissolution. The solution was then extracted three times with 50 mL portions of ethyl acetate. The combined organic phases were dried over anhydrous sodium sulfate, filtered, and the solvent was removed under reduced pressure. The resulting solids were dried in an oven at 60 °C overnight.

0 d: FTIR (Diamond-ATR): ν [cm⁻¹] = 3261 (O–H, w), 2933 (C–H, w), 2837 (C–H, w), 1587 (COO⁻, m), 1507 (C=C, m), 1457 (m), 1412 (m), 1358 (m), 1332 (m), 1292 (s), 1220 (s), 1119 (m), 1086 (s), 1043 (S=O, m), 930 (s), 878 (s), 834 (s), 769 (s), 707 (s), 658 (s), 622 (s).

3 d: FTIR (Diamond-ATR): ν [cm⁻¹] = 3261 (O–H, w), 2930 (C–H, w), 1585 (COO⁻, m), 1508 (C=C, s), 1408 (m), 1357 (s), 1333 (s), 1296 (s), 1240 (s), 1214 (s), 1171 (s), 1123 (m), 1088 (m), 1048 (S=O, m), 930 (s), 834 (s), 768 (s), 707 (s), 659 (s), 621 (s).

5 d: FTIR (Diamond-ATR): ν [cm⁻¹] = 3261 (O–H, w), 2930 (C–H, w), 1585 (COO⁻, m), 1507 (C=C, s), 1407 (m), 1358 (s), 1333 (s), 1296 (s), 1240 (s), 1222 (s), 1171 (s), 1121 (m), 1088 (m), 1047 (S=O, m), 930 (s), 877 (s), 833 (s), 769 (s), 706 (s), 688 (s), 627 (s).

7 d: FTIR (Diamond-ATR): ν [cm⁻¹] = 3271 (O–H, w), 2944 (C–H, w), 1589 (COO⁻, m), 1406 (m), 1357 (s), 1336 (m), 1297 (m), 1214 (s), 1122 (s), 1088 (s), 1048 (S=O, m), 930 (s), 832 (s), 767 (s), 656 (s), 626 (s).

10 d: FTIR (Diamond-ATR): ν [cm⁻¹] = 3270 (O–H, w), 2940 (C–H, w), 1589 (COO⁻, m), 1397 (m), 1335 (s), 1088 (m), 1048 (S=O, m), 833 (m), 780 (s), 669 (s), 625 (s).

Characterization

Nuclear magnetic resonance (NMR). NMR measurements were performed on a BRUKER Avance 400 MHz spectrometer. A 5 mm broadband inverse probe with automatic frequency determination, 5 mm QNP probe, and 5 mm broadband inverse probe were used as probe heads. 16 scans with a delay time of 1 s were accumulated. Chemical shifts are reported relative to tetramethylsilane (TMS). Dimethylsulfoxide-d₆ (Deutero GmbH, 99.8%) and Deuterium oxide (Deutero GmbH, 99.9%) were used as deuterated solvents.

UV-Vis spectroscopy. UV-vis absorption spectra were recorded using a UV-3100PC spectrophotometer (VWR, Radnor, PA, USA) controlled by the program UV-Vis analyst provided by the manufacturer and equipped with quartz cuvettes with a path length of 1 cm. Samples were prepared by dissolving a respective amount of 30–60 μL of sample in 3 mL deionized water. A blank measurement was performed using the pure solvent to correct for background absorbance prior to each measurement. All measurements in a series were performed with the same amount of sample. The amounts were only changed depending on the reaction mixture and respective absorption and then kept the same over the course of the reaction, to get quantitative comparisons over time. The absorbance spectra were recorded over a wavelength range of 200–800 nm at room temperature (25 °C).

Fourier-transform infrared (FTIR). FTIR spectra were recorded using a Nicolet iS5 spectrometer (Thermo Scientific, Waltham, MA, USA), equipped with an iD5 diamond attenuated total reflection (ATR) unit. Spectra were recorded in a range of 4000–400 cm⁻¹ doing 128 scans per spectrum.

Molecular weight distribution. Molecular weight determination of sodium lignosulfonate and treated samples was performed by SEC (Thermo Scientific, Dionex ICS 5000+) using a PSS MCX analytical 100 A + 1000 A + 100 000 A column (8 mm × 300 mm, Thermo Fischer). The analysis was conducted at a temperature of 30 °C, 0.1 mol L⁻¹ NaOH was used as eluent. A flow rate of 0.5 mL min⁻¹ was maintained and a wavelength detector set to a wavelength of 280 nm was used for detection. The system was calibrated using standards from PSS (Polymer Standard Service) with a molecular range of 891–976 000 g mol⁻¹ and vanillin.

Direct infusion (DI) ESI-HRMS. Depolymerized and lignosulfonate samples, extracted via liquid–liquid extraction, were dissolved in methanol, ultrasonicated for 30 min and centrifuged for 10 min at 14 000 rpm. High-resolution mass spectrometry (HRMS) was used as advanced analytical method to gain structural information on the degradation products of sodium lignosulfonate and on the catalyst AQ2S. MS and MSn spectra were obtained using an Orbitrap-IQX high-resolution mass spectrometer (ThermoFisher Scientific, Bremen, Germany), equipped with an ESI source. ESI-MS analyses were carried out in ESI(−) and ESI(+)mode. The solutions were infused into the ESI source via direct infusion (DI) at a rate of 5 μL min⁻¹. Typical spray and ion optics for negative mode conditions were the following: source voltage, 3.0 kV; sheath gas flow rate, 8 arb; capillary temperature, 275 °C; capillary voltage, −50 V; tube lens voltage, −130 V. Fragmentation and interpretation was done based on negative ionization mode, positive ionization mode was used for additional confirmation. Xcalibur version 2.0.7 and Mass Frontier version 8.0 (ThermoFisher Scientific, Bremen, Germany) software were used for data processing and evaluation.

Computational methods. Calculations were performed using Gaussian 09⁶³ within the unrestricted framework, employing the M06-2X⁶⁴ functional, which is designed to accurately predict thermodynamic properties of main group compounds. The 6-31G(d, p) basis set was used for all atoms. Solvent effects were taken into account using the Solvation Model based on Density (SMD)⁶⁵ by considering water as the solvent.

The Hessian matrix was analytically calculated for the optimized structures to confirm the location of correct minima (no imaginary frequencies) or saddle points (only one imaginary frequency), and to estimate the thermodynamics at 298.15 K. Transition states were calculated using a Berny geometry optimization, with force constants calculated analytically for the initial points. The nature of all transition states was investigated by analysis of the vectors associated with the imaginary frequency and confirmed by the Intrinsic Reaction Coordinate (IRC) method.^66–68

Conclusions

We have demonstrated a green photochemical strategy for the depolymerization of sodium lignosulfonate (LS) in both saline media and natural seawater. AQ2S was successfully employed as a photocatalyst for depolymerization in aqueous environments, despite its known instability under prolonged irradiation in such conditions. The presence of salt significantly improved catalyst stability, even allowing for its reuse in consecutive LS batch treatments. DFT investigations indicate that AQ2S deactivation at basic pH primarily results from rapid, competing hydroxylation reactions, leading to the formation of inactive dianionic species. Moreover, NaCl appears to suppress this deactivation by electrostatically hindering OH⁻ access, in a concentration-dependent manner. Additionally, Cl⁻ and H˙ radicals may act as effective scavengers that mitigate radical hydroxylation of ³AQ2Svia radical mechanism. Using the stabilized catalyst, LS depolymerization was monitored via various analytical techniques, including UV-vis spectroscopy, FTIR, NMR, and SEC. Notably, SEC analysis indicated that the final product consists predominantly of low-molecular-weight, narrowly dispersed oligomers, which are well-suited for downstream valorization. These oligomers are more chemically accessible and functionally versatile than native lignin, enhancing their potential in applications such as bio-based adhesives and fine chemicals. Additionally, the lower-molecular-weight fragments obtained may serve as suitable building blocks for new polymeric materials, as such can be obtained from repolymerizing feedstocks produced from depolymerized lignins.^69,70 Moreover, depolymerized lignin feedstocks also hold significant potential as a basis for sustainable fuels following hydrodeoxygenation.^71,72 Further advancements and shorter times could be expected with higher illumination power and flow reactor in order to suppress the parasitic light absorption.

Overall, these findings highlight the potential of light-driven depolymerization as a scalable and environment-neutral method for producing customized lignin-derived building blocks, advancing the development of greener biorefinery processes.

Conflicts of interest

Author Marcella Frauscher is employed by AC2T Research GmbH. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. The authors declare the study received fundings from Austrian COMET program (Project InTribology2, No. 906860). The funder was not involved in the study design, collection, analysis, interpretation of data, the writing of this article or the decision to submit it.

Data availability

The data supporting this article have been included as part of the SI. Supplementary information is available. See DOI: https://doi.org/10.1039/d5gc03734d.

Acknowledgements

We would like to thank the University of Wuppertal for the research support. The authors would also like to acknowledge A. Ristic for the HRMS measurements. We gratefully acknowledge Gerhard Petersen for collecting the seawater used in this study by personally walking into the Baltic Sea and thereby sparing the first author from getting cold feet.

References

1. E. S. Wibowo and B.-D. Park, Molecules, 2023, 28, 2755.
2. N. L. Radhika, S. Sachdeva and M. Kumar, Fuel, 2022, 312, 122935.
3. S. Gonçalves, J. Ferra, N. Paiva, J. Martins, L. H. Carvalho and F. D. Magalhães, Polymers, 2021, 13, 4196.
4. T. Aro and P. Fatehi, ChemSusChem, 2017, 10, 1861–1877.
5. A. Slabon and B. V. M. Rodrigues, Green Chem., 2025, 27, 2178–2183.
6. S. Gillet, M. Aguedo, L. Petitjean, A. R. C. Morais, A. M. da Costa Lopes, R. M. Łukasik and P. T. Anastas, Green Chem., 2017, 19, 4200–4233.
7. E. Subbotina and J. S. M. Samec, Nat. Chem. Eng., 2024, 1, 28–30.
8. M. Zirbes, T. Graßl, R. Neuber and S. R. Waldvogel, Angew. Chem., Int. Ed., 2023, 62, e202219217.
9. J. Haupenthal, A. K. H. Hirsch and K. Barta, RSC Sustainability, 2025, 3, 4039–4048.
10. R. Roy, M. S. Rahman, T. A. Amit and B. Jadhav, Biomass, 2022, 2, 130–154.
11. H. Ben and A. J. Ragauskas, Energy Fuels, 2011, 25, 4662–4668.
12. D. J. Gardner, T. P. Schultz and G. D. McGinnis, J. Wood Chem. Technol., 1985, 5, 85–110.
13. N. Luo, M. Wang, H. Li, J. Zhang, H. Liu and F. Wang, ACS Catal., 2016, 6, 7716–7721.
14. L. Song, X. Zhao, L. Cao, J.-W. Moon, B. Gu and W. Wang, Nanoscale, 2015, 7, 16695–16703.
15. K. Imamura, H. Kato, Y. Wada, K. Makabe, A. Onda, A. Tanaka, H. Kominami, K. Sato and K. Nagaoka, Chem. Commun., 2018, 54, 7298–7301.
16. S. Shao, X. Wang, W. Li, Y. Zhang, S. Liu, W. Xiao, Z. Yue, X. Lu and X. Fan, Catal. Sci. Technol., 2025, 15, 962–987.
17. N. Mittal, M. Kaur, V. Singh, A. Babbar, T. Dua and V. Singh, J. Photochem. Photobiol., A, 2025, 467, 116440.
18. D. Wang, H. Yang, J. Yang, B. Wang, P. Wasnik, B. B. Xu and Z. Shi, Adv. Compos. Hybrid Mater., 2023, 6, 138.
19. D. Kun and B. Pukánszky, Eur. Polym. J., 2017, 93, 618–641.
20. H. Yoo, M.-W. Lee, S. Lee, J. Lee, S. Cho, H. Lee, H. G. Cha and H. S. Kim, ACS Catal., 2020, 10, 8465–8475.
21. J. Xu, M. Li, J. Qiu, X.-F. Zhang and J. Yao, Int. J. Biol. Macromol., 2021, 185, 297–305.
22. G. Han, T. Yan, W. Zhang, Y. C. Zhang, D. Y. Lee, Z. Cao and Y. Sun, ACS Catal., 2019, 9, 11341–11349.
23. Q. Li, X. Luo, F. Liu, J. Cai, W. Han, X. Cao, X. Zhang and Y. Lan, Int. J. Biol. Macromol., 2024, 279, 135492.
24. Y. Niu, C. Guo, X. Cao, J. Li, S. Yang and J. Wang, J. Colloid Interface Sci., 2025, 677, 342–351.
25. H. Liu, H. Li, J. Lu, S. Zeng, M. Wang, N. Luo, S. Xu and F. Wang, ACS Catal., 2018, 8, 4761–4771.
26. P. Zhang, C. Wang, X. Geng, J. Xu, F. Song, H. Pei, N.-N. Zhang, J.-P. Zou, H. Cui and H. Tan, J. Catal., 2025, 445, 116007.
27. X. Liu, Z. Shen, Z. Jiang, M. Jia and W. Zhao, Sep. Purif. Technol., 2025, 354, 129010.
28. T. Zhang, W. Schilling, S. U. Khan, H. Y. V. Ching, C. Lu, J. Chen, A. Jaworski, G. Barcaro, S. Monti, K. De Wael, A. Slabon and S. Das, ACS Catal., 2021, 11, 14087–14101.
29. Y. Zhang, S. Qin, N. Claes, W. Schilling, P. K. Sahoo, H. Y. V. Ching, A. Jaworski, F. Lemière, A. Slabon, S. Van Doorslaer, S. Bals and S. Das, ACS Sustainable Chem. Eng., 2022, 10, 530–540.
30. K. Kaya, A. Atsay, H. Gunduz, A. Slabon and Y. Yagci, Adv. Sustain. Syst., 2024, 8, 2300366.
31. K. Chen, J. Schwarz, T. A. Karl, A. Chatterjee and B. König, Chem. Commun., 2019, 55, 13144–13147.
32. J. D. Nguyen, B. S. Matsuura and C. R. J. Stephenson, J. Am. Chem. Soc., 2014, 136, 1218–1221.
33. Q. Zhang, N. K. Gupta, M. Rose, X. Gu, P. W. Menezes and Z. Chen, Chem. Catal., 2023, 3, 100470.
34. M.-W. Zheng, Y. Lang, X. Han, L. Jia, H. Zeng and C.-J. Li, Cell Rep. Phys. Sci., 2024, 5, 102009.
35. C.-X. Chen, S.-S. Yang, J.-W. Pang, L. He, Y.-N. Zang, L. Ding, N.-Q. Ren and J. Ding, Environ. Sci. Ecotechnol., 2024, 22, 100449.
36. D. Kolb, A. Fedulin and B. König, ChemPhotoChem, 2025, 2500036.
37. C.-X. Chen, S.-S. Yang, J. Ding, L. Ding, R. Wu, L.-M. Liu, J.-W. Pang, L. He, J.-Q. Jiang and N.-Q. Ren, Appl. Catal., B, 2023, 334, 122823.
38. P. R. Maddigapu, A. Bedini, C. Minero, V. Maurino, D. Vione, M. Brigante, G. Mailhot and M. Sarakha, Photochem. Photobiol. Sci., 2010, 9, 323–330.
39. S. J. St. Romain, L. M. Basirico, Y.-H. Kim, E. V. Nolan, W. Xu and K. L. Armbrust, Environ. Toxicol. Chem., 2023, 42, 1721–1729.
40. T. Scapini, C. Dalastra, A. F. Camargo, S. Kubeneck, T. A. Modkovski, S. L. A. Júnior and H. Treichel, Bioresour. Technol., 2022, 344, 126325.
41. T. C. Jackson, J. D. Verrier and P. M. Kochanek, Cell Death Dis., 2013, 4, e451–e451.
42. A. Bedini, E. De Laurentiis, B. Sur, V. Maurino, C. Minero, M. Brigante, G. Mailhot and D. Vione, Photochem. Photobiol. Sci., 2012, 11, 1445–1453.
43. D. Broadbent, Chem. Commun., 1967, 382–383.
44. B. Yuan, D. Mahor, Q. Fei, R. Wever, M. Alcalde, W. Zhang and F. Hollmann, ACS Catal., 2020, 10, 8277–8284.
45. E. Tacchi, G. Rossi, M. Natali, L. Đorđević and A. Sartorel, Adv. Sustain. Syst., 2025, 9, 2400538.
46. A. E. da H. Machado, R. Ruggiero and M. G. Neumann, J. Photochem. Photobiol., A, 1994, 81, 107–115.
47. J. Wang, Y. Deng, Y. Qian, X. Qiu, Y. Ren and D. Yang, Green Chem., 2016, 18, 695–699.
48. R. A. Lee, C. Bédard, V. Berberi, R. Beauchet and J.-M. Lavoie, Bioresour. Technol., 2013, 144, 658–663.
49. J. Hynynen, A. Riddell, A. Achour, Z. Takacs, M. Wallin, J. Parkås and D. Bernin, Green Chem., 2021, 23, 8251–8259.
50. E. A. Karpukhina, D. S. Volkov and M. A. Proskurnin, Agronomy, 2023, 13, 1141.
51. F. Xu, J. Yu, T. Tesso, F. Dowell and D. Wang, Appl. Energy, 2013, 104, 801–809.
52. H. Li and A. G. McDonald, Ind. Crops Prod., 2014, 62, 67–76.
53. F. Fink, F. Emmerling and J. Falkenhagen, Chem. Methods, 2021, 1, 354–361.
54. U.S. Bureau of Reclamation , Water Facts, https://www.usbr.gov/mp/arwec/water-facts-ww-water-sup.html (accessed August 22, 2025).
55. S. B. Kim, S. J. Hong, D. H. Kim, G. M. Lee, Y. Lim, S. Lee, Y. Kwon and H. Y. Bae, Green Chem., 2025, 27, 7960–7972.
56. J.-L. Brédas, D. Beljonne, V. Coropceanu and J. Cornil, Chem. Rev., 2004, 104, 4971–5004.
57. Q. Miao, Z. Cao, Y. Liu, X. Zhou, J. Xie, T. Huo and W. Jiang, Chem. Eng. J., 2025, 508, 160957.
58. K. P. Clark and H. I. Stonehill, J. Chem. Soc., Faraday Trans. 1, 1972, 68, 577–590.
59. A. D. Broadbent and R. P. Newton, Can. J. Chem., 1972, 50, 381–387.
60. A. D. Broadbent, H. B. Matheson and R. P. Newton, Can. J. Chem., 1975, 53, 826–830.
61. V. Maurino, D. Borghesi, D. Vione and C. Minero, Photochem. Photobiol. Sci., 2008, 7, 321–327.
62. E. Atinault, V. De Waele, U. Schmidhammer, M. Fattahi and M. Mostafavi, Chem. Phys. Lett., 2008, 460, 461–465.
63. M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. A. Montgomery, J. E. Peralta, F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, O. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski and D. J. Fox, Gaussian 09, Revision B.01, Gaussian, Inc., Wallingford CT, 2009.
64. Y. Zhao and D. G. Truhlar, Theor. Chem. Acc., 2008, 120, 215–241.
65. A. V. Marenich, C. J. Cramer and D. G. Truhlar, J. Phys. Chem. B, 2009, 113, 6378–6396.
66. C. Gonzalez and H. B. Schlegel, J. Chem. Phys., 1989, 90, 2154–2161.
67. C. Gonzalez and H. B. Schlegel, J. Phys. Chem., 1990, 94, 5523–5527.
68. C. Gonzalez and H. B. Schlegel, J. Chem. Phys., 1991, 95, 5853–5860.
69. H. Wang, G. J. Giardino, R. Chen, C. Yang, J. Niu and D. Wang, ACS Cent. Sci., 2023, 9, 48–55.
70. S. Nanayakkara, A. F. Patti and K. Saito, Green Chem., 2014, 16, 1897–1903.
71. X. Diao, Y. Xiong, Y. Shi, L. Ma, C. Dong, S. Zhang and N. Ji, Green Chem., 2024, 26, 11406–11426.
72. S. Yu, P. Xu and Z. Luo, Ind. Crops Prod., 2024, 222, 119632.

---