Organic photoredox catalysts: tuning the operating mechanisms in the degradation of pollutants

Introduction

Photocatalysis is an emerging area of chemistry that takes advantage of light as the primary source of energy to carry out chemical transformations. Light is considered an environmentally friendly reagent since it is harmless and infinitely available if coming from sunlight. Recently, more and more papers have been published describing photocatalytic processes, not only with applications in synthesis [1], [2], [3], [4], [5], [6], [7], [8], [9], but also in the degradation of recalcitrant organic pollutants [10], [11], [12], [13], [14], [15], [16], [17], [18]. The latter constitutes an example of the application of the techniques known as advanced oxidation processes (AOPs) [19], [20], [21], [22], [23]. In general terms, AOPs are characterized by the generation of highly reactive species, such as hydroxyl radical (OH·), able to oxidize organic compounds, giving rise ideally to carbon dioxide and water as final products. The most emblematic heterogeneous photocatalyst is TiO₂. This semiconductor exhibits great activity, high stability, low cost, and non-environmental impact. However, it is only able to absorb light in the UV-A and UV-B regions of the solar spectrum. Besides, the photogenerated electron/hole pair experience fast recombination, which is only prevented in the presence of appropriate electron acceptors. Finally, despite being heterogeneous, TiO₂ is generally used as a nanopowder, which makes catalyst recovery difficult at a large scale [24, 25]. On the other hand, organic photocatalysts appear as a less explored alternative to treat polluted effluents.

Hence, although organic photocatalysts are not able to generate hydroxyl radical, their photoactivated excited states generated using visible light can act as strong oxidants in most cases, which constitutes an interesting alternative to produce photodegradation of contaminants [13, 26, 27]. In fact, it has been well established that pollutant photooxidation can be produced from an initial electron transfer between an excited state of an organic photocatalyst and the contaminant, generating their respective radical anion and cation (Type I mechanism). Subsequent interaction of reactive pollutant radical cation with other species in the media, such as oxygen, gives rise to oxidation products. However, as most of the organic photocatalysts are able to generate singlet oxygen, pollutant degradation can also be initiated from this oxidative species (Type II mechanism). In this context, organic homogeneous photocatalysis is mainly governed by organic dyes and transition metal complexes, but in terms of sustainability, dyes are economical, harmless to the environment, and easy to use. These advantages convert them into excellent alternatives to the transition metal complexes, which are mainly made of expensive and polluting rare metals such as Ru and Ir. However, some drawbacks of organic dyes such as their low photostability (they can undergo oxidative or reductive processes as pollutants do, or be sensitive to other parameters that influence their photodegradation such as irradiation dose, temperature, solvent, pH, etc.) [28] or their impossibility to be recovered and reused (from homogeneous solutions), are limiting factors for an extensive application.

Nevertheless, their heterogenization seems the straightforward step to boost photostability and facilitate the recovery of the photocatalysts. Different materials can be used to achieve heterogenization, such as nanoparticles [29], [30], [31], but the probable most extended methodology is based on the adsorption of the organic photocatalysts onto a solid inert material, such as zeolites, carbon-based materials or TiO₂ [32], [33], [34], [35], [36], [37]. Even more, covalent derivatization of SiO₂ has also been reported [38], [39], [40], [41], [42], [43]. The heterogenization of a photocatalyst is not a trivial issue, not only for the preparation and characterization of a new photocatalyst, but also in terms of understanding the reactivity of the new photocatalytic material. In fact, it cannot be assumed that the main reactions of a photocatalyst are the same under homogeneous conditions than in heterogeneous media. Herein we have selected Rose Bengal (RB), Riboflavin (RF) and a perylene diimide derivative (PDI) to illustrate different modes of action of these organic photocatalysts under homogeneous/heterogeneous conditions (Fig. 1). The postulated operating mechanisms are mainly based on photophysical experiments. Thus, time resolved and steady state fluorescence, as well as laser flash photolysis are fundamental techniques to evaluated the processes involved in the degradation of pollutants by organic photocatalysts.

Fig. 1:

Chemical structures of selected photocatalysts (PC): Rose Bengal (RB), perylene diimide (PDI), and riboflavin (RF).

Pathways involved in oxidative degradation of pollutants by organic photocatalysts

The participation of all the initial species arising from the photocatalyst ground state (PC), such as their excited states and/or singlet oxygen, can be evaluated in the photodegradation of the pollutants (Q) on the basis of the photophysical features of the photocatalyst (PC). A map of the different alternatives to produce the oxidation of a pollutant upon excitation of an organic photocatalyst is depicted in Scheme 1. Thus, for that purpose, the association constant (K) between the organic photocatalyst and the pollutant (Q), lifetimes (τ) and quantum yields (Φ) of fluorescence and triplet excited state of photocatalyst, as well as its singlet oxygen quantum yield (Φ_Δ) and all quenching rate constants (k) shown in Scheme 1 must be evaluated considering the initial concentration of the pollutant and photocatalyst. The percentage of participation of each species shown in Scheme 1 can be determined using eqs. (1)–(8). Hence, eq. (1) allows determining the percentage of photocatalyst involved in the formation of complexes with the pollutant ( [ P C δ + … . Q δ ′ + ] ). These complexes could, in principle, absorb light giving rise to an excited state from which the oxidation of the pollutant would also take place.

Scheme 1:

Main photocatalytic pathways involved in pollutant oxidative processes. PC, photocatalyst; Q, pollutant; K, association constant between the pollutant and the photocatalyst in its ground state; k _qS, quenching rate constant for the reaction between Q and the ¹PC*; k _qT, quenching rate constant for the reaction between Q and the ³PC*; k _qT-O2, quenching rate constant for the reaction between O₂ and ³PC*; and k _q1O2-Q, quenching rate constant for the reaction between Q and ¹O₂.

In parallel, the percentage (%) of photocatalyst that reaches its singlet excited state (¹PC*) can be calculated according to eq. (2).

Once the free photocatalyst reaches its singlet excited state, an electron transfer can happen in the presence of a pollutant (Q), and the efficiency (in %) of this reaction can be obtained from eq. (3). In competition with this process, ¹PC* suffers intersystem crossing to its triplet excited state (³PC*), and this percentage can be obtained using eq. (4). Photophysical features of PC, such as fluorescence lifetime (τ_S), their fluorescence and intersystem crossing quantum yield (Φ_F and Φ_ISC, respectively), and the ¹PC* quenching rate constant (k _qS) by the pollutant, are needed. Next, the generated ³PC* could react with the pollutants and/or with molecular oxygen; the % of each process is obtained according to eqs. (5) and (6), respectively, where the reactivity of ³PC* with Q and O₂ (k _qT and k _qT-O2, respectively) and the lifetime of the ³PC* (τ_T) are determined in the solvent used in the reaction media. The % of generated ¹O₂ from the reaction between ³PC* and O₂ is determined from eq. (7). Singlet oxygen quantum yield of the photocatalyst (Φ_Δ) is included in this equation as new data. Moreover, the percentage of ¹O₂ that reacts with the pollutant is determined according to eq. (8), using the quenching rate constant of ¹O₂ by Q (k _q1O2-Q) and ¹O₂ lifetime (τ_1O2) in the solvent used in the reaction media.

Rose Bengal

Rose Bengal (Fig. 1) is a well-known dye which exhibits a high singlet oxygen quantum yield (Φ_Δ = 0.76–0.83) [13, 44, 45]. Among other applications, it has been widely used for wastewater remediation in homogeneous phase, where its activity has been attributed to the Type II mechanism [46], [47], [48], [49]. However, based on its redox potential (−0.95 V vs. SCE) it could also be able to react via electron transfer (Type I mechanism), mainly from its triplet excited state [11, 26, 50, 51]. Thus, RB constitutes a good candidate in order to study the competition between Type I vs. Type II reactions.

To rationally quantify the involvement of each pathway, Type I or Type II, a detailed mechanistic kinetic study of the performance of RB as homogeneous photocatalyst in the disappearance of two common pollutants (acetaminophen, ACF and diclofenac, DCF) has been performed [52]. Preparative results confirmed the photodegradation of the pollutants in the presence of RB upon green LED irradiance. Then, photophysical experiments indicated that the photodegradation may happen through both pathways. In fact, ³ RB* is quenched by the pollutants (k _qT ca. 10⁷–10 [8] M⁻¹ s⁻¹) and also by O₂ to generate ¹O₂ (k _qT-O2 ca. 10⁹ M⁻¹ s⁻¹). The ¹O₂ formed is also subsequently quenched by the pollutants (k _1O2-Q ca. 10⁶–10⁷ M⁻¹ s⁻¹). However, attending to the typical concentration of oxygen in aqueous aerobic media [(O₂) = 2.9 × 10⁻⁴ M] [53], the concentration of the pollutants (10⁻⁵ M) and the experimentally determined rate constants, formation of ¹O₂ accounts for the major quenching pathway of ³ RB* (>97 %). However, subsequent quenching of ¹O₂ by the pollutants, in the range of 10⁶–10⁷ M⁻¹ s⁻¹ is very inefficient, which involves that, at the typical contaminant concentration almost all the ¹O₂ formed decays without reacting. Thus, photodegradation of ACF and DCF by homogeneous RB happens mainly through Type I mechanism (Scheme 2) [52]. Analogous results were also found using N-methylquinolinium salt, a UV-absorbing photocatalyst with a high reported Φ_Δ = 0.85 [13, 54].

Scheme 2:

Kinetic analysis of the involved pathways shed light on the main role of Type I (electron transfer to form free radicals, ROS and hydrogen peroxide) vs. Type II (energy transfer to form singlet oxygen) in the degradation of the pollutants acetaminophen (ACF) and diclofenac (DCF) photocatalyzed by Rose Bengal (RB) [52].

In order to improve the prospects of RB, a wide variety of RB-based photocatalysts have been synthesized on heterogeneous supports [41, 55–58]. Recently, several RB-based photocatalysts based on covalent derivatization of nanometric SiO₂ spheres (without or with magnetite core, SiO ₂ -RB and Fe ₃ O ₄ @SiO ₂ -RB, respectively) [59] and on micrometric glass wool (GW) fibers (GW-RB and GW-RB+) [60, 61] have been developed in order to investigate their mechanisms of action in decontamination and disinfection processes.

From the synthetic point of view, RB has been covalently anchored to the silica supports by pre-treatment of the spheres with an aminosilane connector [(3-aminopropyl)triethoxysilane, APTES], or by treating the fibers with previously derivatized APTES-RB (Scheme 3). In all cases, covalent anchoring prevents subsequent leaching of RB into the reaction media [59], [60], [61]. These chemical strategies open the possibility to design further dual photocatalysts that enhance the effect of RB with additional additives. As for example, GW-RB+ was designed to increase the bactericidal effect of RB against Gram-negative bacteria by incorporating cationic chains on the surface of GW, vide infra [60].

Scheme 3:

General procedure for the preparation of new heterogeneous photocatalysts based on derivatization of nanometric SiO₂ spheres (SiO ₂ -RB and Fe ₃ O ₄ @SiO ₂ -RB) and micrometric glass wool (GW) fibers (GW-RB+ and GW-RB) [59], [60], [61].

The photophysical properties of ¹ RB* and ³ RB* were not altered upon covalent binding of RB to NPs with and without magnetite core, and their intersystem crossing quantum yields (Φ_ISC) would be approximately the same as described for RB in homogeneous solutions (Φ_ISC = 0.8–0.98). Also, no differences in signal intensity or ¹O₂ lifetime were observed between homogeneous and heterogeneous systems (SiO ₂ -RB and Fe ₃ O ₄ @SiO ₂ -RB) using D₂O as solvent [59]. In the case of GW-RB and GW-RB+ the generation of ¹O₂ was chemically demonstrated, by oxidation of 9,10-diphenylanthracene (DPA) to its corresponding endoperoxide [60].

The heterogeneous RB-based photocatalysts were evaluated in the photodegradation of different drugs (DCF and ACF) [59, 60]. Heterogeneous photocatalysis resulted in higher photodegradation yields than control irradiations using homogeneous RB. Specifically, the anti-inflammatory drug DCF was completely removed after 3 h green LED irradiation, under aerobic conditions with all RB-heterogeneous photocatalysts, while only 60 % photodegradation was observed under homogeneous RB. In the case of ACF, SiO ₂ -RB and Fe ₃ O ₄ @SiO ₂ -RB achieved more than 70 % degradation after 5 h under aerobic irradiation vs. 30 % under homogeneous conditions. The higher photodegradation efficiency can be attributed to the higher photostability of RB when is supported (Fig. 2) [59, 60].

Fig. 2:

Real pictures showing the color evolution upon irradiation time with green LEDs under air conditions. (a) Homogeneous RB vs. Fe ₃ O ₄ @SiO ₂ -RB and SiO ₂ -RB (5 × 10⁻⁶ M RB in homogeneous media and 10 % mol in heterogeneous media) photodegradation of diclofenac (DCF, 5 × 10⁻⁶ M) in 10 mL of water [59]. (b) Homogeneous RB (1.59 µM) vs. GW-RB+ and GW-RB (45 and 35 mg, respectively) photodegradation of diphenylanthracene (DPA, 23 µM) in 3 mL of CHCl₃ [60].

The static quenching of ¹ RB* by the drugs is very similar for SiO ₂ -RB and Fe ₃ O ₄ @SiO ₂ -RB and in homogeneous conditions. Moreover, the ³ RB* was also quenched by the two drugs. The analysis of the quenching constants involving pollutant degradations (DCF and ACF) has evidenced that electron transfer between RB and drugs, in both homogeneous solutions and heterogeneous media, is the initial step of their oxidations (see Scheme 1). Although, the incorporation of the magnetic core was planned to increase the triplet lifetime and therefore enhance the photodegradation yield, this effect was not observed [59].

Moreover, these heterogeneous silica-based RB photocatalysts were evaluated in disinfection [59], [60], [61]. Both were very efficient in the inactivation of Gram-positive bacteria (similar result was already reported under homogeneous conditions). However, we have observed that their efficiency is dependent on the concentration. In fact, higher concentration of catalyst induces a lower efficiency in the disinfection process, which could be attributed to the formation of aggregates that results in lower singlet oxygen generation [62, 63]. On the other hand, no inhibition was observed for Escherichia coli regardless the concentration tested, neither in the presence of the heterogeneous photocatalysts nor homogeneous RB. It has been widely reported that Gram-negative bacteria are more resistant to photodynamic bactericidal activity than Gram-positive bacteria due to their highly organized outer wall [64]. In addition, the repulsion between the anionic RB and the negative charges on the surface of the outer membrane of Gram-negative bacteria could be associated to the observed inefficiency. Nevertheless, recent studies showed increased bactericidal action of electrostatic RB against Gram-negative Pseudomonas aeruginosa using ion exchange resins as cationic polystyrene supports [65, 66]. To the best of our knowledge, we have reported the first dual photocatalyst (GW-RB+), which includes the cationic chains and RB covalently anchored to glass wool, capable of causing a strong synergistic effect on Gram-negative bacteria (>4 log₁₀ at 120 min) under irradiation [61]. The synergistic effect induces a very effective inactivation of E. coli (as well as the Gram-positive ones) without the need to add disrupting agents to enhance the penetration of the dye to inner cell compartments. The synthesis and design of such dual photocatalysts open the way for bacterial photoinactivation in all types of silica-based materials.

Overall, the heterogenization of RB on different supports offers several advantages. On the one hand, it clearly increases its photostability. Nevertheless, photophysical experiments provided evidence for the type I mechanism as the main operating one in the photodegradation of contaminants regardless of the homogeneous/heterogeneous conditions. On the other hand, heterogenization, together with the incorporation of cationic chains, demonstrated that RB could be an appropriate photocatalyst for the inactivation of Gram-negative bacteria. Although the type II mechanism is the most accepted one for the inactivation of bacteria, and singlet oxygen is also generated from these heterogeneous materials, the participation of Type I could not be disregarded. Covalent attachment of the dye on the support proved to be better than electrostatic attachment preventing leaching and facilitating recovery and reuse.

Perylene diimide

The family of perylene dyes possesses high luminescence efficiency, electron mobility, and thermal, chemical, and photochemical stability. In the last years, several perylene diimide derivatives have been prepared for organic solar cells, organic light-emitting diodes, fluorescent sensors, as well as phototheranostics and bioimaging agents [67–69].

The UV–vis absorption spectra of perylene diimide (PDI) derivatives show three vibronic peaks and typically high fluorescence quantum yields. The energies of the excited states are in the order of 54 kcal mol⁻¹ for the singlet and 27 kcal mol⁻¹ for the triplet excited states [70]. Interestingly, the physicochemical, optical, and structural properties of PDIs can be modified by changing their substituents at the positions R₁ and R₂ (Fig. 3). Unfortunately, notwithstanding their excellent properties, the strong tendency of PDIs to π–π stacking results in poor solubility in common organic solvents limiting their use as photocatalysts in the homogeneous phase.

Fig. 3:

PDI main structure and common modification positions.

Nevertheless, herein, the successful covalent incorporation of PDI (Fig. 1) on silica mesoporous matrices MCM-41 and SBA-15 giving rise to versatile heterogeneous photocatalysts has been demonstrated (MCM-PDI and SBA-PDI, respectively) [71]. The efficiency of the ¹O₂ production from perylene singlet excited state is reported as 0.27 in homogeneous acetonitrile solution [72–79]. Thus, the generation of ¹O₂ by the excitation of heterogeneous PDI-Silicas was determined using dimethyl anthracene as a probe reagent and higher catalytic activity was achieved with SBA-PDI. Therefore, SBA-PDI was evaluated as a photoredox catalyst using molecular oxygen as an oxidant under green light excitation [71]. The oxidative scission of stilbene to form benzaldehyde was selected as a representative example of an oxidation reaction (Scheme 4a). Upon irradiation, the singlet-excited state (SBA- ¹ PDI*) can be readily quenched by t-stilbene to get the radical anion of the SBA-PDI ^.− and the corresponding radical cation derived from t-stilbene (participation of the triplet excited state was discarded according to the low reported intersystem crossing quantum yield, <0.001) [70]. Reactive oxygen species, including O₂, O₂ ^·−, and ¹O₂, were responsible for the subsequent quenching of the radical cation giving rise to the aldehydes. In addition, SBA-PDI ^.− can be quenched by O₂ to close the catalytic cycle. Positive evidence on the formation of the SBA-PDI ^.− was obtained by UV–vis spectroscopy due to the characteristic absorption band at around 700 nm. Furthermore, SBA-PDI was also investigated in the photocatalytic reduction of benzyl bromide, and a plausible mechanism was proposed (Scheme 4b). Upon excitation under inert atmosphere, SBA- ¹ PDI* can experience an electron transfer process with sacrificial electron donors (SED), such as triethylamine, present in the reaction media giving rise to SBA-PDI ^.- . Subsequently, SBA-PDI ^.− transfers an electron to benzyl bromide, leading to benzyl radical prior to the formation of both observed products [71].

Scheme 4:

Postulated mechanisms to explain the photocatalytic oxidation of stilbene (a) and the photocatalytic reduction of benzyl bromide (b) by SBA-PDI (perylene diimide moiety grafted onto silica). SED, sacrificial electron donor [71].

The successful heterogenization of PDI results into the synthesis of two heterogeneous photocatalysts that showed dual photoredox activity to organic oxidations or reductions. A deep mechanistic study of both processes provided evidence for the postulated mechanisms and open the door to further explore the versatility of this kind of photocatalytic materials.

Riboflavin

Riboflavin (RF), a naturally occurring pigment (Fig. 1), is found in living organisms where it acts as an antioxidant and is the precursor of flavin mononucleotide (FMN) and flavin dinucleotide (FAD) [80]. It absorbs in the visible region with a defined band peaking at ca. 450 nm, and in the UVA with a maximum at 370 nm. The combination of the redox potential of the RF (E_red = −0.80 V vs. SCE) together with the energy of its excited states (E_S = 2.48 eV and E_T = 2.17 eV), turns RF into a moderate oxidant from its singlet (¹ RF*) and triplet excited states (³ RF*) [80–83]. Moreover, RF is known for its efficient ability to generate singlet oxygen (Φ_Δ = 0.49) [84]. For this reason, RF has been suggested as the substance responsible for the “naturally” occurring abatement of pollutants [13, 46, 47, 49, 85–92]. From the mechanistic point of view, RF could behave as an endogenous photocatalyst for wastewater remediation acting through Type I and Type II mechanisms, see Scheme 1 [93]. However, RF exhibits low photostability due to the intramolecular H-abstraction from the ³ RF* and the generated ¹O₂ that attacks the heterocycle [94, 95]. To reduce this degradation under light irradiation, a popular strategy is to convert RF into the riboflavin 2′,3′,4′,5′-tetraacetate (RFTA) [96]. The photodegradation of several pollutants mediated by RFTA and visible light has been widely studied in homogeneous media [85, 87, 97]. Kinetics studies revealed efficient degradations, in which participation of different species such as singlet oxygen, superoxide radical anion and/or the singlet and triplet excited states of RF was envisaged [98, 99]. Recent examples demonstrated also the photodegradation of drugs, such as Carbamazepine (CBZ) and Atenolol (ATN), by RFTA under visible light in less than 2 h. Photophysical studies indicated that CBZ and ATN are oxidized by a combination of Type I and Type II mechanisms [100].

Beyond the efficiency of RF in AOPs, the photocatalytic potential of RF has been expanded to the photoreduction of recalcitrant contaminants [14, 101]. In fact, RFTA^•− can act as a reductant to abate organic contaminants in water. To afford RFTA^•− an efficient electron donor such as 1,4-Diazabicyclo [2.2.2]octane (DABCO) is required. DABCO efficiently quenches the singlet and the triplet-excited states of RFTA with diffusion-controlled quenching constants [102]. Therefore, when DABCO is used at molar concentration, naturally occurring O₂ (at sub-milimolar concentration) is not able to compete with DABCO for ³ RFTA*. The photogenerated RFTA^•− is responsible for the observed degradation of recalcitrant pollutants such as benzotriazole-derived solar filters (BUVSs). Photophysical experiments provided evidence for the participation of this species and allowed determining the quenching rate constants that were in good agreement with the observed photodegradation (Scheme 5) [103].

Scheme 5:

Photoreductive degradation of BUVSs (benzotriazole-derived solar filters) under aerobic conditions in the presence of RFTA (riboflavin 2′,3′,4′,5′-tetraacetate). DABCO, 1,4-Diazabicyclo [2.2.2]octane (efficient electron donor) [103].

An additional alternative to enhance the photostability of RF, and at the same time favor its recyclability, is heterogenization. In this regard, we have recently developed a recyclable and robust organic photocatalyst for wastewater remediation based on RF [104]. As a support, SiO₂ particles were selected due to their high mechanical stability and well-known surface chemistry [71, 105]. Then, a large number of RF molecules, previously derivatized by treatment with 3-(triethoxysilyl) propyl isocyanate [106], were covalently anchored, creating a complete shell on the surface of the silica particles as shown from loading experiments.

The photophysical properties of the synthesized SiO ₂ -RF photocatalyst revealed great differences from the ones displayed by homogeneous RF [104]. Thus, the fluorescence quantum yield of SiO ₂ - ¹ RF* (Φ_F = 0.02) and its lifetime (within the lamp pulse, <1 ns) were clearly smaller than those of ¹ RFTA* (Φ_F = 0.4 and τ_S = 7.6 ns) [107]. Furthermore, the emission maximum displayed by SiO ₂ -RF was red-shifted (531 nm vs. 516 nm), indicating an enhanced electronic delocalization for the first singlet excited state of SiO ₂ - ¹ RF*. These singlet excited state features acted as unambiguous evidence of the expected π–π stacking interactions between chromophore dimers. Thus, the most likely deactivation pathway of the first singlet excited state of SiO ₂ - ¹ RF* seems a charge transfer interaction with close ground state RF moieties, an effect that has previously been described when agglomeration of RF molecules gives rise to the formation of RF dimers [108]. The efficient deactivation of SiO ₂ - ¹ RF* prevents the formation of SiO ₂ - ³ RF* and, consequently, the ¹O₂ was not generated, as demonstrated from laser flash photolysis experiments. With this short-lived singlet excited state and no triplet generation, the adsorption capability of SiO ₂ -RF was crucial to eventually act as an efficient photocatalyst. An aqueous mixture of four phenolic pollutants was selected to evaluate this property. Results indicated that more than 50 % of the pollutants were adsorbed on the surface, ensuring a high proximity between the chromophore and the contaminants [104]. Even more, it could be anticipated that other molecules with similar or higher molecular sizes will show similar adsorption affinity. Next, upon visible light irradiation, SiO ₂ -RF produced complete photocatalytic removal of the phenolic pollutants in aqueous aerobic media in less than 4 h. Furthermore, the photodegradation of each pollutant was in good agreement with its adsorption percentage, indicating that the photodegradation is produced on the adsorbed molecules. Moreover, up to 60 % of its activity was observed in the third consecutive run as a further proof of the enhanced photostability. In order to establish a mechanism of action, the quenching of SiO ₂ - ¹ RF* by four phenolic derivatives was evaluated [104]. Static quenching is operating, while no changes were detected for the lifetime of SiO ₂ - ¹ RF*. As the triplet excited state SiO ₂ - ³ RF* is not formed, participation of this excited state and of ¹O₂ is precluded. As conclusion, the main photocatalytic process involved in the degradation of pollutants (Q) photocatalyzed by SiO ₂ -RF is an electron transfer from the pollutants to SiO ₂ - ¹ RF*. In this context, the adsorption of the pollutants on the SiO ₂ -RF surface is the key physical process to accomplish the pollutant photodegradation due to the small lifetime of the SiO ₂ - ¹ RF* species, see Scheme 6 [104].

Scheme 6:

Postulated mechanism to explain the photodegradation of the selected phenolic pollutants (Q) by SiO ₂ -RF (Riboflavin molecules covalently anchored on the surface of the silica particles) under visible light irradiation [104].

Conclusions

Organic molecules absorbing in the visible region of solar spectrum can act as efficient photocatalysts to produce the abatement of pollutants with a wide variety of chemical structures. Herein several examples have been selected to illustrate that photophysical experiments provide crucial information to postulate plausible operating mechanisms. Examples shown illustrated that beyond the oxidative Type I and Type II mechanisms, experimental conditions could be adjusted to achieve also reduction of recalcitrant pollutants. Even more, it is important to highlight that the heterogenization of organic photocatalysts is currently one of the most interesting lines of research because it results in more robust materials, easily recoverable and operating through additional mechanisms. For this reason, a great scientific effort must be made from the synthetic chemical point of view to efficiently address the anchoring of organic photocatalysts to new materials. Moreover, from the photochemical point of view, investigation of the behavior of heterogeneous photocatalysts must be mandatory to postulate their plausible mechanisms that may differ from those under homogeneous conditions.