Oxidative Degradation of Pharmaceuticals: The Role of Tetrapyrrole-Based Catalysts

: Nowadays, society’s widespread consumption of pharmaceutical drugs and the consequent accumulation of such compounds or their metabolites in efﬂuents requires the development of efﬁcient strategies and systems that lead to their effective degradation. This can be done through oxidative processes, in which tetrapyrrolic macrocycles (porphyrins, phthalocyanines) deserve special attention since they are among the most promising degradation catalysts. This paper presents a review of the literature over the past ten years on the major advances made in the development of oxidation processes of pharmaceuticals in aqueous solutions using tetrapyrrole-based catalysts. The review presents a brief discussion of the mechanisms involved in these oxidative processes and is organized by the degradation of families of pharmaceutical compounds, namely antibiotics, analgesics and neurological drugs, among others. For each family, a critical analysis and discussion of the fundamental roles of tetrapyrrolic macrocycles are presented, regarding both photochemical degradative processes and direct oxidative chemical degradation.


Introduction
The increasing presence of pharmaceuticals in wastewaters has become a huge environmental problem due to direct toxicity to living organisms and particularly the development of antibiotic-resistant bacteria. Therefore, it is urgent to develop processes capable of destroying these drug-based pollutants in the environment. Among the multiple oxidative processes described in the literature so far [1][2][3][4][5][6][7], this review intends to cover exclusively the ones based on the use of tetrapyrrolic macrocycle (TPM) catalysts over the last decade. First, for systematic-mechanistic clarification, we present the accepted photo-and nonphoto oxidation mechanisms, and the following sections are focused on the photochemical and non-photochemical degradation of several families of pharmaceutical drugs using TPM-based catalysts.

Oxidation Mechanisms Using Tetrapyrrolic Macrocycle-Based Catalysts
The general photochemical mechanism involved in the degradation of pharmaceutical drugs using photosensitizers, such as tetrapyrrolic macrocycles, is depicted in Scheme 1. The absorption of visible light with an appropriate wavelength promotes the catalyst/photosensitizer excitation from the singlet ground state, S 0 , to a singlet excited state (S 1 , S 2 , ... S n ). Generally, if a transition occurs to singlet excited states of a higher order (S 2 to S n ), a quick transition to S 1 occurs through vibrational relaxation (VR) and internal conversion (IC). Singlet excited states are short-lived and can directly return to the ground state through fluorescence emission or IC. Alternatively, an intersystem crossing (ISC) can occur to a triplet excited state (usually T 1 ), which is a longer-lived state capable of interacting with oxygen under two mechanisms. On the one hand, a direct or indirect electron transfer to molecular oxygen can occur (Type I mechanism), ultimately resulting in the formation of HO • and/or O 2 •− . On the other hand, an energy transfer (Type II mechanism) can occur between the catalyst in T 1 and 3 O 2 , resulting in the concomitant return of the catalyst to the ground state (S 0 ) and the formation of singlet oxygen ( 1 O 2 ) [8,9].
For Type I reactions [10], the photosensitizer (PS) in the triplet excited state ( 3 PS*) can directly interact with substrates present in the medium to original either anionic or cationic radical species (PS •− and/or PS •+ ; Equations (1) and (2)). In particular, the anionic species PS •− can interact with molecular oxygen, forming superoxide anion (O 2 •− ), while returning to its ground state (Equation (3)). Then, the reversible protonation of O 2 •− can lead to the formation of peroxo radicals (HOO • ; Equation (4)), which can then combine and form H 2 O 2 (Equation (5)). Finally, the reaction of H 2 O 2 with O 2 •− can lead to the formation of HO • (Equation (6)), which is considered the most reactive ROS as it possesses a higher one-electron redox potential and thus can oxidize a wider number of substrates [9]. 3 In the cases where there is a presence of either free Fe 3+ /Fe 2+ or the corresponding metal tetrapyrrolic macrocycles complexes [11], a Fenton-type redox reaction of H 2 O 2 with Fe 2+ can occur, leading to the formation of more OH • (Equation (7)). The reaction of the formed Fe 3+ with O 2 •− can regenerate the active Fe 2+ species (Equation (8)).
Additionally, Fenton-type redox reactions can also occur with peroxymonosulfate (PMS) to generate sulfate radicals, also commonly used in several advanced oxidation processes (AOPs). Radiation, such as UV, can efficiently activate PMS. Two activation pathways might occur when using radiation. The first one is the O-O bond fission provoked by the input of energy (Equations (9) and (10)). Furthermore, the radiation might dissociate water molecules (Equation (11)), producing the electron, which activates PMS by electron conduction (Equations (12) and (13)) [12].
For pharmaceutical degradation, the ideal outcome of ROS-promoted degradation would be the total mineralization of the products, as depicted in Equation (14).
It is worth noting that tetrapyrrole-based catalysts for photodegradation are often used in hybrid materials containing semiconductors in order to take advantage of their absorption in the UV region and complementary mechanisms of ROS formation. These are summarized in Scheme 2. Scheme 2. General mechanism for ROS formation using semiconductors.
Light absorption by semiconductors promotes charge dissociation, resulting in the formation of positively charged holes (h + ) in the valence band (VB) and electrons (e − ) in the conduction band (CB). Positively charged holes can promote oxidation reactions of organic molecules (drugs) adsorbed on the surface of the catalyst or generate ROS via water conversion into HO • . On the other hand, electrons in the CB can be transferred to molecular oxygen, generating O 2 •− , or promote reduction in H 2 O 2 to HO • . These semiconductor-promoted oxidation mechanisms are complementary to the ROS formation pathways mentioned above for the TPM-based catalysts and thus potentiate drug degradation. [1,13] Thus, a significant portion of the work developed in this field has been focused on the development of hybrid catalysts comprising TPM immobilized in semiconductors [2,[14][15][16][17][18][19][20][21][22].
In the absence of light, metal complexes of tetrapyrrolic macrocycles can also promote substrate oxidation in the presence of oxidants such as O 2 or H 2 O 2 . In fact, this process plays a crucial role in human drug metabolism, where the heme group containing a Fe(II)/Fe(III) metalloporphyrin is the prosthetic group of the cytochrome P450, one of the most important oxidative enzyme families [23][24][25][26]. Typically, biomimetic drug degradation studies for environmental remediation use water as a matrix, H 2 O 2 as the oxidant and a simplified mechanism, as depicted in Scheme 3. The first step involves the coordination of the central metal with H 2 O 2 and the formation of a peroxo species. It is worth mentioning that in cases where O 2 is used as an oxidant, there are two additional reductive steps to activate it in order to form the peroxo species. Then, the formation of oxo species occurs after H 2 O elimination and oxidation of the central metal [27,28]. In protic solvents, these intermediates can then react with drugs and promote a series of oxidative reactions, namely epoxidations, hydroxylation, dealkylation, deamination, decarbonylation, N-or S-oxidation, among others [29]. Scheme 3. Simplified mechanism for drug oxidation using H 2 O 2 and a tetrapyrrolic-based metal complex.

Degradation of Antibiotics
The term antibiotic derives from the ancient greek αντί (anti) + βιoτικóς (biotikos), which means "against a living being". Therefore, these drugs are exclusively used to treat bacterial infections and have no effect on viral infections. Typically, antibiotics are grouped according to their mechanism of action against specific types of bacteria. The main types of antibiotics are penicillins, tetracyclines, sulfonamides, quinolones, cephalosporins, aminoglycosides and macrolides [30].
The clinical overuse of antibiotics for humans and animals has led to the existence of active pharmaceutical ingredients of antibiotics in the environment, particularly in domestic wastewater (concentrations range from ng L −1 to µg L −1 ) and in hospital and pharmaceutical manufacturing wastewater (concentrations 100-500 mg·L −1 ) [31]. This occurrence has caused a huge human health problem due to the consequential development of multiresistant bacteria. Therefore, the development of AOPs for antibiotic degradation in wastewaters is currently a topic of utmost relevance. The structures of the antibiotics degraded by TMP so far are shown in Figure 1.

Photochemical Degradation
Several TPM-based photocatalysts have been used in the degradation of antibiotics, and their main structures are presented in Figure 2. Antibiotics of the tetracycline family (tetracycline-TC, oxytetracycline-OTC and their corresponding hydrochloride salts-TC·HCl and OTC·HCl, (Figure 1) are among the most prescribed and, fortunately, among the most studied regarding their photodegradation (Table 1, entries 1-12).
Sökmen [45] reported the synthesis of Co(quin) 4 Pc@TiO 2 by impregnation of Co(quin) 4 Pc ( Figure 2) onto semiconducting TiO 2 . The catalyst, in a concentration of 1 g/L, was used in the photodegradation of amoxicilin (AMX, Figure 1), and 40% degradation was obtained after 150 min using a UV light lamp (12 W) at 254 nm. Nevertheless, neat TiO 2 induced 38% degradation under the same conditions, which can be ascribed to the use of the UV light source, which enables TiO 2 photocatalytic properties (Table 1, entry 20).
The authors proposed a degradation pathway for TMP (Scheme 6) and observed that the major processes involved in the degradation of TMP were hydroxylation, oxidation and demethylation, mostly by HO

Degradation of Analgesic Drugs
Analgesics are drugs that relieve different types of pain. Most prescribed analgesics include anti-inflammatory drugs, which reduce inflammation (e.g., ibuprofen, diclofenac, acetaminophen derivatives and salicylic acid, known as nonsteroidal anti-inflammatory drugs), and opioid analgesics, which change the way the brain perceives pain (e.g., tramadol). These types of drugs are commonly encountered in municipal wastewaters since they are consumed in quite high quantities by the general population [55,56].

Photochemical Degradation
The Huang group used the iron complex of perchlorinated phthalocyanine, FePcCl 16 ( Figure 6) ( Table 3, entry 2) [58], as a heterogeneous photocatalyst to promote the degradation of salicylic acid (SA, Figure 5) upon irradiation with visible light from a 500 W halogen lamp. A 70% SA degradation was achieved after 10 h irradiation, with a catalyst concentration of 0.4 g/L, using H 2 O 2 as oxidant (3 × 10 −3 M), with a kinetics of k obs = 1.2 × 10 −1 min −1 . Once again, the mechanism was attributed to the formation of HO • radicals.
Anucha and Altin [15] developed the material CuPc(th) 4 @TiO 2 /ZnO by modifying TiO 2 and ZnO semiconductor blends with thiazol tetrasubstituted copper phthalocyanine (CuPc(th) 4 , Figure 6). This hybrid material was evaluated on the photodegradation of ibuprofen (IBU, Figure 5) in an aqueous matrix using five mercury lamps of 40 W at 365 nm ( Table 3, entry 3). Under these conditions, 80% of IBU photodegradation was achieved in 4 h at pH 6.5 and a degradation rate constant of k obs = 7.0 × 10 −3 min −1 , against 42% by TiO 2 /ZnO semiconducting catalyst. The authors also investigated the catalyst's stability, and a small decline (77%) of IBU degradation started only after the fifth cycle run. The authors put in evidence that a hydroxyl radical (HO • ) and superoxide anion (O 2 •− ) are the main ROS species involved in IBU degradation.  Mlynarczyk also investigated the IBU aqueous photodegradation in the presence of zinc(II) and copper(II) phthalocyanines (ZnPc and CuPc, respectively, Figure 6) embedded onto pure anatase-phase TiO 2 nanoparticles (ZnPc@TiO 2 and CuPc@TiO 2 ) [20]. Catalyst photoactivity was evaluated on a 10 mg/L solution of IBU in water by irradiating with three lasers (20 mW/cm 2 ) under either UV (365 nm) or visible light (665 nm) ( Table 3, entry 4). They obtained 95% IBU degradation with CuPc@TiO 2 under UV light irradiation after 6 h and a rate constant k obs = 3.8 × 10 −1 min −1 , while using ZnPc@TiO 2 , only 55% degradation was observed (Figure 7a). Interestingly, this value is even lower than the one observed when TiO 2 was used as catalyst, which reached nearly 93% photodegradation. On the other hand, negligible degradation was observed for both catalysts under visible-light irradiation, as can be seen in Figure 7b, which may be attributed to low light absorption by the catalyst. Furthermore, Ju and Hou [59] prepared the photocatalyst FePc@ZnO by impregnation of FePc ( Figure 6) onto semiconducting ZnO. The authors studied the photo-Fenton-type degradation of IBU ( Figure 5) in the presence of the catalyst, using H 2 O 2 as the oxidant and a visible-light Xe lamp (330 W) as the irradiation source. A 90% degradation of IBU was achieved in just 10 min, for which HO • was found to be the main oxidation species. The catalysts showed reusability for up to five cycles without significant loss of activity (Table 3, entry 5).
Several groups studied the ability of TPM-based catalysts to degrade diclofenac (DF, Figure 5). Yu's group studied its photodegradation using two different porphyrin-based MOFs. In a first study, they prepared PCN-134, TCPP@Zr-BTB MOF [60] based on a mesotetra(carboxyphenyl)porphyrin (TCPP, Figure 6) metal organic framework. After stirring a DF aqueous solution (30 mg/L) for 30 min in the dark to favor the adsorption-desorption equilibrium, the solution was irradiated with a visible-light Xe lamp (500 W) for 5 h, 99% of DF photodegradation was achieved using PCN-134 as photocatalyst (0.1 g/L; Table 3, entry 6). Moreover, the authors established that the mechanism of PCN-134 photodegradation is type II due to the generation of singlet oxygen ( 1 O 2 ). Additionally, the authors also performed catalyst-reutilization studies for three cycles, providing removal rates > 95%.
In a subsequent study, the same group developed another MOF-type catalyst, TCPP@UiO-66 [61]. Particularly, TCPP (Figure 6) was introduced onto UiO-66 crystals via an in situ solvothermal one-pot reaction, preserving the morphologic characteristics of UiO-66. Then, the authors irradiated a DF aqueous solution (30 mg/L) containing a catalyst concentration of 0.1 g/L, at pH 7, by simulated-sunlight 350 W Xe lamp (290 nm ≤ λ ≤ 1200 nm) ( Table 3, entry 7). Under these conditions, 99% of DF photodegradation was obtained after 4 h (k obs = 8.4 × 10 −3 min −1 ). The catalyst showed good recyclability for four reaction cycles without loss activity. As in the previous study, the authors attributed the photodegradation of DF mainly to the generation of 1 O 2 , along with h + (holes), to a minor extent.

Oxidative Chemical Degradation
The degradation of DF ( Figure 5) was also studied under nonphotochemical conditions. The Nackiewicz group studied the catalytic activity of iron(II) octacarboxyphthalo-cyanine (FeC8Pc, Figure 8) in the homogenous oxidation of an aqueous solution of DF at pH 8, using H 2 O 2 or NaIO 4 as oxidants (Table 4, entry 1) [55]. At a substrate/catalyst 50:1 molar ratio, the authors obtained full DF degradation in 35 min and 25 min when using H 2 O 2 and NaIO 4 , respectively. FeC8Pc also self-degraded completely due to the production of hydroxyl radicals formed by H 2 O 2 . The authors further studied the degradation pathway for DF, observing the generation of an unstable dimeric DF oxidation compound as intermediate, further leading to the final oxidation products.   The authors further proposed a degradation pathway for DF (Scheme 7). DF (m/z = 295) was oxidized by HO • into m/z 311 and m/z 309. Their referred quinone's ability to form charge-transfer complexes through electron transfer from donor to acceptor allowed DF and m/z 309 to be transformed into cation (m/z 295) and anion (m/z = 309) radicals, forming a charge-transfer complex [55]. Shi and Deng studied the catalytic performance of CoC4Pc@CNOMS, prepared by electrostatic linking of cobalt(II) tetracarboxyphthalocyanine (CoC4Pc, Figure 8) to an amino-functionalized manganese octahedral molecular sieve (CNOMS) on the degradation of a 10 mg/L DF aqueous solution, using PMS (3.2 × 10 −4 M) as oxidant (Table 4, Entry 2) [62]. After a 20 min reaction, 99% of DF degradation was achieved, following a pseudo-first-order kinetics (k obs = 1.87 × 10 −1 min −1 ). The authors observed that at pH = 7 and light activation of PMS, the degradation was promoted by both type I and type II mechanisms (HO • , SO 4 •− and 1 O 2 ROS species). The catalyst was also reused in four cycles, and a continuous slight decrease in the activity was observed.

Degradation of Neurological Pharmaceuticals
Neurology drugs manage diseases, disorders and conditions that affect the brain and nervous system. Carbamazepine is the only known neurological-system pharmaceutical evaluated in TPM-based degradation studies (Figure 9). This psychotropic drug is commonly used to treat epilepsy. The mechanism of action involves the blocking of the sodium channels of the neuron's membranes. This drug is commonly found in wastewaters due to its frequent use and incomplete removal by the traditional processing methods. [64]

Oxidative Chemical Degradation
Lu's group extended their exhaustive studies of CBZ degradation, using other phthalocyaninebased catalysts ( Table 6, entries 1-3) [70][71][72] under Fenton-like oxidation conditions, using H 2 O 2 as the oxidant. In a first study, the authors blended iron phthalocyanine (FePc, Figure 12) onto polyacrylonitrile (PAN), obtaining FePc@PAN nanofibers [70]. This catalyst, in a 1 g/L load, was then used in the degradation of CBZ (6 mg/L) in the presence of H 2 O 2 (20 mM) at pH = 3 and temperature = 70 • C ( Table 6, entry 1). The authors proposed that the hydroxyl (HO • ) radicals were the dominant active species in the oxidation catalytic process. The catalyst could be reused eight times without significant loss of activity. Later, the same group tested the iron(II) hexadecafluorinated phthalocyanine µ-oxo dimer (FePcF 16 ) 2 O (Figure 12) as catalyst for CBZ degradation, again using H 2 O 2 as oxidant [71]. In a 0.1 g/L catalyst concentration and 20 mM H 2 O 2 , a 6 mg/L solution of CBZ was fully degraded in 40 min, regardless of the pH tested (Table 6, entry 2). By electron paramagnetic resonance and electrospray ionization-mass spectrometry, the authors proposed that the Fe IV =O is the main active species arising from heterolytic cleavage of Fe III -OOH species. Similar results were obtained later [72] using the same phthalocyanine µ-oxo dimer (FePcF 16 ) 2 O (Figure 12) impregnated onto MWCNTs, providing the hybrid catalyst (FePcF 16 ) 2 O@MWCNT (Table 6, entry 3). The system (FePcF 16 ) 2 O@MWCNT/H 2 O 2 (0.2 g/L and 5 mM, respectively) was also used to degrade CBZ, reaching full degradation in 60 min. Reutilization studies showed that the catalyst is active along 10 reutilization cycles. [70][71][72].
Chetty [75] used the same porphyrin (TCPP, Figure 14) to prepare the hybrid photocatalysts TCPP@ATiNT and TCPP@Si-ATiNT by immobilizing it onto anatase titanium nanotubes (ATiNT) and Si-ATiNT (Table 7, entry 4). The authors irradiated 100 mg/L of FAM aqueous solutions, using both a 150 W UV-light mercury lamp and a 500 W visiblelight halogen lamp containing both catalysts in concentrations of 0.25 g/L. The authors concluded that TCPP@ATiNT was the best catalyst to promote full FAM degradation upon 240 min of irradiation, as demonstrated by its higher kinetic rate, k obs = 21.7 × 10 −3 min −1 against k obs = 8.0 × 10 −3 min −1 , when TCPP@Si-ATiNT was used. This difference was ascribed to the higher recombination rate favored by the silane-linker group present in the TCPP@Si-ATiNT catalyst, with a consequent decrease in electron injection. The authors also suggested that h+ holes formed in the HOMO of TCPP upon photoexcitation may be the most relevant oxidizing agent, rather than 1 O 2 , once photocatalytic experiments under oxygen-deficient conditions showed a similar rate of FAM degradation ( Figure 15). Figure 15. Schematic illustration of the proposed visible-light photocatalysis mechanism for FAM degradation using TCPP@ATiNT and TCPP@Si-ATiNT. Adapted with permission from ref [75]. Copyright 2020 Elsevier.
Regarding the degradation of PRP with PMS (0.2 g/L) as oxidant, Huiping [77] described the synthesis of a catalyst based on a binuclear cobalt carboxyl-substituted phthalocyanine (Co 2 CPc, Figure 14) supported by electrostatic interactions onto aminofunctionalized manganese octahedral molecular sieves (CNOMS) ( Table 7, entry 6). A 93% PRP degradation (k obs = 9.2 × 10 −2 min −1 ) was observed after 30 min, and 47% TOC. The authors proposed that both SO 4 •− radicals and 1 O 2 were the main oxidation species. Furthermore, reutilization was performed, and the catalyst remained active and stable for four cycles, as shown in Figure 16.

Conclusions and Perspectives
Considering the literature herein reviewed, one can generally conclude that the scientific community has been committed to finding more sustainable and alternative catalytic processes for the degradation of pharmaceuticals in the environment, particularly using tetrapyrrole macrocycle-(TPM) based catalysts. The main parameters requiring attention when developing such a catalytic system aiming its transposition at real-world application must include: (i) the TPM, considering its modulability and functionality, including substitution patterns for activity/stability; (ii) the type of support, aiming preferential immobilization at efficient reutilization and/or the holding of suitable semiconducting characteristics; (iii) the light source, when designing a photocatalytic system, preferentially using visible/solar energy; (iv) the oxidants, when designing oxidative chemical systems, giving preference to environmentally benign ones.
The most relevant examples discussed above can be highlighted by, for instance, the TNCuPc@CeO 2 /Bi 2 MoO 6 (copper(II) β-tetranitrophthalocyanine deposited on the surface of semiconducting CeO 2 /Bi 2 MoO 6 nanoflowers), which, when used in 1.5 g/L concentration, showed high catalytic performance in the photodegradation of antibiotic tetracycline (0.05 g/L concentration), with reusability up to four cycles without significant loss of activity, reaching, after 120 min irradiation with 800 W xenon visible light, a TOC removal efficiency of 83%, one of the highest reported so far [21].
For oxidative chemical degradation systems, the use of an MnTDCPPS@N-SiO 2 catalyst (meso-tetrakis(2,6-dichloro-3-sulfophenyl)porphyrinato manganese(III) covalently attached to aminopropyl functionalized silica gel) in the degradation of recalcitrant trimethoprim antibiotic must be highlighted. The immobilized catalyst, notwithstanding its use in quite a low concentration (0.002 g/L) and in the presence of 0.26 mM H 2 O 2 as oxidant, was able to promote the oxidative degradation of trimethoprim (0.13 g/L) in 95% (24% TOC decrease) after 150 min, showing reusability for up to five cycles without losing its activity [54].
As a whole, efforts in the design and larger-scale preparation of efficient catalysts should point to the modulation/functionalization of TPMs holding appropriate substituentimparting catalyst stability (e.g., bearing electron-withdrawing groups at the periphery) and suitable functionalities to promote covalent immobilization to supports, therefore avoiding catalyst leaching upon reutilization. When developing photocatalytic systems, these should aim for longer wavelength absorption and lower-charge carrier recombination (when using semiconductors as supports), where solar energy irradiation sources should be preferred. On the other hand, oxidants should be of primary concern when using catalysts for chemical oxidation, giving preference to benign sources, such as molecular oxygen, hydrogen peroxide or potassium peroxymonosulfate.
Another important challenge concerns the achievement of complete degradation of pharmaceuticals. When only partial/low mineralization occurs (the large majority of the studies herein reported), we consider that it is crucial to evaluate the byproducts generated to ensure their lower toxicity and environmental persistence since they can also contribute to increase the environmental toxicity and, particularly, the development of multi-resistant bacteria when dealing with antibiotics.