Application of visible light active photocatalysis for water contaminants: A review

Abstract Organic water pollutants are ubiquitous in the natural environment arising from domestic products as well as current and legacy industrial processes. Many of these organic water pollutants are recalcitrant and only partially degraded using conventional water and wastewater treatment processes. In recent decades, visible light active photocatalyst has gained attention as a non‐conventional alternative for the removal of organic pollutants during water treatment, including industrial wastewater and drinking water treatment. This paper reviews the current state of research on the use of visible light active photocatalysts, their modified methods, efficacy, and pilot‐scale applications for the degradation of organic pollutants in water supplies and waste streams. Initially, the general mechanism of the visible light active photocatalyst is evaluated, followed by an overview of the major synthesis techniques. Because few of these photocatalysts are commercialized, particular attention was given to summarizing the different types of visible light active photocatalysts developed to the pilot‐scale stage for practical application and commercialization. The organic pollutant degradation ability of these visible light active photocatalysts was found to be considerable and in many cases comparable with existing and commercially available advanced oxidation processes. Finally, this review concludes with a summary of current achievements and challenges as well as possible directions for further research. Practitioner Points Visible light active photocatalysis is a promising advanced oxidation process (AOP) for the reduction of organic water pollutants. Various mechanisms of photocatalysis using visible light active materials are identified and discussed. Many recent photocatalysts are synthesized from renewable materials that are more sustainable for applications in the 21st century. Only a small number of pilot‐scale applications exist and these are outlined in this review.


INTRODUCTION
Water pollution has serious negative impacts on both aquatic environments and associated human activities. The effluent from industry, agriculture, hospital, and households may contain many organic pollutants, such as dyes, paints, pesticides, gasoline, and pathogen (Cunha et al., 2019;Priyanka et al., 2020;Tran et al., 2014;Vatanpour et al., 2019). Some of them are recalcitrant and can only partially be treated through the traditional wastewater treatment process; examples are shown in Table 1. Due to rapid population increase, global industrialization, and urbanization, clean water resources are under pressure due to large increases in demand over the past number of decades. Organic pollutants discharged from such activities and sources have exacerbated the problem and many traditional wastewater treatment methods have been studied and applied to tackle this, including biotreatment, chemical treatment, filtration, and adsorption. However, such traditional wastewater treatment methods have limitations and disadvantages in treating water laden with persistent organic pollutants (Pardeshi & Patil, 2008;Villaluz et al., 2019); hence, advanced oxidation processes (AOPs) have been developed to improve treatment performance.
Among all AOPs, photocatalysis has received increasing attention, due to its lower cost, non-toxic materials, relatively high chemical stability of the catalyst, and efficiency under mild conditions using potential sunlight (Miklos et al., 2018;Xu et al., 2020;Yusuff et al., 2020). In addition, it also has a high potential for complete degradation, destruction, or mineralization of organic pollutants (dos Santos et al., 2019;Gmurek et al., 2019;Katal et al., 2019).
Photocatalysis may be divided into two categories, ultraviolet (UV) active photocatalysis and solar/visible light active photocatalysis, which has become a preferred choice (Shaniba et al., 2020;Sujatha et al., 2020). Although UV photocatalysis generally has better treatment performance, as the intensity of UV is stronger than solar/visible light's (Sujatha et al., 2020), but due to the economics of solar/visible light photocatalysis, it is preferred over UV light photocatalysis and has been considered an environmentally friendly technology for pollutant removal.
Recently, significant progress has been made on the visible light active photocatalysts. This review summarizes the latest developments in visible light active photocatalysis. It starts with the mechanism of photocatalysis, followed by synthesis and doping methods of several common photocatalysts, and the application of visible light active photocatalysis. Special attention has been devoted to pilot-scale tests and introduced separately. Finally, current research deficiencies and prospects for future research are considered. We believe that this review will not only promote the further developments of T A B L E 1 Examples of recalcitrant pollutants

VISIBLE LIGHT ACTIVE PHOTOCATALYSTS
Visible light active photocatalysis is a type of AOP based on the generation of radicals after photoexcitation of a semiconductor material (Zuniga-Benitez & Penuela, 2020). In general, the visible light active photocatalyst consists of semiconductor materials, lightharvesting antennas, and active species (Dong et al., 2015;Waso et al., 2020;Xu et al., 2020). The mechanism and materials are critical information for photocatalysis research. In this section, the mechanism of photocatalytic function and materials have been reviewed.

Mechanism of photocatalytic function
In general, the mechanism of visible light active photocatalysis is using solar/visible energy to create radical and other active species and then degrade pollutants. The series of actions possibly happened at the visible light active photocatalysts due to light absorption for pollutant degradation, which has been intensively reported in much literature (Ali et al., 2019;Bibova et al., 2019;Chong et al., 2010;Dong et al., 2015;Fujishima et al., 2008;Pena et al., 2005) and it summarized as follows: h þ þ OH À ! ÁOH ð3Þ Pollutant þ ÁOH, h þ , ÁOOH, or O 2 If visible light energy (hν, ν is light's frequency, h is called Planck's constant and equal to 6.62608 Â 10 À34 Js) absorbed by the photocatalyst is stronger than its band T A B L E 2 AOP technology and associated advantages and disadvantages

AOP technology
Advantages Disadvantages Reference

Electrochemical oxidation
No chemical compounds created Versatility Scalability Consumes electricity (Krzeminska et al., 2015) ( Garcia-Segura et al., 2018) Sonolysis Low interference from the water matrix Less heat transfer Highly energy intensive Low electrical efficiency (Pang et al., 2011) Microwave Enhance reaction rates reduce selective heating Low electrical efficiency Required cooling devices (Pang et al., 2011) (Miklos et al., 2018) Fenton The high contaminants removal efficiency Chemical consumption (Babuponnusami & Muthukumar, 2012) (Krzeminska et al., 2015) (Bokare & Choi, 2014) Fenton-like oxidation process Low cost Each non-ferrous catalyst has its merits and demerits Limited application (Bokare & Choi, 2014) (Garrido-Ramirez et al., 2010) Ozonization High degradation and mineralization efficiency Not effective for recalcitrant organics (Malik et al., 2020) gap energy (E g ), valence band (VB) electrons (e À ) will be excited to the conduction band (CB) and leave behind photogenerated holes (h + ) at the VB (Equation 1) (Dong et al., 2015;Jung et al., 2020). Then, the produced e À /h + will migrate to the surface and participate in a range of redox reactions shown above (Fujishima et al., 2008;Perovi c et al., 2020). Dong et al. (2015) summarized the three main active species for photocatalytic: h + , hydroxyl radical (ÁOH), and superoxide radical (ÁO 2 À ), where ÁOH has been considered the primary oxidant with scavenging properties (Chong et al., 2010). ÁOH is generated by three routes: (1)  ) has a negative impact on photocatalysis (Fujishima et al., 2008). The photocatalytic reaction mechanism is shown in Figure 1. The mechanism of modified photocatalyst composites, except for the reactions above, usually includes the transfer of e À and h + between different photocatalysts, due to the different band potential (Geng et al., 2019;B. S. Li, Lai, et al., 2019). A good design of composite will lead to optimized parameters (like broadened energy band), efficient separation of photogenerated electron-hole pairs, and charge carrier recombination hindrance (Meena et al., 2020).
To capture solar energy effectively and achieve the above reactions, the semiconductor material for photocatalytic must have several critical properties. First, semiconductor materials should have a narrow band gap to absorb visible light effectively (Gopalakrishnan et al., 2020); the band gap determines the wavelength of light that materials can be absorbed (Casbeer et al., 2012). Secondly, the semiconductor materials should have a low recombination rate of h + -e À pairs to ensure h + and e À migrate to the surface (Shaniba et al., 2020). Moreover, the semiconductor material should have good chemical and structural stability with the ability to operate under normal temperature and pressure (Chong et al., 2010). Other preferable properties include low cost, good stability, non-toxicity, desired environmental acceptance, and easy to be separated . The choice of semiconductor material is crucial for photocatalysis.
Overall, the design of a good photocatalyst involves two key factors: (1) suitable semiconductor materials, preferable characteristics including good chemical and structural stability, solar energy absorption properties, and separation efficiency of charge carriers; and (2) favorable surface areas structure design of the photocatalyst (Berger et al., 2020); generally, the higher surface areas will lead to improved solar absorption and higher reaction rate but may leave traces as residue in water (Sharmila et al., 2020). Two common methods to fabricate better photocatalysts are modifying known materials and developing new semiconductor materials.
F I G U R E 1 Photocatalytic reaction mechanism

Photocatalysis materials
Photocatalysts can be divided into two categories which include heterogeneous and homogeneous catalysts. Heterogeneous photocatalyst is in a different phase from the reactants (e.g., TiO 2 for water treatment) (Ameta et al., 2018). On the contrary, a homogeneous photocatalyst is an assembly of soluble molecular catalysts in the same phase as the reactants, and all reactions are carried out in a single form (e.g., photo-Fenton [PF] reagent for water treatment) (Shwetharani & Balakrishna, 2014;Tahir et al., 2020). Furthermore, heterogeneous catalysts may be divided into metallic compounds including carbon-based compounds and metal-carbon composites.
The metallic compound is very popular as a photocatalyst material. Generally, titanium dioxide (TiO 2 ) and zinc oxide (ZnO) are the most common materials for photocatalysis, which have been widely used in the past few decades, due to their chemical stability, non-toxic nature, abundance, and cost-effectiveness (Priyanka et al., 2020;Shaniba et al., 2020;Xu et al., 2020;Yazdanbakhsh et al., 2019). Moreover, they can be easily prepared by simple and scalable approaches (Serra, Zhang, et al., 2019;Wang et al., 2014). However, TiO 2 can only absorb UV light, which is only approximately 3%-5% of the entire light emitted across the solar spectrum (wavelengths shorter than 400 nm) because of its large band gap energy (3-3.2 eV) and fast charge recombination of h + and e À before reaching the surface. In addition, TiO 2 powder is hard to separate from the reaction system. These defects lead to poor photocatalytic efficiency (Jung et al., 2020;Kavil et al., 2020;Shaniba et al., 2020;Weon et al., 2019;Xu et al., 2020). ZnO also has similar drawbacks; it has a wide band gap energy (3.3 eV), high photocorrosion activity, high electron mobility (200-300 cm 2 /V/s), low quantum efficiency, high h + and e À recombination rate, long electron lifetime (>10 s) (Ani et al., 2018;Hu et al., 2019;, and potential ecotoxicological effects (Serra et al., 2020). In addition, in this particular case, ZnO can cause ecotoxicological effects, especially relevant for some microorganisms, such as microalgae (Djearamane et al., 2018;Shahid et al., 2020).
Through hydrogenation of TiO 2 , it is possible to generate a black TiO 2 solution. Black TiO 2 has recently been developed, first reported in 2011 (X. Chen et al., 2011), and has triggered much research interest in recent years. It displays excellent red-shifted light absorption properties, mainly owing to hydrogenation dramatically changing the structural, chemical, electronic, and optical properties of TiO 2 nanoparticles (NPs), so it has a stronger visible light utilization ability (J. U. Choi et al., 2019;. However, according to the latest review, summarized by Rajaraman et al. (2020), the synthesis of black TiO 2 is quite challenging and, in some cases, compared with pristine TiO 2 , the black TiO 2 actually showed reduced photoactivity.
In recent years, metal-free carbon-based materials have provoked much research interest. Some carbonbased materials such as graphene, graphitic carbon nitride (g-C 3 N 4 ), and graphene oxide (GO) are nonmetal semiconductors and have been studied as cocatalysts due to their eco-friendly composition, lightweight structure, high surface area, tunable band gap stability, and cost effectual process of preparation (Priyanka et al., 2020;Zheng et al., 2019). For instance, g-C 3 N 4 has 1.8-2.7 eV band gaps that allow the harvesting of visible light up to 460-698 nm (Zheng et al., 2016). Such metal-free carbon-based materials can combine to enhance photocatalytic ability, examples including g-C 3 N 4 /GO aerogel (Tong et al., 2015) and g-C 3 N 4 -agar (Tan et al., 2019). The carbon-based material also can combine with metallic compounds, such as sulfur-doped graphene oxide (sGO)/Ag 3 VO 4 (Priyanka et al., 2020) and Ag 3 PO 4 /polyaniline@g-C 3 N 4 . Some cutting-edge carbon-based materials also have been invested as photocatalysts, examples including carbon quantum dots (CQDs) (D. Choi et al., 2018;Rahbar et al., 2019), graphene quantum dots (GQDs) (Ge et al., 2016;Wei et al., 2016), and multiwalled carbon nanotubes (MWCNTs) (Cong et al., 2011;Yan et al., 2011). However, these materials only have been discovered in the last several decades, and the research on their green synthesis and mechanism is still at an early stage (Heng et al., 2021;Kaur et al., 2018) There is less research on the homogenous catalyst, compared to heterogeneous catalysts. The homogenous catalyst usually combines light radiation with chemical oxidizing agents (Stan et al., 2012). PF is one of the most common homogenous catalysts, which usually consists of iron ion (Fe 2+/3+ ) and hydrogen peroxide (H 2 O 2 ) (Moncayo-Lasso et al., 2009;Shwetharani & Balakrishna, 2014).

METHODS OF SYNTHESIS
The synthesis of NP semiconductor materials is a major challenge in photocatalysis research. Synthesis methods have a significant impact on photocatalysts' morphology, structure, and performance (Sahu & Biswas, 2011;Sun et al., 2019;Taherinia et al., 2019).

Sol-gel method
The sol-gel method has a large scope of application in the preparation of inorganic ceramic and glass materials (Czok & Golonka, 2016;Jones, 2013;Wang et al., 2014). Due to its wide range of advantages, including high chemical purity, good uniformity, and controllable morphology (Taherinia et al., 2019;Wetchakun et al., 2012), it has become a common method for photocatalysts synthesis. It is important to mention that a change in molar ratios may lead to different hydrolysis speeds and further different structures and properties (Lu et al., 2013;Wang et al., 2014;You et al., 2012).
A typical example of the sol-gel method is the synthesis of TiO 2 . In this method, TiO 2 is synthesized by alkoxide precursors, like titanium butoxide (Ti(OBu) 4 ) and titanium isopropoxide (TTIP). Initially, the sol-gel method was a one-step process (Long & Yang, 2002), and now, it usually contains two steps. First, dissolve the precursor with/without doping materials into a solvent, usually alcoholic (i.e., ethanol) or acid (i.e., HNO 3 ). Due to hydrolysis, the sol is formed, and under polycondensation and action of density, the sol particles form a three-dimensional network, resulting in the production of gel. Sonication could be applied in this step to enhance the reaction. Second, the gel is calcined in a muffle furnace at 400-700 C and, after this, it is dried and powdered (Kavil et al., 2020;Khataee et al., 2017;Taherinia et al., 2019;Villaluz et al., 2019). A schematic of this method is shown in Figure 2. Wang et al. (2014) have summarized the reaction mechanism as follows: By adjusting the precursor and solvent, sol-gel can also be used to synthesize other photocatalysts, examples including ZnO (Rafie et al., 2019), Bi 2 O 3 , Al 80 Ce 10 Zr 10 (Perez-Osorio et al., 2020), and Ag/g-C 3 N 4 /V 2 O 5 (El-Sheshtawy et al., 2019).
F I G U R E 2 A typical sol-gel method for TiO 2 Hydrothermal method Hydrothermal technology is an important liquid-phase preparation technology, which has extensive application in materials science, chemistry, physics, biology, metallurgy, and earth science (Yang & Park, 2019). It has been applied to produce nano TiO 2 with high purity, crystal symmetry, metastable compounds with unique properties, and narrow particle size distributions (Byrappa & Adschiri, 2007) and is also preferred for ZnO synthesis (Danwittayakul et al., 2020). Concurrently, it has been widely used in photocatalyst synthesis, especially with carbon dopants (i.e., GO and activated carbon) (Shinde et al., 2018;Subramani et al., 2007;Yin et al., 2021;Zhang et al., 2017).
The hydrothermal method is a one-step process. A typical example of its application is the synthesis of TiO 2 . By mixing a titanium precursor with water or NaOH/isopropanol/2-propoanol(peptizer)-water solution, or alternatively TTIP with the acidic ethanol-water environment, with/without doping materials, add them into autoclaves for 16-72 h at 110-180 C. The precipitate is then calcined and dried to gain well-defined TiO 2 or composite TiO 2 photocatalyst (X. Chen & Mao, 2007;Shinde et al., 2018) The hydrothermal method can synthesize not only TiO 2 and ZnO but also many other photocatalysts, including MoS 2 /reduced graphene oxide (rGO) (Gopalakrishnan et al., 2020), Mn 2 O 3 , ZnWO 4 (Gong et al., 2020), FeCo 2 O 4 , and ZnS (Sabaghi et al., 2018).

Miscellaneous methods
The sol-gel method, hydrothermal method, and solvothermal method are popular in the current study. Along with these commonly used synthesis methods, researchers have developed some additional methods and techniques.
The template method entails the synthesis of monodisperse tubular or wire-like nanostructures within the pores of a membrane or nanoporous solid . For example, Nozaki et al. (2018) used commercial silica microspheres as template material, using cerium(III) nitrate hexahydrate as a precursor to synthesize CeO 2 .
The flame aerosol method is a single-step process and allows independent control of the material properties such as particle size, crystallinity, homogeneity, and degree of aggregation (Sahu & Biswas, 2011). In general, by burning precursors with fuel and oxygen, the photocatalyst will grow in the entrained air. Applications of this method include materials such as Cu-TiO 2 (Sahu & Biswas, 2011), C-TiO 2 (Lim et al., 2008), and Cr-TiO 2 (Inturi et al., 2014).
The sonochemical method has triggered much attention in advanced chemical synthesis as well as photocatalyst preparation. Ultrasonic generates alternating rarefaction and compression zones in this liquid, which leads to extremely high local energy densities . Briefly, precursors and capping agents reacted under ultrasonic and synthesized photocatalysts (Zinatloo-Ajabshir et al., 2020). Examples of reactive materials developed through this technique include rGO-V 2 O 5 NPs (Mishra et al., 2020), silver tungstate (Zinatloo-Ajabshir et al., 2020), and graphene oxide@ZnO (Muthukrishnaraj et al., 2020).
The physical vapor deposition (PVD) method shares a similar application with the CVD method. It deposits films or forms coatings that are transported through a vacuum or low-pressure gaseous/plasma environment , such as coating TiO 2 (Shuang et al., 2016).
The electrodeposition method is a type of plating by depositing material on a substrate through electrochemical reduction (Hasan et al., 2019) and has been applied in photocatalyst preparation. Reactive materials developed using this method include Au/ZnO film (da Silva et al., 2013), TiO 2 film (Hachisu et al., 2016), and TiO 2 on a silicon wafer surface (Basheer et al., 2020).
Apart from these physical and chemical methods above, the biosynthesis process also has been developed. Ahmed et al. (2020) used A. Chinensis fruits extract with FeCl 2 Á4H 2 O to prepare Fe NPs. Similarly, Ghazal et al. (2020) synthesize NiO NPs through the use of Cydonia oblonga extract.

Composite photocatalyst synthesis
Specific methods have been developed to synthesize composite photocatalyst. Many researchers do not synthesize common photocatalysts like TiO 2 and ZnO in the laboratory but instead purchase them with specific specifications of surface area and particle size (Davididou et al., 2018;Jung et al., 2020;Sujatha et al., 2020;Vela et al., 2018c;Zuniga-Benitez & Penuela, 2020). Methods introduced in previous sections may also apply to composite photocatalyst synthesis, but some unique methods are introduced below.
The chemical bath deposition (CBD) method is used for film or array photocatalyst preparation. Briefly, metal slides/flakes are placed in a chemical bath and photocatalytic deposition is encouraged forming a photocatalyst film or array. Specifically, the slides/flakes can be ordinary supporting inorganic material, like glass slides, or semiconductor material that involved photocatalysis, like TiO 2 nanoflakes (Chai et al., 2014;Kite et al., 2020;Zhou et al., 2019). Examples include WO 3 /TiO 3 array , Zn 0.2 Cd 0.8 S films (Chai et al., 2014), ZnS thin films (Y. Chen et al., 2012), and FeSe thin films (Sohrabi & Ghobadi, 2019).

WATER CONTAMINANT TREATMENT
Textile and organic wastewater treatment Textile wastewater often contains highly toxic and chemically stable organic compounds, including dyes, surfactants, oils, acid, alkali, solvents, and some metal salts, which are graded as one of the foremost pollutants in all industrial sectors Kavil et al., 2020;Yusuff et al., 2020). Globally, it has been estimated that 280,000 tons of textile dyes alone are produced every year (Perez-Osorio et al., 2020). Without appropriate treatment, before discharge, it has the potential to cause serious environmental contamination for receiving aquatic ecosystems.
TiO 2 composites TiO 2 composites are a common form of photocatalyst, which have been widely investigated for textile and organic wastewater treatment. The commercial TiO 2 -P25 already shows strong photocatalytic performance. Adamek et al. (2019) immobilized TiO 2 -P25 on a glass fiber mat under sunlight and achieved decolorization of an anionic dye (Acid Orange 7; AO 7) within 20 min in comparison with commercial TiO 2 -P25 alone.
The main purpose of modified TiO 2 for visible light active photocatalysis is the enhancement of the visible light absorption capacity of the TiO 2 composite. Lim et al. (2019) combined nanocubic-like TiO 2 with N-doped graphene quantum dots (N-GQDs) and proved N-GQDs can provide light-harvesting ability, especially in visible light and near-infrared region, which completely removed BPA after 30 min. Similar results were also reported by Kavil et al. (2020) whereby a comparison of methylene blue (MB) removal was investigated from seawater against more conventional TiO 2 , C/TiO 2 , and Cu-C/TiO 2 catalysts. The Cu-C/TiO 2 had the best performance, which achieved maximum efficiency removing 100% MB in 45 min. Xu et al. (2020) prepared porous polymers/TiO 2 /Cu (PPTC) through a two-step method and proved it performed better than TiO 2 and TiO 2 /Cu. Feng et al. (2019) synthesized Z-scheme Mn-CdS/MoS 2 / TiO 2 ternary photocatalyst, proving it had 3.16 times better performance than TiO 2 alone when treating methyl orange (MO) and 9-anthracenecarboxylic acids (9-AC). It degraded N,N-dimethylformamide (DMF), MO, MB, and phenol of 73.7%, 97.8%, 100%, and 98.6%, respectively, under 3 h of xenon light exposure. Gmurek et al. (2019)  ) and reported their performance under sunlight is better than under UVA. Among these TiO 2 composites, TiO 2 -Pd had the best performance, which totally degraded paraben in 120 min. Other composites' performance under sunlight and UV is shown in Table 3.
To modify TiO 2 , many studies have applied metal materials. Scott et al. (2019) coated ultrathin MgO overlayer on Ag/TiO 2 nanorods to treat phyenol and reported the thickness of MgO coating has a significant impact on photocatalysis. Under optimum conditions, the degradation efficiency is up to 95% in 120 min. Ghanbari et al. (2019) prepared a new N-F-codoped TiO 2 /SiO 2 nanocomposite (NFTS) via the sol-gel method to treat a mix of three azo dyes (Basic red 29, Basic blue 41, and Basic yellow 51). This achieved a 58.5% decrease in TOC under solar irradiation, in addition to reducing 100% of Cr(VI). Yuvaraj M. Hunge (2017) synthesized WO 3 /TiO 2 thin films through spray pyrolysis and the sonochemical method, which degraded 66% benzoic acid (BA) after 320 min.
Carbon-based material also has been applied for TiO 2 photocatalysts. Das and Mahalingam (2019) immobilized TiO 2 , rGO, and g-C 3 N 4 in a polystyrene film under sunlight for the removal of Remazol Turquoise Blue (RTB), which achieved 60% decolorization and 51.43% degradation after 90 min. Also with g-C 3 N 4 , J. U. Choi (Shahid et al., 2020). Recently, Bora et al. (2022) synthesized Ag 2 CO 3 /ZnO via a simple precipitation route and reported it can achieve a totally TOC removal for MB wastewater in 30 min and 32% disinfection of Escherichia coli in an hour.
Along with the photocatalytic material, the structure of the photocatalytic material has a significant impact on photocatalysis performance. Serra, Zhang, et al. (2019) synthesized ZnO-based biomimetic fern-like microleaves and found ZnO@ZnS micro/nanoferns had the best photo-remediation performance for persistent organic pollutants compared with ZnO, Ag-ZnO, and Ni-ZnO, which increased over sixfold for pollutant degradation rate capacity compared with pristine ZnO catalyst. The following research using the same bioinspired ZnO@ZnS photocatalyst achieved nearly 97% of MB after 60 min with mineralization of >98% of a mixture of MB, 4-nitrophenol (4-NP), and RhB after 210 min and the removal of nearly 65% of Cr(VI) after 180 min . Recently, Bora et al. (2022) used a waste material-ground granulated blast furnace slag (GGBFS) as a low-cost geopolymer to hybridize with ZnO. They found that, due to the increased surface area, the discoloration efficiency of textile wastewater is twice better than normal ZnO (Table 4) Similarly, an all-day-active photocatalyst was synthesized by employing Ag@AgI NPs decorated with Ag 3 PO 4 cubes (C-Ag 3 PO 4 @Ag@AgI) designed by Cai et al. (2019), which completely degraded RhB and removed 80% of BPA in 80 min under sunlight. They also reported that this photocatalyst can maintain photocatalytic activity even on a cloudy day. Bora and Mewada (2017) synthesized Ag 2 CO 3 /SiC through a simple precipitation route and found that SiC improves Ag 2 CO 3 photoactivity by inducing a charge transfer between SiC and Ag 2 CO 3 mimicking the Z-scheme in photosynthesis, which removes 98% of MB in 4 h. X. Han et al. (2020) 2019) used a microwave solvothermal method to apply BiOCl with oxygen vacancies and its removal efficiency was 2.7 and 33.8 times higher than that of BiOCl fabricated by the conventional solvothermal and precipitation methods (Song et al., 2017). In addition, synthesis methods can also affect photocatalysts' performance. Y. F. Li, Zhong, et al. (2018) compared Bi 2 Ti 2 O 7 photocatalytic ability prepared by co-precipitation and solvothermal method and found that the Bi 2 Ti 2 O 7 photocatalyst via co-precipitation method had better performance (240 min, 92.8% removal). Bismuth-based catalysts can also be modified with other compounds. T. Liu et al. (2014) compared a pure Bi 2 Mo 3 O 12 sample with Bi 2 Mo 3 O 12 /MoO 3 , due to the excellent adsorption behavior of MoO 3 , Bi 2 Mo 3 O 12 /MoO 3 showed an obviously enhanced photocatalytic activity, which removes 92% of MB.
Carbon-based materials also have a variety of applications in visible light active photocatalysis.  synthesized sulfuric acid-treated graphitic carbon nitride (SA-g-C 3 N 4 ) by thermal polymerization and chemical exfoliation methods embedded within a porous cellulose network (CN/CA film) (T. Li et al., 2013;Xu et al., 2013), which removed $99% RhB in 150 min and reduced 95% of Cr(VI) in 100 min. Mishra et al. (2020) used RGO-V 2 O 5 nanocomposite, which degraded 71% of MB in 20 min. A complete MB degradation was achieved by a hybrid of Zr-based metal-organic framework (UiO-66) with graphitic carbon nitride (g-C 3 N 4 ) nanosheets (UiO-66/g-C 3 N 4 sheets) photocatalysts within 240 min . A similar result from V 2 O 5 /S-g-C 3 N 4 was achieved by Chegeni et al. (2019) for MB and phenol treatment, which achieved a 99% and 89% removal rate within 60 min, respectively ( Figure 3). Also, El-Sheshtawy et al. (2019) immobilized Ag on g-C 3 N 4 /V 2 O 5 F I G U R E 3 Schematic of the photocatalyst g-C 3 N 4 (Chegeni et al., 2019) surface to enhance its photocatalytic activity, which can totally reduce p-nitrophenol (NP) within 8 min under sunlight. They also reported that this photocatalyst only needs 60 min to totally reduce 4-NP and 4-AP in dark. Priyanka et al. (2020) synthesized sulfur-doped graphene oxide (sGO/Ag 3 VO 4 /Ag) through Hummers's method (Hummers & Offeman, 1958) with a one-pot method that can degrade above 99% cationic dyes, 75%-80% anionic dyes, and 90% organic carbon in 1 h under sunlight. Sharma et al. (2019) used activated carbon-supported strontium/cerium bimetallic nanocomposite (Sr/Ce/AC BNC) to degrade RhB and reach 91% total degradation in 120 min. The wastewater mixture of MB and RhB was treated by MoS 2 /rGO/Cu 2 O grown on etched carbon paper, and 95% of MB and RhB are degraded in 45 min (Gopalakrishnan et al., 2020). The complete removal of MB was achieved by flexible graphene composites (FGCs) with Al 2 O 3 :Eu 3+ and SrAl 2 O 4 : Bi 3+ catalysts, respectively, after 180 and 270 min (Oliva et al., 2018).
Apart from the heterogeneous photocatalyst covered above, Sharmila et al. (2020) prepared mixed Spirulina platensis cultivated water (Spcw) as a homogeneous photocatalyst motivated by its benign chemical composition and eco-friendly properties. This catalyst was shown to remove a mixture of 5 ppm of MB, 70 ppm of malachite green (MG), and 6 ppm of congo red (CR) within 3 h (Table 5).

Pharmaceutical wastewater
Pharmaceutical wastewater typically contains persistent organic pollutants that contain a high concentration of organic matter, microbial toxicity, and high salt content, which is difficult to biodegrade, and municipal water resource recovery facility cannot treat pharmaceutical wastewater effectively (Ahmad et al., 2017;Guo et al., 2017;Keshvadi et al., 2020). Pharmaceuticals and their metabolites, even at low concentrations, have potentially fatal effects on natural ecosystems and human health Keshvadi et al., 2020). For example, the occurrence of antibiotics in an aquatic environment can develop antibiotic resistance (Ben et al., 2019;Walsh, 2013;Wang, Chen, et al., 2020). In addition, it has been shown that 58%-68% of consumed common mild analgesic medicine acetaminophen (ACE) is released into the environment, which may transform to N-acetyl-p benzoquinone-imine causing protein denaturation, lipid peroxidation, and DNA damage to organisms (Behravesh et al., 2020;Tobajas et al., 2017).
TiO 2 composites are also common for pharmaceutical wastewater treatment. Shaniba et al. (2020) synthesized TiO 2 /nitrogen-doped holey graphene (TiO 2 /NHG) nanocomposite via hydrothermal and calcination methods to treat antibiotic cefixime and achieve complete mineralization in 25 mg/L concentration within 90 min. Using similar catalysts involving nitrogen-doped TiO 2 (N-TiO 2 ), Keshvadi et al. (2020)  TiO 2 also can combine with carbon-based material to treat pharmaceutical wastewater. Cunha et al. (2019) synthesize TiO 2 /AC to treat benzodiazepine drugs via the impregnation method and reported that TiO 2 /AC10% (w/ w) had the best removal efficiency of >97.5% within 60 min. Shinde et al. (2018) synthesized Pt-rGO-TiO 2 to treat pharmaceutical pollutant β blocker Propranolol and showed a 20-fold increase in removal (COD removal rate: 94%) under simulated solar light compared with TiO 2 alone under UV exposure. Notably, the contaminant degradative performance of TiO 2 composites may not always be better than a simple conventional TiO 2 catalyst. Palma et al. (2020) treated antibiotic chloramphenicol (CAP) and paracetamol (N-(4-hydroxyphenyl)acetamide) by TiO 2 and PbS/TiO 2 . The results showed that TiO 2 alone can remove 5% more CAP in 240 min than the PbS/TiO 2 composite and can completely remove paracetamol in 235 min whereas PbS/TiO 2 only removes 93% in 240 min. Moreover, Behravesh et al. (2020) compared the removal ability between zeolite-supported TiO 2 (TiO 2 -Z) and ZnO (ZnO-Z) to degrade ACE and codeine and reported that ZnO-Z has better removal efficiency.
Other metallate photocatalysts also have been invested to remove organic contaminants. To remove the antibiotic tetracycline hydrochloride (TC), Z. J. Liu et al. (2019) Domingues et al. (2020) used red mud considered waste from the metal production industry to treat the organic waste compounds of CBZ, LRZ, and SMX and achieved 58%, 62%, and 51% removal efficiency, respectively (Table 6).

Disinfection
Worldwide, it is estimated that 844 million people still do not have access to basic drinking water services, and 159 million people in rural areas use untreated drinking water, which potentially may expose them to health risks from contaminated water sources (World Health Organization & Unicef, 2017;Yan et al., 2020). For example, 4 billion cases of diarrhea each year are caused by inadequate hygiene and sanitation drinking water (Danwittayakul et al., 2020;World Health Organization & Unicef, 2015). Compared with traditional drinking water disinfection methods, like adsorption, coagulation, chemical, and physical disinfection, solar disinfection (SODIS) is much cheaper and easier to access in most areas, and it can be environmentally favorable to inactivate microorganisms in natural surface waters and drinking water (Garcia-Gil et al., 2020;Malato et al., 2009;Rodriguez-Chueca et al., 2019;Zeng et al., 2020). In recent years, photocatalytic disinfection has sparked much attention and established a trend to replace traditional solar disinfection (Djellabi et al., 2020;Vivar et al., 2020;Yan et al., 2020). Just like the mechanism of photocatalytic degradation of pollutants, photocatalysts can produce hydroxyl radical (ÁOH), h + , and superoxide radical (ÁO 2 À ), which also can deactivate pathogens' destroying cell membrane structure to achieve disinfection (Ge et al., 2016;Zeng et al., 2020). TiO 2 composite catalysts are very common for water disinfection applications. Yan et al. (2020) used an anodic oxidation method to fabricate MoS 2 /TiO 2 nanotube arrays and investigated its inactivation ability by treating   (2015) developed TiO 2 -rGO via a sonochemical method. Importantly, they proved that the concentration of TiO 2 -rGO may not be proportional to disinfection efficiency. At 500 mg/L of TiO 2 , only 10 min of solar treatment can completely inactivate E. coli. On the contrary, the best F. solani inactivation efficiency was observed for 10 mg/L, requiring 30 min of treatment for complete inactivation. Waso et al. (2020) also used the same material by the same preparation method to treat rainwater and removed all Klebsiella pneumoniae (from 2.00 Â 109 CFU/ml to below the detection limit [BDL]) with 120 min of natural sunlight exposure after pretreatment by Bdellovibrio bacteriovorus. ZnO composites have also been extensively investigated. Danwittayakul et al. (2020) synthesized ZnO nanorods on cellulose and polyester substrates, which achieved nearly total disinfection within 15 min. A similar result was achieved by Yadav et al. (2019), where ZnO NPs were applied to enhance solar disinfection for fecal coliforms with compound parabolic concentrators. The result showed that there was complete inactivation within 15 min, which was a 50% improvement without using ZnO (Table 7).

Photocatalyst
To reduce Cr(VI), Zhang et al. (2019) synthesized Iron(III)-alginate (Fe-SA) hydrogel beads, which reduced Cr(VI) up to 100% within 150 min. Nitrogen-phosphorus-doped fluorescent carbon dots (NP-CD) show high efficiency for Cr(VI) in a linear range from 10 ppm (in approximately 10 min) to 2000 ppm (in approximately 320 min) (Bhati et al., 2019). To treat nitrate (NO 3 À -N) and ammonia (NH 4 + -N), Zhao et al. (2018) synthesized 3D/2D Mn 2 O 3 /g-C 3 N 4 and achieved high removal efficiency of 94.5% and 97.4% for NO 3 À -N and NH 4 + -N (  [2018][2019][2020][2021][2022]. In this section, this research that has been successfully tested on a pilot scale will be introduced. Compound Parabolic Collector (CPC) (Figure 4) is widely applied in many photocatalyst pilotscale tests. Other equipment like falling film photoreactor (FFR) (Figure 5), tubular reactor, and self-made reactor are also applied. Majority of pilot tests carried out in the last 5 years used TiO 2 with its composites as the photocatalyst. Just using the commercial TiO 2 with CPC, Zuniga-Benitez and Penuela (2020) tested their photodegradation ability for Benzophenone-1 and Benzophenone-2 (BP1 and BP2). The pilot scale was tested in the city of Medellín, Colombia, using a solar compound cylinder-parabolic collector; greater than 90% of BP have been removed after 6 h of reaction (between 10:00 and 16:00 h). Using TiO 2 with CPC for carbapenem antibiotics imipenem and meropenem degradation, they can remove 75% of imipenem after 60 min and 75% of meropenem after 45 min (Cabrera-Reina et al., 2019). For degradation of diphenhydramine hydrochloride (DPH), Lopez et al. (2018) achieved 14,889 mg TOC/kWh demineralization efficiency but under black blue lamps has better efficiency (21,141 mg TOC/kWh). On the contrary, also using TiO 2 -P25, Haranaka-Funai et al. (2017) reported that the CPC has the best removal performance for valproic acid sodium salt (VA) treatment (91% VAC removal efficiency and 50% TOC degradation rate) compared with three artificial irradiation (black light blue lamps, UVC, and Xe lamp). Also applied CPC and TiO 2 -P25, Grilla et al. (2019) found that the TMP degradation is proportional to SPS concentration and inversely proportional to water matrix complexity. Diaz-Angulo et al. (2020) also used them to remove DCF; by applying MR as the photo-sensitizer,  (Chavez et al., 2020) (Continues) the mineralization rate can be increased up to 65%, which removes 99% of DCF and 99% of MB. Also applied CPC, to treat pesticides in wastewater, Vela et al. (2018b) used two commercial TiO 2 nanopowders (Degussa P25 and Kronos vlp 7000) for degradation of a mixture of six pesticides (malathion, fenotrothion, quinalphos, vinclozoline, dimethoate, and fenarimol). After optimized operational conditions under laboratory conditions (Vela et al., 2015), they set up a polite scale test at Murcia, SE Spain (3000 h sun per year). The result showed that the use of TiO 2 alongside an electron acceptor like Na 2 S 2 O 8 can strongly enhance the degradation rate and the best result was achieved by TiO 2 Kronos vlp 7000 with Na 2 S 2 O 8 , which achieved 90% degradation (DT 90 ) in 32 min. These photocatalysts also have been tested to treat endocrine disruptors (EDs) at a pilot plant scale. Different from pesticides test results, TiO 2 -P25/ Na 2 S 2 O 8 has the best performance, which achieved a 3 J/cm 2 half-fluence (H 50 ). They also found that TiO 2 (mainly P25) in tandem with Na 2 S 2 O 8 can avoid recombination of e À /h + pairs (Vela et al., 2018c). Except for TiO 2 , they also used ZnO/Na 2 S 2 O 8 at the pilot plant scale. It can achieve an 83% removal rate for dissolved organic carbon (DOC) for EDs removal (Vela et al., 2018a) and min DT 90 for several fungicides and insecticides removal . Some similar pilot-scale tests using TiO 2 /ZnO with Na 2 S 2 O 8 also have been completed Kushniarou et al., 2019).
TiO 2 composites have also been invested under pilot tests. For carbon-based TiO 2 composites, Luna-Sanguino et al. (2020) synthesized two TiO 2 -rGO from commercial TiO 2 (P25 and Hombikat UV100, HBK) for several pesticides' treatments (methomyl, pyrimethanil, isoproturon, and alachlor) and run a pilot-scale test in a 3.2 m 2 CPC located in the Plataforma Solar de Almería (PSA-CIE-MAT). The TiO 2 -rGO (P25) has a better performance, which removes all pesticides in 210 min. They also reported that the use of H 2 O 2 can speed up the removal. Also, TiO 2 /H 2 O 2 has been proved efficient for Curvularia sp. deactivation by using CPC, which achieved completed disinfection in 300 min (Aguas et al., 2017). Shaban et al. (2016) used carbon-modified titanium oxide (CM-n-TiO 2 ) to treat polychlorinated biphenyls (PCBs) in seawater on a pilot-plant scale by a solar falling film reactor (SFFR), which reached completed degradation after 75 min. Vatanpour et al. (2019) modified TiO 2 by urea for Reactive Orange 29 (RO29) removal and tested it in a continuous pilot-scale submerged photocatalytic membrane reactor (SPMR). They found that modified TiO 2 using urea with a 1:6 ratio (  (Zhao et al., 2018) result showed that UV/H 2 O 2 is the best treatment for dye and TiO 2 achieved a better elimination of coliform bacteria. It is important to mention that TiO 2 showed a detrimental effect on the overall elimination of dyes. Ahmadpour et al. (2020) compared ZnFe 2 O 4 @TiO 2 / Cu nanocomposites removal ability for recalcitrant drug NPX in batch and large-scale systems. In a large-scale experiment, 1000 ml of NPX pollutant solution was treated and reported a 63.14% degradation rate in 120 min, compared with 80.73% for the batch experiment. After optimizing the photocatalyst design in a laboratory-scale experiment, Bibova et al. (2019) designed two pilot-scale tests for SiO 2 /TiO 2 /calcined Liapor. One is in Czech Republic to simulate the remediation of contaminated water in rural areas, and another is in Vietnam for a comparative solar experiment. In the Czech Republic, oxalic acid (OA; 3.3 Â 10 À3 mol/dm 3 ) is chosen as a model compound and achieved an 82.1% TOC removal rate in 3 days. In Vietnam, MB as a model compound is used and achieved around 58% degradation in 6 days. Graywater is a highly reclaimable water source. Saran et al. (2019) (Booshehri et al., 2017) of E. coli in 120 min. Mohsenzadeh et al. (2019) used the composition of polyaniline (PAni) and TiO 2 on glass beads to remove 1,2-dichloroethane (1,2-DCE) in a pilot-scale packed bed recirculating photocatalytic reactor, which achieves complete degradation after 360 min. Saran et al. (2018) tested TiO 2 -Ag for rainwater disinfection using a pilot-scale solar photocatalytic fixed bed tubular reactor and achieved complete disinfection in addition to COD removal within 120 min. Y. Li et al. (2016) fabricated foam concrete incorporated with nitrogen-doped TiO 2 (N-TiO 2 /FC) and tested its removal performance for trichlorfon in a pilot-scale experiment, which reported that 6% N-TiO 2 /FC can remove almost 100% of trichlorfon in 2 weeks. However, the TiO 2 composite is not always better than commercial TiO 2 . Arce-Sarria et al. (2018) modified TiO 3 by WO 3 and tested its photocatalytic performance for antibiotic amoxicillin by CPC, which achieved maximum 64.4% removal rate but still underperformed compared with the commercial TiO 2 -P25. Except for the TiO 2 composite, some other composites also have been invested under pilot-scale tests. For example, Dhatshanamurthi et al. (2017) coated ZnO on fevicol and found that it is more efficient than the catalysts made by commercial ZnO and TiO 2 -P25 catalysts for really dye industry effluent treatment in a pilot-scale FFR, which achieved around 95% COD removal rate in 4 h. Booshehri et al. (2017) synthesized Ag-modified BiVO 4 for E. coli, E. faecalis, and spores of F. solani inactivation and tested its photocatalytic disinfection by CPC, which achieved almost total degradation for E. coli after around 160 min, although its performance is still worse than TiO 2 -P25 (Table 9).

Summary
Currently, only very limited researchers have explored the pilot-scale application of visible light photocatalysis. Although there are many visible light photocatalysts that have been developed and tested in the laboratory, the majority of the pilot-scale test still focused on TiO 2 and its composites.
The most common reactor used in pilot-scale tests is CPC. It is a non-imaging concentrator for solar energy collection (Kalogirou, 2014). Some studies use very simple reactors without proper design; some of them even put the water directly into a giant water vat. These reactors, including CPC, need further optimizations for realworld visible light photocatalysis treatments: (1) the attachment of photocatalyst; (2) the mixture and separation of photocatalyst; (3) energy consumption of the reactor; and (4) durability.

CONCLUSION AND PERSPECTIVES
Visible light active photocatalyst for water treatment has been widely examined due to its unique advantages. In this review, we have reviewed the current state of the art with a focus on materials, application, and pilot-scale investigations. In addition, we have outlined the mechanisms of photocatalysis and reviewed a variety of synthesis, doping, and modification methods.
Significant progress in visible light active photocatalyst has been made during the past number of years. Many composite materials have been designed as well as the development of some new facile synthesis methods. Many studies have proved this technology as a promising advanced oxidation process (AOP) that has extensive application in urban, industrial, agricultural, and pharmaceutical wastewater treatment as well as natural water and drinking water treatment with potential for future further research. Visible light active photocatalyst can remove persistent organic pollutants, in addition to metallic ion reduction and microorganism degradation. Concurrent to water treatment applications, the visible light active photocatalyst is also used in air purification (J. U. Choi et al., 2019;Pichat, 2019), hydrogen production (Lee et al., 2019;Reddy et al., 2020), and chlorine production (Chehade et al., 2020). Notwithstanding, visible light active photocatalyst still faces many challenges that have to be addressed in future and are summarized below: 1. There is a lack of theoretical research on the photocatalyst modification mechanism. It is common knowledge that some widely used photocatalysts (i.e., TiO 2 ) are not suitable for visible light active photocatalysis.
Researchers have tried to modify it by shifting the absorption spectrum, reducing band gap energy, and slowing down the speed of charge recombination. Common methods include optimizing the photocatalyst's structure, doping other materials, and using support materials. However, the lack of a theoretical guide leads to unorganized modification studies, with some of them even having reported modified photocatalysts to be less effective than previous. 2. Research in photocatalysis-enhanced solar disinfection is relatively limited despite millions of people having to use untreated drinking water in parts of the world with extensive solar exposure, and most solar disinfection studies did not apply photocatalyst. The majority of research in photocatalysis enhanced solar disinfection is only applied to common TiO 2 and ZnO photocatalyst composite materials and very few of these have progressed to the pilot-scale phase. 3. Many new visible light photocatalysts have been developed over the last few decades, but few have made it to the pilot-scale application stage. Following laboratory and bench-scale proof of concept, ability, and function, pilot-scale experimental trials should be carried out before eventual application. Only a small number of pilot-scale experimental trials have been documented in the literature with the majority using commercial TiO 2 or other common composites. More pilot-scale experimental trials are needed to expand the currently laboratory-scale results for the full-scale application of this promising sustainable technology.

ACKNOWLEDGMENTS
This research did not receive any specific grant from funding agencies in the public, commercial, or not-forprofit sectors. Open access funding provided by IReL.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.