Organic Degradation Potential of Real Greywater Using TiO2-Based Advanced Oxidation Processes

In keeping with the circular economy approach, reclaiming greywater (GW) is considered a sustainable approach to local reuse of wastewater and a viable option to reduce household demand for freshwater. This study investigated the mineralization of total organic carbon (TOC) in GW using TiO2-based advanced oxidation processes (AOPs) in a custom-built stirred tank reactor. The combinations of H2O2, O3, and immobilized TiO2 under either dark or UVA irradiation conditions were systematically evaluated—namely TiO2/dark, O3/dark (ozonation), H2O2/dark (peroxidation), TiO2/UVA (photocatalysis), O3/UVA (Ozone photolysis), H2O2/UVA (photo-peroxidation), O3/TiO2/dark (catalytic ozonation), O3/TiO2/UVA (photocatalytic ozonation), H2O2/TiO2/dark, H2O2/TiO2/UVA, H2O2/O3/dark (peroxonation), H2O2/O3/UVA (photo-peroxonation), H2O2/O3/TiO2/dark (catalytic peroxonation), and H2O2/O3/TiO2/UVA (photocatalytic peroxonation). It was found that combining different treatment methods with UVA irradiation dramatically enhanced the organic mineralization efficiency. The optimum TiO2 loading in this study was observed to be 0.96 mg/cm2 with the highest TOC removal (54%) achieved using photocatalytic peroxonation under optimal conditions (0.96 mg TiO2/cm, 25 mg O3/min, and 0.7 H2O2/O3 molar ratio). In peroxonation and photo-peroxonation, the optimal H2O2/O3 molar ratio was identified to be a critical efficiency parameter maximizing the production of reactive radical species. Increasing ozone flow rate or H2O2 dosage was observed to cause an efficiency inhibition effect. This lab-based study demonstrates the potential for combined TiO2-AOP treatments to significantly reduce the organic fraction of real GW, offering potential for the development of low-cost systems permitting safe GW reuse.


Introduction
Global water resources are under increasing pressure due to rising demand caused by climate change [1], urbanization, and the growing world population [2]. In order to tackle those challenges, there is a considerable interest in the diversification of water sources, with the use of reclaimed greywater (GW) highlighted as a crucial potential water source [3]. GW is defined the untreated domestic wastewater, excluding toilet, bidet and urinals [4], and, based on some reports, the wastewater generated in the kitchen can also be excluded [3]. Reclaiming GW could meet the needs of up to 75% of household water consumption [5]. In keeping with the circular economy approach, reclaiming GW will not only reduce household demand for freshwater [6], but will also reduce the volume of generated wastewater [7] and the associated energy required for collection/distribution. Given the pressure on freshwater resources in developing countries, and indeed the need to control pollution of water bodies, where around 80% of untreated sewage is discharged to [8], considering options to treat and reclaim GW offer significant promise.
Reclaiming GW is common in villages and rural areas in Jordan [9], where it is used for irrigation without any treatment [10], however the introduction of standards by the Jordan Standards and Metrology Organization (JSMO) in 2013 (JS1776:2013) aims to control GW quality for irrigation and toilet-flushing purposes [11].
Although there is a long list of traditional GW treatment methods, a limited number of studies have considered the use of advanced oxidation processes (AOPs) for GW treatment. AOPs involve the formation of hydroxyl radicals (OH • ) in sufficient quantity to remove pollutants [12][13][14]. TiO 2 photocatalysis has gained significant attention [15][16][17] primarily due to the catalysts' ability to utilize solar and near UV radiation, chemical and photochemical stability in water, wide availability, and low cost [12,18]. In most photocatalytic studies, TiO 2 was utilized in suspended form [15], with limited research investigating the perhaps more scalable immobilized TiO 2 system [12,19].
To enhance AOP degradation efficiency, primarily by boosting the production of reactive oxygen species, several studies have considered combining photocatalysis with ozone [14,[20][21][22] or hydrogen peroxide [23][24][25][26]. However, the application of enhanced photocatalysis for GW treatment is limited and usually focused on laboratory-simulated GW, with few studies comparing different AOPs using the same real GW source. The objective of this study was to systematically examine the suitability of enhanced immobilized TiO 2 based photocatalytic systems to treat real GW, focusing specifically on the removal of the organic fraction.

Greywater Selection and Characteristics
Domestic GW characteristics are greatly affected by the generating source (shower, kitchen sink, wash hand basin), the number, age, lifestyle and income level of residents, and the location of property [27]. In order to reduce the variation in GW characteristics, GW quality from individual houses in different Jordanian cities was monitored for several months, and one house in Irbid was selected based upon the consistency of biological load. The homeowner installed the GW collection system in the house during construction, with the wastewater generated from the washing machine, hand-washing basins and showers collected into a 200 L tank housed outside. The GW collection tank was fitted with a submersible pump connected to the sewer to ensure the tank did not overflow. The average GW retention time in the collection tank was roughly 48 h, before use for irrigation purposes or disposal to the sewer. The GW characteristics in this house during the monitoring period are summarized in Table 1 and compared with Jordan's GW reuse standard JS1776:2013 [11], showing parameters as a function of reuse option. GW analysis was carried at the Royal Scientific Society laboratories according to the international standards [28] within 48 h of collection.

Greywater Collection and Storage
GW was collected manually from the top of the home-owners GW tank in a 20 L polyethylene container, transferred to the laboratory within 1-2 h, and stored at 3-5 • C to minimize any changes that might occur in GW properties before the experiments. Each set of experiments was conducted with a single batch of collected GW, which was manually shaken prior to addition to the reactor.

Materials
All chemicals used were of high purity grade and sourced from Sigma Aldrich and BDH. Potassium hydrogen phthalate, for TOC standards preparation, was supplied by Nacalai Tesque Inc. Titanium dioxide was a mixture (50:50) of Anatase (25 nm particle size, 99.7% trace metals basis) and Rutile

Analytical Methods
To evaluate organic mineralization, total organic carbon (TOC) analysis was preferred over chemical oxygen demand (COD) to avoid H 2 O 2 interference with COD measurement [29]. TOC was analyzed using Shimadzu 5000 TOC/V with solid sample combustion unit SSM-5000A employing regular sensitivity Pt catalyst. The solid samples procedure was used to avoid a reduction in TOC via pre-filtration solids removal. All samples in this study were analyzed in triplicate unless otherwise stated.

TiO 2 Immobilization
Borosilicate glass plates (135 × 135 mm) were cleaned by sonication in hot water with detergent for 30 min and thrice rinsed in distilled water before drying and weighing. A (5% w/v) suspension of TiO 2 in methanol was prepared and sonicated for 30 min before coating the glass substrates. The catalyst suspension was sprayed over the cleaned borosilicate glass plates at a fixed spraying rate using compressed air (Jialile Air Compressor, 0.8 MPa). The films were dried during the spray coating process using an IR lamp (A I.R.R OSRAM infrared lamp, 275 watts). Once coated with the desired loading, plates were annealed at 450 • C for 1 hour in the furnace (Stuart Muffle Furnace, SF7/s model) to achieve cohesion and adhesion of the particles and then allowed to naturally cool to ambient temperature. TiO 2 powder was observed to be uniformly coated across the surface of the glass plate (supplementary materials Figure S1). With respect to mechanical strength, powder was not removed from the surface following use in the STR, but could be removed by low levels of direct mechanical force (such as a scratch with a soft pencil or press/rub of a finger). We have previously reported that such annealed films are highly fractured at micron scale, with the porous nature giving rise to an increase in the surface area available for the reaction [30]. The catalyst loading was determined gravimetrically by considering the weight difference and the coated area. Six representatives coated plates with catalyst loadings of 0.19 to 1.98 mg TiO 2 /cm 2 were used in the experiments. The coated films were characterized by measuring the absorbance over the range of 200-700 nm of the coated films (Lambda 11 UV-vis spectrophotometer; Perkin Elmer, USA) with films examined visually (Eclipse polarizing microscope Nikon, LV100 POL).

Stirred Tank Reactor
Experiments were performed using a custom-built stirred tank reactor (STR) based on a previous study [12], as shown in Figure 1 and specifications reported in Table 2. The STR was made from food-grade stainless steel (Type 304) in the form of an open-top-bottom cylinder ( Figure S1a) with baffles on the side walls. The bottom of the cylinder was fastened to a borosilicate glass plate ( Figure S1b), either coated or uncoated with TiO 2 , to form the reactor chamber. A stainless steel propeller driven at 2000 rpm by variable speed overhead homogenizer motor (JJ-1 electrical stirrer, Zhengzhou Boke Instrument Equipment Co., Ltd.) was used to achieve optimal mixing conditions during experiments. For experiments under illumination, the STR was illuminated from below using two 11W UVA lamps (Actinic BL TL 11W/10, Philips lamp, Holland) or two 11W fluorescent lamps (MASTER PL-S 11W/840/4P 1CT, Philips Lamps, Holland) positioned in a wooden box with an air fan for cooling. Details of the spectral output of the lamps are shown in Figure S2.

Stirred Tank Reactor
Experiments were performed using a custom-built stirred tank reactor (STR) based on a previous study [12], as shown in Figure 1 and specifications reported in Table 2. The STR was made from foodgrade stainless steel (Type 304) in the form of an open-top-bottom cylinder ( Figure S1a) with baffles on the side walls. The bottom of the cylinder was fastened to a borosilicate glass plate ( Figure S1b), either coated or uncoated with TiO2, to form the reactor chamber. A stainless steel propeller driven at 2000 rpm by variable speed overhead homogenizer motor (JJ-1 electrical stirrer, Zhengzhou Boke Instrument Equipment Co., Ltd.) was used to achieve optimal mixing conditions during experiments. For experiments under illumination, the STR was illuminated from below using two 11W UVA lamps (Actinic BL TL 11W/10, Philips lamp, Holland) or two 11W fluorescent lamps (MASTER PL-S 11W/840/4P 1CT, Philips Lamps, Holland) positioned in a wooden box with an air fan for cooling. Details of the spectral output of the lamps are shown in Figure S2.

Experiments Operating Conditions
Depending on the purpose of the experiments, a coated or uncoated borosilicate glass plate was fastened to the bottom of the reactor. For experiments with the TiO 2 , coated borosilicate glass plates with different TiO 2 loadings were used. Air or air with ozone was continuously bubbled into the reactor from an ozone generator (OZ-3G, Ozonefac Ltd., China) that has a constant air flow rate at 5 L/min and a variable ozone flow rate. Hydrogen peroxide was added as a single dosage to the greywater at the beginning of each experiment with different concentrations. All experiments were carried out for three hours, and samples (8 mL) were periodically withdrawn using a syringe for TOC analysis. Table 3 describes the systematic approach to the experimental conditions.

Calculations
TOC percentage removal was calculated as follows Equation (1) where the (0) notation represents the initially measured values before the treatment.
The kinetics of TOC reduction were fitted using the pseudo-first-order rate equation as follows Equation (2)

Control Experiments
In the control experiments, GW was stirred for three hours under different conditions to examine the TOC removal by aeration alone, direct photolysis (UVA with and without aeration), and adsorption of organics onto the TiO 2 coated films in the dark (using catalyst loading 0.19-1.98 mg TiO 2 /cm 2 , with and without aeration). In all controls, the TOC reduction reached a maximum of 3%, with results implemented on the related figures throughout the paper. A 2% TOC reduction was observed in the aeration control, indicating a low fraction of volatile organic compounds in GW samples being collected and transported to the lab. Direct photolysis reduced the sample TOC by 2.3% and 1.3%, with and without aeration, respectively, which was attributed to the low energy of UVA photons and/or low concentration of natural photosensitizers in the GW limiting reactive oxygen species (ROS) production [32]. Whilst direct photolysis may cause different structural and chemical changes to the organic compounds in water, no significant reduction in organic content is commonly reported [33,34]. This trend has been reported with a range of different water sources and specifically for GW, with Chin et al. [33] reporting negligible COD reduction by direct photolysis using a photoreactor fitted with a higher intensity UV source (two 15 W UVC lamps). Similarly, Gulyas et al. [35] reported negligible TOC removal using UVA irradiation of biologically treated GW in the absence of a photocatalyst. The adsorption of organics onto the TiO 2 films, regardless of the catalyst loading, was not observed (<1.0% TOC reduction). Organic adsorption onto the catalytic surface is regarded as one of the critical parameters for efficient catalytic and photocatalytic treatment [36], which could perhaps be a limitation within the STR system.

TiO 2 Photocatalysis
The photocatalytic efficiency as a function of catalyst loading (0.19 to 1.98 mg TiO 2 /cm 2 ) was examined under UVA illumination with aeration at a flowrate of 5 L/min ( Figure 2). For all catalyst loadings, mineralization was observed to increase as a function of treatment time; no plateau was reached. The photocatalytic degradation kinetics followed a pseudo-first-order model [37] with good correlation between ln (TOC/TOC 0 ) and time (R 2 ≥ 0.95). The maximum percentage TOC removal and apparent pseudo-first kinetic rate constants (k) for photocatalytic treatment using different catalyst loadings are depicted in Table 4. An optimal catalyst loading, at which the maximum TOC reduction (≈31%) and the highest apparent rate constant (22.10 × 10 −4 min −1 ) was achieved, was 0.96 mg TiO 2 /cm 2 ( Figure 2 and Table 4). Above the optimal catalyst loading, the TOC removal reduced, as is commonly reported, due to screening of the photons by the increased immobilized catalyst mass [16,19], or the presence of unirradiated layers of photocatalyst [38].  Figure S3).  [39]. The photocatalysis mechanism is widely explained in the literature [40], but, in summary, when a semiconductor such as titanium dioxide is illuminated by a light of energy (ℎ ) greater than or equal to the band gap energy ( ) electrons are excited to the conduction band ( ), leaving a positive hole in valance band ( ) (Equation 5) [41].
Electron-hole pairs can migrate to the catalyst surface, where they either recombine or participate directly in the oxidation of or reduction in pollutants and assist in the production of the very reactive hydroxyl radical ( • ) and other reactive oxygen species (ROS) such as superoxide radical anions ( • ) and hydroperoxyl radicals ( • ) (Equations 6-10) [39] ℎ ( )  Figure S3). The reduction in TOC content can be attributed to the photocatalytic generation of ROS [39]. The photocatalysis mechanism is widely explained in the literature [40], but, in summary, when a semiconductor such as titanium dioxide is illuminated by a light of energy (hv) greater than or equal to the band gap energy (E gap ) electrons are excited to the conduction band (CB), leaving a positive hole in valance band (VB) Equation (5) [41].
Electron-hole pairs can migrate to the catalyst surface, where they either recombine or participate directly in the oxidation of or reduction in pollutants and assist in the production of the very reactive hydroxyl radical (OH • ) and other reactive oxygen species (ROS) such as superoxide radical anions (O •− 2 ) and hydroperoxyl radicals (HO • 2 ) Equations (6)-(10) [39] h + (VB) When using back-face illumination of immobilized TiO 2 films, increased catalyst loading (TiO 2 film thickness) to the optimal loading increases photon absorption, and as such increased the concentration of generated ROS [12,19]. The ability of ROS to migrate to the water-catalyst interface also plays a critical role in the production of an efficient immobilized system [12]. For thick films, mass transfer of the generated ROS to the catalyst-water interface is often an issue [42,43]. As can be seen in Figure 3, an increase in absorbance was observed as a function of catalyst loading; however, the system does not consider scattering losses [44].
Water 2020, 12, x FOR PEER REVIEW 8 of 19 When using back-face illumination of immobilized TiO2 films, increased catalyst loading (TiO2 film thickness) to the optimal loading increases photon absorption, and as such increased the concentration of generated ROS [12,19]. The ability of ROS to migrate to the water-catalyst interface also plays a critical role in the production of an efficient immobilized system [12]. For thick films, mass transfer of the generated ROS to the catalyst-water interface is often an issue [42,43]. As can be seen in Figure 3, an increase in absorbance was observed as a function of catalyst loading; however, the system does not consider scattering losses [44].    Figure S4 and S5).

Ozonation and Catalytic Ozonation
Ozonation alone (Figure 4a: 0 mg TiO2/cm 2 ) resulted in significant TOC reduction (maximum of 16% with an ozone flow rate of 41.7 mg O3/min). Ozone has a high oxidation potential and can attack organic compounds directly, however, mineralization (and therefore TOC reduction) is mainly reported to result from ozone decomposition into hydroxyl radicals [36,45] as described by [46] and references there in Equations 11-14 If pH < ≈ 8 If pH > ≈ 8 According to the literature, an inhibitory effect is often observed with high doses of ozone [5,47], where excess ozone reacts with the hydroxyl radicals and produces the less reactive hydroperoxyl radicals [48], (Equation 15 [5]).
As shown in Figure 4a, catalytic ozonation (O3/TiO2/dark) enhanced TOC removal compared to ozonation alone as previously observed in a range of water sources [49,50]. The maximum TOC removal observed was ≈19.5%, with an ozone flow rate of 41.7 mg O3/min and 1.98 mg/cm 2 TiO2 loading. Unfortunately, regardless of the expansive list of the catalysts reported to enhance ozonation, the underpinning mechanisms are not well understood [21,22]. The generally proposed mechanism involves either the chemisorption of ozone or organic molecules or both on the catalyst surface, which may lead to either the formation of active species [36] or the formation of metalorganic complexes with higher reactivity toward molecular ozone [51]. In this work, no initial organic adsorption on the bare TiO2 was observed in the dark control experiment, however, the intermediates generated by direct ozone attack, known to transform organic compounds and functional groups into  Figures S4 and S5).

Ozonation and Catalytic Ozonation
Ozonation alone (Figure 4a: 0 mg TiO 2 /cm 2 ) resulted in significant TOC reduction (maximum of 16% with an ozone flow rate of 41.7 mg O 3 /min). Ozone has a high oxidation potential and can attack organic compounds directly, however, mineralization (and therefore TOC reduction) is mainly reported to result from ozone decomposition into hydroxyl radicals [36,45] as described by [46] and references there in Equations (11)-(14) According to the literature, an inhibitory effect is often observed with high doses of ozone [5,47], where excess ozone reacts with the hydroxyl radicals and produces the less reactive hydroperoxyl radicals [48], (Equation (15) [5]).
As shown in Figure 4a, catalytic ozonation (O 3 /TiO 2 /dark) enhanced TOC removal compared to ozonation alone as previously observed in a range of water sources [49,50]. The maximum TOC removal observed was ≈19.5%, with an ozone flow rate of 41.7 mg O 3 /min and 1.98 mg/cm 2 TiO 2 loading. Unfortunately, regardless of the expansive list of the catalysts reported to enhance ozonation, the underpinning mechanisms are not well understood [21,22]. The generally proposed mechanism involves either the chemisorption of ozone or organic molecules or both on the catalyst surface, which may lead to either the formation of active species [36] or the formation of metal-organic complexes with higher reactivity toward molecular ozone [51]. In this work, no initial organic adsorption on the bare TiO 2 was observed in the dark control experiment, however, the intermediates generated by direct ozone attack, known to transform organic compounds and functional groups into smaller molecular weight saturated species, could be adsorbed on the catalyst [45]. To investigate this hypothesis, dark adsorption experiments were carried out post-ozonation (at 41.7 mg O 3 /min) and post-catalytic-ozonation (41.7 mg O 3 /min and 1.98 mg TiO 2 /cm 2 ) treatment (data not shown), with no change in TOC observed, contrary to reports by Garcia and coworkers [50]. As such, the mechanism described in Equations (16)- (19) relating to the chemisorption of ozone and subsequent formation of OH • radicals on the TiO 2 surface sites (S) is considered to be viable [22].
Although catalytic ozonation enhanced the treatment efficiency compared to ozonation alone, increasing the catalyst loading to greater than the lowest used value (0.19 mg TiO 2 /cm 2 ) resulted in minor increases in TOC removal. According to Ikhlaq et al. [21], this behavior implies that the catalyst acts as an adsorbent for the ozone and does not initiate the formation of ROS. Ma and Graham [52] also reported a low increase in enhancement for homogenous catalytic ozonation of atrazine using a Mn(II) based catalysts. Further work is required to identify the mesoporous nature of the immobilized TiO 2 to aid in the design of a high catalytically active surface area.

Ozone Photolysis and Photocatalytic Ozonation
Ozone photolysis (Figure 4b: 0 mg TiO 2 /cm 2 ) was more effective in reducing the organic content of GW in comparison to the dark scenario (Figure 4a) (removal maximum of 24% with an ozone flow rate of 41.7 mg O 3 /min). This effectiveness is attributed to ozone's ability to absorb photons between 200 and 360 nm [53], and generate hydroxyl radicals Equation (20) [53,54].
In the photocatalytic ozonation experiments (Figure 4b), increasing the ozone flowrate did not exhibit an increased mineralization as a function of catalyst loading; (i) for the lowest catalyst loading (0.19 mg TiO 2 /cm 2 ), TOC removal increased with increasing the ozone flowrate, (ii) for the middle range catalyst loading (0.53-1.35 mg TiO 2 /cm 2 ), TOC removal increased to a maximum, then decreased, and (iii) for high catalyst loading (1.61 and 1.98 mg TiO 2 /cm 2 ) ozone showed an inhibitory effect, as explained within within Equation (15) and supporting text.
In addition to the ozone and photocatalytic degradation mechanisms explained above, Wang et al. [55] and Agustina et al. [18] suggested that ozone could be adsorbed on TiO 2 surfaces during photocatalytic ozonation reacting with photogenerated electrons Equations (21) and (22) to reduce charge carrier recombination at the catalyst surface, and, as such, result in the increased production of hydroxyl radicals. Moreover, the resultant generation of the ozonide radical (O •− 3 ) could play a further role in degradation mechanism.
We postulate that for low catalyst loadings, increasing ozone flowrate boosts organic degradation through generation of additional hydroxyl radicals via ozone decomposition, ozone photolysis, and catalytic ozonation. For higher catalyst loading, a smaller UVA photon flux penetrates through the TiO 2 film, as can be observed from the brightness of the microscope images of the coated TiO 2 films ( Figure 5), hence the hydroxyl radical concentration generated from ozone photolysis was decreased. Similarly, reduced ozone decomposition by photolysis results in additional ozone being available for hydroxyl radicals scavenging [5,48] and, when coupled with increasing ozone flow rate, reduces efficiency.

H2O2/dark and H2O2/UVA
TOC reduction by peroxidation (H2O2/dark) and photo-peroxidation (H2O2/dark) was investigated across a range of H2O2 dosages (1.45-53.01 mg/L). The sole use of hydrogen peroxide with or without exposure to UVA resulted in less than a 2.7% reduction in TOC (data not shown).
Lamsal, Walsh and Gagon [56] observed a similar effect with the low oxidation potential of H2O2 [54] and the absence of hydroxyl radicals, attributed to low rates of mineralization [57]. Although hydrogen peroxide can be activated by UV photons to generate hydroxyl radicals [29], wavelengths below 280 nm are required for H2O2 decomposition [54]-much higher energy photons than the UVA lamps employed in this study [48]. Figure 6 shows TOC reduction by H2O2 combined with TiO2 investigated under different irradiation conditions (dark, visible and UVA) using a range of catalyst loadings (0.19-1.98 mg TiO2/cm 2 ) and H2O2 dosages (1.45-53.01 mg/L). In those experiments, H2O2/TiO2 molar ratios (0.0-0.7) were adopted from the H2O2 to O3 ratios used in peroxonation (H2O2/O3) studies [14,54,58] specifically to prevent excess H2O2 hindering treatment efficiency [59].  Lamsal, Walsh and Gagon [56] observed a similar effect with the low oxidation potential of H 2 O 2 [54] and the absence of hydroxyl radicals, attributed to low rates of mineralization [57]. Although hydrogen peroxide can be activated by UV photons to generate hydroxyl radicals [29], wavelengths below 280 nm are required for H 2 O 2 decomposition [54]-much higher energy photons than the UVA lamps employed in this study [48].  This synergistic effect has been documented in several studies [24,60,61], where reactive oxygen species (ROS) were detected in dark hydrogen peroxide/titanium dioxide suspensions. Janson et al. [25] reported an 11% decrease in the concentration of rhodamine B using 0.5 g/L mixture of Anatase and Rutile particles with 1.9 M H2O2 in the dark compared to less than 2% reduction with H2O2 alone.

H 2 O 2 Combined with TiO 2 in the Dark, under Visible and UVA Irradiation
For the studied catalyst loadings (0.19-1.98 mg TiO2/cm 2 ), a greater TOC removal percentage was observed with increasing H2O2/TiO2 molar ratio. The TOC removal at the highest molar ratio (0.7) were comparable and ranged from 12.25% at 0.19 mg TiO2/cm 2 to 15.42% at 1.98 mg TiO2/cm 2 . For catalyst loading greater than 0.98 mg TiO2/cm 2 and an H2O2/TiO2 molar ratio greater than 0.3, a plateau in TOC reduction was observed (Figure 6a). Sánchez et al. [24] described similar behavior by an apparent saturation of the TiO2 sites at high molar ratios; additionally, scavenging of free radicals by H2O2 (Equation 24) could play a role [23,57]. Figure 6b shows an increase in TOC reduction through irradiation of the H2O2/TiO2 system using visible radiation (H2O2/TiO2/Vis). Increased TOC removal was observed as a function of H2O2/TiO2 molar ratio. The maximum TOC removal was ≈ 25% with a catalyst loading of 0.98 mg TiO2/cm 2 and 0.7 H2O2/TiO2 molar ratio. In the work of Liu et al. [26], methylene blue (MB) degradation was found to be enhanced by increasing H2O2 to TiO2 molar ratio under visible light irradiation, with the authors This synergistic effect has been documented in several studies [24,60,61], where reactive oxygen species (ROS) were detected in dark hydrogen peroxide/titanium dioxide suspensions. Janson et al. [25] reported an 11% decrease in the concentration of rhodamine B using 0.5 g/L mixture of Anatase and Rutile particles with 1.9 M H 2 O 2 in the dark compared to less than 2% reduction with H 2 O 2 alone.

H2O2/TiO2 under Visible Irradiation
For the studied catalyst loadings (0.19-1.98 mg TiO 2 /cm 2 ), a greater TOC removal percentage was observed with increasing H 2 O 2 /TiO 2 molar ratio. The TOC removal at the highest molar ratio (0.7) were comparable and ranged from 12.25% at 0.19 mg TiO 2 /cm 2 to 15.42% at 1.98 mg TiO 2 /cm 2 . For catalyst loading greater than 0.98 mg TiO 2 /cm 2 and an H 2 O 2 /TiO 2 molar ratio greater than 0.3, a plateau in TOC reduction was observed (Figure 6a). Sánchez et al. [24] described similar behavior by an apparent saturation of the TiO 2 sites at high molar ratios; additionally, scavenging of free radicals by H 2 O 2 Equation (24) could play a role [23,57].  Figure 6b shows an increase in TOC reduction through irradiation of the H 2 O 2 /TiO 2 system using visible radiation (H 2 O 2 /TiO 2 /Vis). Increased TOC removal was observed as a function of H 2 O 2 /TiO 2 molar ratio. The maximum TOC removal was ≈ 25% with a catalyst loading of 0.98 mg TiO 2 /cm 2 and 0.7 H 2 O 2 /TiO 2 molar ratio. In the work of Liu et al. [26], methylene blue (MB) degradation was found to be enhanced by increasing H 2 O 2 to TiO 2 molar ratio under visible light irradiation, with the authors describing that H 2 O 2 was adsorbed onto the TiO 2 surface, forming a peroxo-titania complex (> Ti-OOH) which can be photo-activated even under visible light according to Equation (25).
The reaction of photogenerated charge carriers with H 2 O 2 results in enhanced hydroxyl radical production and reduces surface recombination Equations (26) and (27) [23,62].
Whilst mineralization increased as a function of TiO 2 loading, TOC removal at catalyst loadings of 1.61 and 1.98 mg TiO 2 /cm 2 was not enhanced, potentially due to the saturation of the active sites and also the limited visible light penetration through the thick films [63], as demonstrated in Figure 5 [62] and Janson et al. [25], however, Beltrán and Rodríguez reported no increase in removal efficiency of their H 2 O 2 /TiO 2 system under UV as opposed to visible light irradiation [64].
The efficiency enhancement for H 2 O 2 /TiO 2 under UVA illumination is primarily due to additional ROS generated via a series of mechanisms, including the reactions described for H 2 O 2 /TiO 2 in the dark/visible illumination, and other reaction pathways based on multi-phase transport mechanisms at the catalyst surface or in the solution. Additional reaction pathways include the generation of HO • 2 and OH • catalyzing peroxide decomposition at oxidized (S + ) and the reduced (S) TiO 2 surface sites Equations (28) and (29) [23], reaction of H 2 O 2 with the photogenerated electrons and holes (Equations (26) and (27) [23,62] or with superoxide radical (O •− 2 ) reactions that can lead to the formation of OH • radical and singlet oxygen ( 1 O 2 ) Equations (30)-(33) [62].
The effect of H 2 O 2 /TiO 2 molar ratio was more significant at low catalyst loadings; at high catalyst loading, there was no measured enhancement. Similar findings have been reported by Sakkas et al. [63], who mainly attributed the effect to the radical scavenging [23,57] and the reduction in light penetration when higher catalyst loadings were used. In catalytic and photocatalytic peroxonation, the optimal TiO 2 loading from the previous experiment (0.96 mg TiO 2 /cm 2 ) was used (Figure 7). the range of 8.3-25 mg O3/min to avoid the O3 inhibition effect observed in photocatalytic ozonation experiments. In catalytic and photocatalytic peroxonation, the optimal TiO2 loading from the previous experiment (0.96 mg TiO2/cm 2 ) was used (Figure 7).  Figure S9-S12).
A synergistic effect was observed when combining hydrogen peroxide and ozone treatments, in accordance with other works [32,65] where additional hydroxyl radicals were proposed to be generated via Equation 34 [66].
A greater synergetic effect was observed at higher ozone flowrates due to the lack of competition between the organic content of GW and H2O2 for the ozone. According to Gulyas et al. [15], at low doses of ozone, organic compounds in water can compete with hydrogen peroxide for the ozone, resulting in low rates of ROS formation. Further enhancement was observed with UV peroxonation treatment, as has been previously reported [16,65].
In experiments without TiO2, peroxonation and UVA-peroxonation (Figure 7a and Figure 7c), the enhancement decreased with high H2O2/O3 molar ratios, with an optimal ratio observed between 0.3 and 0.5 in line with previous studies [14,54,58]. Excess hydrogen peroxide has been reported to lead to hydroxyl radical scavenging [39,57]. We observed that a maximum TOC removal with an ozone flow rate of 25 mg/min (30.31%) was achieved by peroxonation (at 0.5 H2O2/O3 molar ratio) with UVA-peroxonation (at 0.3 H2O2/O3 molar ratio) increasing TOC removal to 34.97%.
TOC removal via catalytic and photocatalytic peroxonation is shown in Figure 7b and 8d. Treatment efficiency was observed to improve with increasing H2O2/O3 ratio to a maximum of 33.42% TOC removal via catalytic peroxonation at an ozone flow rate of 25 mg/min and 0.7 H2O2/O3 molar ratio. Organic mineralization of GW was significantly enhanced to 54.12% through photocatalytic peroxonation. This increased oxidation can be accredited to the generation of additional hydroxyl radicals from surface state reactions (as described above), reducing scavenging of both hydrogen peroxide and ozone.

Conclusion
The data presented represent one of the most comprehensive and systematic studies of enhanced photocatalysis for the treatment of real greywater. We demonstrate that in a stirred tank reactor under optimal catalyst loading for an immobilized system, the ROS generated via photocatalysis can A synergistic effect was observed when combining hydrogen peroxide and ozone treatments, in accordance with other works [32,65] where additional hydroxyl radicals were proposed to be generated via Equation (34) [66].
A greater synergetic effect was observed at higher ozone flowrates due to the lack of competition between the organic content of GW and H 2 O 2 for the ozone. According to Gulyas et al. [15], at low doses of ozone, organic compounds in water can compete with hydrogen peroxide for the ozone, resulting in low rates of ROS formation. Further enhancement was observed with UV peroxonation treatment, as has been previously reported [16,65].
In experiments without TiO 2 , peroxonation and UVA-peroxonation (Figure 7a,c), the enhancement decreased with high H 2 O 2 /O 3 molar ratios, with an optimal ratio observed between 0.3 and 0.5 in line with previous studies [14,54,58]. Excess hydrogen peroxide has been reported to lead to hydroxyl radical scavenging [39,57]. We observed that a maximum TOC removal with an ozone flow rate of 25 mg/min (30.31%) was achieved by peroxonation (at 0.5 H 2 O 2 /O 3 molar ratio) with UVA-peroxonation (at 0.3 H 2 O 2 /O 3 molar ratio) increasing TOC removal to 34.97%.
TOC removal via catalytic and photocatalytic peroxonation is shown in Figure 7b,d. Treatment efficiency was observed to improve with increasing H 2 O 2 /O 3 ratio to a maximum of 33.42% TOC removal via catalytic peroxonation at an ozone flow rate of 25 mg/min and 0.7 H 2 O 2 /O 3 molar ratio. Organic mineralization of GW was significantly enhanced to 54.12% through photocatalytic peroxonation. This increased oxidation can be accredited to the generation of additional hydroxyl radicals from surface state reactions (as described above), reducing scavenging of both hydrogen peroxide and ozone.

Conclusions
The data presented represent one of the most comprehensive and systematic studies of enhanced photocatalysis for the treatment of real greywater. We demonstrate that in a stirred tank reactor under optimal catalyst loading for an immobilized system, the ROS generated via photocatalysis can mineralize >30% of the organic content of GW. Either single or a combination of additional oxidants, H 2 O 2 and/or ozone, under conditions of optimal molar ratio, enhanced the remediation efficiencywe observed 54.12% TOC removal through photocatalytic peroxonation. As we move to co-design AOP systems for use by members of communities in Jordan with whom we work [67], the data showing both UVA and visible light enhancement of the H 2 O 2 /TiO 2 system are of particular interest, demonstrating potential for enhanced photocatalysis as a viable option to reduce the organic fraction of GW and permit local water reuse. Further work is underway to examine and quantify synergistic effects between AOPs and to demonstrate the potential of the above systems to disinfect GW, therefore moving towards attainment of the Jordanian GW reuse standard for indoor use (toilet flushing) as opposed to the lower requirements for vegetable irrigation.