Advances in Treatment of Brominated Hydrocarbons by Heterogeneous Catalytic Ozonation and Bromate Minimization

The formation of carcinogenic bromate ions is a constraint when ozone is used for the remediation of water containing brominated organic materials. With its strong oxidizing ability, ozone rapidly transforms bromide in aqueous media to bromate, through a series of reactions involving hydroxyl radicals. Several strategies, such as limiting the ozone concentration, maintaining pH < 6, or the use of ammonia or hydrogen peroxide were explored to minimize bromate generation. However, most of the above strategies had a negative effect on the ozonation efficiency. The advanced oxidation processes, using catalysts together with ozone, have proven to be a promising technology for the degradation of pollutants in wastewater, but very few studies have been conducted to find ways to minimize bromate formation during this approach. The proposed article, therefore, presents a comprehensive review on recent advances in bromate reduction in water by catalytic ozonation and proposes reaction mechanisms associated with the catalytic process. The main aim is to highlight any gaps in the reported studies, thus creating a platform for future research and a quest to find environment friendly and efficacious catalysts for minimizing bromate formation in aqueous media during ozonation of brominated organic compounds.


Introduction
The need to reduce environmental pollution is currently receiving urgent attention around the world. The rapid increase in the human population, coupled with growing demands from industrial and other sectors, has triggered the large-scale usage of diverse non-biodegradable chemicals, leading to extensive pollution of water systems. Since these polluted waters pose a serious threat to the environment, ongoing research is conducted to explore cost effective treatment methodologies for the removal of varied toxic chemicals from the water systems. An alternative to chlorination and adsorption agents for water purification is ozonation, which is becoming a useful methodology for improving the quality of water. The use of ozone has proven to be excellent for microorganism destruction and biological contaminant removal from water [1], but is not effective for degrading recalcitrant organic pollutants in water. The presence of bromide (Br − ) in polluted waters poses a serious problem during ozonation. Bromide is rapidly oxidized to toxic bromate BrO − 3 during ozone treatment. Bromide is usually present in low concentrations of between 10 4 and 10 6 ppb in wastewaters and approximately 67 × 10 3 ppb in seawater [2]. Relatively low amounts were found in rainwater, ranging from 0 to 110 ppb [3], but in groundwater, between 10 and 2 × 10 3 ppb were detected [4]. Higher bromide Figure 1. Bromate formation pathways. [14],.
The use of suitable heterogeneous catalysts has proven to be beneficial to enhance the efficiency of the ozonation process and minimize the generation of toxic by-products [16]. Studies have shown that hydroxyl radicals generated during ozonation in the presence of metal oxides could increase bromate formation [17]. This review presents a comprehensive assessment on recent advances on bromate reduction in water by heterogeneous catalytic ozonation.

Bromate Minimization Strategies
The following mechanism was proposed by von Gunten and Hoigne´ for the conversion of Br to BrO during ozonation [18]:
The use of suitable heterogeneous catalysts has proven to be beneficial to enhance the efficiency of the ozonation process and minimize the generation of toxic by-products [16]. Studies have shown that hydroxyl radicals generated during ozonation in the presence of metal oxides could increase bromate formation [17]. This review presents a comprehensive assessment on recent advances on bromate reduction in water by heterogeneous catalytic ozonation.

Bromate Minimization Strategies
The following mechanism was proposed by von Gunten and Hoigne' for the conversion of Br − to BrO − 3 during ozonation [18]: They concluded that the direct oxidative conversion of Br − to HOBr/OBr − was mainly controlled by molecular ozone, while further oxidation of HOBr/OBr − to BrO • radicals was influenced by HO • radicals. The unstable BrO • radicals disproportionate to BrO − 2 . The dissolved ozone in the water then rapidly oxidizes BrO − 2 to BrO − 3 [1]. Limited studies have been conducted to establish the effects of catalytic ozonation on bromate formation. The most recent studies are discussed below.

MCM-48, CeO 2 and Ce x -MCM-48
Li et al. [19] reported on catalytic ozonation of bromide containing waters with MCM-48, CeO 2 and combined mesoporous sieve Ce x -MCM-48 (cerium combined with MCM-48) with various Si/Ce molar ratios (Ce 30/66/100/200 -MCM-48). All catalysts were able to considerably impede BrO − 3 formation in comparison to ozonation alone. After 30 min of ozone treatment, the inhibition efficiencies of MCM-48 and CeO 2 were 78.6% and 63.9%, respectively. When MCM-48 was doped with Ce, a marked improvement in BrO − 3 minimization was observed. When the Ce content was increased from x = 200 to x = 66, BrO − 3 yield decreased, giving a maximum inhibition efficiency of 91% after 30 min of ozonation. However, an additional increase of Ce to x = 30, resulted in an increase in BrO − 3 concentration and an inhibition efficiency of 78%. Their explanation for this trend was that doping MCM-48 with Ce resulted in the generation of more surface hydroxyl groups, which successively enhanced decomposition of O 3 on the active sites of the catalyst surface. However, doping beyond x = 66 blocked the active sites, leading to a destruction of the mesoporous structure of MCM-48, hence leading to poor catalyst activity.
Li et al. [19] proposed a bromate reduction pathway for Ce 66 -MCM-48 with the aid of bromine mass balance studies. Their results revealed that Ce 66 -MCM-48 did not adsorb Br − , HOBr/OBr − and BrO − 3 , the main bromine-containing species present in the water solution.  [20] according to the following pathway: Ce 3+ also reacts with H 2 O 2 to form Ce 4+ [21]: An alternative pathway produces H 2 O 2 from aqueous O 3 decomposition Ce 3+ is regenerated by HO − 2 , which converts Ce 4+ to Ce 3+ [21]: T. Zang et al. [22] investigated the effect of a number of metal oxides, such as α − FeOOH, α − Fe 2 O 3 , γ − FeOOH and CeO 2, on bromate production during ozone treatment of bromide in water. The catalytic reactions with α − Fe 2 O 3 produced more BrO − 3 relative to ozonation alone, whereas the reactions with α − FeOOH, γ − FeOOH and CeO 2 minimized bromate formation. However, CeO 2 was most active in reducing bromate production. They determined simultaneously the concentrations of Br − and HOBr/OBr − for uncatalysed ozonation and CeO 2 catalysed ozonation. They found that the Br − amounts in catalytic ozonation was lower with ozone treatment alone before 15 min, and remained similar thereafter. The HOBr/OBr − amount in CeO 2 catalytic ozonation was always significantly higher in comparison to ozone treatment alone. According to von Gunten [1], HOBr/OBr − is an essential intermediary for BrO − 3 production during ozonation, therefore, its accumulation in CeO 2 catalytic ozonation suggests that CeO 2 considerably inhibits the conversion of HOBr/OBr − to BrO − 3 . The formation of H 2 O 2 was detected in both ozonation alone and ozonation with CeO 2 . The results showed that the amount of H 2 O 2 with CeO 2 was poorer compared to single ozonation. Studies have shown that the surface of CeO 2 can initiate the decomposition H 2 O 2 generating oxygen in water [23]. Therefore, the lesser H 2 O 2 amount in CeO 2 catalytic ozonation can be attributed to its concurrent disintegration on the surface of CeO 2 . One study mentioned that low amounts of hydrogen peroxide can promote BrO − 3 formation, arising from hydroxyl radical formation from the interaction of HO − 2 with O 3 [1], and other studies discussed that hydrogen peroxide at high amounts (H 2 O 2 /O 3 molar ratio >1:2) is likely to reduce HOBr/OBr − to Br − , hence minimizing BrO − 3 formation [17,18,24]. According to Zang et al. [22], the enhanced BrO − 3 minimization in CeO 2 catalytic ozonation is primarily due to the lower H 2 O 2 amounts. Since CeO 2 catalytic ozonation produced a lower amount of H 2 O 2 than single ozonation, the HO • amount is expected to be moderately lower, hence resulting in a lower oxidation rate of HOBr/OBr − to BrO • . Furthermore, BrO • can be reduced to HOBr/OBr − by Ce 3+ , which is a temporary reductive state of surface Ce 4+ in catalytic decomposition of H 2 O 2 [25]. Thus, an additional pathway for BrO − 3 minimization is the reduction of BrO • to HOBr/OBr − on the CeO 2 surface. Both BrO − 3 reduction routes require the involvement of surface active Ce 4+ sites. It has been reported that SO 2− 4 ions, when combined with metal oxides, have a strong attraction for their surface sites [26]. Zang et al. [22], therefore, added various concentrations of SO 2− 4 to the bromide containing solutions to ascertain its affinity for surface active Ce 4+ sites, and the impact on BrO − 3 . They found that the difference in bromate formation between ozonation alone and CeO 2 catalytic ozonation decreased as SO 2− 4 amounts increased from 0 to 5 mM. The diminishing effectiveness of CeO 2 to minimize BrO − 3 formation is ascribed to surface Ce 4+ − SO 2− 4 co-ordination, thus indicating that surface Ce 4+ sites account for most of the BrO − 3 minimization during CeO 2 catalytic ozonation.

Nano-Metal Oxides, SnO 2 and TiO 2
Wu et al. [27] conducted simulation studies to investigate the influence of nano-metal oxides, SnO 2 and TiO 2 on bromate generation in pure water during ozone treatment. Their results showed that ozonation in the presence of nano-metal oxides (SnO 2 and TiO 2 ) as catalysts, minimized BrO − more O 3 to HO • radicals. The lower ozone concentration results in lower HOBr/OBr − , hence minimizing BrO − 3 formation. Furthermore, HO • radicals can rapidly combine to generate H 2 O 2 , which can reduce HOBr/OBr − to Br − [28,29]. The presence of humic acid influenced bromate generation. Increasing the humic acid concentration from 0 to 3.0 ppm resulted in a decrease in bromate formation. Humic acid reacts readily with O 3 and hydroxyl radicals, which also reacts with Br − and HOBr/OBr − [16,30]. Therefore, a lower concentration of HOBr/OBr − leads to lesser bromate formation [22].

Mn Incorporated MCM-41
Xue et al. [31] employed mesoporous Mn incorporated MCM-41 to hinder bromate production during catalytic ozonation of waters containing bromide. A comparison of the three temperature ramping rates (0.5 K min −1 , 1 K min −1 and 2 K min −1 ) during calcination of Mn X -MCM-41 (X = 40, 80, 100 and 120, the molar ratio of Si/Mn), revealed that Mn 100 -MCM-41 with ramping rate of 1 K min −1 showed superior surface characteristics and the greatest bromate inhibition efficiency. A 96.7% inhibition efficiency was achieved after 60 min when compared to ozonation alone. XPS data revealed that Mn 100 -MCM-41 (1 K min −1 ) has more oxygen vacancies, which has tendency to adsorb and dissociate H 2 O to surface active species [32]. Ozone readily reacts with these surface-active species, resulting in less ozone exposure for Br − oxidation to HOBr/OBr − , hence minimizing bromate formation. The higher fraction of Mn 2+ and Mn 3+ in Mn-MCM-41 enhanced bromate inhibition efficiency.
Xue et al. revealed that the concentration of HOBr/OBr − during Mn 100 -MCM-41 ozonation was lower than single ozonation. They explained that Mn 100 -MCM-41 adsorbs H 2 O and dissociates to form surface active species. Ozone then readily reacts with these surface-active species, hence leading to low ozone exposure for Br − oxidation HOBr/OBr − . Furthermore, hydrogen peroxide was detected in both uncatalysed and Mn 100 -MCM-41 catalysed ozonation. The concentration of H 2 O 2 increased steadily in Mn 100 -MCM-41 ozonation, but decreased in uncatalysed ozonation, signifying that more reactive oxygen species [32] is formed in the presence of Mn 100 -MCM-41. These species are capable of consuming HOBr/OBr − and preventing bromate formation. To verify the role of hydroxyl radicals, TBA (a potential HO • radical scavenger) was introduced in both single ozonation and Mn 100 -MCM-41 ozonation. The bromate yield decreased for both processes, thus confirming that HO • was primarily responsible for BrO − 3 production. In ozonation alone, the decrease in bromate yield is mainly attributed to the decrease in hydroxyl radicals. In Mn 100 -MCM-41/O 3 process, the decreased bromate yield is due to the decrease in both hydroxyl radicals and residual ozone. A similar phenomenon was evident with Fe-Cu-MCM-41 [33]. They attributed the bromate reduction to ozone decomposition by the catalysts, resulting in a reduced amount of ozone for bromate generation [19]. The higher bromate yield in Fe-Cu-MCM-41/O 3 than in Fe-MCM-41/O 3 and Cu-MCM-41/O 3 systems, is due to more HO • presence in the Fe-Cu-MCM-41/O 3 system. The presence of both the redox couples, Fe 3+ /Fe 2+ and Cu 2+ /Cu + on the catalyst surface (confirmed by XPS analysis) further accelerated ozone decomposition into HO • radicals. As illustrated in Figure 2, bromate is produced through both the direct and indirect oxidation of Br − by O 3 /HO • [34].
After the addition of the catalyst, more ozone is consumed, resulting in a hindrance of the direct oxidation of Br − to HBrO/BrO − by ozone (a key intermediate reaction for bromate generation), and additional oxidation of HBrO/BrO − to BrO − 3 [19]. The superior efficiency of Fe-Cu-MCM-41, causes an abundance of hydroxyl radicals. A greater HO • concentration results in an impediment of pathway 1, thus resulting in a higher bromate build-up [35].  After the addition of the catalyst, more ozone is consumed, resulting in a hindrance of the direct oxidation of Brto HBrO/BrOby ozone (a key intermediate reaction for bromate generation), and additional oxidation of HBrO/BrOto BrO 3 - [19]. The superior efficiency of Fe-Cu-MCM-41, causes an abundance of hydroxyl radicals. A greater HO • concentration results in an impediment of pathway 1, thus resulting in a higher bromate build-up [35]. The addition of t-butanol (TBA) to the Brsubstrate solution, generated less bromate in both single ozonation and ozonation with Fe-Cu-MCM-41. As reported, the bromate formation requires the presence of both ozone and hydroxyl radicals [36]. Bromide is first oxidized by ozone directly to HBrO/BrO -. Thereafter, the HBrO/BrOis oxidized by HO • to BrO 3 -. Thus, in single ozonation, since the HO • radicals are scavenged by TBA, bromate formation is primarily due to molecular ozone. In the Fe-Cu-MCM-41/O3 process, the ozone concentration in the water significantly decreases due to the surface reactions, and the generated HO • radicals are also scavenged by TBA. Both actions result in the suppression of the bromate formation pathway, hence, lowering bromate yield. Bromate production was also inhibited in both ozonation alone and Fe-Cu-MCM-41 catalytic ozonation with the addition of PO 4 3-. Bromide yields were found to increase with an increase in PO 4 3dosage. As proposed by Huang, PO 4 3accelerates the generation of H 2 O 2 , which reduces HBrO/BrOto Br , hence constraining BrO 3 generation [37].

Fe-Al LDH Supported on Mesoporous Al 2 O 3
Nie et al. [38] prepared Fe-Al layered double hydroxides (Fe-Al LDH, the molar ratio of Fe : Fe = 1:10) supported on mesoporous Al 2 O 3 and showed its effectiveness to minimize bromate formation. The BrO 3 concentration rapidly increased during the uncatalysed ozonation reaching 20 ppb after 60 min of ozone treatment. However, ozonation with Fe-Al LDH/Al2O3 completely inhibited BrO 3 formation. Furthermore, even when the initial Brconcentration and ozone dose were increased, the BrO 3 yield after 60 min of catalytic ozonation stayed below the allowable limit of 10 ppb. Fe-Al LDH/Al2O3 in the presence of a mixture of phenazone (PZ) and BrO 3 only, revealed that approximately 45% of BrO 3 was adsorbed on Fe-Al LDH/Al2O3 and 18% of Brwas generated. They ascribed the BrO 3 reduction to Fe formed during Fe-Al LDH/Al2O3 preparation, which was confirmed by XPS analysis [39]. However, 82% of BrO 3 was converted to Brduring Fe-Al LDH/Al2O3 ozonation of the PZ/BrO 3 mixture. The reduction of BrO 3 to Brincreased with the ozone dose and BrO 3 concentration. In contrast, the PZ/O3 system could not reduce BrO 3 to Br -.
Furthermore, when phosphate was added to the Fe-Al LDH/Al2O3/O3 system, BrO 3 reduction was completely suppressed. The presence of phosphate permanently blocked the active surface sites of the catalyst, resulting in the replacement of surface hydroxyl groups and the formation of complexes with Fe within the catalyst, thereby decreasing catalytic activity [40,41]. The adsorption of BrO 3 and the interaction of O3 with Fe-Al LDH/Al2O3 was suppressed, therefore, poor BrO 3 reduction is expected. Further investigations indicated that BrO 3 reduction to Br by surface Fe is responsible for complete inhibition of  The addition of t-butanol (TBA) to the Br − substrate solution, generated less bromate in both single ozonation and ozonation with Fe-Cu-MCM-41. As reported, the bromate formation requires the presence of both ozone and hydroxyl radicals [36]. Bromide is first oxidized by ozone directly to HBrO/BrO − . Thereafter, the HBrO/BrO − is oxidized by HO • to BrO − 3 . Thus, in single ozonation, since the HO • radicals are scavenged by TBA, bromate formation is primarily due to molecular ozone. In the Fe-Cu-MCM-41/O 3 process, the ozone concentration in the water significantly decreases due to the surface reactions, and the generated HO • radicals are also scavenged by TBA. Both actions result in the suppression of the bromate formation pathway, hence, lowering bromate yield.
Bromate production was also inhibited in both ozonation alone and Fe-Cu-MCM-41 catalytic ozonation with the addition of PO 3 [46]. Therefore, Mn 2+ is responsible for promoting O 3 to eliminate organic pollutants and also assist in inhibiting BrO − 3 formation. The proposed reactions on MnOx/Al 2 O 3 in the presence of ozone occurs as follows [42]: Reaction (14) proposes the generation of H 2 O 2 in both uncatalysed and catalytic ozonation. The results showed that H 2 O 2 concentration was remarkably lower in uncatalysed ozonation than in MnOx/Al 2 O 3 catalytic ozonation. This trend suggests that in catalytic ozonation, reaction (14) is suppressed, since more HO • 2 is used up by reactions (10) and (11), hence leading to increased generation of Mn 2+ . This confirmed that the presence of different oxidation states of manganese is responsible for controlling BrO − 3 generation.

Ce x Zr x-1 O 2 Mixed Oxides
Yang et al. [47] prepared mixed oxides Ce x Zr x−1 O 2 (x = 0.16, 0.50, 0.75, 0.9) and CeO 2 to study BrO − 3 reduction during ozonation of Br − containing filtered water. The results indicated that catalytic ozonation with Ce x Zr x−1 O 2 and CeO 2 minimized bromate formation better than ozonation alone. They concluded that the Ce x Zr x−1 O 2 mixed oxides and CeO 2 effectively suppressed the oxidation of Br − by O 3 and HO • radicals. Furthermore, the Ce 0.75 Zr 0.25 O 2 mixed oxide displayed the best catalytic activity for BrO − 3 minimization, with 53% of BrO − 3 formation being reduced after 20 min of ozonation. The adsorption of Br − and BrO − 3 on catalyst surface were not detected, since anions have no affinity for the neutral or negatively charged oxide surface. Furthermore, the catalyst material exhibited good stability, since no leaching of metal ions were detected during the ozonation process.
To confirm the role of O 3 and HO • radicals in BrO − 3 inhibition, p-chlorobenzoic acid (pCBA), a HO • scavenger was introduced to monitor HO • radicals. HPLC analysis revealed that pCBA concentration decreased rapidly with ozone treatment time, and its concentration was considerably lower in Ce 0.75 Zr 0.25 O 2 ozonation than in single ozonation. This indicates that Ce 0.75 Zr 0.25 O 2 mixed oxide significantly promoted the decomposition of O 3 to HO • radicals during the catalytic ozonation process. Their results also showed that BrO − 3 formation and O 3 decomposition was extremely rapid during the first 5 min of ozonation, further confirming that HO • radicals play a major role during BrO − 3 formation. The organic compounds in water favours organic/HO • reactions more than Br − /HO • reactions, since the rate of reaction for oxidative degradation of organic compounds by HO • radicals is faster than that for oxidizing Br − by HO • radicals [48]. Since the HO • radicals facilitate the efficient degradation of organic substituents, therefore, the suppression of the oxidation of Br − is favoured, leading to the minimization of BrO − 3 yield.

TiO 2
Parrino et al. [49] investigated simultaneous ozonation and photocatalysis for purifying wastewater containing formic acid/4-nitrophenol and bromide ions. The initial ozonation experiments performed on formate and bromide ions in the presence and absence of TiO 2 , showed similar degradation rates, suggesting that reactions occurring on the TiO 2 surface did not contribute to the degradation of the target compounds [50]. It was also observed that the oxidation of formate was not affected by the presence of bromide ion and the oxidation of bromide to bromate occurred only after the consumption of formate ions. Bromide ions reacted with hydroxyl radicals generated during photocatalysis, according to the following reaction scheme: Lastly, the photoelectrons generated on the photocatalyst surface reduced the hypobromite species to bromide.
As illustrated, these pathways eventually lead to the recovery of bromide ion, Equation (18). Furthermore, if solution pH is in the range 6-8, a secondary pathway facilitates the conversion of hypobromous acid to bromide. The generated HOBr, as shown in Equation (16), primarily exists in its protonated form, and H 2 O 2 generated during the photocatalytic reaction, acts as a scavenger for hypobromite, by reducing it to bromide [51].
From this outcome, they concluded that bromate generation can be prevented by interrupting the ozone treatment as soon as the oxidation of the organic species is almost complete. Furthermore, reducing bromate is also a more practical way to minimize its accumulation, and as per the previous reports, photocatalysis alone is efficient to convert bromate to bromide [51]. When 4-nitrophenol was substituted in the place of the formate ion, the formation of bromate, took place once again only after the disappearance of 4-nitrophenol, and was found to be faster than with formate ion. This implies that the type of organic contaminant in the water plays a decisive role in the amount of bromate formed.
Results also showed that no Fe 2+ was formed when β − FeOOH/Al 2 O 3 was present in water alone, however, a small amount of surface Fe 2+ was observed when β − FeOOH/Al 2 O 3 in water was ozonated.

Perovskite-Type Oxides, LaFeO 3 and LaCoO 3
Y. Zhang et al. [54] synthesized two perovskite-type oxides, LaFeO 3 and LaCoO 3 , and examined their capacity to degrade benzotriazole (BZA) and minimize BrO − 3 formation in water during ozonation. The ozonation of an aqueous mixture of BZA and Br − generated the most amount of BrO − 3 . The bromate yield increased sharply for the first 20 min of ozonation and then showed a decreasing trend up to 120 min. The bromate yield decreased significantly after the addition of catalyst, especially during the first 30 min of ozonation, but the conversion of Br − was faster with LaCoO 3 compared with LaFeO 3 . The concentration of HBrO/BrO − was found to be higher in LaCoO 3 ozonation than with LaFeO 3 , which explains its superior BrO − 3 minimization ability. The production of HO •− 2 /O •− 2 resulted in the generation of H 2 O 2 , which also contributed to the reduction of BrO − 3 to HBrO/BrO − . Y. Zhang et al. [54] further illustrated the reaction mechanism of LaFeO 3 and LaCoO 3 facilitated ozonation of benzotriazole (BZA) and BrO − 3 minimization. They concluded that LaFeO 3 did not catalytically promote molecular ozone decomposition to reactive oxygen species (ROS), which is needed for BZA degradation, but instead rapidly reduced BrO − 3 .  Further studies on HZSM-5 (Si/Al = 300) showed that its high efficiency for bromate minimization is related to its affinity to adsorb OBr − , a major intermediate in bromate formation [1]. The results have shown that HZSM-5 had no affinity to adsorb of Br − , BrO − 3 and HOBr, therefore no direct electron transfer reaction is expected on HZSM-5. However, the majority of OBr − was rapidly adsorbed onto HZSM-5 within 0.5 min. They then concluded that the specific adsorption of OBr − on the HZSM-5 prevents the oxidation of OBr − to BrO − 3 in water. Their results also detected the presence of H 2 O 2 in both single ozonation and ozonation with HZSM-5. Considerably higher yields of H 2 O 2 were detected in single ozonation than in O 3 /HZSM-5 process, and the HZSM-5 neither adsorbed nor decomposed H 2 O 2 in water. The lower H 2 O 2 concentration in O 3 /HZSM-5 leads to lower bromate yields.
2.14. FeO X /CoO X Gounden et al. [56], conducted a study on the degradation of hazardous halohydrin, 2,3-dibromopropan-1-ol (2,3-DBP) in water by ozonation alone and ozonation with Co loaded on Fe prepared by co-precipitation (Co-ppt) and a simple physical mixing method (Mixed). Their results showed that debromination of 2,3-DBP produced large quantities of Br − and BrO − 3 ions. The Fe:Co (Mixed) catalyst was found to be more effective in suppressing the generation of bromate than the Fe:Co (Co-ppt) catalyst. The presence of Fe:Co (Mixed) lowered the solution pH from 6.8 to 5.7, which was an ideal condition for inhibiting bromate formation. The reaction pathway for conversion of Br − to BrO − 3 was described in the presence of Fe-Co (Mixed) catalyst. Firstly, since pH of the initial solution (5.7), is higher than the pZc value (5.1) of the Fe-Co (Mixed) catalyst, its surface can comprise mostly of negative Fe − Co − active sites (Scheme 1). These sites repel the negatively charged bromide ions, thus preventing electron transfer reactions on the catalyst surface, resulting in a lower bromate yield. yield increased with time in a single ozonation, O3/HZSM-5 and O3/CeO2. The bromate concentration in O3/HZSM-5 was significantly lower than in single ozonation and in O3/CeO2. The HZSM-5 with Si/Al ratios of 300 and 25 showed superior capacity for bromate minimization and reduced approximately 58% bromate formation potential after 20 min of ozone treatment, while CeO2 only reduced 22%. Further studies on HZSM-5 (Si/Al = 300) showed that its high efficiency for bromate minimization is related to its affinity to adsorb OBr -, a major intermediate in bromate formation [1]. The results have shown that HZSM-5 had no affinity to adsorb of Br -, BrO 3 and HOBr, therefore no direct electron transfer reaction is expected on HZSM-5. However, the majority of OBrwas rapidly adsorbed onto HZSM-5 within 0.5 min. They then concluded that the specific adsorption of OBron the HZSM-5 prevents the oxidation of OBrto BrO 3 in water. Their results also detected the presence of H 2 O 2 in both single ozonation and ozonation with HZSM-5. Considerably higher yields of H 2 O 2 were detected in single ozonation than in O3/HZSM-5 process, and the HZSM-5 neither adsorbed nor decomposed H 2 O 2 in water. The lower H 2 O 2 concentration in O3/HZSM-5 leads to lower bromate yields.

FeOX/CoOX
Gounden et al. [56], conducted a study on the degradation of hazardous halohydrin, 2,3dibromopropan-1-ol (2,3-DBP) in water by ozonation alone and ozonation with Co loaded on Fe prepared by co-precipitation (Co-ppt) and a simple physical mixing method (Mixed). Their results showed that debromination of 2,3-DBP produced large quantities of Brand BrO 3 ions. The Fe:Co (Mixed) catalyst was found to be more effective in suppressing the generation of bromate than the Fe:Co (Co-ppt) catalyst. The presence of Fe:Co (Mixed) lowered the solution pH from 6.8 to 5.7, which was an ideal condition for inhibiting bromate formation. The reaction pathway for conversion of Brto BrO 3 was described in the presence of Fe-Co (Mixed) catalyst. Firstly, since pH of the initial solution (5.7), is higher than the pZc value (5.1) of the Fe-Co (Mixed) catalyst, its surface can comprise mostly of negative Fe-Coactive sites (Scheme 1). These sites repel the negatively charged bromide ions, thus preventing electron transfer reactions on the catalyst surface, resulting in a lower bromate yield.
Fe-Co-OH 2 + Secondly, since the pH of the initial solution is much lower than the pKa (8.8) of the HOBr/OBrsystem, an equilibrium shift occurs to the left, thus favoring a higher yield of HOBr and lower OBr -. As ozone is more reactive towards HOBr than OBr -, a decrease in bromate yield is anticipated (Scheme 2). Scheme 1. Reaction pathway for formation of protonated/deprotonated Fe-Co surface in water.
Secondly, since the pH of the initial solution is much lower than the pKa (8.8) of the HOBr/OBr − system, an equilibrium shift occurs to the left, thus favoring a higher yield of HOBr and lower OBr − . As ozone is more reactive towards HOBr than OBr − , a decrease in bromate yield is anticipated (Scheme 2).

Effect of Initial Solution pH
Previous studies have shown that lowering of pH to below 7, preceding ozonation results in a decrease in bromate formation [57]. A decrease of one pH-unit results in 50-63% reduction in BrO formation [58]. This decrease has been attributed to two factors: (i) At pH < 7, oxidized bromide is likely to primarily be found as hypobromous acid (HOBr , resulting in limited amounts of hypobromite (OBr available for reaction with ozone [18,59]: Pathway for inhibition/formation of bromate in catalytic ozonation.

Effect of Initial Solution pH
Previous studies have shown that lowering of pH to below 7, preceding ozonation results in a decrease in bromate formation [57]. A decrease of one pH-unit results in 50-63% reduction in BrO formation [58]. This decrease has been attributed to two factors: (i) At pH < 7, oxidized bromide is likely to primarily be found as hypobromous acid (HOBr , resulting in limited amounts of hypobromite (OBr available for reaction with ozone [18,59]: Reaction pathways for bromate formation during Fe-Co catalysed ozonation and degradation of 2,4,6-TBP in water.

Effect of Initial Solution pH
Previous studies have shown that lowering of pH to below 7, preceding ozonation results in a decrease in bromate formation [57]. A decrease of one pH-unit results in 50-63% reduction in BrO − 3 formation [58]. This decrease has been attributed to two factors: (i) At pH < 7, oxidized bromide is likely to primarily be found as hypobromous acid (HOBr), resulting in limited amounts of hypobromite (OBr − ) available for reaction with ozone [18,59]:

HOBr
OBr − + H + pK a = 8.7 As the solution pH is increased, the concentration of OBr − increases, hence promoting BrO − 3 production, since OBr − is more reactive with ozone than HOBr [1]. The main oxidant for bromate formation in natural water is the hydroxyl radical. At a lower pH, the conversion of molecular ozone to hydroxyl radicals is low, therefore, the amount of bromate formed through the hydroxyl radical pathway is limited. At lower pH, the ratio of hydroxyl radical to ozone tends to be lower than at higher pH. The lowering of pH can also be problematic because it can result in poor or incomplete degradation of organic substrates, which can lead to the formation of various hazardous brominated organic compounds. Furthermore, for high alkalinity wastewaters, the lowering of pH is not economically feasible.
Li et al. [19] studies confirmed that bromate formation increased significantly in ozonation alone as pH was increased from 6.3 to 9.5. This can be due to fact that in alkaline medium (i) OH − shifts the acid/base equilibria of HOBr (pK a = 8.8) towards OBr − , which reacts readily with both O 3 and HO • [1], and (ii) OH − decomposes O 3 to HO • radicals, which enhances BrO − 3 formation. Their Ce 66 -MCM-48/O 3 system minimized BrO − 3 formation and was also pH dependant. For pH range of 7.6-8.6, a higher minimization efficiency of 87-91% was attained by Ce 66 -MCM-48 after 10 min of ozonation. With a decrease in pH to 6.3, the inhibition efficiency decreased to 76%. When the pH was increased to 9.5, the minimization efficiency of Ce 66 -MCM-48 reduced to 82%. At high pH, OBr − is the major species. It reacts rapidly with both O 3 and HO • to form significant amounts of BrO − 3 . The experiments conducted by T. Zhang et al. [22] at controlled pH revealed that BrO − 3 yield increased rapidly in both single ozonation and in the O 3 /CeO 2 system as the pH was increased from 5.5 to 8.9. An 84% reduction in BrO − 3 yield was achieved at pH 6.2. They attributed the catalytic activity and BrO − 3 reduction to the surface charge of CeO 2 and intermediary HOBr/OBr − speciation, which are pH dependent. When the pH of the solution is close to the pH pzc of CeO 2 (6.6), its surface is not charged. If solution pH is below the pH pzc of CeO 2 its surface becomes positively charged, due to protonation of its surface hydroxyl sites by water. This condition increases the proportion of HOBr, hence minimizing BrO − 3 formation. If solution pH is above the pH pzc of CeO 2 its surface becomes negatively charged due to deprotonation of surface hydroxyl sites, thus continuously increasing the quantity of OBr − , which favours the formation of bromate ion.
Wu et al. [27] monitored BrO − 3 formation at different pH values during single ozonation and ozonation with nano−TiO 2 . Their results indicated that ozonation with nano−TiO 2 favoured the formation of BrO − 3 as solution pH increased initially from 6.0 to 7.9. They also concluded that at high pH, rapid ozone decomposition is favoured, hence increasing production of hydroxyl radicals, resulting in higher BrO − 3 formation. A higher proportion of OBr − is present at pH 7.9, which would also promote BrO − 3 formation. The increasing pH led to a slight decrease in BrO − 3 formation rate from 73.75% to 71.32%, displaying a reduced activity for nano−TiO 2 .
Xue et al. [31] observed that the initial solution pH had a significant influence on bromate formation during ozonation in the presence of Mn 100 -MCM-41(1 K min −1 ). The inhibition efficiencies for bromate formation were 96.7%, 83.4% and 68.2% at pH 6.5, 7.5 and 9.5 respectively. The increase in bromate formation with pH, is influenced by the equilibrium of HOBr/OBr − and the stability of ozone in aqueous media. The increasing pH favours the formation of more OBr − ions, which readily decomposes O 3 to form HO • radicals, therefore, accelerating bromate formation. In acidic conditions, ozone is stable and more HOBr is present, therefore, bromate formation is suppressed [60].
Chen et al. [33] observed that by increasing the initial solution pH from 3.0 to 9.0 increased bromate formation for both uncatalysed and Fe-Cu-MCM-41 catalysed ozonation, however, for the entire pH range Fe-Cu-MCM-41/O 3 process generated lower bromate yield. As the pH increased to 9.0, bromate steadily accumulated, reaching 913 ppb in single ozonation and 335 ppb in Fe-Cu-MCM-41 ozonation. At the acidic condition, HBrO is favoured (pH < pK a ), and since O 3 is more stable, fewer HO • radicals are formed. As HBrO predominantly reacts with HO • , the oxidation pathway 2 in Figure 2 is suppressed and a reduced amount of bromate is formed. Under basic conditions, the equilibrium shifts towards BrO − , which is highly reactive towards both O 3 and HO • , resulting in accelerated bromate production [35].
Zhang et al. studied the influence of pH on bromate formation for the O 3 /HZSM-5 system [55]. In ozonation alone, it was observed that as solution pH increased from 6.6 to 9.3, the bromate yield increased rapidly from 4.9 ppb to 27 ppb. In catalytic ozonation with HZSM-5, the bromate yield increased more steadily from 2.8 ppb to 9.4 ppb. They attributed the drop in bromate formation to the adsorption of BrO − on HZSM-5 at different pH levels. Considering the equilibrium constant of 10 -9 for HOBr/OBr − , the fraction of BrO − in HOBr/OBr − at pH 8.0 and pH 9.3 is approximately 14% and 76%, respectively. This would mean that higher amounts of BrO − can be adsorbed on HZSM-5 at pH 9.3 than at pH 8.0, so would the bromate reduction efficiency. However, their results have shown that the percent reduction in bromate formation increased only by 7.6%, when the solution pH was raised from 8.0 to 9.3. Since BrO − is more reactive towards ozone than HOBr, and the HO • /OBr − reaction rate is approximately two times that of HO • /HOBr [1]. Therefore, the increase in pH leads to a substantial increase in bromate yield in single ozonation. In HZSM-5 ozonation, O 3 and HO • compete with HZSM-5 for BrO − , thus resulting in lower bromate formation at higher pH.
Kishimoto and Nakamura [61] concluded from their studies that hydroxyl radicals are more crucial than molecular ozone in bromate production. They demonstrated that in ozonation alone, BrO − 3 yield increased as Br − concentration decreased at neutral pH in the absence of 4-chlorobenzoic acid (4-CBA). However, BrO − 3 yields considerably decreased compared to Br − removal at acidic pH and in the presence of 4-CBA. Although acidic pH decreased BrO − 3 generation, it limited the oxidation capacity of ozone for successful 4-chlorobenzoic acid degradation. Therefore, the acidification during ozonation is favorable for BrO − 3 minimization, but it has the disadvantage of affecting the removal efficiency of organic pollutants from water.

Effect of Initial Bromide Concentration
Several studies have shown that the presence of small quantities of bromide ion can result in the generation of significant amounts of bromate ion during single ozonation. Bromate ion yield increased as bromide ion concentration increased. A few studies were conducted to investigate the influence of initial bromide concentration on the bromate formation during catalytic ozonation.
Wu et al. [27] examined BrO − 3 formation for various initial Br − concentrations during single ozonation and ozonation with nano−TiO 2 . The data indicated that in single ozonation BrO − 3 yield increased rapidly as a function of initial Br − concentration, however in ozonation with nano−TiO 2 the BrO − 3 yield was significantly lower. When initial Br − concentration increased from 0.4 ppm to 1.2 ppm, the reduction rate of BrO − 3 decreased from 67.22% to 47.11%, suggesting that the activity of nano−TiO 2 is severely inhibited with an increase in initial Br − concentration.
The experiments conducted by T Zhang et al. [22] to study the influence of initial bromide concentration on bromate production showed that in single ozonation BrO − 3 yield increased rapidly from 0.5 ppm to 2 ppm, as the concentration of bromide ion increased. In CeO 2 catalysed ozonation, BrO − 3 formation was significantly suppressed for Br − concentrations ≤ 1.0 ppm, however, for Br − concentrations > 1.0 ppm, BrO − 3 yield started to increase rapidly. The BrO − 3 yield in CeO 2 catalysed ozonation was always lower than that obtained with uncatalysed ozonation.

Effect of Ozone Dosage
Sufficient availability of ozone showed an increase in the bromate ion formation, until all bromide ion was converted to bromate ion [58]. von Gunten and Hoigne [18] have introduced a standard measure for the ozone concentration (C) as a function of reaction time (t), which is defined as the Ct value (mg/L·min) for ozone exposure. An increase in the quantity of ozone improves the Ct value during ozone treatment of water. Wu et al. [27] demonstrated that BrO − 3 yield kept on increasing as ozone concentration was increased in both single ozonation and nano−TiO 2 ozonation, that is, for all experiments BrO − 3 formation increased linearly as the 'Ct value increased. When ozone dosage was increased from 2.22 ppm to 4.62 ppm, an improvement in the BrO − 3 reduction rate from 62.94% to 75.66% was observed. The BrO − 3 formation rates in single ozonation were found to be much higher than in catalytic ozonation, however, no explanation was given for this trend.
Zhang et al. [55] showed that bromate yield increased rapidly from 7.8 ppb to 95 ppb in single ozonation as the ozone concentration was increased from 0.38 ppm to 1.16 ppm. In catalytic ozonation with HZSM-5, bromate yield increased much slower (from 4.3 ppb to 21 ppb) for the same increase in ozone dose. HZSM-5 may have depleted the concentrations of ozone and/or intermediate species, which are needed for bromate formation.

Influence of Temperature Changes
The increasing temperature generally increases bromate ion production in water during ozonation. The effects of temperature are due to the following facts: (i) Ozone decomposition into HO • radicals is favoured at higher temperatures; (ii) an increase in temperature enhances the reaction rate and (iii) the pK a of the HOBr/OBr − system is temperature dependent.
The experimental data showing the influence of solution temperature on bromate minimization efficiency indicated that in the temperature range of 15°C to 30°C Ce 66 -MCM-48 catalytic ozonation showed nearly the same minimization efficiency as that of single ozonation [19]. This temperatureindependent feature of Ce 66 -MCM-48 is advantageous for water treatment by ozonation.
The influence of solution temperature on BrO − 3 formation showed that, in single ozonation, the BrO − 3 yield increased moderately when the temperature was increased from 5 • C to 15 • C, and increased more sharply when raised from 15 • C to 25 • C. The generation of BrO − 3 in CeO 2 ozonation was found to be similar to single ozonation, however much less BrO − 3 was produced in CeO 2 ozonation [22].

Influence of Catalyst Dosage
Generally, the bromate yield increases as a function of catalyst dose. For example, bromate production with increasing nano−TiO 2 dosage (0 to 200 ppm) investigated by Wu et al. [27] showed that when nano−TiO 2 dose was increased from 0 to 100 ppm, the BrO − 3 reduction rate increased from 0% to 72.59%. However, when nano−TiO 2 dose increased from 100 to 200 ppm, the BrO − 3 reduction rate only went up to 74.27%. The nanoparticles have extremely high surface area, therefore, increasing nano−TiO 2 dosage would result in more active catalytic sites for surface reactions. However, in aqueous solution, ozone concentrations are limited, hence the marginal increase in BrO − 3 reduction rate.

Conclusions and Recommendations
The literature indicates that catalytic ozonation using appropriate catalyst materials is a better solution for bromate minimization than uncatalysed ozonation. However, there is still a need for more efficient and practically applicable catalysts to be explored for complete elimination of bromate formation during ozonation. All catalysts reported were able to significantly minimize BrO − 3 formation in comparison to ozonation alone, however, only few were able to minimize bromate formation below the 5 ppb limit. The following bromate inhibition strategies/mechanisms during catalytic ozonation of bromide containing waters were proposed: • Increasing the number of hydroxyl groups on the catalyst surface resulted in enhanced ozone decomposition to HO • radicals, thus limiting the contribution of direct O 3 for the sequential oxidation of Br − → HOBr/OBr − → BrO − 3 . The formation of excess HO • is beneficial for removal of organic pollutants from the water.

•
Redox reactions on the catalyst surface causes inhibition of Br − → HOBr/OBr − and in some cases reduction of BrO • to HOBr/OBr − , thus limiting bromate formation. The lesser HOBr/OBr − concentration leads to lesser BrO − 3 .

•
The generation of hydrogen peroxide was detected in most catalytic ozonation systems, but was found to be lower than in ozonation alone. The lesser H 2 O 2 means lesser HO • radicals, therefore, the oxidation rate of HOBr/OBr − to BrO • to BrO − 3 is diminished. Contrary to this, some authors observed an increase in H 2 O 2 , which they attributed to the reactive oxygen species, which are capable of consuming HOBr/OBr − . Further work on the relationship between H 2 O 2 generation and bromate inhibition is therefore needed.

•
The presence of phosphate and humic acid had a tendency to limit bromate formation, however, high levels of phosphate and humic acid can result in poor water quality.

•
The limited studies on photocatalytic ozonation of bromide containing waters showed that the concentration of hypobromite species can be minimized by the photoelectrons generated on the photocatalyst surface, thus contributing to bromate reduction.

•
Bromate reduction was enhanced in the presence of certain organic compounds, due to electron transfer reactions on the catalyst surface. • Some catalysts have an affinity to adsorb critical intermediate species (OBr − ) needed for bromate formation.

•
Mixed metal oxides were found to effectively minimize bromate formation by simply lowering the initial solution pH to more acidic levels.
Author Contributions: For this review article, while all the required literature material was collected and draft compiled by A.N.G., the final manuscript was prepared and edited by S.B.J.

Funding:
No funding is received for the preparation of this review article.