The Oxidation Mechanism of Aniline by Ozone Water and Ozone Micro-nano Bubble Water and Its Inuencing Factors

9 Aniline is a kind of refractory contaminant that is difficult to be degraded by 10 microorganisms. Ozone is a green and efficient reagent to oxidize aniline, while the 11 ozone oxidation efficiency is restricted by the low ozone mass transfer rate. Micro-nano 12 bubble ozonation has been developed as a new method to significantly improve the 13 ozone utilization rate, while the characteristics of ozone micro-nano bubble when 14 compared with dissolved ozone is not clear. The paper carried out batch experiments to 15 research the oxidation effect of aniline by ozone water (OW) and ozone micro-nano 16 bubble water (OMNBW), and found that the degradation rate of aniline by OMNBW 17 was 2.8~5.9% higher than that by OW. The increase of pH had a negative effect on the 18 degradation of aniline by OW and OMNBW. SO 42- , Cl - , HCO 3- and Mg 2+ could inhibit 19 the degradation efficiency by 0.04%, 0.99%, 0.44% and 10.4% for OW, while the ratios 20 were 1.1%, 6.4%, 4.1% and 1.5% for OMNBW. The addition of humic acid and fulvic 21 acid could decrease the oxidation rate of aniline by 35% and 49% for OW, while the 22 ratios were 41% and 62% for OMNBW. Through quenching experiment, it was found 23 that the direct oxidation by ozone molecules and the indirect oxidation by superoxide 24 radicals were main pathways for aniline oxidation by OW and OMNBW. This work 25 provided a practical guide for the application of OMNBW in wastewater and groundwater treatment process.


30
Aniline is an important chemical material and widely used in dyes, pesticides,  Environmental Protection Agency (USEPA) (Trautwein et al. 2015). In addition, aniline 37 is difficult to degrade and easy to maintain high toxicity even at low concentrations 38 (Orge et al. 2017). Therefore, it is necessary to develop water treatment technologies to 39 degrade aniline in order to meet the relative standards. 40 The remediation technology of polluted water included physical adsorption, 41 biological process and chemical oxidation. However, for wastewater containing 42 complex components, the efficiency of physical adsorption was very limited (Liu et al. 43 2009). For relatively persistent pollutants such as aniline, biological methods were 44 inefficient and usually spent more time than chemical methods (Ikehata et al. 2008). 45 Chemical oxidation was one of the most commonly used methods for water treatment 46 by different reagents, such as Fenton reagent, potassium permanganate, persulfate and 47 ozone. However, the oxidation efficiency of Fenton process was highly dependent on 48 pH and was easily affected by the various substances in actual water, such as chloride 49 and bicarbonate (Neyens and Baeyens 2003). In addition, hydrogen peroxide was 50 unstable with a short half-life (Clarizia et al. 2017). Persulfate must be activated before 51 it could be used for water treatment, but both catalyst and sulfate would produce a large 52 number of by-products, resulting in secondary pollution (Hou et al. 2012). Chemical 53 oxidation by permanganate would produce MnO2, which might also lead to secondary 54 pollution (Li and Schwartz 2005). 55 Compared with other treatment technologies, ozone had been widely applied in 56 the treatment of drinking water and reclaimed water (Miao et al. 2015, Shen et al. 2008, 57 because it produced fewer secondary pollutants than other oxidants. When ozone was 58 applied to oxidize contaminants in water, ozone gas was often injected directly by gas 59 sparger at the designed place. In this situation, the oxidation effect could only be 60 significant in its impact radius, which was near the injection point ( MNBs made it more stable in water (Shangguan et al. 2018). In addition, MNBs had 70 large specific surface area and high interior gas pressure, which was related to the high 71 mass transfer rate of ozone from the gas phase to the liquid phase as well as the high

73
When MNBs shrunk in water, charged ions quickly concentrated and enriched on a very 74 narrow bubble interface, so MNBs had higher zeta potential (Takahashi et al. 2007).

75
MNBs also had stronger migration ability, which had the potential ability to overcome        The pseudo first-order kinetic model could be used to show the difference of self-165 decomposition rate between OMNBW and OW as Eq. (1).
where C was the dissolved ozone concentration (mg/L), k1 was the first-order 168 decomposition coefficient (s -1 ), t was the reaction time (s). Eq.
(1) could be integrated 169 to obtain the following equation: where Ct was the dissolved ozone concentration at time t (mg/L), C0 was the initial The k1 value for OMNBW and OW were 0.016 min -1 and 0.033 min -1 respectively, 177 which showed that the self-decomposition of ozone in OMNBW was much slower than   When the initial aniline concentration was 0.5 mg/L, aniline was completely 199 degraded within 30 s by both OMNBW and OW. When the initial aniline concentration 200 was 1 mg/L, the degradation rate of aniline by OMNBW reached 79.8% in 5 min, which 201 was 5.9% higher than that by OW. When the initial aniline concentrations were 2 mg/L 202 and 4 mg/L, the degradation rate by OMNBW were 51.1% and 32%, which were 2.8% 203 and 3.1% higher than that by OW respectively. The results showed that although aniline 204 could react with ozone molecules directly, the removal rate of aniline by OMNBW was 205 still higher than that by OW when the dissolved ozone concentration was the same.  pollutants. Therefore, the degradation rate of aniline by OMNBW when pH was 9 was 231 only 1.3% lower than that when pH was 8 (Fig. 4b), which indicated that the oxidation 232 of aniline by OMNBW was less affected by the solution pH when compared with OW.

267
The influence of HA and FA on aniline oxidation by OMNBW and OW was shown 268 in Fig. 6, which showed that both HA and FA had great inhibition effects on the 269 oxidation of aniline by OMNBW and OW. When HA was added in the solution, the 270 degradation rate of aniline by OW and OMNBW decreased by 35% and 41%, while FA 271 inhibited the aniline degradation rate by 49% and 62% respectively, indicating that FA 272 with small molecules would react with oxidants more easily (Wang et al. 2017). 273 According to previous research, HA and FA could be oxidized by ozone, which 274 competed with aniline and reduced the aniline oxidation rate as well ).

275
The ozone oxidation mechanism by OMNBW was mainly divided into the direct 276 oxidation by molecular ozone dissolved in water and the "interface degradation" after  290 Previous data in Section 3.2 showed that at the same dissolved ozone concentration,

291
OMNBW had a stronger removal rate of aniline than OW, which might be caused by 292 the free radicals that was formed in OMNBW. TBA was very easy to react with •OH 293 (kTBA-•OH = 6.0×10 8 M −1 s −1 ), but it hardly reacted with ozone (kTBA-O3 = 3.0×10 -3 M -1 s -294 1 ) (Nawrocki 2013, Nawrocki and Kasprzyk-Hordern 2010). Therefore, TBA was 295 applied to identify the existence of •OH in OMNBW and OW. Fig. 7 showed that the 296 aniline oxidation rate by OW only decreased by 0.30% after the addition of TBA as 297 scavenger, while the oxidation rate by OMNBW only decreased by 0.15%. It indicated 298 that indirect pathway by •OH took a very small fraction in the aniline oxidation process.

299
In addition, previous studies reported that TBA tended to disperse in aqueous solution 300 and was not efficient in capturing radicals near bubble surface (Yao et al. 2020).