Removal of phenol from wastewater by high-gravity intensified heterogeneous catalytic ozonation with activated carbon

In this study, the high-gravity technique is used to intensify the heterogeneous catalytic ozonation with activated carbon (AC) as the catalyst for removal of phenol from wastewater in a rotating packed bed (RPB), and the effects of high-gravity factor, inlet O3 concentration, liquid–gas ratio, and initial pH on the degradation and mineralization of phenol at room temperature are investigated. It is revealed that the degradation rate of phenol reaches 100% at 10 min and the removal rate of total organic carbon (TOC) reaches 91% at 40 min under the conditions of high-gravity factor β = 40, inlet O3 concentration = 90 mg·L−1, liquid flow rate = 80 L·h−1, and initial pH = 11. Compared with the bubbling reactor (BR)/O3/AC and RPB/O3 systems, the mineralization rate of phenol by the RPB/O3/AC system is increased by 24.78% and 34.77%, respectively. Free radical quenching experiments are performed using tertiary butanol (TBA) and benzoquinone (BQ) as scavengers of ·OH and O2−, respectively. It is shown that the degradation and mineralization of phenol are attributed to the direct ozonation and the indirect oxidation by ·OH generated from the decomposition of O3 adsorbed on AC surface, respectively. ·OH and O2·− are also detected by electron paramagnetic resonance (EPR). Thus, it is concluded that AC-catalyzed ozonation and high-gravity technique have a synergistic effect on ·OH initiation, which in turn can significantly improve the degradation and mineralization of organic wastewater.


Introduction
Phenol is a toxic organic compound with numerous applications in coal chemical, petrochemical, paper-making, dyeing, and pharmaceutical industries (Yang et al. 2019a;Zhang et al. 2020a). However, it is highly carcinogenic, teratogenic, and mutagenic in nature that may cause severe damage to both humans and animals even in a low amount (Naguib and Badawy 2019). Approximately 0.6 tons of phenol wastewater would be produced per ton of phenol obtained, and 0.16-0.24 million tons of phenol wastewater is released globally in 2021 (Huang et al. 2015). Now, phenol has been listed as a priority pollutant in the USA and China.
Therefore, it is of great significance to seek an economical and effective method to treat phenol wastewater (Bhausaheb and Korake 2016;Zhang et al. 2020a, b).
Catalytic ozonation has attracted considerable attention as an efficient wastewater treatment method. In catalytic ozonation, ·OH is produced from catalytic decomposition of O 3 , which has strong oxidative capacity (E 0 = 2.8 eV) and can react non-selectively with organic pollutants in wastewater (Jaafarzadeh et al. 2017). Importantly, most catalysts used in heterogeneous catalytic ozonation have the advantages of high catalytic activity (Ahmed et al. 2010), easy separation from the liquid phase, high reusability, and no secondary pollution to the environment (Nawrocki 2013;Wang and Chen 2020). Commonly used heterogeneous catalysts include carbon materials, metal oxides, and supported metals or metal oxides (Afzal et al. 2020(Afzal et al. , 2016Xta et al. 2020). Activated carbon (AC) is a desired carbon-based heterogeneous catalyst for its unique advantages such as high efficiency, high stability, large specific surface area, and low cost (Ma et al. 2016). In heterogeneous catalytic ozonation, AC can act not only as a catalyst but also as an adsorbent (Abdedayem et al. 2015;Wang et al. 2019a, b). Jans and Hoigné (1998) found that AC promoted the decomposition of dissolved O 3 into hydroxyl radicals. Mousavi et al. (2017) compared ozonation and AC-catalyzed ozonation for reducing chlorophyll a, turbidity, UV 254 absorbance, dissolved organic carbon (DOC), and color from eutrophic water. It was found that chlorophyll a was reduced from 49.5 to 9.7 and 4.7 µg L −1 by ozonation and AC-catalyzed ozonation, respectively; ozonation led to a 36% increase in DOC, while AC-catalyzed ozonation led to the removal of 76.0% of DOC; and the decline in water color and turbidity was more significant in AC-catalyzed ozonation than in ozonation. However, the degradation of organic pollutants is limited in conventional stirred reactors, because the mass transfer of O 3 from the gas phase to the liquid phase, which is the control step of heterogeneous catalytic ozonation, is low because of insufficient gas-liquid contact (Wang et al. 2019a, b).
There is evidence that ultrasonic, electric, and high-gravity fields can act synergistically with catalysts to improve the decomposition and mass transfer of O 3 by increasing the gas-liquid-solid contact area (Qian et al. 2017). The rotating packed bed (RPB) is a novel gas-liquid reactor with excellent micromixing performance, high mass transfer efficiency, and small size (Jiao et al. 2016). A high-gravity field is created by the centrifugal force resulting from the high-speed rotation of the packing, so that liquid can be sheared into films, threads, and droplets in order to improve gas-liquid dispersion and mixing, reduce mass transfer resistance, and improve mass transfer efficiency. In short, the high-gravity field has the potential to intensify the heterogeneous catalytic ozonation (Jiao et al. 2010;Wu et al. 2017). Li et al. (2020a, b) used RPB instead of a conventional reactor for FeOOH-catalyzed ozonation of nitrobenzene in aqueous solution, and the results showed that RPB could improve the mass transfer rate and oxidative activity of O 3 . However, there have been few studies on AC-catalyzed ozonation for removal of phenol from wastewater in a high-gravity field.
In this study, AC is used as the packing of RPB to provide contact and reaction sites for gas, liquid, and solid phases and to catalyze the decomposition of O 3 . The effects of high-gravity factor (β), liquid-gas ratio (Q L /Q G ), inlet O 3 concentration (C O3 ), initial phenol concentration, and initial pH on the removal rate of phenol and TOC are investigated. In order to understand the synergistic effects of RPB and heterogeneous catalytic ozonation with AC, the degradation and mineralization of phenol by RPB/O 3 , RPB/O 3 /AC, BR/O 3 /AC, and RPB/AC systems are compared. In order to understand the mechanism of heterogeneous catalytic ozonation with AC in RPB, free radical quenching experiments are performed using TBA and BQ as the scavengers of ·OH and O 2 − , respectively.

Experimental setup
The experiment was carried out by semi-batch operation at an ambient temperature of 25 °C ± 2 °C in RPB, as shown in Fig. 1. Briefly, 2 L of phenol solution with a concentration of 100 mg/L was pumped into RPB at a flow rate of 40-100 L/h and sprayed on the inner edge of the rotor through the liquid distributor. After that, the liquid was sheared into microelements as it flowed in the circumferential direction under the action of centrifugal force. Then, 25-90 mg/L ozone was fed from the bottom of the RPB at a flow rate of 60 ~ 120 L/h and fully contacted with the liquid in a cross-flow manner. The reacted phenol solution flowed back to the storage tank under the action of gravity for recycling, and the effluent gas was absorbed by 5% KI solution. Samples were collected at different times for further analysis and characterization. Control experiments were conducted using Pall ring packing in a conventional bubbling reactor.

Analytical methods
The gaseous ozone concentrations at the inlet and outlet were measured by a gaseous ozone detector (3S-J5000; China). The solution pH was measured by a pH meter (PHS-3C; China). The concentrations of phenol and intermediate products were determined by high-performance liquid chromatography (HPLC; Dionex UltiMate 3000) with a C18 reversed column, where the wavelength of the UV detector was set at 276 nm and the mobile phase was the CH 3 OH/DI water (60: 40) with a flow rate of 0.9 mL min −1 . The TOC concentrations were determined by a TOC analyzer (Aurora, O.I. Analytical). The removal efficiency of phenol and TOC was evaluated using the following equation (Jain et al. 2020).
where (%) is the removal efficiency of phenol (or TOC), and C 0 and C are the concentrations of phenol (or TOC) at time t = 0 or t (mg L −1 ), respectively.

Characterization of AC
The surfaces and structural properties of AC before and after ozonation were characterized by scanning electron microscopy (SEM) and Fourier transform infrared spectroscopy (FTIR), as shown in Fig. 2. It is seen from Fig. 2A that the AC surface is spongy with many porous structures and irregularly distributed cavities, which indicates that AC has a high porosity (Banat et al. 2010;Rivera-Utrilla et al. 2011). After ozonation, the AC surface is relatively smooth with fewer protruding particles (Fig. 2B), which may be due to the influence of hydroxyl oxidation and the presence of weak acid phenol solution on the surface of AC that may cause the shedding of ash from the surface. However, it is still as fluffy and porous as the raw material, and no significant changes are observed in the surface structure, which indicates that AC has good stability and durability. In Fig. 2C, the basic characteristics of AC before and after ozonation are as follows: the peak at 3425 cm −1 is attributed to the presence of phenolic hydroxyl groups and carboxylic acid on the AC surface; the peaks at 2924 cm −1 and 2847 cm −1 are attributed to the symmetric and asymmetric stretching vibration of C-H in -CH 2 , respectively; the peak at 1627 cm −1 is attributed to the stretching vibration of C = C, while the peak at 1044-1049 cm −1 is attributed to the stretching vibration of CO (Lei et al. 2007;Singh et al 2020). O 3 molecules can react with oxygen-containing groups on the AC surface to form a five-membered ring structure via hydrogen bonds, which in turn is decomposed into free radicals via electron transfer. However, a new intense peak is observed at 1381 cm −1 after ozonation compared with AC before use, which is attributed to the oxide species resulting from the interaction of O 3 with active sites. The intensity of the characteristic peak at 3425 cm −1 and 1044 cm −1 is increased, because new surface hydroxyl groups are formed from the interaction between the catalyst and aqueous O 3 (Wu et al. 2018). These results indicate that AC provides a catalytic environment for decomposition of O 3 into ·OH and consequently enhance the degradation of phenol in wastewater.

The effects of high-gravity factor
The high-gravity factor β is a dimensionless measure of the intensity of the high-gravity field, and it is defined as the ratio of inertial acceleration to gravitational acceleration (Yang et al. 2019a, b): (2) = G g = r 2 g = N 2 r 900 where ω is the rotational angular velocity of the rotor (s −1 ), r is the average radius of the rotor (m), g is the gravitational acceleration (9.8 m 2 ·s −1 ), and N is the rotational speed of the rotor (rpm). Figure 3A shows that the removal rate of phenol increases with the increase of high-gravity factor, and at 10 min it reaches 100% at β = 40 but only 80% at β = 10. It is clear that a higher high-gravity factor can make the droplets smaller and the films thinner in RPB, and thus increase the volume mass transfer coefficient and the decomposition rate constant of O 3 . Because of this, more O 3 is dissolved in the solution, resulting in an increase in liquid-phase O 3 concentration and subsequently AC-catalyzed decomposition of O 3 into •OH (Qin et al. 2018). However, it should be noted that at β = 40, the removal rate of phenol reaches 100% at 10 min, while that of TOC is only 30% due to the generation of intermediate products in large quantities during the degradation of phenol (Fig. 3B). Thus, the mineralization rate of phenol should also be taken into consideration. It is revealed that as the high-gravity factor increases from 10 to 50, the mineralization rate of phenol at 40 min is increased from 31 to 78%. This is because the gas-liquid-solid contact area is increased as the wastewater is sheared into finer droplets, threads, and films at a higher centrifugal force. As the high-gravity factor increases, the gas-liquid-solid interface can also be renewed faster, which leads to the transfer of more O 3 into the solution and consequently the generation of more ·OH from the decomposition of O 3 . It is noted that the mineralization rates After use C Fig. 2 A, B The SEM image and C FTIR spectra of AC before and after ozonation of phenol are similar between β = 40 and β = 50, because the gas-liquid-solid contact area has reached the peak. Thus, a high-gravity factor higher than β = 40 does not necessarily guarantee high mineralization efficiency of phenol. Given the high energy consumption and cost at high high-gravity factor, the optimal high-gravity factor is set to β = 40. Figure 4A shows that as the inlet O 3 concentration is increased from 25 to 90 mg·L −1 , the removal rate of phenol at 40 min is increased from 44 to 100%. Figure 4B shows that at a given inlet O 3 concentration, the mineralization rate of phenol by the RPB/O 3 /AC system is increased as the reaction proceeds. It is also seen that the mineralization rate of phenol at 40 min is 41% at C O3 = 25 mg·L −1 but 91% at C O3 = 90 mg·L −1 . According to Henry's law, the equilibrium concentration of O 3 in water ∫ t 0 O 3 dt will increase with increasing inlet O 3 concentration, and thus more ·OH will be produced from AC-catalyzed decomposition of O 3 . Elovitz and Gunten (1999) have revealed that the ·OH concentration is proportional to ∫ t 0 [OH] and ∫ t 0 O 3 dt . Increasing the gasphase O 3 concentration can also increase its partial pressure and thus the dissolution of O 3 , and the dissolved O 3 can contact sufficiently with the functional groups on the AC surface to produce ·OH (Álvarez et al. 2005;Shao et al. 2021). Thus, both the degradation and mineralization rates of phenol are significantly improved at higher inlet O 3 concentrations.

Figure 5
Initial ) pH is expected to have an effect on the decomposition of O 3 into ·OH by affecting the charges of hydroxyl groups present on the catalyst surface. In this study, the effects of initial pH on the removal rate of phenol and TOC by the RPB/O 3 /AC system are investigated at pH = 3.0 ~ 11.0. As shown in Fig. 6A and B, the removal rate of phenol first increases rapidly with the increase of pH, and it reaches 100% within 20 min at a phenol concentration of 100 mg/L, which is in good agreement with previous results. It is also seen that the removal rate under alkaline conditions is much higher than that under acidic conditions. This is because O 3 molecules can react with hydroxide to form ·OH under alkaline conditions that can facilitate the ozonation process (Staehelin and Hoigne 1982); while under acidic conditions, O 3 molecules are less likely to be decomposed into free radicals, and only direct ozonation occurs for degradation of organic pollutants by breaking the double bond or the aromatic ring (Turhan and Uzman 2008;Xiong et al. 2020). In addition, ·OH has higher oxidative capacity than O 3 as its redox potential (E 0 = 2.80 V) is higher than that of O 3 (E 0 = 2.07 V). Zeng et al. (2013) found that the reaction rate of phenol with ·OH (k = 2.10 × 10 9 -4.50 × 10 9 M −1 s −1 ) was much higher than that with O 3 (k = 1.6 × 10 3 M −1 s −1 ). Thus, an alkaline environment is more favorable for ACcatalyzed ozonation. Figure 6B shows that the removal rate of TOC varies in the range of 70 ~ 82%, and thus it is concluded that initial pH has smaller effects on the mineralization of phenol than on the degradation of phenol by the RPB/O 3 /AC system. This is because large amounts of small carboxylic acid molecules are produced during the degradation process of phenol, such as oxalic acid, maleic acid, and fumaric acid (Zhang et al. 2021a, b), which make the reaction system more acidic. Figure 6A shows that the phenol concentration decreases rapidly within the first 5 min and then stabilizes gradually probably due to the decrease of pH. Thus, the RPB/O 3 /AC system could not take full advantage of its high oxidative capacity under alkaline conditions. The adsorption capacity of AC also plays a role in the removal of TOC. However, Miao et al. (2013) found that an alkaline environment was not conducive to the adsorption of phenol by AC. For these reasons, initial pH has a significant effect on the removal of phenol by the RPB/O 3 /AC system but no obvious effect on the mineralization of phenol.

The effects of liquid-gas ratio
As the mass transfer of O 3 is liquid film controlled, there is a need to increase the turbulence of the liquid in order to enhance the mass transfer of O 3 . It is known that the higher the liquid flow rate is, the high the volume mass transfer coefficient of O 3 and ∫ t 0 O 3 dt will be (Chen et al. 2005). However, the residence time of the liquid in RPB that depends heavily on the liquid-gas ratio would be greatly reduced in case of high turbulence, which can adversely affect the contact between O 3 and the liquid and subsequent ozonation of phenol. In this experiment, the gas flow rate is set to Q G = 60 L·h −1 and the liquid flow rate is varied to obtain different liquid-gas ratios. Figure 6A and B show that the removal rates of phenol and TOC increase with the increase of Q L /Q G , indicating that the benefits brought about Fig. 5 The effects of initial pH on A the removal rate of phenol and B the mineralization rate of TOC (T = 25 °C, C B = 100 mg·L −1 , m s = 25 g·L −1 , Q G = 60 L·h −1 , β = 40, C O3 = 75 mg·L −1 , Q L = 80 L·h −1 ) Fig. 6 The effects of liquid-gas ratio on A the removal rate of phenol and B the mineralization rate of TOC (T = 25 °C, C B = 100 mg·L −1 , m s = 25 g·L −1 , Q G = 60 L·h −1 , β = 40, C O3 = 75 mg·L −1 , pH = 6) by turbulence outweigh the negative impacts brought about by the short residence time. As the Q L /Q G ratio increases from 40/60 to 100/60, the mineralization rate of phenol at 40 min is significantly increased from 49 to 88%, which can be attributed to the excellent dispersion of liquid in the packing and the high mass transfer efficiency because of the high liquid flow rate in the cross flow . Thus, the optimal liquid-gas ratio is set to 100/60 in this experiment.

Comparison of different technologies
The removal rates of phenol and TOC by BR/O 3 /AC, PRB/ O 3 /AC, BR/O 3 , and PRB/O 3 systems are compared. Figure 7A shows that for the RPB/AC system, the removal rate of phenol reaches equilibrium after 20 min and reaches a saturation state (30%) after 40 min, indicating that AC is capable of adsorbing phenol. Thus, AC can act not only as an excellent adsorbent but also as an excellent catalyst. For the RPB/O 3 system, the removal rate of phenol reaches 100% after 30 min, because O 3 can easily undergo electrophilic substitution reactions with unsaturated aromatic compounds like phenol (Xiong et al. 2020) due to its high oxidative capacity. For the RPB/O 3 /AC and BR/O 3 /AC systems, the removal rate of phenol reaches 100% after 20 min, which is attributed to the ability of AC to catalyze the degradation of intermediate products. It should be noted that the removal rate of phenol by the RPB/O 3 /AC system is higher than that by the BR/O 3 /AC system, which implies that RPB is better than BR to enhance the mass transfer of O 3 and consequently the catalytic ozonation of phenol. The rapid renewal of the liquid-solid interface in RPB provides more adsorption sites of O 3 and thus contributes to the generation of •OH and subsequent catalytic ozonation of phenol (Yang et al. 2019a, b). Thus, the highest removal rate of phenol is obtained with the use of the RPB/O 3 /AC system. Figure 7B shows the mineralization rates of phenol by BR/O 3 /AC, PRB/O 3 /AC, BR/O 3 , and PRB/O 3 systems. It is evident that the removal of TOC lags behind that of phenol.
At 40 min, the mineralization rate of phenol by the RPB/ AC system is only 23%, which indicates that AC-catalyzed ozonation rather than adsorption is the major contributor to the mineralization of phenol. The low mineralization rate of phenol by the RPB/O 3 system (41%) is probably due to the selective reaction of O 3 with organic pollutants. However, a higher mineralization rate is obtained using BR/O 3 /AC and RPB/O 3 /AC systems, because •OH is generated from AC-catalyzed decomposition of O 3 and it has a much higher reaction rate constant with phenol (2.1 × 10 9 M −1 ·s −1 ) compared with that of O 3 (Ku et al. 2011). As expected, the mineralization rate of phenol by the RPB/O 3 /AC system is 24% higher than that by the BR/O 3 /AC system, confirming again that RPB is better than BR to enhance the mass transfer and decomposition of O 3 to •OH, and AC-catalyzed ozonation and high-gravity technique act synergistically to improve the mineralization of phenol.

Free radical scavengers
During AC-catalyzed ozonation, many reactive oxygen species (ROS) are formed from the decomposition of O 3 , such as ·OH, O 2 − , and 1 O 2 with a redox potential of 2.8 V, 1.35 V, and 2.22 V, respectively (Li et al. 2020a, b;Zhang et al. 2021a, b). In the quenching experiment, two free radical scavengers, TBA and BQ, are added to eliminate ·OH and O 2 − (Li et al. 2019) in order to understand their unique contributions to the degradation of phenol by the RPB/O 3 /AC system. Figure 8 shows that the addition of excess TBA results in a 9% decrease in the degradation rate of phenol from 100 to 91% at 40 min, indicating that direct ozonation plays a more important role in the degradation of phenol than indirect oxidation by ·OH. However, the addition of excess BQ results in a 23% decrease in the degradation rate of phenol. BQ can scavenge O 3 , ·OH, and O 2 − , and as O 2 − accounts for only a small portion, O 3 also plays a main role in the degradation of phenol (Yu et al.). However, as the mineralization rate of phenol by RPB/O 3 and RPB/O 3 /AC systems is 41% and 75% at 40 min, respectively, it is inferred that the degradation and mineralization of phenol are attributed to the direct ozonation and the indirect oxidation by ·OH from the decomposition of O 3 adsorbed on AC surface, respectively.

Types of free radicals by EPR
In order to determine the reactive species for the degradation of phenol by the RPB/O 3 /AC system, the types of free radicals are determined by electron paramagnetic resonance (EPR). In Fig. 9A, the DMPO-OH signal is clearly detected with an intensity ratio of 1:2:2:1 in the RPB/O 3 /AC system (Qiao et al. 2019). In Fig. 9B, the O 2 − signal is also detected when the aqueous phase is replaced with ethanol solvent (k ·OH,EtOH = (1.2-2.8) × 109 M −1 ·s −1 ) to inhibit ·OH under the same conditions (Bing et al. 2015). This verifies that the AC

Conclusions
In this study, the high-gravity technique is used to enhance the heterogeneous catalytic ozonation with AC as the catalyst for removal of phenol from wastewater in RPB. As expected, RPB can significantly enhance the mass transfer and decomposition of O 3 to ·OH and, as a consequence, the degradation and mineralization of phenol are increased. Under the same conditions, the mineralization rate of phenol by the RPB/O 3 /AC system at 40 min (74.8%) is 24.78% higher than that by the BR/O 3 /AC system. The quenching experiment reveals that the addition of excess TBA and BQ results in 9% and 23% decrease in the degradation rate of phenol at 40 min, respectively. It is concluded that the degradation and mineralization of phenol are attributed to the direct ozonation and the indirect oxidation by ·OH from the decomposition of O 3 adsorbed on AC surface, respectively. Thus, the combination of high-gravity technique with ACcatalyzed ozonation allows for rapid degradation and deep mineralization of phenol. The free radical quenching experiment confirms that O 2 ·− and ·OH are the main reactive species involved in the mineralization of phenol. In conclusion, AC-catalyzed ozonation and high-gravity technique act synergistically to induce the formation of •OH, which in turn significantly improves the degradation and mineralization of organic wastewater.
Author contribution Shengjuan Shao and Xin Ding were in charge of the experiment. Youzhi Liu, Zhixing Li, and Jiaxin Jing analyzed the experimental data. Jingwen Zhang and Weizhou Jiao were major contributors in writing the manuscript. All authors read and approved the final manuscript. All authors of this manuscript have directly participated in the planning, execution, and analyses of this study. Data availability All data generated or analyzed during this study were included in this article.

Declarations
Ethics approval and consent to participate Not applicable.

Competing interests
The authors declare no competing interests.