Yangbo Huang*ab, Zeyu Guana, Qiang Lia, Qian Lic and Dongsheng Xiaab aSchool of Environmental Engineering, Wuhan Textile University, Wuhan, Hubei 430073, China. E-mail: [email protected] bEngineering Research Center for Clean Production of Textile Dyeing and Printing, Ministry of Education, Wuhan Textile University, Wuhan, Hubei 430073, China cChina Three Gorges Corporation, Wuhan, 430014, China

First published on 12th June 2023

Catalytic ceramic membranes (CMs) integrated with different metal oxides were designed and fabricated by an impregnation-sintering method. The characterization results indicated that the metal oxides (Co3O4, MnO2, Fe2O3 and CuO) were uniformly anchored around the Al2O3 particles of the membrane basal materials, which could provide a large number of active sites throughout the membrane for the activation of peroxymonosulfate (PMS). The performance of the CMs/PMS system was evaluated by filtrating a phenol solution under different operating conditions. All the four catalytic CMs showed desirable phenol removal efficiency and the performance was in order of CoCM, MnCM, FeCM and CuCM. Moreover, the low metal ion leaching and high catalytic activity even after the 6th run revealed the good stability and reusability of the catalytic CMs. Quenching experiments and electron paramagnetic resonance (EPR) measurements were conducted to discuss the mechanism of PMS activation in the CMs/PMS system. The reactive oxygen species (ROS) were supposed to be SO4˙− and 1O2 in the CoCM/PMS system, 1O2 and O2˙− in the MnCM/PMS system, SO4˙− and ·OH in the FeCM/PMS system, and SO4˙− in the CuCM/PMS system, respectively. The comparative study on the performance and mechanism of the four CMs provides a better understanding of the integrated PMS-CMs behaviors.

However, porous ceramic membranes often exhibit a relatively low rejection rate for many small molecular organics.8 In addition, membrane fouling is still a critical obstacle that affects membrane performance and limits the wide application of ceramic membranes.9 To solve these problems, the technology that combines advanced oxidation processes (AOPs) with ceramic membrane separation has recently drawn much attention.10,11 AOPs are generally conducted with the presence of strong oxidizing species. Common AOPs mainly include photocatalytic oxidation, electro-chemistry oxidation, or activation of superoxides (hydrogen peroxide, ozone and persulfate, etc.).12–14 In the presence of nanocatalysts, peroxides are activated to generate reactive oxygen species (ROS) with high oxidizing potential, which can achieve the catalytic degradation of recalcitrant organic pollutants.15 The catalysts can be anchored on the surface of membrane or embedded in the pore walls of the internal porous channels. The resultant catalytic membrane has dual functions of filtration and catalytic oxidation, which greatly enhances the removal efficiency of pollutants.16,17 For one thing, organic contaminants on the membrane surface can be directly oxidized, which significantly alleviates membrane fouling and even achieves in situ cleaning of the fouled membrane.18,19 For another thing, under forced filtration, fluid can bring the reactants to the active catalyst surface, which can enhance mass transfer and achieve a higher catalytic efficiency.8,20

Thereby, many researches based on catalytic membrane for wastewater treatment have been reported recently. Wang et al. verified that a newly synthesized silicate-based ceramic membrane generated hydroxyl radicals in ozonation process, and this combined process increased the removal rate of p-chloronitrobenzene by 50% compared with the ozone-alone process.21 Park et al. developed a catalytic membrane by doping iron oxide nanoparticles on a commercial ceramic membrane and combined the ozonation process for As(III) removal. The catalytic ozonation by the reactive ceramic membrane effectively oxidized As(III) to As(V), which significantly improved the removal rate of As(III) in the hybrid process.22 Zhu et al. sintered a TiO2 active layer on the Al2O3-based ceramic membrane and used the resultant membrane to treat the secondary effluent of the sewage plant. They found that compared with the original membrane, the catalytic membrane showed a higher TOC removal rate and a lower membrane fouling trend.23 In addition, other excellent studies have also discussed fabrication methods and performance of nanocomposite ceramic membranes for industrial wastewater treatment (such as bisphenol A, dyes, humic acid, etc.).24–28 Researchers generally believe that ceramic membranes with dual functions of catalysis and filtration prepared by modifying the surface of membrane or membrane pore channels have a good application prospect.10,29

Among various AOPs, peroxymonosulfate (PMS)-based oxidation technologies have been recently acknowledged as a potential solution for removing toxic and refractory organic pollutants in water.30 PMS can be easily activated to generate sulfate radical (SO4˙−), which is a highly reactive radical with high redox potential (2.5–3.1 V) and long lifetime (30–40 μs).31 Recent studies have also discovered a non-radical oxidation pathway of PMS, which has high selectivity to specific pollutants and can reduce the production of harmful by-products.32–34 These enable PMS to effectively react with the target organic pollutants including dyes, perfluorinated compounds, antibiotic and phenol.35–38 Therefore, PMS advanced oxidation coupled ceramic membrane filtration technology is a promising wastewater treatment method. However, the activation mechanism of PMS by different catalytic membranes and their performance for specific pollutants still need in-deep research.

Until now, some energy-based activation methods (e.g., heat, ultraviolet light, ultrasound and transition metal ions) and heterogeneous activators (metal oxide or carbonaceous materials) were studied to activate PMS.39 Among them, transition metal oxides have been proved to be efficient, economical and designable catalysts for the degradation of various organic pollutants. However, to the best of our knowledge, few works have reported the systematically comparation of the efficiency and mechanism of transition metal oxides combined ceramic membranes. In this study, a series of metal oxides (Co, Mn, Fe and Cu) modified ceramic membranes were prepared by a facile impregnation-sintering method. Subsequently, the morphology and properties of the resultant catalytic CMs were characterized. Then phenol was selected as the target pollutant to evaluate the performance of different catalytic CMs, and the degradation efficiency of phenol under different membranes and operation conditions was studied. The stability and reusability of the CMs were also investigated. Finally, the main active species for phenol degradation were discussed and the removal pathways were clarified. This work can provide theoretical guidance for the design of efficient metal or bimetal oxide integrated ceramic membranes.

Before conducting the phenol degradation experiments, the phenol-containing solution was circulated for 30 min to stabilize the system. Subsequently, PMS was added to the solution and start timing. At pre-determined time intervals, 3 mL sample was collected from the feed tank to a PE tube with 100 μL methanol and the concentration of phenol was determined immediately.

In addition, it was reported that a pseudo-first-order equation could be used to describe the kinetics of catalytic degradation.40 The pseudo-first-order kinetic constant (kobs) is defined as:

The initial pH of the solution was adjusted using 0.05 M NaOH and H2SO4. When evaluating the reusability of the CMs, the membrane was rinsed with 500 mL water after each experiment and dried at 60 °C before next use.

The hydrophilicity of membrane surfaces was characterized by static contact angle and the results were shown in Fig. 3a. The contact angle on the surface of all the four metal oxide modified CMs decreased from 40° to around 22°. In comparison, the modified membranes all exhibited better wettability than the pristine CM, owing to the hydroxyl groups covered on the surface of metal oxide particles.41

The zeta potential of membrane surface was measured at pH ranging from 3 to 12. As shown in Fig. 3b, they gradually decreased with increasing the pH value, which were consistent with the similar trend of metal oxide nanoparticles reported elsewhere.42,43 Compared with the pristine CM, the zeta potentials of all the modified CMs changed more obviously with pH. This could be also attributed to the hydroxyl groups (Me–OH) covered on the surface of metal oxide particles, which experienced a protonation–deprotonation process under different pH conditions. The point of zero charge (pHpzc) of pristine CM is about 6.5, while that of the modified CMs were ranging from 5.3 to 9.3.

In order to analyze the component and crystalline structure of the modified CMs, the FTIR spectrums and XRD tests were measured and CoCM was taken as an example in Fig. 4. As shown in Fig. 4a, the absorption peaks at 3300 cm−1 and 1630 cm−1 corresponded to O–H stretching and bending vibrations, which also suggested the presence of hydroxyl groups on membrane surfaces.43 And the stronger peak intensity on CoCM membrane indicated that more hydroxyl groups covered on the modified membrane surface. This was consistent with the results of the contact angle measurements. Fig. 4b showed the XRD patterns for surface component of pristine CM and CoCM. The characteristic peaks at 25.6, 35.1, 37.8, 43.4, 52.6, 57.5, 66.5, 68.2, and 76.9° could be ascribed to (012), (104), (110), (113), (024), (116), (214), (300) and (119) crystal planes of Al2O3 (PDF#78-2426). This was because the main component of the CMs was Al2O3. However, no significant signals of Co were observed in both FTIR and XRD spectrums. This may be due to the low content and high dispersion of catalysts loadings. Therefore, to clarify the specific composition of the metal oxides on the membrane, catalyst particles were synthesized under the same conditions as the preparation of CMs. Fig. S2† showed the XRD patterns of the four metal oxides. It could be seen that the catalysts coating on the four CMs were Co3O4 (PDF#74-2120), MnO2 (PDF#81-2261), Fe2O3 (PDF#79-1741) and CuO (PDF#48-1548), respectively. In addition, it should be noted that the catalysts on MnCM may also contain small amounts of Mn3O4 (PDF#76-0150) due to the characteristic peak at 33.0°.

The cross section of the CMs were also characterized by SEM-EDS and the results were depicted in Fig. 5. The SEM images clearly showed that the prepared CMs consist of two layers: a dense layer and a support layer. The thickness of the dense layer was about 15 μm. Compared to the pristine CM, the cross-section morphologies of modified CMs changed significantly. It appeared that some catalysts were coated onto the inner structure of the membranes, which was also confirmed by the EDS mapping. The EDS mapping images showed that all the four catalysts were successfully and uniformly integrated to the porous channels of the membranes, which was critical for the catalytic degradation of organics as the activation of PMS could be facilitated within the pores of the catalytic membrane. Table S3† showed the cross-section element compositions of the CMs. The catalyst content inside the membrane was much higher than the surface (Table S2†), which was mainly due to the larger surface area of the porous structure inside the membrane. Therefore, the well-distributed metal oxide particles would play an important role in the CM/PMS system to degrade refractory organics.

Furthermore, although the catalysts were loaded into the membrane pores, the water permeation performance of the membrane was not negatively affected. As shown in Fig. 6, the performance of different CMs in filtration of phenol solution was evaluated. The results showed a slight increase in water flux of the modified CMs, which may be attributed to the improvement of hydrophilicity. However, all the CMs showed a very low phenol retention rate (below 7%) and the retention rate could be mainly attributed to the adsorption of phenol by the membrane. The results also illustrated the necessity of combining the microporous ceramic membrane with the advanced oxidation process to enhance the removal of organic matters.

Fig. S5† compared the phenol and TOC removal rate of the four membrane systems. Comprehensively, it could be concluded that under the same operation conditions, the degradation efficiency of phenol in different CMs/PMS systems was in order of CoCM > MnCM > FeCM > CuCM. The phenol removal rate was 100%, 99%, 87%, 83% for CoCM, MnCM, FeCM and CuCM, while the corresponding TOC removal rate was 88%, 82%, 74% and 70%, respectively. The difference in membrane catalytic performance may be mainly attributable to the types and production mechanism of ROS, which will be analyzed in detail in the activation mechanism section. Besides, comparison of the literature with this work on degradation efficiency of phenol by catalytic membranes is summarized in Table S4.† It can be seen that the four CMs in this work have advantages in phenol wastewater treatment.

CoCM was used as an example to discuss the change of reaction rate with pH value. The results presented in Fig. 7a demonstrated that as the pH decreased from 11 to 3, the reaction rate constant kobs increased from 0.074 to 0.196 min−1 and phenol was completely oxidized under acid conditions within 30 min. However, under alkaline conditions, approximately 12% of the phenol was still not be removed after 60 min. In general, the CoCM/PMS system showed good applicability at pH value from 3 to 11.

It was reported that the relationship between ln(kobs) and logarithms of experimental parameters could reflect which factor was more sensible to phenol removal efficiency by CMs/PMS system.48,49 Fig. S6† showed that the lnCPMS and lnCphenol had good linearity with ln(kobs), and the corresponding slopes were 0.909 and 0.767, respectively. Larger slope means it was a more important factor determining the degradation rate for CMs/PMS system. These results indicated that the influences of the two parameters were in the order of CPMS > Cphenol. Thus, in practical applications, when the CMs/PMS system is used to treat high-concentration phenol wastewater, the dosage of PMS should also be increased accordingly.

Additionally, take CoCM as an example, the effect of metal ions leaching on the performance of the modified CMs was studied. As shown in Fig. 10b, the CoCM could still maintain a high catalytic activity after multiple runs. The phenol removal efficiency declined about 5.5% throughout six cycles. This may be due to the shielded active sites by contaminants during degradation.8 Besides, the variation of metal leakage during several cycles of uses was also studied. It can be seen from Fig. 10c that the maximum amount of metal leakage was occurred at the first cycle, then gradually decreased. In all, the properties of low metal ion leaching, high catalytic activity and good reusability indicate that the catalytic CMs has a promising prospect in degradation of organic pollutants.

Furthermore, EPR technology was used to verify the ROS in the CMs/PMS systems by introducing DMPO and TEMP as spin-trapping agent. As shown in Fig. 11c and d, three characteristic peaks were observed in the CoCM/PMS system, which corresponded to DMPO–SO4˙−, DMPO–·OH and TEMP-1O2, respectively. Additionally, due to the short half-life of O2˙−, MeOH was used instead of water as the solvent to capture O2˙− by DMPO. The resultant quadruple peak in Fig. 11c corresponded to DMPO–O2˙−, indicating the existence of O2˙− in the CoCM/PMS system.

Based on the results and detail analysis above, the possible mechanism for PMS activation by CoCM was proposed in eqn (3)–(7). Co(III) preliminarily reacted with PMS to produce SO5˙− and Co(II), and then the converted Co(II) could further activate PMS to generate SO4˙− and ·OH with the regeneration of Co(III).

Thus, the two experiments of quenching and EPR jointly corroborated that the reactive oxygen species of SO4˙−, ·OH, O2˙−and 1O2 were generated during degradation of phenol by CMs. Among them, 1O2 and SO4˙− were considered to be the main active species for phenol degradation in the CoCM/PMS system with the order of 1O2 > SO4˙−. These results were similar to Li's report on the mechanism of PMS activation.45

Besides, it could be concluded from Fig. S7† that the dominating ROS for phenol degradation in the other three membrane system were quite different from the CoCM/PMS system. The dominating ROS were speculated to be 1O2 and O2˙− in the MnCM/PMS system, SO4˙− and ·OH in the FeCM/PMS system, SO4˙− in the CuCM/PMS system, respectively. Thus, the oxidation mechanism is quite complex in different CM/PMS systems. For example, some researcher also reported that the activated PMS on CuO surface through direct electron transfer contribute a lot in the CuO/CMs system.53,54 The possible catalytic mechanism for PMS activation by other CMs were listed in ESI Text S1.†