An alarming large production of undesired organic contaminants in the textile industry has become a long-standing issue that deserves special attention (I. Ali et al., 2012; Lapworth et al., 2012; Salah et al., 2022; Verma et al., 2022). Moreover, the diverse class of organic contaminants produced by the textile industry triggered a complexity that led to adverse effects on the human body. This undesired byproduct is highly harmful to environmental conditions (I. Ali et al., 2012; Gendy et al., 2022). Both the large volume of effluents and the type of chemical compositions, to some extent, create a tip of an iceberg issue that could eventually grow into complex problems. In the past years, nearly 1 million tons of toxic dyeing wastewater were produced annually, of which 200,000 tons were disposed as effluent without proper treatments (Holkar et al., 2016; Siddique et al., 2017). Therefore, tactical solutions and preventive measures in handling such environmental issues are highly pursued to facilitate safe downstream ecosystems and sustainable human health.

Several approaches have been introduced within the past 2 decades focusing on the treatment of organics-contaminated wastewater, and the common method in the current applications is chemical oxidation (Asghar et al., 2015; Sun et al., 2015). Among other chemical oxidation processes, AOP is categorized to be a versatile approach in wastewater technology due to the in situ formation of reactive oxygen species (ROS) such as hydroxyl radicals (OH•) (Yan et al., 2022; Yasasve et al., 2022; Aljedaani et al., 2023).

In the recent decade, different studies have formulated that the characteristic outcome of ozone-based AOP method is influenced by the chemical and physical parameters, such as pH of the solution (Deng and Zhao, 2015; Boczkaj and Fernandes, 2017; Koc-Jurczyk and Jurczyk, 2019). There is an ongoing discussion that pH is an essential parameter since it plays an important role during O₃ decomposition process (Temesgen et al., 2017). The ozonation combined with UV-radiation (O₃/UV) is an efficient catalytic wastewater system for the degradation of refractory pollutants. In principle, the catalytic process is started by ozone photolysis, followed by ROS production (Emam, 2012). Despite renewed interest of the AOP operational method in the last two decades (Zhang et al., 2021), few reports have compared to degrade various textile wastewaters using different AOP approaches.

Nanobubbles (NB) technology has been widely used to degrade organic contaminants in solutions (Ali et al., 2023; Aluthgun Hewage et al., 2021; Wang et al., 2019). The physicochemical concept of NB technology is slightly comparable to the AOP method. At a glance, NB facilitates the formation of a physical barrier such that a contaminant layers can be encapsulated by the bubble surface (Pan et al., 2021; Suvira and Zhang, 2021; Inoue et al., 2022; Selihin and Tay, 2022) thereby generating ROS efficiently (Fan et al., 2023). The formation of ROS plays a vital role in degrading organic compounds that are present in the wastewater (Mishra et al., 2017). However, the underlying mechanism in the NB technology is based on the realization of gas bubbling rather than a chemical treatment approach. Due to this reason, the ozone-based NB produced by the NB generator could extend its lifetime in the water column for a longer time compared to its counterparts (Hu and Xia, 2018). According to Meegoda and coworkers (J. Meegoda and Batagoda, 2016), when dissolved ozone interacts with the pollutants, the treated water remains in the saturated ozone. Therefore, NB retains its physical properties within the water column for several months under normal pressure. To the best of our knowledge, a few combinations of the NB applications and other AOP are introduced. For example, the work carried out by Tasaki et al. (2009) indicated that oxygen microbubbles (MB) could be implemented for the methyl orange removal in the presence of a photocatalyst under UV irradiation. In this study, we employ a combination of ozone-based AOP and NB to degrade textile wastewater samples during 300 min of residence time. To assess the water quality parameters, such as chemical oxygen demand (COD), ammonia content (NH₃), and total suspended solids (TSS)are compared to find the integrated system performance. The NB formation is then analyzed using Dynamic Light Scatering (DLS) method to find the size, stability, and surface charge of our NB.

To represents the significance of our combined method, the decoloration of textile wastewater is displayed in Figure 3. We find that the textile wastewater progressively changed its color solution as a function of residence time during the treatment process. Within the initial treatment of 60 and 120 min, we found that visual observation of the treated wastewater in tank one was significantly different compare to the tank 2. It was noticeable that within 120 min of residence time, the water color was changed from brownish to the pale-green, indicating an improvement in the water quality was achieved via extending the residence time and continuous circulation of the water line along the recovery water. At the end of the process (300 min), the textile wastewater color was unambiguously improved, showing a semi-transparent solution. Furthermore, the changes in several parameters control, namely, COD, NH_3, and TSS are depicted in Figure 4.

The evolution of COD values as a function of residence time is presented in Figure 4. According to the findings, the COD concentration is reduced along the residence time with the final state reach approximately 80% removal. Further details of COD evolution during 300 min of the treatment process can be found in Supplementary Table S1.

We compare the COD parameter of wastewater that was treated by the combined method of ozone based AOP + NB and ozone-based AOP alone, as depicted in Figure 5. In general, both systems have demonstrated their potential for lowering COD levels in water. As a result of the ozone exposure, the AOP system alone exhibited a sufficient removal rate to degrade the organic contaminants within the range of ∼10% upon 300 min treatment time. Interestingly, the combined method of NB + AOP system yielded substantial improvement in which the COD reduction rate was significantly greater than the AOP reduction (shown in red lines). Here, we postulated that such a high COD reduction rate might be attributed due to the ozone size modification, as those compounds are subsequently through the NB compartment. Ozone has an undesired physical characteristic such as poor solubility in the water. It led to a condition in which ozone is easily escape into the atmosphere from its solution phase. Therefore, one solution is required to enhance the persistence time by reducing the material size or shape. For instance, the size modification of a nanobubble with a smaller size could improve to a longer persistent time (Tekile et al., 2016; Nirmalkar et al., 2018).

Figure 6A Shows the bubble size findings that acquired using the DLS instrument according to the number-distribution data. The peak distribution value, approximated from 99.94% of the ozone-based NB was 216.9 nm in size. Here we noted that the quantity and size of NB could affect their mass transfer efficiency. A significant NB production is desirable in a large total surface area that would enhance the mass transfer flow from the bubbles toward the wastewater solution. However, we should also consider that non-negligible bubbles with a smaller size could arise under the higher internal pressure and larger specific area (Peng and Yu, 2015; Ulatowski et al., 2019; Rahmawati et al., 2021). Here, we revisit the contact principle used in this study in which multiphase gas-liquid flow was employed to generate NB. In order to produce a multiphase flow, we therefore consider that the water flow was injected with a pressurized ozone (Levitsky et al., 2015; Rahmawati et al., 2021). When the multiphase flow passed through the honeycomb structure of NB generator, then the ozone NBs are subsequently formed due to the multiphase flow breakdown (Ren et al., 2018; Rahmawati et al., 2021).

The zeta potential in the final residence time of 300 min is reported to be −21.08 ± 0.35 mV with the corresponding distribution is depicted in Figure 6B. The negative value here is assigned to the generated ROS at the bubble interface, which leads to the bubble stability (Ushikubo et al., 2010; Selihin and Tay, 2022; Zhou et al., 2022). Typically, zeta potential value can be associated to the charge of the bubble rather than its density in the water. The amount of zeta potential indirectly reflects how stable a colloidal system is. The electrically-charged surface causes particles to repel one another, preventing the bubbles from emerging on the other bubble surface (J. N. Meegoda et al., 2018; Ushikubo et al., 2010). The details of NB’s zeta potential in 300 min of the process can be found in Table S5.

To interpret such COD reduction in ozone based AOP and NB system, we propose that this finding could be driven by the enhanced synergistic effects promoting an efficient chemical degradation reaction pathway to remove the organic contaminants from the water.

The pollutant degradation mechanism can be understood in two consecutive processes. The first stage of the water recovery process is related to accumulating ozone NB (see arrow red in Figure 7), which leads to the first reaction with the incoming inlet wastewater. Ozone NB has higher surface contact compared to macro-bubble, so it would be more persistent in the solution. It also strongly influence both the pollutant degradation and ROS generation at the following treatment stage. Here, we found that fine control of ozone exposure under pH-neutral media accommodates the organic contaminants’ physical degradation. Our finding regarding COD reduction that operated at neutral pH in agreement with the previous reports (Lucas et al., 2009; Takahashi et al., 2012; Yang et al., 2012; Temesgen et al., 2017).

The second recovery stage corresponds to the utilization of ozone-based AOP as indicated by the blue arrow in Figure 7. In this section, the generation of ROS or negatively-charged NB dominantly contributed to degrade the organic contaminants. We revisit the conventional thermodynamics of bubble collapse phenomenon to describe the ROS generation (Ghadimkhani et al., 2016) and ozone breakdown (Ikhlaq et al., 2013; Oh and Nguyen, 2022). In the ozone decomposition, we attribute the ROS is a chain reaction comprised of initiation, propagation, and termination (Khuntia et al., 2015).

Furthermore, ROS via collapsing cavity bubbles can be described based on the Young-Laplace equation (Kimura et al., 2004; Takahashi et al., 2007; Ghadimkhani et al., 2016). P = P 1 + 4 σ / d (3) where P denotes gas pressure, P _l denotes liquid pressure, σ denotes liquid surface tension, and d denotes the bubble diameter. The internal pressure of a bubble is substantially determined by its size, according to Eq. 3. Since the d is inversely proportional to the gas pressure, the impact on the internal pressure for the NB differs significantly compared to the macro-bubbles. Thus, once NB is defragmented into molecular gas, thus the additional surface energy can rupture the hydrogen-oxygen bond and therefore generate ROS (Bandala and Rodriguez-Narvaez, 2019).

In addition, we present a semiempirical approach concerning the complete chemistry of the degradation pathway in our system. Based on our understanding, the chemical process in this study related to the concerted scheme of direct and indirect oxidation routes. Here, we consider that a direct route is related to a simultaneous chemical reaction between pollutant target and ozone, while the indirect route utilized ROS to degrade the targeted contaminants.

In terms of temporal consideration, a direct reaction of ozonation between specific compounds and functional groups lasted in relatively slow reaction time (a constant rate in the range of K_d = 1.0–10⁶ M⁻¹ s⁻¹) (Gottschalk et al., 2010). Ozone reacts with a faster degradation rate when it interacts to certain types of aromatic and aliphatic compounds.

The indirect oxidation route can be attributed to the formation of ozone radical chains that comprise of three different steps. The first step involving a decay process of ozone which is accelerated by initiators. In our scenario, the hydroxyl anion (OH⁻) is converted into ROS constituent such as OH•. Based on their kinetic argument, the chemical specificity to the targeted pollutant is non-selective with the rate of K_d = 10⁸–10¹⁰ M⁻¹ s⁻¹ (Gottschalk et al., 2010). Let us consider that the OH• regain their missing electron by removing an electron from the hydrogen atom of the target pollutant molecules. This in turns converted to targeted water product (H₂O). Consequently, the targeted molecules converted to ROS that propagates the chain reactions. However, another scheme could also arise in which a radical reacts to the second radical and terminated the reaction. A formulated pathway (Voigt et al., 2020) is described in Figure 8.

We revisit the previous report that the removal contaminant rate can be maximized if the magnitude of ROS and targeted pollutants are similar, as previously reported by Bui and coworkers (Bui and Han, 2020). We note that in this study ROS was not analyzed quantitatively, nevertheless we predicted that our system able to produce ROS numbers that higher than the number of pollutants. It is an interesting fact that the COD removal during the first 60 min of the process shown a significant reduction around 40% removal. It is noticeable that our system was able to complete half of the final reduction within 60 min of the treatment process. In addition, this phenomenon can be realized with the stability and number of ROS. Upon simultaneous radicals’ formation, when the ROS number is higher than the pollutant number, the ROS will react with others through a terminating pathway to generate secondary ROS (with low reduction potential) and hydrogen peroxide (H₂O₂). Those secondary ROS then predicted it would dominate the oxidation process rather than the primary radicals (Esfahani et al., 2019), while H₂O₂ will contribute to reducing pH value, as seen in Figure 9.

By altering the residence time, the pH solution gradually decreased from 7.93 to 7.46, as shown in Supplementary Table S2 Based on previous study, pH in the initial state influences the rate of ozone oxidation (Patel et al., 2021; Jabesa and Ghosh, 2022; Jesus et al., 2022; John et al., 2022). When the pH of a solution is raised, the direct consequence is that the oxidation rate increases. However, there is a drawback in which the ozone concentration would decrease (Patel et al., 2021; Jabesa and Ghosh, 2022; Jesus et al., 2022; John et al., 2022). Furthermore, aqueous solutions containing high pH induced ozone autodecomposition since it led to indirect ozone breakdown, resulting in ROS such as peroxy radicals, HO₂•, and OH• (Jabesa and Ghosh, 2022; Jesus et al., 2022). However, when the pH is more than 8, this eventually hampers the process with the decay of ozone half-lives (Gardoni et al., 2012). Furthermore, we consider that our UV in the AOP-Ozone combination promotes the chemical breakdown of organic pollutants, as previously described by Rekhate et al. (Rekhate and Shrivastava, 2020) and Wu et al. (Wu, 2008).

Similarly, the elimination of TSS and NH₃ resulted in a significant shift throughout the residence time. In TSS removal, the large contaminants tended to adhere to the NB surface while floating to the outmost water surface (Ali et al., 2021). It will cluster the solid suspension in solution. As a result, the removal process of suspension would last rapidly (F. Ali et al., 2021). Detailed information regarding TSS values in different sampling regimes during 300 min can be found in Supplementary Table S3.

On the other hand, the pH of the solution had a considerable influence on NH₃ decomposition. At pH 9, direct oxidation of NH₃ occurs. In our study, the pH of the solution was determined to be 7-8, which is comparable to the initial conditions. In this pH range, we propose that direct oxidation is likely to occur. At low pH (lower than 7), the quantity of free NH₃ was minimal, resulting in a sluggish rate of NH₃ oxidation. As the pH gradually increases, the number of NH₃ generated is continuously enhanced. Consequently, the availability of free NH₃ increased the oxidation rate (Khuntia et al., 2013). The maximum reduction of NH₃ in textile wastewater is achieved at 300 min residence time; further details can be found in Supplementary Table S4.

We summarized that approximately 81.1% of COD recovery in the water purification system is achieved by utilizing the combination oxidation methods of AOP and NB for textile wastewater samples. Here we emphasize the role of NB method in this investigation is essential to improve the gas–liquid phase efficiency processes. Moreover, the realization of ROS generation mediates the chemical conversion of the polluted water to the environmentally-safe effluent with low level of COD, NH_3, and TSS. To statistically quantify the NB size of our system, we reported that 99.94% of the ozone bubble population is recorded with the size of 216.9 nm upon 300 min. The zeta potential values unambiguously shed some light on the existence of ROS at −22.43 ± 0.34 mV. An abrupt reduction of the COD is observed with the removal percentage of around 40% within the first 60 min residence time. Here, we consider that the initial reaction kinetics takes place in both processes facilitating the faster reaction rate compared to the individual stage of AOP or NB method.