Assessment of the Performance of Ozone Nanobubble Technology to Enhance Water Treatment Performance of a Constructed Floating Wetland

Abstract

Small-scale decentralised wastewater treatment facilities are essential to provide services to remote regional communities. This study presents an innovative and sustainable approach to wastewater treatment by integrating ozone nanobubble technology (ONBT) with constructed floating wetlands (CFWs). Effluent from a community wastewater treatment plant was used in two sets of twelve 170-litre tanks, each with different ONBT–CFW treatment combinations, and monitored for key water quality parameters over an eleven-week study. The experiment results indicated that the combined ONBT–CFW system, particularly with higher ozone doses, achieved substantial reductions in total nitrogen (>70%), BOD (>43%), and E. coli (100%). ONBT alone showed limited effectiveness on nutrient removal, while CFWs performed well in reducing nutrients and controlling E. coli. However, phosphorus removal was modest (~12%), suggesting the need for complementary strategies. Overall, the hybrid ONBT–CFW system demonstrated superior performance compared to individual treatments, offering strong potential for improving wastewater quality and treatment.

1. Introduction

Community Wastewater Management Systems (CWMSs), or generally known as decentralised wastewater treatment systems, are essential for providing cost-effective and environmentally sustainable wastewater treatment options to the regional communities. These systems are typically designed for a shorter life span, and many current CWMS in Australia, especially South Australia, are aging and nearing their capacity limits. The feasible option is to rebuild the systems and extend their capacity by adopting innovative and cost-effective solutions while upgrading the services [1]. The wastewater treatment industry has seen significant advancements in recent years, driven by innovative technologies. Among these, ozone nanobubble technology (ONBT) and constructed floating wetlands (CFWs) have emerged as promising approaches to improve wastewater treatment efficiency [2].

CFWs utilise natural biological processes [3,4], harnessing the filtration and contaminant absorption capabilities of plants to treat wastewater. CFWs are cost-effective, passive biological systems that improve water quality using emergent vegetation (macrophytes) planted on buoyant platforms, and pollutants are assimilated into the plant body [3,4,5,6]. Previous research demonstrated that CFWs can extend the operational capacity of CWMSs [1,7,8,9]. The submerged plant roots play a critical role in filtering suspended solids, removing nutrients, and facilitating microbial activity for additional pollutant breakdown [5,9,10]. CFWs work similarly to hydroponic systems, where the roots freely grow in water, providing a large surface area for biofilm growth [5]. Through diversifying the living environment and catalysing chemical and biochemical reactions, they promote water purification [11]. These systems offer numerous benefits, including low operational costs, the absence of chemical dosing or pumping, easy retrofitting into water environments, and the promotion of ecosystem services, such as wildlife habitats and enhanced landscape aesthetics. Despite the aforementioned benefits, CFWs can provide habitat for mosquitoes; therefore, they should be far from communities [12]. Also, negative environmental impacts like odour and the release of contaminants can occur if constructed wetlands are not run properly [13].

Biological processes alone often struggle to degrade recalcitrant organic compounds due to the absence of necessary metabolising enzymes. Integrating an oxidation step prior to biological treatment can improve the breakdown of refractory organics, transforming them into smaller, more manageable molecules that are easier to adsorb and metabolise. Research results [14,15] highlight the potential for performance improvements through the integration of complementary technologies such as nanobubble ozonation.

Hydrodynamic cavitation is a process of vaporisation, bubble generation, and bubble implosion. The pressure inside the bubble will increase as the bubble shrinks, which finally causes the bubble to collapse [16]. Shrinkage of nanobubbles raises the electric charge density of the electric double layer on the bulb surface. During the collapse, the ions accumulated on the bubble surface release chemical energy. Also, water molecules inside the bulb have a high temperature as a result of compression; this produces various active oxygen species like ^oOH during the collapse [16]. With a high oxidation potential, ^oOH radicals are thermodynamically very active in unselectively attacking organic compounds and their removal.

Ozone oxidation involves direct and indirect mechanisms where oxidation is performed by ozone itself or the ^oOH radicals produced during ozonation, respectively [17]. Ozone reaction with the organic compounds is selective, while the ^oOH radical non-selectively attacks organic and inorganic compounds and the water matrix [17]. Ozone breaks down large organic molecules to more biodegradable compounds and therefore promotes the efficiency of the biological processes. Ozone nanobubbles are strong oxidation agents because of the synergy effect between the substantial oxidation property of ozone and that of the bubble cavitation. Ozone nanobubbles can degrade the resistant organics and increase the dissolved oxygen (DO) level in water while adding nothing to TDS in water. ONBT offers a chemical-free oxidation process, leveraging highly reactive and long-lasting ozone nanobubbles to effectively remove organic and inorganic pollutants.

While both ONBT and CFWs have shown substantial individual potential, their combined effects remain underexplored [7,18]. Building on from these foundations, this study aims to trial a hybrid water treatment approach by integrating ONBT with CFWs and seeks to achieve a substantial improvement in the quality and safety of treated wastewater by enhancing the removal of contaminants such as nitrogen compounds, BOD, and E. coli.

The results of this study will help tackle challenges of wastewater treatment and recycling in rural and regional communities by applying proven technologies in a novel way to deliver sustainable and resilient solutions that enhance water security for the agricultural and horticultural sectors. Ozone treatment in this study was to investigate how varying doses of ozone, with and without the presence of plants, affects the water’s chemical and biological properties.

2. Materials and Methods

The experiment trials in this project spanned over an 11-week period from August to November 2024. The water sample used in this study was sourced from the settlement lagoon at a regional wastewater treatment plant (WWTP) in South Australia. This lagoon was selected because the treated wastewater represents the typical discharge quality from community WWTPs in the region, enabling the research team to assess the efficacy of ONBT and CFWs under conditions that closely simulate real-world operational settings.

2.1. Experiment Setup

The water sample was collected from a sampling point located outside the fenced area of the WWTP. This point is connected via an underground pipe to the settlement lagoon, which is situated beyond the fenced area and receives treated wastewater from the WWTP. The collected sample was transported to the trial site using two pre-cleaned 1000-litre Intermediate Bulk Containers, labelled as Tank 1 and Tank 2. Upon arrival, the water from each tank was discharged separately into two distinctive clusters of experimental tanks. Each tank was thoroughly cleaned before being filled with its respective water sample. No mixing occurred between the samples from the two Intermediate Bulk Containers, ensuring complete separation. This deliberate segregation allowed the research team to account for any differences in initial water quality during data analysis.

Both clusters of experimental tanks were subjected to identical treatment regimes, which included the application of ONBT, CFWs, or a combination of the two (i.e., CFW-ONBT). By maintaining the same experimental conditions across the two clusters, this study ensured that any observed differences in treatment efficacy could be attributed to the potential differences in the initial water quality rather than inconsistencies in the experimental setup. The experimental setup involved filling twelve 170-litre tanks with the collected water, arranged into two clusters and corresponding to the Intermediate Bulk Containers used for water transportation. The treatments comprised combinations of conditions with or without plants (Phragmites australis) and different ozone dosages. Accordingly, the tanks were arranged to reflect each treatment condition as follows, with two replicates per treatment option as below:

• No Plant and No Ozone (Control)
• Plants Only
• Ozone Dose 1 with the nominal ozone concentration of 0.017 g O₃/L (5 min ozone injection)
• Ozone Dose 2 with the nominal ozone concentration of 0.03 g O₃/L (10 min ozone injection)
• Ozone Dose 1 and Plants
• Ozone Dose 2 and Plants

Initial aeration (4 h) was applied for all tanks, except for the two controls (i.e., ‘No Plant and No Ozone’). The primary aim was to aerate and ensure uniformity of the water prior to the experiment while the ‘No Plant and No Ozone’ was left as control.

2.2. Plant and Pot Selection CFW Modules

To simulate CFW treatments, the tanks are configured with CFW modules installed as depicted in Figure 1, containing selected plants, medium and pot to accommodate and sustain the plants, and floating platform as buoyant to support the pot.

Phragmites australis, a plant known for its efficacy in phytoremediation [5], was selected for this study. A total of 24 plants were evenly allocated across six 13.5 L pots, each with a diameter of 300 mm and a depth of 270 mm. These were placed in the outdoor designated tanks to ensure optimal growth conditions with adequate light exposure and water interaction.

Custom-designed buoyancy supports were used to float the pots on the surface of the water. The supports were constructed in the form of ring-shaped floating islands using expanded polystyrene foam, each featuring a central hole with a diameter reduced to 250 mm to control water levels inside the pots. This setup was crucial in maintaining plant stability while optimising their interaction with the water surface.

Water levels were consistently monitored and maintained at 150 mm below the top edge of each drum, a level determined to balance light exposure and minimise overflow risks. Concrete aggregates (14 mm) in the pots were chosen as the growing environment for plants because of their structural stability and effectiveness in supporting plant growth while allowing for necessary filtration processes.

2.3. Ozonation

Depending on the experimental conditions, tanks received different dosages of ozone, with the ozonation process lasting either 5 min for the lower dosage (D1, 0.017 g O₃/L) or 10 min for the higher dosage (D2, 0.03 g O₃/L). This setup was crucial for understanding the interaction between ozone and water contaminants, and it played a significant role in the overall experimental methodology. The schematic of the ozonation system is shown in Figure 2.

The ozonation process in the experiment was characterised by the vigorous bubbling and mixing of water, for thorough distribution of ozone. This was performed by using a ESP-150 Dissolved Oxygen Nanobubble Generator/Ozone Generator (Hydro2050, Adelaide, South Australia, Australia) for producing and supplying ozone gas to the experimental tanks, with a production capacity of 30 g/h. The generator was equipped with a digital control panel for adjusting the ozone output and a pressure gauge to monitor the gas flow, ensuring precise delivery of ozone at the required dosages. Dissolved oxygen (DO) level was measured using a DO meter (Prosolo Optical Dissolved Oxygen Meter, YSI, Ohio, USA). Additionally, the system generates ultra-fine bubbles, each less than 1 micron in size, at a volume of 18,000 litres per hour. These ultra-fine bubbles were critical for maximising the contact area between ozone and water, enhancing the efficiency of the treatment process.

2.4. Analyses

The data collection process was designed to evaluate the effectiveness of the treatment methods by gathering comprehensive data on key water quality parameters and plant growth metrics from all 12 experimental tanks. An in-depth assessment of both water quality and the conditions of the plants used in the CFWs was conducted. In the first week of the experiment, baseline samples were collected from all 12 experimental tanks before any ozonation was applied. These initial samples served as a reference point to assess water quality before treatment. A second set of samples was collected one hour after ozonation from the tanks that received ozone treatment. This sampling process was designed to capture the immediate effects of ozonation on water quality and plant condition. A total of five scheduled sampling events were carried out during the experiment period, occurring in Weeks 2, 5, 8, and 11. In Week 2, two samplings, before and after ozonation, were performed for the tanks with ozonation. Each sampling event followed the same procedure to ensure consistency.

Plant growth was monitored by the measurements of shoot length (mm) and number of shoots (counts). The pH was measured using a method based on APHA 4500-H⁺ B. The method for the measurement of Electrical conductivity was based on APHA 2510 B. Nitrite and nitrate (as N) concentration measurement was based on APHA 4500-NO₃⁻ F. An APHA 4500-Norg D-based method was used to determine Total Kjeldahl Nitrogen (TKN, as N). Total nitrogen (as N) was determined according to APHA 4500-Norg and 4500-NO₃⁻. Total phosphorus (as P) was determined based on APHA 4500-P H. Biochemical Oxygen Demand (BOD) was analysed on the basis of APHA 5210 B. Microbiological analysis for E. coli and Total Coliforms was performed using the Colilert (MPN) method (MM514). All measurements were performed using Standard Methods [19] in a third-party accredited laboratory (ALS) [20]. The water quality analyses were conducted by ALS Scoresby, a NATA-accredited laboratory (No. 992). Additionally, normalised water quality values are used, where applicable, to assess treatment effectiveness regardless of initial differences in water sample conditions. Normalisation is performed by dividing the measured concentration of each sample (Ci) by its corresponding initial concentration (C0), represented as Ci/C0.

3. Results and Discussion

The DO readings immediately after ozonation varied across the tanks, reflecting differences in water conditions prior to the treatments. For instance, the ozone-dosed tanks exhibited DO levels ranging between 33.4 ppm and 36.9 ppm at water temperatures between 18.2 and 18.7 °C. These values indicated a relatively high oxygen level available for microbial activity, which would play a role in organic matter decomposition.

3.1. Plant Growth

This study monitored plant growth over a nine-week period which was tracked through two key metrics, i.e., shoot length and number of shoots per plant, in addition to water quality measurement. These metrics provide insights into the effects of different pre-treatments on the development of Phragmites australis. Environmental conditions, specifically temperature and rainfall [21], were also recorded to assess their impact on plant growth. Temperature and rainfall data provide essential context for the environmental conditions affecting the tanks, with temperature generally rising and rainfall fluctuating throughout the period (Table S1). Both clusters of experimental tanks were subjected to identical treatment regimes, which included the application of ONBT, CFWs, or a combination of the two (i.e., ONBT-CFW). By maintaining the same experimental conditions across the two clusters, this study ensured consistencies in the experimental setup. This approach not only enhanced the reliability and validity of the results but also provided valuable insights into how hybrid treatment systems perform under slightly varying initial conditions in wastewater quality.

Figure 3 depicts how plant growth actively evolved in shoot length (Figure 3a) and number of shoots (Figure 3b) across a nine-week period. The results indicate that Phragmites australis, the plant species used in this study, demonstrates adaptability to ONBT, contributing to the enhanced performance of CFWs.

Plant growth can be influenced by multiple interacting parameters, making it complex to analyse based on just one or two factors. As shown in the figures above, growth patterns varied between weeks for both plant height and shoot numbers, showing fluctuations. To better assess plant growth, shoot length and number of shoots were combined as a single measure. The multiplication of shoot length and shoot number was considered as an applicable measure of plant growth. The average combined results of ‘Plants Only’ and different ozone doses, i.e., Dose 1 (0.017 g O₃/L) and Dose 2 (0.03 g O₃/L), with plants are reported in Figure 4. This figure indicates the steady growth of the plants in all conditions.

The ‘Plants Only’ samples with 34,518 mm showed the highest growth after nine weeks. ‘Ozone D2 and Plants’ and ‘Ozone D1 and Plants’ samples with 30,745 mm and 22,740 mm were in the second and third places, respectively. Ozonation through degrading large organic molecules and oxidation of organic nitrogen to inorganic nitrogen changes the bioavailability of the nutrients. It also increases DO concentration in water and sterilises the root medium. On the other hand, producing oxidised species in water can increase the stress borne by the plants. The conclusion about the effect of treatment conditions on plant growth is not straightforward. As earlier mentioned, plant growth is a complicated function of various environmental conditions and genetic traits interacting with each other. For this reason, finding a clear connection between plant growth and those parameters needs a detailed investigation. While ozone application appeared to damp the overall plant growth, it is obvious that Phragmites australis can thrive in an ozonated water environment. This is a very significant finding that will help in the utilisation of this plant for further studies in developing CFW and its combined methods for wastewater treatment.

3.2. Water Quality

The measures of initial water quality in each tank, based on the first round of sample testing, are detailed in

Table 1. The starting values of quality parameters of the water in test tanks from initial sample testing (ND: Not Detected. The two results in each category refer to the results from each cluster).

Water Quality Measure	Plants Only		Ozone Dose 1		Ozone Dose 2		Ozone Dose 1 and Plants		Ozone Dose 2 and Plants		No Ozone No Plant (Control)
pH (pH Unit)	8.1	7.8	8.2	8.1	8.3	7.6	8.1	7.6	7.9	7.4	7.7	7.8
Electrical Conductivity @ 25°C (µS/cm)	2200	1920	2050	1820	2050	2480	2040	2340	1370	1790	2560	2560
Nitrite plus Nitrate (mg/L)	8.4	3.8	7.6	6.1	7.7	0.0	7.7	0.2	4.1	0.9	5.7	4.8
Total Kjeldahl Nitrogen (mg/L)	11.3	7.9	11.0	8.8	11.2	8.1	10.9	7.7	5.8	5.4	10.2	10.1
Total Nitrogen (mg/L)	19.7	11.6	18.6	14.9	18.9	8.1	18.6	7.9	9.9	6.3	15.9	14.9
Total Phosphorus (mg/L)	4.2	2.9	4.2	3.3	4.4	3.8	3.9	3.4	2.0	2.4	4.9	4.7
BOD (mg/L)	2	3	2	2	2	11	2	9	2	5	52	69
Escherichia coli (MPN/100 mL)	ND	ND	ND	ND	ND	ND	ND	ND	16	ND	ND	ND
Total Coliforms (MPN/100 mL)	120	1600	100	24	ND	130	5	2400	390	2400	140	6
DO (mg/L) immediately after ozonation	–	–	34.48	33.97	37.68	33.37	38.23	33.08	36.9	38.09	–	–

as the baseline for analysis and comparisons.

The unavoidable uncertainty of wastewater quality in the pipe to the sampling point may contribute to the difference in the samples from the two Intermediate Bulk Containers (used for water transport); this was observed by foaming due to turbulence induced by pumping through the hose, which likely contributed to variations in wastewater quality characteristics in the two tanks/two clusters. The elevated initial BOD levels in the ‘No Plant and No Ozone’ tanks (52 mg/L and 69 mg/L, as per Table 1) were due to the absence of aeration (as control), whereas other tanks received aeration during the first week. Aeration enhances the activity of aerobic microorganisms, accelerating the breakdown of organic matter and facilitating BOD reduction. As a result, by the end of Week 2, BOD levels in the aerated tanks had decreased significantly compared to the two control tanks (No Plant and No Ozone), which lacked this treatment. To account for the potential impact of these variables, the average values of the two clusters were used for analysis. Additionally, normalised water quality values are applied, where feasible, to evaluate the effectiveness of treatments, irrespective of initial variations in the water sample conditions. Normalisation is calculated by dividing the measured concentration of each sample (C_i) by its initial concentration (C₀), expressed as C_i/C₀.

3.3. Using E. coli as a Water Quality Indicator

Escherichia coli (E. coli) is a widely recognised indicator of microbial contamination in water, particularly linked to faecal contamination. The presence of E. coli in water signifies the potential risk of harmful pathogens, making its removal essential in any water treatment system designed for reuse or environmental discharge. In this research, E. coli was monitored across various treatment setups, including ozone-treated tanks, plant filtration systems, and control tanks, to assess the effectiveness of each method in reducing microbial loads.

Given its role as a proxy for microbial contamination, E. coli reduction serves as a key benchmark for determining the success of water treatment technologies. In this study, we evaluated how various treatment methods—including ozone nanobubbles, plant filtration, and no active intervention—impacted E. coli levels, providing insights into the effectiveness of each system in achieving safe and clean water. As shown in Figure 5, the increased presence of E. coli in the ‘Ozone D2 and Plants’ and ‘Plants Only’ samples can be attributed to environmental contamination after treatment or inadequate sterile conditions during sampling. However, E. coli levels in these samples eventually reached zero. Across all samples, E. coli concentrations remained well below the local allowable practical limit of 100 MPN/100 mL [22]. Regardless of the pollution source in the ‘Ozone D2 and Plants’ and ‘Plants Only’ samples, the results clearly demonstrate the effectiveness of these treatments in eliminating E. coli by the end of the trial. While the ‘Plants Only’ treatment performed well, the ‘Ozone D2 and Plants’ treatment exhibited greater efficiency by reducing a higher initial microorganism concentration to zero.

3.4. Biochemical Oxygen Demand (BOD)

Biochemical Oxygen Demand (BOD) measures the oxygen needed by microbes to break down organic matter, providing insight into the effectiveness of various wastewater treatments [23]. In this study, BOD was measured across various tanks with different treatments, providing insights into the effectiveness of these treatments in reducing organic pollutants.

The BOD levels were tracked across five sampling periods to assess the effectiveness of each method. As shown in Figure 6, the initial BOD level was higher (61 mg/L) in the ‘No Plants No Ozone’ sample which was reduced to 6–9 mg/L naturally, but it increased to 21 mg/L in the last measurement in Week 11. The initially high BOD levels in the ‘No Plant No Ozone’ tanks resulted from a lack of aeration in Week 1, which inhibited aerobic degradation of organic matter. In contrast, other tanks were aerated during this period. Aeration enhances the activity of aerobic microorganisms, promoting organic matter decomposition and BOD reduction, which explains the lower BOD levels in the other tanks by the end of Week 2 compared to the ‘No Plant No Ozone’ tanks. In the ‘No Plant No Ozone’ tanks, anaerobic oxidation gradually reduced BOD levels by Week 8. Other samples also exhibited an increase in BOD, with the ‘Ozone D2 and Plants’ treatment showing a rise of up to 214%, which was later corrected by the system.

All ozonated tanks exhibited an initial increase in BOD following ozonation, a well-documented effect caused by the breakdown of organic particles and recalcitrant organic molecules, which generates reaction intermediates and temporarily raises BOD levels [24]. Over time, further sorption by plants or reactions involving these intermediates helped reduce BOD, stabilising it by the end of the 11-week trial. Overall, BOD levels remained within the allowable limit of 20 mg/L (5 days) [22]. The ‘Plants Only’ tanks also showed an initial rise in BOD, likely due to the introduction of plant material, which contributed debris, dead tissues, and released organic matter into the solution. Additionally, plants create a favourable environment for microbial growth, further increasing BOD. In this study, plants were introduced at the end of Week 2, after ozonation had been performed in the relevant tanks. This timing explains why all tanks—except for the ‘No Plant No Ozone’ tank—experienced a rise in BOD following ozonation.

All tanks demonstrated acceptable efficiency in BOD removal during Week 11. The BOD measurement method is not sufficiently sensitive to detect small differences between samples. Consequently, the final BOD levels in all tanks (excluding the ‘No Plant No Ozone’ tank) are nearly identical. This conclusion is supported by the error bars.

These results validate the effectiveness of the combined CFW-ONBT systems. These variations are more pronounced by comparing the normalised values as illustrated in Figure 6b, which shows final BOD removal efficacies of 64% for ‘Ozone D1 and Plants’ and 43% for ‘Ozone D2 and Plants’ treatments. This result was likely attributed to the interaction between ozone and the plants when they were introduced to the tanks. The low initial BOD concentration in the wastewater samples led to similar treatment outcomes across all systems, except for the ‘No Plant No Ozone’ setup. Given the active role of both ozone and plants, as discussed earlier, the final BOD values in all investigated tanks were generally comparable. However, the effectiveness of the ‘Ozone D1 and Plants’ and ‘Ozone D2 and Plants’ systems remained significant. These findings highlight the potential of CFW combined with ONBT as a promising option for enhancing wastewater purification.

3.5. Nitrogen

Figure 7 shows the levels of nitrogen species. Nitrogen plays a key role in water ecology but can lead to eutrophication, toxicity, and even ecosystem imbalance when present at substantial levels [25]. Nitrate and nitrite result from converting organic nitrogen to inorganic one via ozone oxidation. Figure 7a,b show that the inorganic nitrogen concentration goes up with time in ‘Ozone D2’ and ‘Ozone D1’ treatments, and the concentrations remain high at the end of the study period (3.8 mg/L to 10.9 mg/L and 6.8 mg/L to 14.6 mg/L, respectively) due to the oxidation by ozonation. For ‘Plants Only’, ‘Ozone D1 and Plants’, and ‘Ozone D2 and Plants’, the initial increase was followed up with a reduction in its concentration to produce a total inorganic nitrogen reduction of 80% and 73% for ‘Plants Only’ and ‘Ozone D2 and Plants’, respectively. In those tanks treated with ozonation, the initial increase in nitrate and nitrite concentrations resulted from oxidation by ozone, as previously discussed. When plants were introduced after ozonation, nitrogen uptake by the plants contributed to the subsequent decline in nitrate and nitrite levels. In the ‘Plants Only’ tanks, the initial rise in inorganic nitrogen may be attributed to the nitrification process or plant activity that releases nitrogen compounds into the solution. The initial concentration of nitrate and nitrite in the sample with ‘No Plant No Ozone’ (control) fell to zero from Week 5 and remained the same up to the end of the study. This trend could be interpreted as a consequence of denitrification, supported by the anaerobic conditions within the tanks. This observation could be seen as extra information from the experiment.

The organic part of nitrogen (total Kjeldahl nitrogen, TKN) considerably dropped in all samples (Figure 7c,d) due to biosorption by plants, oxidation by ozone, or microbial activities. At the end of 11 weeks, an 80% reduction in TKN was obtained for the ‘Ozone D1 and Plants’ treatment. Other samples presented similar amounts of reduction in TKN. For the ‘No Plant No Ozone’ sample, an increase was found in Week 8 from 3.3 mg/L to 6.3 mg/L, which was somewhat compensated up to Week 11 with a decrease to 4.4 mg/L. The increase observed in Week 8 should be considered as an outlier due to sampling error or contamination by an organic source, introducing both nitrogen and phosphorus into the wastewater. The higher TKN concentration in Week 11 compared to Week 5 suggests that contamination was the most likely cause. The initial decline in TKN concentration up to Week 5 might result from the conversion of organic nitrogen to ammonium, followed by its reaction with nitrite, leading to nitrogen removal as N₂ gas. While this reaction typically requires controlled conditions, the low nitrogen concentration could have enhanced its effect in this case. Similar trends can be seen for C_i/C₀ graphs.

The total nitrogen (TN), representing the sum of organic and inorganic nitrogen, exhibited distinct trends across the studied samples (Figure 7e,f). In the ‘Ozone D1’ and ‘Ozone D2’ samples, which had identical initial and final concentrations (16.8 mg/L for ‘Ozone D1’ and 13.5 mg/L for ‘Ozone D2’), only minor variations were observed between these points. This stability was due to a compensatory effect, where the reduction in organic nitrogen was offset by an increase in the inorganic fraction. In contrast, other samples displayed an overall downward trend. The ‘Plants Only’ treatment achieved the highest reduction, with TN decreasing by 80%. The ‘Ozone D1 and Plants’ and ‘Ozone D2 and Plants’ treatments showed TN reductions of 38% and 75%, respectively. This general decline in TN was primarily attributed to nitrogen sorption by plants. The ‘No Plant No Ozone’ sample exhibited fluctuations, initially decreasing to 3.3 mg/L by Week 5, followed by an increase in the subsequent weeks, reaching 4.4 mg/L in Week 11. This pattern closely resembled the trend observed for TKN, likely due to sampling errors or contamination from an organic source introducing nitrogen and phosphorus into the wastewater. The higher TN concentration in Week 11 compared to Week 5 further supports contamination as the more probable cause. The initial decline by Week 5 was likely due to reductions in both the inorganic and organic nitrogen fractions during this period.

Previous research [26] shows successful application of nanobubble technology in removing ammonia and nitrogenous compounds, making it versatile for various water treatment applications. Compared to traditional ozonation, nanobubbles maintain higher dissolved ozone concentrations for longer, improving the degradation of microbial contaminants and organic compounds [27]. The efficiency of the treatment depends on the optimal ozone dosage and contact time. Research recommends dosages between 0.1 and 1.0 mg/L, with 0.2–0.5 mg/L effective for pollutants like ammonia. Contact times of 10–30 min are typical, with 20 min at 0.3 mg/L shown to be optimal for ammonia removal [26,28]. These findings highlight the potential of ozone nanobubbles for efficient, scalable water treatment across diverse applications. The reported results again highlight the significant role and capability of combined ozone and plant systems to reduce the total amount of nitrogen in wastewater and protect the aqueous ecosystems against the eutrophication threat. Once again, the same trends can be discovered in C_i/C₀ graphs.

3.6. Total Phosphorus (TP) Analysis

Total phosphorus (TP) is presented in wastewater due to several factors including detergents and is a leading cause of eutrophication [29]. Considering the important responsibility of phosphorous in the eutrophication phenomenon, its removal from the wastewater stream should be considered to meet the allowable limits. The starting concentrations of TP in various samples were observed from 2.2 mg/L to 4.8 mg/L. Figure 8 indicates that the variations in total phosphorus concentration are minimal across all samples, suggesting that ozone and plants do not significantly influence its treatment.

The highest TP reduction efficiency was 17% with one exclusion (‘No Plant No Ozone’ sample). An increase in TP value from 4.8 mg/L to 5.6 mg/L was observed for the ‘No Plant No Ozone’ sample in Week 8 which was later compensated to 2.5 mg/L, showing a total decrease of 47%. A likewise behaviour was seen for TN and TKN for the same sample. The fluctuation in TP concentration can be ascribed to microbial activity inside wastewater and phosphorus exiting the wastewater by the settlement process. Whilst no significant phosphorus removal was found in this work for combined ozone and plant treatment systems, they are still regarded as promising options for phosphorus removal in wastewater because phosphorus is an essential nutrient that plants absorb from their environment. Therefore, it is believed that further investigation can reveal the efficiency of CFW-ONBT systems in phosphorus removal. The increase in TP in Week 8 coincided with the rise in TKN and TN, likely because of sampling error or contamination of tank contents by an organic source that can add both nitrogen and phosphorus to the wastewater.

3.7. pH and Electrical Conductivity (EC) Analysis

Figure 9a,b show pH values for all tanks. As can be seen in this figure, the pH trend in all tanks is generally upward, which could be due to water evaporation and increasing the concentration of alkaline salts. The initial pH value is around 8 which can rise up to 20%. The highest rise in pH was observed for the ‘No Plant No Ozone’ sample. The changes in the electrical conductivity of the samples are shown in Figure 9c,d. As the figure indicates, the electrical conductivity in all samples generally increased. Electrical conductivity can also be used to measure the effectiveness of floating wetlands, as plant roots are capable of absorbing the salts that are accountable for conductivity in treated wastewater [30]. However, here, ozonation by oxidizing organic matter, e.g., organic nitrogen, to inorganic salts, enhances the conductivity of water. The evaporation can also raise the conductivity through increasing ionic compound concentration. As a result of the above-mentioned parameters, the EC of all samples increased, up to 74% in the ‘Ozone D2 and Plants’ sample.

4. Conclusions

This study investigated an innovative approach to improving wastewater treatment by combining ozone nanobubble technology (ONBT) with constructed floating wetlands (CFWs). It examined the synergistic interaction between ONBT’s powerful oxidative capabilities and the natural filtration and nutrient uptake functions of CFWs and highlighted the effectiveness of this hybrid treatment approach. However, phosphorus removal remained limited (approximately 12%), indicating the need for additional or complementary treatment strategies to address this particular challenge.

Overall, the findings underscore the advantages of combining ozone treatment with plant-based systems (i.e., CFW-ONBT systems), particularly at higher ozone dosages, as demonstrated in the ‘Ozone D2 and Plants’ tanks. The synergy between the oxidative power of ozone and the bioremediation capacity of plants significantly enhanced wastewater quality. Furthermore, the elevated oxygen concentration resulting from ozone decomposition likely played a beneficial role in the treatment process. This study offers valuable foundational insights into the integration of ONBT and CFWs, contributing to the advancement of sustainable and efficient wastewater treatment technologies.

5. Future Research

Drawing upon the findings of this study, future research should focus on scaling up CFW-ONBT systems and testing them over longer periods to assess their effectiveness in real-world conditions. Key priorities include optimising ozone dosage and exposure time to maximise treatment efficiency while minimising energy use and plant stress, exploring the scalability and automation potential of the system, incorporating a wider range of plant species to enhance performance across diverse conditions, and developing real-time monitoring systems to enable dynamic control of ONBT dosing and ensure system stability.