Do Gas Nanobubbles Enhance Aqueous Photocatalysis? Experiment and Analysis of Mechanism

: The performance of photocatalytic advanced oxidation must be improved in order for the technology to make the jump from academic research to widespread use. Research is needed on the factors that cause photocatalysis to become self-limiting. In this study, we introduced, for the ﬁrst time, nanobubbles continuously into a running photocatalytic reactor. Synthetic air, O 2 , and N 2 bubbles in the size range of 40 to 700 nm were added to a reaction system comprising P25 TiO 2 photocatalyst in stirred aqueous solution excited by UV-A lamps, with methyl orange as a target contaminant. The removal of methyl orange was tested under conditions of changing pH and with the addition of different radical scavengers. Results indicated that the oxygen and air nanobubbles improved the photocatalytic degradation of methyl orange—the removal efﬁciency of methyl orange increased from 58.2 ± 3.5% (N 2 aeration) to 71.9 ± 0.6% (O 2 aeration). Dissolved oxygen (DO) of 14.93 ± 0.13 mg/L was achieved using O 2 nanobubbles in comparison to 8.43 ± 0.34 mg/L without aeration. The photodegradation of methyl orange decreased from 70.8 ± 0.4% to 53.9 ± 0.5% as pH increased from 2 to 10. Experiments using the scavengers showed that O 2 − was the main reactive species in photocatalytic degradation under highly dissolved oxygen conditions, which also accounted for the observation that the removal efﬁciency for methyl orange decreased at higher pH. However, without photocatalyst, nanobubbles alone did not improve the removal of methyl orange, and nanobubbles also did not increase the degradation of methyl orange by only photolysis. These experiments show that oxygen and air nanobubbles can act as environmentally friendly catalysts for boosting the performance of photocatalytic water treatment systems. /HO 2 • , while in low dissolved oxygen solution, h + and • OH were the main reactive species. This study suggests that nanobubbles may be a novel method for enhancing the photocatalytic degradation of organic pollutants. W.Y.; W.Y.;


Introduction
Despite decades of research, photocatalysis has not made the jump from academic research to widespread use in water treatment and consumer products. Further progress requires analysis of the factors limiting the performance with the goal of enabling the technology transfer. In use, photocatalytic pollution control relies on the irradiation of a photocatalyst resulting in the production of an electron-hole pair. These primary products react to produce a cascade of secondary reactive species. Many of the details concerning these species, such as their rate of production, concentration, lifetime, and mutual interactions, are only known in broad outline. Going into detail, in the idealized oxidative view, conduction band electrons (e − cb ) react with O 2 to form O 2 •− , and valence band holes (h + ) react with H 2 O to form •OH [1]. Then, O 2 •− and •OH react with organic compounds. While h + and e − cb could react directly with pollutants, dissolved oxygen and especially water are likely to be at much higher concentrations than other components and will react first. Reductive photocatalysis can occur in anoxic solution, wherein reactions of e − cb are first. Reductive photocatalysis can occur in anoxic solution, wherein reactions of e − cb are the main drivers of contaminant degradation [2]. The loss of reactive radicals to side and self-reactions (Figure 1), including the recombination of e − cb and h + , is the largest factor limiting the performance and efficiency of photocatalysts [3,4]. Researchers have tried many methods to prevent the recombination of e − cb and h + with some success; for example, composite materials have been developed that join wide and narrow band gap photocatalysts [5][6][7][8][9] with the different regions acting as sources and reservoirs of e − cb. In addition, defects on the surface of photocatalysts can trap e − cb and prevent them from diffusing to stationary h + [10,11]. One strategy to prevent the recombination of e − cb and h + is to trap the excitation in another form. The requirements were to generate species with sufficiently long lifetime and reactivity with target compounds. The strategy would be to prevent recombination of e − cb and h + by converting them into other species, for example, O2 trapping e − as superoxide, O2‧ -.
In addition to chemical species, physical structures including phase boundaries should be included in the discussion. Neppolian et al. [12] showed that reactions on the surfaces of bubbles, including ‧OH reactions, accelerate the photocatalytic degradation of dodecylbenzensulfonate. It had also been suggested that the higher the O2 concentration in the solution, the higher the pollutant degradation efficiency. Fotiou et al. [13] reported that when O2 was saturated in the solution, the performance of photocatalysts improved. Wöhrle et al [14] showed that DO was the main oxygen source in the photocatalytic oxidation of benzene to phenol. Zhang et al. [15] showed that O2 plays an important role in the photocatalytic oxidation of alcohols. Overall, on the basis of these studies, there is good evidence that dissolved oxygen is a key component of photocatalytic activity [16]. However, the DO level in water is limited not just by its equilibrium with the atmosphere, but also transport into the bulk solution. As a result, most previous studies have been performed within a narrow range of DO values, and no studies have tested the effect of DO on the photocatalytic oxidation under supersaturation conditions. Nanobubbles introduce a unique chemical environment that can be used to test theories regarding the mechanisms and bottlenecks of photocatalysis. Nanobubbles provide abundant surface interface for a small amount of material, and in addition change the chemical composition of the bulk solution [17]. The Kelvin equation describes the vapor Researchers have tried many methods to prevent the recombination of e − cb and h + with some success; for example, composite materials have been developed that join wide and narrow band gap photocatalysts [5][6][7][8][9] with the different regions acting as sources and reservoirs of e − cb . In addition, defects on the surface of photocatalysts can trap e − cb and prevent them from diffusing to stationary h + [10,11]. One strategy to prevent the recombination of e − cb and h + is to trap the excitation in another form. The requirements were to generate species with sufficiently long lifetime and reactivity with target compounds. The strategy would be to prevent recombination of e − cb and h + by converting them into other species, for example, O 2 trapping e − as superoxide, O 2 •− . In addition to chemical species, physical structures including phase boundaries should be included in the discussion. Neppolian et al. [12] showed that reactions on the surfaces of bubbles, including •OH reactions, accelerate the photocatalytic degradation of dodecylbenzensulfonate. It had also been suggested that the higher the O 2 concentration in the solution, the higher the pollutant degradation efficiency. Fotiou et al. [13] reported that when O 2 was saturated in the solution, the performance of photocatalysts improved. Wöhrle et al [14] showed that DO was the main oxygen source in the photocatalytic oxidation of benzene to phenol. Zhang et al. [15] showed that O 2 plays an important role in the photocatalytic oxidation of alcohols. Overall, on the basis of these studies, there is good evidence that dissolved oxygen is a key component of photocatalytic activity [16]. However, the DO level in water is limited not just by its equilibrium with the atmosphere, but also transport into the bulk solution. As a result, most previous studies have been performed within a narrow range of DO values, and no studies have tested the effect of DO on the photocatalytic oxidation under supersaturation conditions. Nanobubbles introduce a unique chemical environment that can be used to test theories regarding the mechanisms and bottlenecks of photocatalysis. Nanobubbles provide abundant surface interface for a small amount of material, and in addition change the chemical composition of the bulk solution [17]. The Kelvin equation describes the vapor pressure of a curved surface such as a bubble of gas in a liquid [18]. Moreover, because of surface tension, the pressure inside a bubble increases with curvature and increases as the inverse of bubble radius. Very high pressures can be achieved. For example, according to the Young-Laplace equation, the pressure inside a 100 nm bubble in water is 1.54 MPa or over 15 bar [19]. Nanobubbles have minimal buoyancy due to their small volume and can remain suspended in solution indefinitely, allowing their dispersion throughout the bulk solution.
The rate of oxygen transfer into water depends on the mass transfer coefficient and the area of the gas-water interface [17]. Nanobubbles have a higher interfacial area per volume of gas than micrometer or larger bubbles, which results in higher rate of transfer into water. Research has shown that by using nanobubbles, the mass transfer rate into water for low solubility gases such as O 2 and ozone can double [20,21]. Unique chemistry occurs on a bubble surface, and high gas pressure can counteract local depletion in DO [17]; equilibrium between the gas and the solution is rapid, occurring in under a second.
The goal of this paper was to test the limits and mechanism of photocatalytic efficiency by using nanobubbles to impact the recombination of e − cb and holes, due to increased trapping of e − cb as O 2 •− as a consequence of enhanced DO, or because of the trapping of e − cb or other reactive species on bubble surfaces. The experiments used nano-particulate titanium dioxide dispersed in water solution containing an organic indicator dye (methyl orange). Conditions such as pH, addition of scavengers, and presence and composition of nanobubbles were varied. Photochemistry was initiated using UV-A lamps, and the concentration of methyl orange was measured at regular intervals to establish photocatalytic activity.

Bubble Size Distribution and DO Level
Several methods are available for making nanobubbles including hydrodynamic cavitation, particle cavitation, acoustic sonication, electrochemical cavitation, and mechanical agitation. In this study, hydrodynamic cavitation resulted as a consequence of high flow velocity caused by the release of a high-pressure gas through a small orifice. As described by Bernoulli's equation, the total pressure is conserved and if the dynamic pressure is increased the static pressure decreases accordingly. If it is lower than the vapor pressure of the liquid, cavitation results.
where P s is the static pressure, ρ the density, V the flow velocity, ρV 2 /2 the dynamic pressure, and P 0 the total pressure to remain constant [22]. Bubbles generated by the ceramic diffuser were mainly from 40 to 700 nm. Oxygen gas was seen to form smaller nanobubbles (40-700 nm) than air (200-700 nm). The size distributions are shown in Figure 2. The bubbles generated in this work had size distributions comparable with the bubbles generated in other studies using water circulation techniques in bubble generators (Table 1). It should be noted that most previous studies have prepared nanobubble solutions via generators that require circulation of the solution within the system, which can run only in batch-mode. In this study, however, the use of the ceramic diffuser is less complex and allows the continuous addition of nanobubbles to a running reaction.  Using different gases, the solutions showed different DO levels. The highest DO was in the O2 aerated solution, which was 14.93 mg/L, while the least saturated DO was generated by N2 bubbling (6.53 mg/L) (the air equilibrated O2 concentration was 8.18 mg/L). Compared to nanobubbles generated by generators with hydrodynamic cavitation method, the ceramic diffusor that is used in this study could not result in as high of a pressure as the pump provides (0.6 MPa), and thus it did not reach as high of a DO concentration as with the nanobubble generator; however, it provides a lower cost (Price: ≈ 166 USD, 0 W) compare to nanobubble generators (Price: ≈ 2000 USD, ≥3 kW) and continuous bubbling. Figure 3 shows the results of methyl orange photodegradation catalyzed by TiO2 with different gas types bubbling continuously and at neutral pH conditions. Control tests were performed using only TiO2 or with air, O2, or N2 bubbles without TiO2. The pseudofirst-order rate was calculated using the following equation:  Using different gases, the solutions showed different DO levels. The highest DO was in the O 2 aerated solution, which was 14.93 mg/L, while the least saturated DO was generated by N 2 bubbling (6.53 mg/L) (the air equilibrated O 2 concentration was 8.18 mg/L). Compared to nanobubbles generated by generators with hydrodynamic cavitation method, the ceramic diffusor that is used in this study could not result in as high of a pressure as the pump provides (0.6 MPa), and thus it did not reach as high of a DO concentration as with the nanobubble generator; however, it provides a lower cost (Price: ≈166 USD, 0 W) compare to nanobubble generators (Price: ≈2000 USD, ≥3 kW) and continuous bubbling. Figure 3 shows the results of methyl orange photodegradation catalyzed by TiO 2 with different gas types bubbling continuously and at neutral pH conditions. Control tests were performed using only TiO 2 or with air, O 2 , or N 2 bubbles without TiO 2 . The pseudo-first-order rate was calculated using the following equation:

Effect of Applied Gasses on the Degradation
where k 1 is the pseudo-first-order rate constant (min −1 ) of methyl orange photodegradation (apparent first order reaction rate constant), k c is the intrinsic rate constant (min −1 ) [27] for methyl orange photodegradation, and K O 2 is Langmuir adsorption constant of oxygen over TiO 2 surface (L mg −1 ).
With the assistance of O 2 nano-bubbles, the photocatalytic performance of TiO 2 increased to 14.7% compared to no bubbling, and with N 2 aeration, the removal of methyl orange was slightly decreased (−13.7%). The observed results could be due to the following factors: (a) dissolved oxygen helps increase photocatalytic oxidation; (b) nanobubble collapse increases photocatalytic oxidation; (c) the nanobubble surfaces modify the concentrations of reactive species acting both as a reservoir and a catalyst. On the basis of the evidence, we can strike option 'b' from the list, while research has shown that during the process of nanobubble collapse, reactive radicals such as •OH are generated [29]. However, in our study, O 2 nanobubbles alone did not degrade methyl orange. Thus, there was no evidence that nanobubbles themselves have oxidation potential, but they can act as co-catalysts.
Reactive  [32]. In our study, without photocatalysts, there was no removal of methyl orange by using only nanobubbles and light irradiation observed; however, with the presence of air and O 2 nanobubbles, the photocatalytic degradation of methyl orange increased (with TiO 2 and light irradiation).
DO concentrations decreased during the experiments due to the consumption of dissolved oxygen by the reaction (Figure 3B). The initial DO for air aeration and no aeration were similar; however, with air nanobubble aeration, the removal efficiency of methyl orange, 64.7 ± 0.3%, was higher than for no aeration, 60.3 ± 3.4%. This result showed that introducing nanobubbles into the system enhanced the photocatalytic efficiency of degrading methyl orange. The depletion rate of dissolved oxygen was calculated by linear fitting. The absolute value of the depletion rate of dissolved oxygen was 6.16 × 10 −2 , 2.60 × 10 −2 , 4.30 × 10 −2 mg/(L min) in O 2 -, N 2 -, and air-saturated solution, respectively. Without aeration, the rate was 3.86 × 10 −2 mg/(L min). With neither TiO 2 nor aeration, or without TiO 2 but with O 2 aeration, the depletion rate was close to 0. This result indicates there was no consumption of DO. The O 2 -saturated solution showed fastest DO depletion rate along with the largest k 1 . The consumption rate of dissolved oxygen and removal efficiency of methyl orange were correlated (Pearson's correlation coefficient 0.9977). This indicates that dissolved oxygen plays an important role in methyl orange removal.

The Role of Reactive Species in the Degradation Mechanisms
According to the established mechanisms of photocatalysis, the key reactive species are the photogenerated h + and e − cb , and reactive radical species such as O 2 •− and •OH [33]. The contributions of the separate reactive species for pollutant removal can be identified using specific scavengers. All experiments were performed at neutral pH. The characterization of the reactive species was performed by introducing radical scavengers into the system.
Potassium iodide (KI) [34,35] was used as the h + and •OH radical scavenger. Sodium nitrate NaNO 3 was used as the e − cb scavenger [36], tert-butyl alcohol (TBA) [37,38] was the •OH scavenger, and p-benzoquinone (pBQ) [39,40] was the O 2 •− scavenger. During the tests, a 5.5 mM aliquot of a single scavenger was added into solution with 20 mg L −1 methyl orange and 2 g L −1 TiO 2 to monitor the change of methyl orange removal efficiency compared with the results from tests without scavengers.
As shown in Figure 4, the inhibition effects of KI, TBA, NaNO 3 , and pBQ without aeration were 28.2%, 47.0%, 26.5%, and 24%, respectively. As shown in Table 2, the pseudo-first-order reaction rate constant k1 decreased to 0.0017 min −1 when pBQ was added, thus indicating the O2 ·− played a predominant role in the air nanobubble photocatalytic degradation of methyl orange. With the addition of NaNO3, k1 decreased to 0.0026 min −1 , which we interpret as demonstrating e − eb is another major active species of air nanobubble photocatalytic degradation of methyl orange. With h + and ‧OH together, k1 decreased to 0.0042 min −1 while with only ‧OH, k1 was 0.0066 min −1 . With N2 aeration, h + and ‧OH play a dominant role due to the low dissolved oxygen in the solution. Without aeration, O2 ·− and e − cb are dominant, making similar contributions to the methyl orange removal. With these conditions, they are more important than h + and ‧OH.
The large decrease in the methyl orange removal efficiency when using pBQ in air aeration and O2 aeration shows that the dominant reactive species of methyl orange degradation in an oxygenated environment was O2 ·− , in accordance with other reports (Table  3). As shown in Table 2, the pseudo-first-order reaction rate constant k 1 decreased to 0.0017 min −1 when pBQ was added, thus indicating the O 2 •− played a predominant role in the air nanobubble photocatalytic degradation of methyl orange. With the addition of NaNO 3 , k 1 decreased to 0.0026 min −1 , which we interpret as demonstrating e − eb is another major active species of air nanobubble photocatalytic degradation of methyl orange. With h + and •OH together, k 1 decreased to 0.0042 min −1 while with only •OH, k 1 was 0.0066 min −1 . With N 2 aeration, h + and •OH play a dominant role due to the low dissolved oxygen in the solution. Table 2. Comparison of different gas aeration methods and decrease in the apparent first-order reaction rate constant k 1 (/min) for different scavengers (20 mg L −1 methyl orange, 2 g L −1 TiO 2 , UVA). The large decrease in the methyl orange removal efficiency when using pBQ in air aeration and O 2 aeration shows that the dominant reactive species of methyl orange degradation in an oxygenated environment was O 2

Target Reactive Species
•− , in accordance with other reports (Table 3). Therefore, we conclude that photogenerated h + and •OH together make the main contribution towards oxidizing methyl orange, both with and without nano-bubbling, while O 2 •− was the main reactive oxidizing reagent in a dissolved oxygen-rich environment. In N 2 nanobubble photocatalytic degradation, •OH contributed most to the degradation of methyl orange. A study by Temesgen et al. showed that the surfaces of nanobubbles are negatively charged with OH − from the water molecules [22], which might provide the reactive species in a low dissolved oxygen environment.
Wang et al. [25] reported over 60% of oxytetracycline (OTC) was directly photodegraded (k = 3.92 × 10 −3 min −1 ) in the presence of only O 2 nanobubbles with sizes in the range from 100 to 500 nm after 240 min of visible light. In their study, •OH was shown to be the main reactive species of OTC degradation, and 20% of OTC would degrade with only photolysis. The main reactive species for degrading methyl orange was found to be O 2 •− in the present work; our study showed no degradation with only O 2 nanobubbles and photolysis. This might suggest that different pollutants undergo different degradation mechanism, and that adding O 2 nanobubbles for water treatment may not affect the degradation of all pollutants similarly.

Effect of pH
The pH value had a critical effect on the efficiency of photocatalytic reactions. The influence of solution pH on the process was determined using a constant concentration of methyl orange at 20 mg L −1 , 2 g L −1 TiO 2 with O 2 nanobubbles, and without bubbling by varying pH from 2 to 10. The results are shown in Figure 5 and Table 4. The lowest removal efficiency for methyl orange was detected at pH = 10 with and without bubbling, and the highest removal efficiency was found at pH = 2 and pH = 5 for no bubbling and air bubbling, respectively.
Catalysts 2021, 11, x FOR PEER REVIEW 9 of 14 and the highest removal efficiency was found at pH = 2 and pH = 5 for no bubbling and air bubbling, respectively. The methyl orange (MO) adsorption efficiency of TiO2 is significantly impacted by the changing electrostatic interaction between TiO2 and MO under the different pH conditions [43]. The adsorption efficiency values at different pH values are summarized in Table 5. The highest adsorption capacity (31.0 ± 4.2%) was observed at pH = 5, while no MO adsorption occurred on the surface of TiO2 at pH = 10. These differences resulted from the changes in the charge of MO and the surface charge of TiO2 particles. The point of zero charge of TiO2 is at pH = 5.8 ± 0.1 [44]; if the pH is more acidic than this value, the surface is positively charged, TiO2 + H + →TiO2H + , and if pH is above this value, it is negatively The methyl orange (MO) adsorption efficiency of TiO 2 is significantly impacted by the changing electrostatic interaction between TiO 2 and MO under the different pH conditions [43]. The adsorption efficiency values at different pH values are summarized in Table 5. The highest adsorption capacity (31.0 ± 4.2%) was observed at pH = 5, while no MO adsorption occurred on the surface of TiO 2 at pH = 10. These differences resulted from the changes in the charge of MO and the surface charge of TiO 2 particles. The point of zero charge of TiO 2 is at pH = 5.8 ± 0.1 [44]; if the pH is more acidic than this value, the surface is positively charged, TiO 2 + H + →TiO 2 H + , and if pH is above this value, it is negatively charged, TiO 2 + OH − → TiO 2 OH − [45]. Thus, at acidic pH values, negatively charged nanobubbles will be attracted to the surface. The methyl orange indicator has a pKa value of 3.47. At pH values above this value, the molecule is negatively charged [46]. In this form there is electrostatic repulsion between the negatively charged TiO 2 and negatively charged sulfate group on the azo-form methyl orange, leading to no methyl orange adsorption on the TiO 2 surface. Under conditions with pH below pKa, methyl orange is better adsorbed since the RSO 3 − is attracted by positively charged surface functionalities. According to the scavenger test the main reactive species for degrading methyl orange is O 2 •− , while at high pH, the high concentration of OH − may limit the separation of electron-hole pairs, which will further inhibit photocatalytic activity. Since pKa of the O 2 •− /HO 2 • was 4.8, HO 2 may act as a stronger oxidant than O 2 •− , which could be contributing to the faster MO degradation at pH = 2. Compared to other studies (Table 5), at lower pH, the photocatalytic reaction presented the highest removal efficiency in most cases. The trend of the effect of pH did not change because of higher oxygen dissolved in solution.
At pH = 10, the removal decreased compared to no aeration. Nanobubbles were proved to be negatively charged from pH 2 to 12 [22], and the magnitude of the negative zeta potential of nanobubbles increased monotonically as a function of pH. At pH = 10, the zeta potential of oxygen nanobubbles could reach to −32 mV, resulting in a negative charge on nanobubble's surface. At low pH, nanobubbles are expected to have a short lifetime due to the lack of OH − ions needed to form a stable electric layer around the bubble surface, while in alkaline solutions, the bubbles are more stable and have a longer lifetime [47]. Agrawal et al. [48] showed that lower pH enhances the generation of free radicals. Our experimental results confirm these findings.

Chemical and Reagents
UV-A light from 340 to 400 nm was generated using 1 PL-L 24W/10/4P lamp with light intensity of 4.9 W of UV-A output from Philips with a maximum intensity at 365 nm. The concentration of methyl orange (pH 3.0-4.4, Sigma-Aldrich, St. Louis, MO, USA) was 20 mg/L. The P25 TiO 2 was purchased from Sigma-Aldrich (St. Louis, MO, USA, 21 nm primary particle size, ≥99.5% purity on a trace metals basis) and added at a nominal concentration of 2 g/L. O 2 (Air Liquide, Alphagaz−1, 99.5%, Taastrup, Denmark), and N 2 (Air Liquide, Alphagaz−1, 99.995%) were used without further purification. Filtered compressed technical air was used as supplied by university facilities. Ceramic micro bubble diffusers (PMBD 75, Point Four Systems, Airoxi, India) were used to produce uniform fine bubbles from 100 to 500 nm, with gas pressures from 1.7 to 2.4 bar [52].

Sample Characterization
The gas bubble size distribution was measured using a Zetasizer (Brookhaven Instruments, Holtsville, NY, USA). The DO concentration and pH were measured using the Multiparameter Meter edge ® from Hanna Instruments, Kungsbacka, Sweden.

Experiment Setup
The photocatalytic activity of TiO 2 nanoparticles was quantified by monitoring the photodegradation of methyl orange in aqueous solution with continuous supply of nanobubbles, under UVA irradiation. Photocatalytic tests were carried out using a beaker filled with an aqueous suspension (2 L) containing methyl orange, TiO 2 , and the nanobubble generator with different types of gas; without aeration, pure N 2 , pure O 2 , and with clean technical air.
Initially, 20 mg/L of methyl orange was put into a beaker with Milli-Q water and stirred for 10 min. The pH of the solution was adjusted using NaOH or HCl and measured with the pH meter. TiO 2 (2 g/L) was added into a stirred beaker under dark conditions for 30 min. Keeping it in the dark, the nanobubble generator with the selected gas (pure N 2 , pure O 2 , or clean technical air; flow rate: 100 mL min −1 ) generates nanobubbles with inherently different DO levels. After 30 min, the system was irradiated under UVA light until the initial dissolved oxygen concentration reached a constant level, keeping the solution in a water bath at 23 ± 2 • C with continuous magnetic stirring. The UV lamps were immersed in the solution with a light path of 12 cm. Next, samples were taken at specific time intervals. After centrifugal separation, twice at 5000 rpm for 20 min (Labofuge 200, Heraeus Sepatech GmbH, Oesterode, Germany ), the absorption spectra of methyl orange in supernatant were analyzed with a UV-VIS spectrophotometer (UV−2401 PC, Shimadzu Inc., Columbia, MD, USA).

Conclusions
We studied the synergistic effect of nanobubbles and TiO 2 photocatalysis on the degradation of methyl orange under UVA light for the first time. By adding oxygen nanobubbles, we increased photocatalytic degradation from 58.2 ± 3.5% (N 2 aeration) to 71.9 ± 0.6%, whereas without aeration, the removal efficiency was 60.3 ± 3.4%. This may have been due to increased DO, increased oxygen gas/water surface area, or the presence of oxygen bubbles on the photocatalyst surface. Methyl orange could not be directly degraded by nanobubbles without irradiation or without the photocatalyst, while during the photocatalytic reaction, the consumption of DO may enhance the generation of reactive species, thus improving the removal efficiency. The removal efficiency was greater at acidic pH. Scavenger tests showed that h + , •OH, O 2 •− , and e − cb were the reactive species in the methyl orange degradation. The dominant active species in the reaction with high dissolved oxygen condition was O 2 •− /HO 2 • , while in low dissolved oxygen solution, h + and •OH were the main reactive species. This study suggests that nanobubbles may be a novel method for enhancing the photocatalytic degradation of organic pollutants.