Stability and Free Radical Production for CO2 and H2 in Air Nanobubbles in Ethanol Aqueous Solution

In this study, 8% hydrogen (H2) in argon (Ar) and carbon dioxide (CO2) gas nanobubbles was produced at 10, 30, and 50 vol.% of ethanol aqueous solution by the high-speed agitation method with gas. They became stable for a long period (for instance, 20 days), having a high negative zeta potential (−40 to −50 mV) at alkaline near pH 9, especially for 10 vol.% of ethanol aqueous solution. The extended Derjaguin, Landau, Verwey, and Overbeek (DLVO) theory was used to evaluate the nanobubble stability. When the nanobubble in ethanol alkaline aqueous solution changed to an acidic pH of around 5, the zeta potential of nanobubbles was almost zero and the decrease in the number of nanobubbles was identified by the particle trajectory method (Nano site). The collapsed nanobubbles at zero charge were detected thanks to the presence of few free radicals using G-CYPMPO spin trap reagent in electron spin resonance (ESR) spectroscopy. The free radicals produced were superoxide anions at collapsed 8%H2 in Ar nanobubbles and hydroxyl radicals at collapsed CO2 nanobubbles. On the other hand, the collapse of mixed CO2 and H2 in Ar nanobubble showed no free radicals. The possible presence of long-term stable nanobubbles and the absence of free radicals for mixed H2 and CO2 nanobubble would be useful to understand the beverage quality.


Introduction
There have been many reports on bulk nanobubbles or nanoparticles in ethanol aqueous solution. In this research, the object of nanobubbles in ethanol solution was hydrogen in argon and carbon dioxide gas nanobubbles. However, there are some reports on the exsistence of imputities in bulk ethanol nanobubbles. Therefore, they were first introduced and then the other reports on the stability of nanobubbles in ethanol were discussed.
The structure and properties of ethanol aqueous solutions were reported in the literature [1][2][3][4]. For nanobubble production in ethanol aqueous solution, some works have reported existing impurities. Rak and Sedlák, 2020 showed the existence of hydrophobic ing microbubbles under strongly acidic conditions without any dynamic stimulus such as ultrasound or a large pressure difference [25]. Liu et al., 2016 reported that ·OH and superoxide ion (·O 2 − ) can be produced in an aqueous solution with bulk nanobubbles, and investigated the germination processes for plant seeds [26]. The ·OH could be generated directly from bursting nanobubbles or simple hypoxic sediment/water oxygenation [27].
On the other hand, the free radical could not be generated by the self-collapse of air micro/nano bubbles in pure water produced by fiber membrane filter, and the ·OH peak was observed with a weak supersonic wave [28]. The results of the numerical simulations suggested that no ·OH was produced from a dissolving nanobubble. It was suggested that the signals reported experimentally did not originate in ·OH, but rather in H 2 O 2 produced during hydrodynamic cavitation in the production of bulk nanobubbles [26]. The radical production by ultrasonic wave irradiation becomes more important to produce the radicals, especially ·OH in water, by comparing no irradiation and irradiation [29]. Fujita et al., 2021 reported the ·OH scavenging and the ·O 2 diminishing by mixing the CO 2 nanobubbles after hydrogen nanobubble blowing in water and alcohol aqueous solution [30]. As described above, the ethanol and water solutions containing nanobubbles have various characteristics and there is a possibility to produce the free radicals. Soda drinks with carbon dioxide need long stability and the prevention of free radical production in consideration of our health.
In this study, the characteristics of hydrogen in argon and carbon dioxide nanobubbles in ethanol aqueous solution were investigated from the point of view of stability for 20 days of nanobubbles according to the surface charge on bubbles and radical production by controlling the bubble surface charge to near zero by changing pH and with the application of small ultrasonic wave and ultraviolet irradiation. When the absolute value of zeta potential of nanobubbles is low, there is a possibility of the collapse of nanobubbles by Brownian motion and the production of the free radical. The presence or absence of the free radical is useful to know the beverage quality, among others.

Materials
Deionized (DI) water with a resistivity of 18.2 MΩ·cm prepared by the Classic Water Purification System from Hitech instruments CO., Ltd. (Shanghai, China) was used. Ethanol with a purity higher than 99.7% produced by Guangdong Guanghua Sci-Tech Co., Ltd., Guangzhou, Guangdong, China was used. The ethanol percentages of ethanol aqueous solution mixtures were 0, 10, 30, and 50 vol%. The gases of 8% H 2 in Ar and CO 2 were supplied from a tank produced by Guangdong Huate Gas Co., Ltd., Guangzhou, Guangdong, China. The pH adjusters of ethanol aqueous solution were sodium hydroxide (NaOH) aqueous solution (Guangdong Guanghua Sci-Tech Co., Ltd., Guangzhou, Guangdong, China) and hydrochloric acid (HCl) aqueous solution (Chengdu Chron Chemicals Co., Ltd., Qionglai, China). The solubility of H 2 gas [31] and CO 2 gas [32] in ethanol was about 100 times in H 2 and 10 times in CO 2 larger than those gases in water.

Nanobubble Preparation and Measurements
The nanobubbles were prepared using mechanical high-speed cavitation equipment (self-made equipment), as shown in Figure 1. The gases of 8% in Ar and CO 2 were fed from the gas tank through the gas inlet, and the propeller mixing speed was 20,000 rpm using a 7.2 cm diameter of the blade. The gas mixture nanobubbles were prepared by initially blowing 8% H 2 in Ar gas and then CO 2 gas. There are several bulk nanobubble preparation methods, such as the utilization of high-speed cavitation, pressure difference with circulation, ultrasonic wave, and passing ultrafine pores [33,34]. In this study, the equipment shown in Figure 1 was utilized to produce a large amount of nanobubbles in liquid in a fast manner. There are some methods to measure the nanobubble size; for example, dynamic light scattering (DLS), particle trajectory, resonant mass, and laser diffraction methods [34]. In particular, it was convenient to measure the particle/bubble size distribution by the DLS and particle trajectory methods discussed in previous studies of this group [35]. In this study, the nanobubble size distribution was measured by the DLS system (NanoBook Omni, Brookhaven Instruments, Holtsville, NY, USA). The nanobubble number density was obtained through outsourcing from NanoSight, NS300, Malvern (Worcestershire, UK). The zeta potential values were measured through the micro-electrophoresis method by the phase analysis light scattering method (NanoBrook Omni, Brookhaven Instruments, Holstville, NY, USA.).

Radical Preparation and Measurement by ESR
The solution pH was adjusted to pH 9 for the three kinds of nanobubbles (i.e., 8% H2 in Ar, CO2, and a mixture of CO2 after 8% H2 in Ar) in ethanol aqueous solutions, and they were kept for 20 days. After 20 days, the pH of ethanol aqueous solutions was decreased to pH 5 by adding HCl aqueous solution, and the solutions were set in the ultrasonic wave vessel (SIBATA SCIENTIFIC TECHNOLOGY, Tokyo, Japan, SU,40 kHz, 500 W) for 30 s; then, rapidly, the radicals were measured. The produced radicals were measured by the following procedure. A spin-trapping reagent, sc-5-(5,5-dimethyl-2-oxo-1,3,2-dioxaphosphinan-2-yl)-5 methyl-1-pyrroline N-oxide (G-CYPMPO)24, was used by adding its solution. G-CYPMPO could spin-trap •OH in UV (4 W, OHM ELECTRIC INC., Tokyo, Japan) illuminated condition and •O2 − [36][37][38]. A JEOL JES-TE25X ESR spectrometer (Tokyo, Japan) was used to obtain ESR spectra of free radicals of •OH and •O2 − . The measured peaks produced by nanobubble collapse were compared with eight kinds of peak positions of standard •OH and •O2 − .

Determination of Diameter of Nanobubbles
The stability of nanobubbles in water was reported in the literature [35], and there are reports on the stabilization of bubbles by ion adsorption [39]. The nanobubbles displaying a higher absolute zeta potential value showed a good stability and constant nanobubble diameter for a long period. On the other hand, with a low zeta potential value, the nanobubble size quickly increased and disappeared. The zeta potentials of 8% H2 in Ar, CO2, and the mixture of 8% H2 in Ar and CO2 gas nanobubbles in water as a function of pH are shown in Figure 2 [35], where the isoelectric point (IEP) is in between 5 and 6. As shown in Figure 2, in the ethanol aqueous solution of less than 50 vol%, the zeta potentials of nanobubbles are also close to zero at pH 5. To maintain the stability of nanobubbles, There are several bulk nanobubble preparation methods, such as the utilization of highspeed cavitation, pressure difference with circulation, ultrasonic wave, and passing ultrafine pores [33,34]. In this study, the equipment shown in Figure 1 was utilized to produce a large amount of nanobubbles in liquid in a fast manner. There are some methods to measure the nanobubble size; for example, dynamic light scattering (DLS), particle trajectory, resonant mass, and laser diffraction methods [34]. In particular, it was convenient to measure the particle/bubble size distribution by the DLS and particle trajectory methods discussed in previous studies of this group [35]. In this study, the nanobubble size distribution was measured by the DLS system (NanoBook Omni, Brookhaven Instruments, Holtsville, NY, USA). The nanobubble number density was obtained through outsourcing from NanoSight, NS300, Malvern (Worcestershire, UK). The zeta potential values were measured through the micro-electrophoresis method by the phase analysis light scattering method (NanoBrook Omni, Brookhaven Instruments, Holstville, NY, USA.).

Radical Preparation and Measurement by ESR
The solution pH was adjusted to pH 9 for the three kinds of nanobubbles (i.e., 8% H 2 in Ar, CO 2 , and a mixture of CO 2 after 8% H 2 in Ar) in ethanol aqueous solutions, and they were kept for 20 days. After 20 days, the pH of ethanol aqueous solutions was decreased to pH 5 by adding HCl aqueous solution, and the solutions were set in the ultrasonic wave vessel (SIBATA SCIENTIFIC TECHNOLOGY, Tokyo, Japan, SU, 40 kHz, 500 W) for 30 s; then, rapidly, the radicals were measured. The produced radicals were measured by the following procedure. A spin-trapping reagent, sc-5-(5,5-dimethyl-2-oxo-1,3,2-dioxapho-sphinan-2-yl)-5 methyl-1-pyrroline N-oxide (G-CYPMPO)24, was used by adding its solution. G-CYPMPO could spin-trap ·OH in UV (4 W, OHM ELECTRIC INC., Tokyo, Japan) illuminated condition and ·O 2 − [36][37][38]. A JEOL JES-TE25X ESR spectrometer (Tokyo, Japan) was used to obtain ESR spectra of free radicals of ·OH and ·O 2 − . The measured peaks produced by nanobubble collapse were compared with eight kinds of peak positions of standard ·OH and ·O 2 − .

Determination of Diameter of Nanobubbles
The stability of nanobubbles in water was reported in the literature [35], and there are reports on the stabilization of bubbles by ion adsorption [39]. The nanobubbles displaying a higher absolute zeta potential value showed a good stability and constant nanobubble diameter for a long period. On the other hand, with a low zeta potential value, the nanobubble size quickly increased and disappeared. The zeta potentials of 8% H 2 in Ar, CO 2 , and the mixture of 8% H 2 in Ar and CO 2 gas nanobubbles in water as a function of pH are shown in Figure 2 [35], where the isoelectric point (IEP) is in between 5 and 6. As shown in Figure 2, in the ethanol aqueous solution of less than 50 vol%, the zeta potentials of nanobubbles are also close to zero at pH 5. To maintain the stability of nanobubbles, the pH of ethanol aqueous solution containing nanobubbles can be adjusted to assign higher absolute zeta potential values on the bubble surface, for instance, at alkaline pH 9 or acidic Nanomaterials 2022, 12, 237 5 of 15 pH 3. In this experiment, pH 9 was selected to examine the stability of bubbles and fits with the water quality standard pH (from 6.5 to 9.5 in EU directive [40]). The nanobubble stability was also evaluated by extended DLVO theory calculation, which will be discussed in the following Section 3.3.
higher absolute zeta potential values on the bubble surface, for instance, at alkaline pH 9 or acidic pH 3. In this experiment, pH 9 was selected to examine the stability of bubbles and fits with the water quality standard pH (from 6.5 to 9.5 in EU directive [40]). The nanobubble stability was also evaluated by extended DLVO theory calculation, which will be discussed in the following Section 3.3.
When CO2 nanobubbles are added in 8% H2 in Ar nanobubbles containing aqueous solution, the CO2 solubility is high in aqueous solution and HCO3 − ion is produced according to the acidity constant. The HCO3 − ions are adsorbed on the positively charged bubbles. While in the alkaline region, CO3 2− ion is also produced according to equilibrium constant; however, anions are not adsorbed on the negatively charged bubbles. This agreed with the CO2 solubility phenomena explained in the literature [35]. Figure 2. Zeta potential as a function of pH for 8% H2 in Ar, CO2, and a mixture of 8% H2 in Ar and CO2 gas nanobubble in water [35].
The nanobubble mean diameters at pH 9 after 1 and 20 days are shown in Figure 3. The diameters of 8% H2 in Ar nanobubbles in ethanol aqueous solutions with various ethanol vol.% in the first day by controlling pH 9 are small and do not change significantly (between 300 and 750 nm). On the other hand, CO2 and CO2 after 8% H2 in Ar had noticeably large diameters at 50 vol.% ethanol aqueous solution (2000-3500 nm). The CO2 nanobubbles undergo mass loss at a higher pH, corresponding to the mass transfer process owing to the concentration gradient at the surrounding nanobubbles, and their mean diameter decreased [41]. Figure 3 shows that the CO2 diameter decreased after 20 days, in 0 vol.% (from 1300 to 600 nm) and 50 vol.% ethanol (from 3500 to 2200 nm). The nanobubble size at the CO2 after 8% H2 in Ar was larger at 10 and 30 vol% ethanol on the first day (1300 nm) than at 20 days (250-600 nm). It can be explained by Ostwald ripening [42], increasing the nanobubble size, and the size decreased after 20 days. After 20 days, the three kinds of nanobubble mean diameters existed from 300 to 600 nm in 10 and 30 vol% ethanol aqueous solution. Figure 2. Zeta potential as a function of pH for 8% H 2 in Ar, CO 2 , and a mixture of 8% H 2 in Ar and CO 2 gas nanobubble in water [35].
When CO 2 nanobubbles are added in 8% H 2 in Ar nanobubbles containing aqueous solution, the CO 2 solubility is high in aqueous solution and HCO 3 − ion is produced according to the acidity constant. The HCO 3 − ions are adsorbed on the positively charged bubbles. While in the alkaline region, CO 3 2− ion is also produced according to equilibrium constant; however, anions are not adsorbed on the negatively charged bubbles. This agreed with the CO 2 solubility phenomena explained in the literature [35].
The nanobubble mean diameters at pH 9 after 1 and 20 days are shown in Figure 3. The diameters of 8% H 2 in Ar nanobubbles in ethanol aqueous solutions with various ethanol vol.% in the first day by controlling pH 9 are small and do not change significantly (between 300 and 750 nm). On the other hand, CO 2 and CO 2 after 8% H 2 in Ar had noticeably large diameters at 50 vol.% ethanol aqueous solution (2000-3500 nm). The CO 2 nanobubbles undergo mass loss at a higher pH, corresponding to the mass transfer process owing to the concentration gradient at the surrounding nanobubbles, and their mean diameter decreased [41]. Figure 3 shows that the CO 2 diameter decreased after 20 days, in 0 vol.% (from 1300 to 600 nm) and 50 vol.% ethanol (from 3500 to 2200 nm). The nanobubble size at the CO 2 after 8% H 2 in Ar was larger at 10 and 30 vol% ethanol on the first day (1300 nm) than at 20 days (250-600 nm). It can be explained by Ostwald ripening [42], increasing the nanobubble size, and the size decreased after 20 days. After 20 days, the three kinds of nanobubble mean diameters existed from 300 to 600 nm in 10 and 30 vol% ethanol aqueous solution.

Effect of Ethanol Ratio in Zeta Potential and pH of Nanobubbles
Zeta potential and pH for 8% H2 in Ar nanobubble solution as a function of ethanol percentage in ethanol aqueous solution mixture after 1 and 20 days are shown in Figure  4. The natural pH of nanobubble solutions was around pH 6 to 7 after 1 and 20 days. Once the solution pH was adjusted to pH 9 by adding NaOH aqueous solution after the first day; it decreased to pH around 8 after 20 days, and the absolute value of the negative zeta potential decreased. The hydrogen solubility in ethanol is explained by Henry's law [31]. The pH solutions at around pH 8 were adjusted to 5 by adding HCl aqueous solution. The zeta potential of 8% H2 in Ar nanobubbles at pH 5 was positive of a few mV, regardless of ethanol percentage. In particular, the zeta potential deviation was the largest (between −45 and 5 mV) at 10 vol% ethanol aqueous solution during the above-mentioned conditioning procedures. Thus, the radical production by 8% H2 in Ar nanobubbles in 10 vol% ethanol aqueous solution was investigated by changing the pH from 9 to 5 after 20 days, and the results will be discussed in Section 3.5.
Zeta potential and pH of CO2 nanobubble solution as a function of ethanol percentage in ethanol aqueous solution mixtures after 1 and 20 days are shown in Figure 5. The natural pH of nanobubble solutions was around pH 4 to 5 after 1 and 20 days. Once the solution pH was adjusted to pH 9 by adding NaOH aqueous solution on the first day, the solution pH increased to between 9 and 10 after 20 days, and the absolute value of negative zeta potential increased 5 to 10 mV at 10 to 30 vol% ethanol aqueous solution. Our results agreed with the literature. The pH of the aqueous solution containing CO2 gas nanobubbles slightly increased after several days compared with the pH under the initial condition [35]. Dalmolin et al. [32] showed that the CO2 solubility in ethanol aqueous solution increased by increasing the ethanol mole fraction and pressure and decreasing the temperature.

Effect of Ethanol Ratio in Zeta Potential and pH of Nanobubbles
Zeta potential and pH for 8% H 2 in Ar nanobubble solution as a function of ethanol percentage in ethanol aqueous solution mixture after 1 and 20 days are shown in Figure 4. The natural pH of nanobubble solutions was around pH 6 to 7 after 1 and 20 days. Once the solution pH was adjusted to pH 9 by adding NaOH aqueous solution after the first day; it decreased to pH around 8 after 20 days, and the absolute value of the negative zeta potential decreased. The hydrogen solubility in ethanol is explained by Henry's law [31]. The pH solutions at around pH 8 were adjusted to 5 by adding HCl aqueous solution. The zeta potential of 8% H 2 in Ar nanobubbles at pH 5 was positive of a few mV, regardless of ethanol percentage. In particular, the zeta potential deviation was the largest (between −45 and 5 mV) at 10 vol% ethanol aqueous solution during the above-mentioned conditioning procedures. Thus, the radical production by 8% H 2 in Ar nanobubbles in 10 vol% ethanol aqueous solution was investigated by changing the pH from 9 to 5 after 20 days, and the results will be discussed in Section 3.5.  The pH of solutions with pH around 8 were adjusted to pH 5 by adding HCl aqueous solution. The zeta potential of CO2 nanobubbles at pH 5 was a few negative mV. In particular, the zeta potential deviation was the largest (between −45 and 5 mV) at 10 vol% ethanol aqueous solution during the above-mentioned conditioning procedures. This can Zeta potential and pH of CO 2 nanobubble solution as a function of ethanol percentage in ethanol aqueous solution mixtures after 1 and 20 days are shown in Figure 5. The natural pH of nanobubble solutions was around pH 4 to 5 after 1 and 20 days. Once the solution pH was adjusted to pH 9 by adding NaOH aqueous solution on the first day, the solution pH increased to between 9 and 10 after 20 days, and the absolute value of negative zeta potential increased 5 to 10 mV at 10 to 30 vol% ethanol aqueous solution. Our results agreed with the literature. The pH of the aqueous solution containing CO 2 gas nanobubbles slightly increased after several days compared with the pH under the initial condition [35]. Dalmolin et al. [32] showed that the CO 2 solubility in ethanol aqueous solution increased by increasing the ethanol mole fraction and pressure and decreasing the temperature. The pH of solutions with pH around 8 were adjusted to pH 5 by adding HCl aqueous solution. The zeta potential of CO2 nanobubbles at pH 5 was a few negative mV. In particular, the zeta potential deviation was the largest (between −45 and 5 mV) at 10 vol% ethanol aqueous solution during the above-mentioned conditioning procedures. This can be explained by the bubble collapse. Therefore, the radical production by CO2 nanobubbles in 10 vol% ethanol aqueous solution was investigated by changing the pH from 9 to 5 after 20 days, and the results will be discussed in Section 3.5. Zeta potential and pH of the mixture of CO2 and 8% H2 in Ar nanobubble solution as a function of ethanol percentage in ethanol aqueous solution mixtures after 1 and 20 days are shown in Figure 6. The natural pH of nanobubble solutions was around pH 5 to 6 after 1 and 20 days. On the other hand, once the solution pH was adjusted to pH 9 by adding NaOH aqueous solution on the first day, the solution pH increased to between 9 and 9.5 after 20 days, and the absolute value of negative zeta potential increased from 10 to 30 vol.% ethanol aqueous solution (−40 to −50 mV). The pH of solutions with pH around 8 were adjusted to pH 5 by adding HCl aqueous solution. The zeta potential of CO 2 nanobubbles at pH 5 was a few negative mV. In particular, the zeta potential deviation was the largest (between −45 and 5 mV) at 10 vol% ethanol aqueous solution during the above-mentioned conditioning procedures. This can be explained by the bubble collapse. Therefore, the radical production by CO 2 nanobubbles in 10 vol% ethanol aqueous solution was investigated by changing the pH from 9 to 5 after 20 days, and the results will be discussed in Section 3.5.
Zeta potential and pH of the mixture of CO 2 and 8% H 2 in Ar nanobubble solution as a function of ethanol percentage in ethanol aqueous solution mixtures after 1 and 20 days are shown in Figure 6. The natural pH of nanobubble solutions was around pH 5 to 6 after 1 and 20 days. On the other hand, once the solution pH was adjusted to pH 9 by adding NaOH aqueous solution on the first day, the solution pH increased to between 9 and 9.5 after 20 days, and the absolute value of negative zeta potential increased from 10 to 30 vol.% ethanol aqueous solution (−40 to −50 mV). in Section 3.5.
The three kinds of well stabilized nanobubbles (8%H2 in Ar, CO2, and CO2 after 8%H2 in Ar) at pH 9 in the 10 vol.% ethanol solution displayed decreases in the zeta potential to near zero when adjusting pH to 5, and during such pH adjustment, there was a possibility to produce the radicals originated from the nanobubble collapse. This point will be further discussed in the following Section 3.3 (nanobubble stability), Section 3.4 (nanobubble number), and Section 3.5 (radical production).

Nanobubble Stability Evaluation Using Extended DLVO Theory
Among many kinds of stabilization models for nanobubbles, Tan et al., 2021 suggested that the charge stabilization model can provide reasonable and consistent explanations [15].
In this study, bubble stabilization was evaluated by the extended DLVO theory using our experimental results of bubble size ( Figure 3) and zeta potential (Figures 4-6). Two bubble interactions can be expressed by extended DLVO theory [35,43,44]. Here, the extended DLVO theory is utilized as qualitative analysis. When two same radii (a) of nanobubbles are set at the surface-to-surface distance (h) between them in the ethanol aqueous solution, the total potential energy (VT) can be the sum of van der Waals interaction energy (VA), hydrophobic interaction energy (VH), and the electrostatic interaction energy (VR). VT is described in Equations (1) and (2), normalized by the absolute temperature (T) and Boltzmann constant (kB).

= + +
(1) where the Hamaker constant is A and hydrophobic constant is K. The pH of the solutions at around pH 9 were adjusted to pH 5 by adding HCl aqueous solution. The zeta potential of nanobubbles at pH 5 was slightly positive (i.e., about 5 mV), regardless of ethanol percentage. In particular, the zeta potential deviation was the largest (between −50 and 5 mV) at 10 vol% of ethanol aqueous solution during the above-mentioned conditioning procedures, similar to other results shown in Figures 4 and 5. The radical production by CO 2 and 8% H 2 in Ar nanobubbles in ethanol aqueous solution was investigated by changing the pH from 9 to 5 after 20 days, and the results will be discussed in Section 3.5.
The three kinds of well stabilized nanobubbles (8%H 2 in Ar, CO 2 , and CO 2 after 8%H 2 in Ar) at pH 9 in the 10 vol.% ethanol solution displayed decreases in the zeta potential to near zero when adjusting pH to 5, and during such pH adjustment, there was a possibility to produce the radicals originated from the nanobubble collapse. This point will be further discussed in the following Section 3.3 (nanobubble stability), Section 3.4 (nanobubble number), and Section 3.5 (radical production).

Nanobubble Stability Evaluation Using Extended DLVO Theory
Among many kinds of stabilization models for nanobubbles, Tan et al., 2021 suggested that the charge stabilization model can provide reasonable and consistent explanations [15].
In this study, bubble stabilization was evaluated by the extended DLVO theory using our experimental results of bubble size ( Figure 3) and zeta potential (Figures 4-6). Two bubble interactions can be expressed by extended DLVO theory [35,43,44]. Here, the extended DLVO theory is utilized as qualitative analysis. When two same radii (a) of nanobubbles are set at the surface-to-surface distance (h) between them in the ethanol aqueous solution, the total potential energy (V T ) can be the sum of van der Waals interaction energy (V A ), hydrophobic interaction energy (V H ), and the electrostatic interaction energy (V R ). V T is described in Equations (1) and (2), normalized by the absolute temperature (T) and Boltzmann constant (k B ).
where the Hamaker constant is A and hydrophobic constant is K. The Hamaker constant A for air in water (air-water-air value) is 3.7 × 10 −20 J [45]. As the Hamaker constant A is proportional to the surface tension of solvent, 2.2 × 10 −20 , 1.7 × 10 −20 , and 1.4 × 10 −20 J were used as A for air in 10, 30, and 50 vol.% ethanol aqueous solution, respectively. Hydrophobic constant K was estimated at 10 −17 J in the absence of salt and 10 −19 J in a 1 mM NaCl aqueous solution [44]. In a 1 M ethanol aqueous solution, K is about 3 to 7 × 10 −17 J, which is 3 to 7 times larger than K in water [46]. In this article, at 10 vol.% ethanol aqueous solution at pH 9 and 0.01 mM, 1 × 10 −17 J was used as K.
When the surface charge of nanobubbles is Ψ, κ is the Debye-Hückel parameter; r and o are relative permeability and space permeability, respectively; and V R is shown in Equations (3) and (4).
where n is the concentration of anions or cations in the solution and is equal to 1000 N A C (N A is the Avogadro's number and C is concentration in mol/L), z is the valence of ion, e is the electron charge, and the thickness of the electric double layer of the nanobubble is Debye length = 1/κ. The total potential energy based on electrostatic interaction energy, van der Waals interaction energy, and hydrophobic interaction energy as a function of the bubble distance at pH 9 and pH 5 of three kinds of gas nanobubbles (i.e., 8%H 2 in Ar, CO 2 , and CO 2 after 8%H 2 in Ar)) is shown in Figure 7. The total potential energy barriers at pH 9 of three types of nanobubbles exist with a high negative zeta potential. The various potential energies in the −38, −45, and −50 mV zeta potential of 8% H 2 in Ar, CO 2 , and CO 2 after 8% H 2 in Ar nanobubbles at pH 9 are shown (A, B, C) in Figure 7, respectively. The extended DLVO theory was utilized as qualitative explanation between the bubble's stability in this paper. The retardation in van der Waals potential can be estimated at least at a longer separation distance than about 15 nm according to Israelachvili, 1985 [47]. Between hydrophobic surfaces, very-long-range attraction can be observed in Figure 7, as also reported for separation [48].
The long-distance hydrophobic forces are evident. In contrast, the total potential is larger than 20 k B T at more than 100 nm distance and maintains the stability of bubbles owing to the strong repulsion explained by high electrostatic interaction energy. The maximum total potential energy V T 15 k B T would be the boundary to determine coagulation or dispersion [49]. The total potential energy barrier appeared at 8% H 2 in Ar, CO 2 , and CO 2 and 8% H 2 in Ar nanobubbles at 10 vol.% ethanol aqueous solution at pH 9 ( Figure 7). On the other hand, at pH 5, the absolute zeta potential value becomes less than 5 mV for three kinds of gas nanobubbles and total potential energy barrier disappeared owing to negligible electrostatic repulsion (V R , Figure 7). Therefore, at pH 5, nanobubbles would break owing to the low bubble surface charge, and thus make larger bubbles by bubble coalescence. Zhang et al., 2011 suggested that a higher than 20 vol.% ethanol solution may remove the nanobubbles and cause them to disappear, and is related to the long-range hydrophobic force with ethanol contents [18]. At 50 vol.% ethanol aqueous solution, the zeta potentials for three kinds of nanobubbles initially controlled at pH 9 were about −10 mV after 20 days, and bubble size was large-around 2000 nm-as shown in Figure 3, and not stable.

Number of Nanobubbles
The nanobubble numbers for three kinds of gas (8% H2 in Ar, CO2, and CO2 after 8% H2 in Ar) in 10 vol.% ethanol aqueous solution mixtures were investigated at pH 9 and pH 5 by nano site after 20 days, and the results are shown in Figure 8. The nanobubble number decreases from pH 9 to pH 5 for each nanobubble solution. In particular, the

Number of Nanobubbles
The nanobubble numbers for three kinds of gas (8% H 2 in Ar, CO 2 , and CO 2 after 8% H 2 in Ar) in 10 vol.% ethanol aqueous solution mixtures were investigated at pH 9 and pH 5 by nano site after 20 days, and the results are shown in Figure 8. The nanobubble number decreases from pH 9 to pH 5 for each nanobubble solution. In particular, the nanobubble number of 8%H 2 in Ar decreased the largest from 4 × 10 8 to 1 × 10 8 bubbles/mL. As the nanobubble number decreased from pH 9 to pH 5, the nanobubbles became larger by their coalescence and were broken. When the nanobubbles were broken, there was a possibility to produce radicals. Takahashi et al., 2021, reported that ·OH was generated from the collapsing microbubbles, including oxygen with 2 vol% ozone in 1 mM FeSO 4 aqueous solution under strongly acidic conditions without ultrasound or a large pressure difference [25]. The following Section 3.5 will report and discuss the experimental results on radical observation as a function of solution pH. Nanomaterials 2022, 12, x 11 of 15 nanobubble number of 8%H2 in Ar decreased the largest from 4 × 10 8 to 1 × 10 8 bubbles/mL. As the nanobubble number decreased from pH 9 to pH 5, the nanobubbles became larger by their coalescence and were broken. When the nanobubbles were broken, there was a possibility to produce radicals. Takahashi et al., 2021, reported that •OH was generated from the collapsing microbubbles, including oxygen with 2 vol% ozone in 1 mM FeSO4 aqueous solution under strongly acidic conditions without ultrasound or a large pressure difference [25]. The following Section 3.5 will report and discuss the experimental results on radical observation as a function of solution pH.  Figure 9 shows the two standard data points (i.e., superoxide anion on the top, hydroxyl radical on the bottom) and three data points from our gas bubble solutions (i.e., 8% H2 in Ar, CO2, and CO2 after 8% H2 in Ar). The •O2 − and •OH and standard eight peaks by spin trap reagent G-CYPMPO appear under different magnetic fields [38]. They are plotted in the top and bottom of Figure 9, respectively. The peaks appearing of 8% H2 with Ar, CO2, and the mixture of CO2 and 8% H2 with Ar nanobubbles in 10% ethanol aqueous solutions are plotted as the second, third, and fourth curves from the top in Figure 9, respectively. The peaks corresponding •O2 − and •OH − are marked by blue and red circles, respectively. In this experiment, small peaks were observed in each position owing to smaller nanobubble numbers after 20 days passed.

Radical Observation by Changing the pH
Radical observation by changing the solution pH from 9 to 5 of three kinds of gas nanobubbles in 10 vol.% ethanol aqueous solution mixture is shown in Figure 9. The pH of the 8% H2 in Ar nanobubbles by changing to 5 (second data point from the top) showed a small amount of superoxide anion peaks. The reaction is considered as Equation (5): The CO2 nanobubbles, by changing to pH 5 (third data point from the top), showed a small amount of hydroxyl radical peaks. The reaction is shown as follows:

Radical Observation by
Changing the pH Figure 9 shows the two standard data points (i.e., superoxide anion on the top, hydroxyl radical on the bottom) and three data points from our gas bubble solutions (i.e., 8% H 2 in Ar, CO 2 , and CO 2 after 8% H 2 in Ar). The ·O 2 − and ·OH and standard eight peaks by spin trap reagent G-CYPMPO appear under different magnetic fields [38]. They are plotted in the top and bottom of Figure 9, respectively. The peaks appearing of 8% H 2 with Ar, CO 2 , and the mixture of CO 2 and 8% H 2 with Ar nanobubbles in 10% ethanol aqueous solutions are plotted as the second, third, and fourth curves from the top in Figure 9, respectively. The peaks corresponding ·O 2 − and ·OH − are marked by blue and red circles, respectively. In this experiment, small peaks were observed in each position owing to smaller nanobubble numbers after 20 days passed.
Radical observation by changing the solution pH from 9 to 5 of three kinds of gas nanobubbles in 10 vol.% ethanol aqueous solution mixture is shown in Figure 9. The pH of the 8% H 2 in Ar nanobubbles by changing to 5 (second data point from the top) showed a small amount of superoxide anion peaks. The reaction is considered as Equation (5): The CO 2 nanobubbles, by changing to pH 5 (third data point from the top), showed a small amount of hydroxyl radical peaks. The reaction is shown as follows: On the other hand, the mixture of CO 2 and 8% H 2 in Ar (fourth data point from the top) showed neither ·O 2 nor ·OH peaks.
When the acid was added to the nanobubble solution, 8% H 2 in Ar nanobubble solution produces ·O 2 − , while the CO 2 nanobubble solution produced ·OH. On the other hand, the gas mixture (8% H 2 in Ar and CO 2 ) nanobubbles were prepared by 8% H 2 in Ar gas blowing followed by CO 2 gas blowing. At first, the reaction in Equation (7) occurred, and then the reaction in Equation (8) When the acid was added to the nanobubble solution, 8% H2 in Ar nanobubble solution produces •O2 − , while the CO2 nanobubble solution produced •OH. On the other hand, the gas mixture (8% H2 in Ar and CO2) nanobubbles were prepared by 8% H2 in Ar gas blowing followed by CO2 gas blowing. At first, the reaction in Equation (7) occurred, and then the reaction in Equation (8) occurred and reduced the radicals. This reaction agreed with the literature. The existing •OH and •O2 − scavenging was reported by blowing CO2 nanobubbles after blowing H2 nanobubbles [30]. The formation of a monolayer by ethanol [12] and the arrangement of ethanol molecules on interfaces stabilize the bulk nanobubbles [13,14]. The water molecules were arranged to maximize the hydrogen bonding between the oriented ethanol and the adjacent water molecules [16]. Our model of nanobubble increases with coagulation by changing pH from 9 to 5; therefore, the bubble becomes easily breakable, and the production of radicals is shown in Figure 10. Here, the weak ultrasonic wave and ultraviolet light are irradiated. There is a report that the •OH peak was observed with a weak supersonic wave [28]. Moreover, •OH was generated from the collapsing microbubbles under strongly acidic conditions without any dynamic stimulus such as ultrasound [25]. The CO2 nanobubbles after 8% H2 in Ar nanobubbles can exist in 10 vol.% ethanol aqueous solution mixture at pH 9 for a long period such as 20 days (Figure 3). The alkaline 10 vol.% ethanol aqueous solution with CO2 nanobubbles after 8% H2 in Ar nanobubble did not show noticeable free radicals by changing the pH to acidic (i.e., pH 5). The phenomena studied The formation of a monolayer by ethanol [12] and the arrangement of ethanol molecules on interfaces stabilize the bulk nanobubbles [13,14]. The water molecules were arranged to maximize the hydrogen bonding between the oriented ethanol and the adjacent water molecules [16]. Our model of nanobubble increases with coagulation by changing pH from 9 to 5; therefore, the bubble becomes easily breakable, and the production of radicals is shown in Figure 10. Here, the weak ultrasonic wave and ultraviolet light are irradiated. There is a report that the ·OH peak was observed with a weak supersonic wave [28]. Moreover, ·OH was generated from the collapsing microbubbles under strongly acidic conditions without any dynamic stimulus such as ultrasound [25]. The CO 2 nanobubbles after 8% H 2 in Ar nanobubbles can exist in 10 vol.% ethanol aqueous solution mixture at pH 9 for a long period such as 20 days (Figure 3). The alkaline 10 vol.% ethanol aqueous solution with CO 2 nanobubbles after 8% H 2 in Ar nanobubble did not show noticeable free radicals by changing the pH to acidic (i.e., pH 5). The phenomena studied and discussed in this article would be useful to prepare a soda alcohol beverage, among others. Nanomaterials 2022, 12, x 13 and discussed in this article would be useful to prepare a soda alcohol beverage, am others. Figure 10. Model of nanobubble breakage by changing pH from 9 to 5 and the production of rad by decreasing the bubble zeta potential absolute value.

Conclusions
The 8% hydrogen (H2) in argon (Ar) and carbon dioxide (CO2) gas nanobubbles ethanol were produced at 10, 30, and 50 vol.% ethanol aqueous solution by the high s agitation method with gas injection, and the main findings were as follows: • The prepared nanobubbles were stable for 20 days owing to a high negative zeta tential at alkaline pH 9.

•
When the pH of ethanol alkaline aqueous solution with nanobubbles was adjust acidic at around pH 5, the zeta potential of nanobubbles was almost zero. The n bers of nanobubble decreased at almost zero charge (pH 5) were identified by m uring their numbers using the particle trajectory method (Nano site).

•
The extended Derjaguin, Landau, Verwey, and Overbeek (DLVO) theory was us evaluate the nanobubble stability (repulsion between bubbles) in alkaline condit and its instability (attraction between bubbles) in acidic conditions.
The collapsed nanobubbles at zero charge generated slight free radicals detected ing G-CYPMPO spin trap reagent in electron spin resonance (ESR) spectroscopy. The duced free radicals were superoxide anions at collapsed 8% H2 in Ar nanobubbles hydroxyl radicals at collapsed CO2 nanobubbles. On the other hand, the collapsed m CO2 and H2 in Ar nanobubbles showed no free radicals.
Based on this study, a schematic model of nanobubble breakage and the produ of radicals by changing solution pH was proposed. These phenomena and their un standing would be useful to formulate healthy beverages, for example.

Conclusions
The 8% hydrogen (H 2 ) in argon (Ar) and carbon dioxide (CO 2 ) gas nanobubbles with ethanol were produced at 10, 30, and 50 vol.% ethanol aqueous solution by the high speed agitation method with gas injection, and the main findings were as follows:

•
The prepared nanobubbles were stable for 20 days owing to a high negative zeta potential at alkaline pH 9.

•
When the pH of ethanol alkaline aqueous solution with nanobubbles was adjusted to acidic at around pH 5, the zeta potential of nanobubbles was almost zero. The numbers of nanobubble decreased at almost zero charge (pH 5) were identified by measuring their numbers using the particle trajectory method (Nano site).

•
The extended Derjaguin, Landau, Verwey, and Overbeek (DLVO) theory was used to evaluate the nanobubble stability (repulsion between bubbles) in alkaline conditions, and its instability (attraction between bubbles) in acidic conditions.