Preparation and Properties of CO 2 Micro-Nanobubble Water Based on Response Surface Methodology

Featured Application: This article discusses the preparation of CO 2 micro-nanobubble water using the dissolved gas release method. Taking the CO 2 content as the optimization target, micro-nanobubble water was prepared with the optimal content based on a mathematical model constructed using the response surface method. Meanwhile, the characteristics of the micro-nanobubbles in the water provided a theoretical reference for applying CO 2 in agriculture, which will assist in achieving a quantitative increase in the CO 2 enrichment of crops. Abstract: Carbon dioxide (CO 2 ) enrichment in an agricultural environment has been shown to enhance the efﬁciency of crop photosynthesis, increasing crop yield and product quality. There is a problem of the excessive use of CO 2 gas when the CO 2 is enriched for crops, such as soybean and other ﬁeld crops. Given the application of micro-nanobubbles (MNBs) in agricultural production, this research takes CO 2 as the gas source to prepare the micro-nanobubble water by the dissolved gas release method, and the response surface methodology is used to optimize the preparation process. The results show that the optimum parameters, which are the gas–liquid ratio, generator running time, and inlet water temperature for the preparation of CO 2 micro-nanobubble water, are 2.87%, 28.47 min, and 25.52 ◦ C, respectively. The CO 2 content in the MNB water prepared under the optimum parameters is 7.64 mg/L, and the pH is 4.08. Furthermore, the particle size of the bubbles is mostly 255.5 nm. With the extension of the storage time, some of the bubbles polymerize and spill out, but there is still a certain amount of nanoscale bubbles during a certain period. This research provides a new idea for using MNB technology to increase the content and lifespan of CO 2 in water, which will slow the release and increase the utilization of CO 2 when using CO 2 enrichment in ranges of 70–80 nm and 700–900 nm. The size of the bubbles tended to decrease with an increase in the generator circulation time, which was consistent with the conclusion of Wang et al. [21]. According to the understanding of the generation mechanism for MNBs, it can be concluded that the size of the generated MNBs was determined by a range of factors, including the physicochemical qualities of the gas, the pressure at which the MNBs were formed, and the duration of the reaction.


Introduction
CO 2 is the primary substrate for photosynthesis, and it plays an essential role in crop growth, development, and productivity. Using an anthropogenic method to increase the CO 2 concentration in an enclosed greenhouse called CO 2 fertilization [1,2] increases the growth and yield of the crop plants [3]. As an essential gas fertilizer, CO 2 has been widely used in greenhouses for the last century [4,5]. Because of the characteristics of CO 2 diffusion, CO 2 fertilizer has mainly been used in protected cultivation facilities. However, The MNBs were produced using a self-suction MNB generator developed by Hangzhou Xiyue New Material Technology Co., Ltd. (Hangzhou, China). This MNB generator is based on the principle of adding pressure to dissolve a gas and then releasing the pressure to release the gas (the dissolved gas release method). More specific details of this principle are shown in Figure 1.
This device comprises a high-pressure pump, dissolved gas cylinder, throttle-type releaser, gas flow meter, pressure gauge, etc. The MNB generator uses a high-pressure self-priming air pump and gas flow meter to control the gas flow. When the gas reaches the high-pressure pump impeller, the impeller first mechanically breaks up the gas to form tiny bubbles. The large gas bubbles dissolve in the water in the dissolved gas cylinder because of the high back pressure developed by the throttle-type releaser hole. Finally, when passing through the throttle-type releaser, the gas dissolved in the water is released in large amounts as a result of the increased flow velocity and sudden decrease in pressure, forming the MNBs.
In this study, the source of the gas was a CO 2 cylinder connected to the generator. The other experimental materials and reagents and the main instruments used for the data measurement are listed in Tables 1 and 2, respectively.
Appl. Sci. 2021, 11, x FOR PEER REVIEW 3 of 1 Figure 1. Schematic illustration of the MNB generator, where 1 represents the gas supply (CO2 cyl inders); 2 represents the gas flow meter; 3 represents the diaphragm pump; 4 represents the dis solved gas tank; 5 represents the hydraulic pressure gauge; 6 represents the throttling nozzle for th released gas.  In this experiment, the CO2 MNB water was prepared using the dissolved gas releas method. The effects of the gas-liquid ratio (v/v, %), generator running time (min), and Schematic illustration of the MNB generator, where 1 represents the gas supply (CO 2 cylinders); 2 represents the gas flow meter; 3 represents the diaphragm pump; 4 represents the dissolved gas tank; 5 represents the hydraulic pressure gauge; 6 represents the throttling nozzle for the released gas.  In this experiment, the CO 2 MNB water was prepared using the dissolved gas release method. The effects of the gas-liquid ratio (v/v, %), generator running time (min), and inlet water temperature ( • C) on the content of the CO 2 gas in the MNB water were found to determine the optimal MNB water preparation technique.
(1) Impact of Gas-liquid Ratio on CO 2 Content When preparing the CO 2 MNB water, the MNB generator running time (30 min) and inlet water temperature (25 • C) were made constant. At the same time, the gas-liquid ratio (v/v, %) was adjusted to values of 1.3%, 2%, 2.7%, 3.4%, and 4.1%. The best CO 2 gas-liquid ratio was determined based on the CO 2 content measured by the carbon dioxide meter.
(2) Impact of Equipment Running Time on CO 2 Content By fixing the gas-liquid ratio (2.7%) of the MNB generator and inlet water temperature (25 • C), the CO 2 contents of the MNB water were measured for generator running times of 10, 20, 30, 40, and 50 min, and used to define the optimal range of generator running times.
(3) Impact of Inlet Temperature of Water on CO 2 Content By fixing the generator running time and gas-liquid ratio at 30 min and 2.7%, respectively, the CO 2 contents of the MNB water with different inlet temperatures (25,30,35,40, and 45 • C) were measured to define the optimal range of inlet water temperatures.

Optimal CO 2 MNB Water Preparation Based on Response Surface Methodology
Based on the single-factor experimental results, the gas-liquid ratio (%, A), generator running time (min, B), and inlet water temperature ( • C, C) were selected as independent variables, and the CO 2 content of the MNB water was the dependent variable. Design-Expert software was used for a three-level Box-Behnken experimental design (BBD) [22] with three factors to optimize the response values (mg/L, R). The design factor levels are listed in Table 3.

) Standard Solution Preparation
The standard solution 1 (0.2 M NaCl + 0.01 M NaHCO 3 ) was prepared by first weighing 1.68 g of NaHCO 3 and 11.69 g of NaCl, and they were then added to purified water not containing CO 2 to obtain a constant volume of 1000 mL.
Standard solution 2 (0.2 M NaCl + 0.002 M NaHCO 3 ) was prepared by weighing 10.52 g of NaCl, mixing it into 100 mL of standard solution 1, and using purified water not containing CO 2 to dilute it to obtain 1000 mL.
(2) Standard Sample Calibration First, 50 mL of standard solution 1 was transferred to a flask, and five drops of phosphoric acid were added. While mixing at 300 rpm, the carbon dioxide measuring device was used to calibrate standard solution 1. Standard solution 2 was calibrated using a similar procedure.
(3) Sample Measurement The sample being tested was placed on the magnetic stirrer. At a moderate mixing speed, the calibrated electrode was used for measurements. The measured value was recorded as the concentration of CO 2 in the sample being tested. In this experiment, the CO 2 MNB water was prepared under the optimal preparation parameters based on the response surface methodology. It used dynamic light scattering (DLS, Zetasizer Nano ZS, Malvern), performed at a scattering angle of 173 • , and a temperature of 25 • C to analyze the size and distribution of the MNBs in the test sample. The refractive index of the material was fixed at 1 to correspond to air, and the refractive index of the water was 1.33 [23,24]. All the samples were analyzed in triplicate, and the results were averaged.
(2) Stability of CO 2 MNBs The stability of the CO 2 MNBs was measured based on the changes in the CO 2 content of the MNB water. In this experiment, 2 L of CO 2 MNB water was prepared under the optimal preparation parameters based on the response surface methodology. Then, the lid was opened, it was left for 48 h at room temperature, and the CO 2 content was measured every 4 h. The average results of the multiple measurements were used to reduce the error in the experiment and finally obtain the variation curve for the CO 2 in the MNB water.
(3) pH of CO 2 MNB Water When preparing CO 2 MNB water under the optimal preparation parameters based on the response surface methodology, the pH value of CO 2 MNB water was measured every 3 min during the generator running time. Measurements of the samples were performed five times, and the average value was obtained. The variation curve for the pH of the MNB water under the optimal preparation conditions could present the property of CO 2 MNB water.

Statistical Analysis
The results of the single-factor experiments for researching the properties of CO 2 MNBs were used in Origin 2021b for the plot analysis of the experimental data. We used Design-Expert 8.0.6 to design the orthogonal experiment, along with ANOVA analyses and drew an interactive three-dimensional surface plot and contour diagram of the responses and input variables [25].

Single-Factor Experimental Result Analysis
(1) Analysis of Gas-Liquid Ratio Effect on CO 2 Concentration in MNB Water Figure 2 shows the change curve for the CO 2 content in the MNB water under different gas-liquid ratios. As shown in Figure 2, the CO 2 content in the MNB water gradually increased with the gas-liquid ratio. The CO 2 content in the MNB water tended to be stable when the gas-liquid ratio reached 2.7%, which showed that the CO 2 in the water has reached the saturation level.
(2) Impact of Generator Running Time on CO 2 content in MNB Water As the running time of the generator increased, the quantity of CO 2 MNBs in the reaction vessel gradually increased, which became a suspension of the CO 2 MNBs. Therefore, the generator cycle running time is an essential factor influencing the CO 2 content in the MNB water. Figure 3 shows the resulting change in the CO 2 content in the MNB water under a fixed gas-liquid ratio and inlet water temperature. The results show that the CO 2 content gradually increased with the generator cycle running time. In particular, the trend of the changes in the CO 2 content became smaller when the time reached 30 min. When the running time was 40 min, the CO 2 content was at its maximum value and subsequently decreased as a result of some MNBs bursting and escaping from the water. Therefore, the optimal cycle running time for CO 2 MNB water is 30-40 min. The experimental results are shown in Figure 3. (2) Impact of Generator Running Time on CO2 content in MNB Water As the running time of the generator increased, the quantity of CO2 reaction vessel gradually increased, which became a suspension of the CO2 fore, the generator cycle running time is an essential factor influencing the C the MNB water. Figure 3 shows the resulting change in the CO2 content in th under a fixed gas-liquid ratio and inlet water temperature. The results show content gradually increased with the generator cycle running time. In partic of the changes in the CO2 content became smaller when the time reached the running time was 40 min, the CO2 content was at its maximum value and decreased as a result of some MNBs bursting and escaping from the water. optimal cycle running time for CO2 MNB water is 30-40 min. The experimen shown in Figure 3. (2) Impact of Generator Running Time on CO2 content in MNB Water As the running time of the generator increased, the quantity of CO2 reaction vessel gradually increased, which became a suspension of the CO2 fore, the generator cycle running time is an essential factor influencing the C the MNB water. Figure 3 shows the resulting change in the CO2 content in th under a fixed gas-liquid ratio and inlet water temperature. The results show content gradually increased with the generator cycle running time. In partic of the changes in the CO2 content became smaller when the time reached the running time was 40 min, the CO2 content was at its maximum value and decreased as a result of some MNBs bursting and escaping from the water. optimal cycle running time for CO2 MNB water is 30-40 min. The experimen shown in Figure 3. (3) Impact of Inlet Water Temperature on CO2 MNB content The solubility of gas has an impact on the gas-liquid equilibrium bala reaction vessel. Furthermore, temperature is an important factor influencing [26,27]. As a result, an investigation of the effect of the inlet water tempe quired. Figure 4 depicts the effect of the inlet water temperature on the CO2 (3) Impact of Inlet Water Temperature on CO 2 MNB content The solubility of gas has an impact on the gas-liquid equilibrium balance inside the reaction vessel. Furthermore, temperature is an important factor influencing gas solubility [26,27]. As a result, an investigation of the effect of the inlet water temperature was required. Figure 4 depicts the effect of the inlet water temperature on the CO 2 content of the MNB water. The amount of CO 2 dissolved in the water gradually decreased as the incoming water temperature increased, as seen in the results. MNB water. The amount of CO2 dissolved in the water gradually decreased ing water temperature increased, as seen in the results.

Response Surface Methodology Results and Analysis
(1) Orthogonal Experimental Design and Results The orthogonal experimental design was conducted based on the singl iment results, taking the gas-liquid ratio (A), generator running time (B), a temperature (C) as independent variables, and the CO2 content (R) in the M the response value. The results are listed in Table S1. (

2) Response Surface Model Fitting and Analysis
This research compared the fitting results of the first-order model and model. Table S2 lists the ANOVA results of these two types of models. W that the second-order model fits the experimental data better than the first In addition, a regression model with the CO2 content (R) as the response va lished using a second-order model, as shown in Formula 1.
The ANOVA results of the regression model are listed in Table S3. From this regression model's F-value was 284.30, and the p-value that was less t cates a high significance. Therefore, this model was statistically significant. the p-value associated with the partial regression coefficient of A, B, and C w 0.05, indicating that the gas-liquid ratio, generator cycle time, and tempe cantly impacted the CO2 content in the MNB water. Within the selected facto the degrees of influence of the three had the following order: A > B > C. T between two of the three influencing factors significantly impacted the CO2 MNBs in the significance tests of AB, AC, and BC (p < 0.05). The quadrati

(1) Orthogonal Experimental Design and Results
The orthogonal experimental design was conducted based on the single-factor experiment results, taking the gas-liquid ratio (A), generator running time (B), and inlet water temperature (C) as independent variables, and the CO 2 content (R) in the MNB water as the response value. The results are listed in Table S1.

(2) Response Surface Model Fitting and Analysis
This research compared the fitting results of the first-order model and second-order model. Table S2 lists the ANOVA results of these two types of models. We can observe that the second-order model fits the experimental data better than the first-order model. In addition, a regression model with the CO 2 content (R) as the response value was established using a second-order model, as shown in Formula 1.
The ANOVA results of the regression model are listed in Table S3. From these results, this regression model's F-value was 284.30, and the p-value that was less than 0.05 indicates a high significance. Therefore, this model was statistically significant. In this model, the p-value associated with the partial regression coefficient of A, B, and C was lower than 0.05, indicating that the gas-liquid ratio, generator cycle time, and temperature significantly impacted the CO 2 content in the MNB water. Within the selected factor level ranges, the degrees of influence of the three had the following order: A > B > C. The interaction between two of the three influencing factors significantly impacted the CO 2 content in the MNBs in the significance tests of AB, AC, and BC (p < 0.05). The quadratic terms A 2 , B 2 , and C 2 had extremely significant effects on the response value (p < 0.05). Further error statistical analyses were performed on the regression model, and the results are listed in Table 4. The R 2 value was 0.9973, which is close to 1, indicating that the stage two model had an excellent fitting effect. The signal-to-noise ratio of 65.718 was much greater than 4. For the same reason, the reliability and accuracy of the model were high. A comparison between the model's predicted values and the actual values is shown in Figure 5.
The R 2 value was 0.9973, which is close to 1, indicating that t an excellent fitting effect. The signal-to-noise ratio of 65.718 was m the same reason, the reliability and accuracy of the model were tween the model's predicted values and the actual values is show The results showed that the values predicted by the mode actual measured values and contained only minor errors, which the model had a high degree of fit and was the ideal model. The results showed that the values predicted by the model were very close to the actual measured values and contained only minor errors, which further confirmed that the model had a high degree of fit and was the ideal model.     The interaction between the gas-liquid ratio and the generator cycle running time had a more apparent impact on the CO2 content in the MNB water than the other two factors. The gas-liquid ratio had the most significant influence on the CO2 content. The interaction between the gas-liquid ratio and the water temperature had a weaker impact on the CO2 content. Figure 7 shows that the effect of the water temperature was smaller than that of the gas-liquid ratio, which was consistent with the previous single-factor experimental results. Figure 8 illustrates the effect of the interaction between the water temperature and the generator cycle running time. Compared with the interaction between the water temperature and the gas-liquid ratio, the water temperature and cycle time had a more significant effect on the CO2 content.

(4) Model Optimization and Validation Results
In this study, the Design-Expert software was used to select the maximum response value, R, as the optimal condition from the range of experimental factor values. The optimal preparation parameters obtained from the optimization were a gas-liquid ratio of 2.87%, a generator cycle running time of 28.47 min, and an inlet water temperature of 25.52 °C, which resulted in a CO2 content of 7.64 mg/L in the MNB water.
According to the optimal experimental conditions obtained above, the existing experimental platform was controlled to meet the aforementioned optimal parameter conditions to conduct a verification experiment. Finally, the average value of CO2 content in the MNB water under the optimized conditions was 7.76 mg/L, and the relative error was 1.3%, which was within a reasonable error range. This showed that the optimization method for the CO2 MNB water preparation process based on the response surface method was feasible. The model constructed in this study has significance because it provides guidance for optimizing the CO2 content in MNB water.

Particle Size and Distribution Properties of CO2 MNBs from Dissolved Gas Release Method
The measurements of the particle size and concentration distribution of the MNBs produced under different cycle running times in this experiment are shown in Figure 9. The preparation of the MNBs using the dissolved gas release method can be divided into two processes: the pressurization to dissolve the gas and the decompression to release it  The interaction between the gas-liquid ratio and the generator cycle running time had a more apparent impact on the CO 2 content in the MNB water than the other two factors. The gas-liquid ratio had the most significant influence on the CO 2 content. The interaction between the gas-liquid ratio and the water temperature had a weaker impact on the CO 2 content. Figure 7 shows that the effect of the water temperature was smaller than that of the gas-liquid ratio, which was consistent with the previous single-factor experimental results. Figure 8 illustrates the effect of the interaction between the water temperature and the generator cycle running time. Compared with the interaction between the water temperature and the gas-liquid ratio, the water temperature and cycle time had a more significant effect on the CO 2 content.

(4) Model Optimization and Validation Results
In this study, the Design-Expert software was used to select the maximum response value, R, as the optimal condition from the range of experimental factor values. The optimal preparation parameters obtained from the optimization were a gas-liquid ratio of 2.87%, a generator cycle running time of 28.47 min, and an inlet water temperature of 25.52 • C, which resulted in a CO 2 content of 7.64 mg/L in the MNB water.
According to the optimal experimental conditions obtained above, the existing experimental platform was controlled to meet the aforementioned optimal parameter conditions to conduct a verification experiment. Finally, the average value of CO 2 content in the MNB water under the optimized conditions was 7.76 mg/L, and the relative error was 1.3%, which was within a reasonable error range. This showed that the optimization method for the CO 2 MNB water preparation process based on the response surface method was feasible. The model constructed in this study has significance because it provides guidance for optimizing the CO 2 content in MNB water.

Particle Size and Distribution Properties of CO 2 MNBs from Dissolved Gas Release Method
The measurements of the particle size and concentration distribution of the MNBs produced under different cycle running times in this experiment are shown in Figure 9. The preparation of the MNBs using the dissolved gas release method can be divided into two processes: the pressurization to dissolve the gas and the decompression to release it [28]. According to the results, the sizes of the MNBs showed two peaks, with peak particle size ranges of 70-80 nm and 700-900 nm. The size of the bubbles tended to decrease with an increase in the generator circulation time, which was consistent with the conclusion of Wang et al. [21]. According to the understanding of the generation mechanism for MNBs, it can be concluded that the size of the generated MNBs was determined by a range of factors, including the physicochemical qualities of the gas, the pressure at which the MNBs were formed, and the duration of the reaction.
an increase in the generator circulation time, which was consistent with the co Wang et al. [21]. According to the understanding of the generation mechanism it can be concluded that the size of the generated MNBs was determined by factors, including the physicochemical qualities of the gas, the pressure at MNBs were formed, and the duration of the reaction.
The particle size distribution and concentration information of the CO2 M the optimized conditions are listed in Table S4 and shown in Figure S1. The av ticle size of the bubbles was 134.9 nm, and the bubble diameters correspondin lative distributions of 10%, 50%, and 90% were 33.8 nm, 105.6 nm, and 255.5 n tively. The concentration was 6.8 × 10 6 pieces/mL.

Properties of CO2 Content and Lifetime in MNB Water
The CO2 content in the MNB water changed with the generator cycle ru and the storage time at room temperature, as shown in Figure 10. This shows th content in the water increased with the cycle time and then increased in a st fashion. The CO2 content in the water reached the maximum value of approx mg/L in the linear increase stage when the generator running time was 30-40 that, the CO2 content in the water was basically in a stable state. This result was with the variation of the dissolved oxygen content in the MNB water in [29], w fied that MNB technology can increase the solubility of gas in water.  Table S4 and shown in Figure S1. The average particle size of the bubbles was 134.9 nm, and the bubble diameters corresponding to cumulative distributions of 10%, 50%, and 90% were 33.8 nm, 105.6 nm, and 255.5 nm, respectively. The concentration was 6.8 × 10 6 pieces/mL.

Properties of CO 2 Content and Lifetime in MNB Water
The CO 2 content in the MNB water changed with the generator cycle running time and the storage time at room temperature, as shown in Figure 10. This shows that the CO 2 content in the water increased with the cycle time and then increased in a stable linear fashion. The CO 2 content in the water reached the maximum value of approximately 6.5 mg/L in the linear increase stage when the generator running time was 30-40 min. After that, the CO 2 content in the water was basically in a stable state. This result was consistent with the variation of the dissolved oxygen content in the MNB water in [29], which verified that MNB technology can increase the solubility of gas in water.
After the 60 min experiment, the MNB water sample was stored at room temperature to investigate the storage property of CO 2 MNBs. The change in the CO 2 content in the MNB water with the storage time is shown in Figure 11. It shows that the CO 2 content had no obvious decrease within 48 h, which indicates that the lifespan of the MNBs prepared by the generator was longer. The CO 2 content fluctuated in the range of 6.3-6.6 mg/L. This proved that the MNBs increased the CO 2 content in the water. After the 60 min experiment, the MNB water sample was stored a to investigate the storage property of CO2 MNBs. The change in the MNB water with the storage time is shown in Figure 11. It shows that t no obvious decrease within 48 h, which indicates that the lifespan of t by the generator was longer. The CO2 content fluctuated in the range o proved that the MNBs increased the CO2 content in the water.   After the 60 min experiment, the MNB water sample was stored at room to investigate the storage property of CO2 MNBs. The change in the CO2 MNB water with the storage time is shown in Figure 11. It shows that the CO no obvious decrease within 48 h, which indicates that the lifespan of the M by the generator was longer. The CO2 content fluctuated in the range of 6.3proved that the MNBs increased the CO2 content in the water.   Figure 12 shows the pH change of the CO2 MNB water prepared und preparation conditions. CO2 is a slightly soluble gas in water [30]. It read water to produce carbonic acid. Thus, the pH of the MNB water decreased erator ran longer because a portion of the CO2 was dissolved in the water when the temperature and pressure remained constant at normal levels, th the CO2 in the water did not change. As a result, the pH of the MNB w essentially stable as the cycle duration increased. Figure 12 shows the ex  Figure 12 shows the pH change of the CO 2 MNB water prepared under the optimal preparation conditions. CO 2 is a slightly soluble gas in water [30]. It readily reacts with water to produce carbonic acid. Thus, the pH of the MNB water decreased when the generator ran longer because a portion of the CO 2 was dissolved in the water. Furthermore, when the temperature and pressure remained constant at normal levels, the solubility of the CO 2 in the water did not change. As a result, the pH of the MNB water remained essentially stable as the cycle duration increased. Figure 12 shows the experimental results.  The MNB generator increased the quantity of CO2 dissolved in the w normal conditions, and the pH value was considerably lowered during the the cycle. When the water achieved saturation, the CO2 in the MNB water w CO2 hydrate, H2CO3, and bubbles in the water [31]. At the end of the exper of the deionized water used had dropped from 5.62 to 3.99, or approximate

Conclusions
The dissolved gas release method was utilized in this work to create M CO2 gas. The procedure for preparing and characterizing the CO2 MNB tailed, and the optimum parameters were also obtained using the response odology. According to these results, the method proposed in this paper quantifying the content of CO2 in micro-nanobubble water. Therefore, thi vides new theoretical support for the application of micro-nanobubble tec riculture. However, the application scope of the optimization model propo per is limited to some extent due to the limitations of experimental conditio ment. In the following research, different MNB generation principles will At the same time, the method will be applied to actual agricultural plantin best way to verify the importance of the research results.
Supplementary Materials: The following are available online at www.mdpi.com/x Orthogonal experiment design and results; Table S2: Results of statistical models; T sion model statistical analysis results; Table S4, The result of particle distribution of der optimal conditions; Figure S1, The particle size distribution of the CO2 MNBs mized conditions. The MNB generator increased the quantity of CO 2 dissolved in the water relative to normal conditions, and the pH value was considerably lowered during the early stage of the cycle. When the water achieved saturation, the CO 2 in the MNB water was mainly free CO 2 hydrate, H 2 CO 3 , and bubbles in the water [31]. At the end of the experiment, the pH of the deionized water used had dropped from 5.62 to 3.99, or approximately by 29%.

Conclusions
The dissolved gas release method was utilized in this work to create MNB water with CO 2 gas. The procedure for preparing and characterizing the CO 2 MNB water was detailed, and the optimum parameters were also obtained using the response surface methodology. According to these results, the method proposed in this paper is feasible for quantifying the content of CO 2 in micro-nanobubble water. Therefore, this method provides new theoretical support for the application of micro-nanobubble technology in agriculture. However, the application scope of the optimization model proposed in this paper is limited to some extent due to the limitations of experimental conditions and equipment. In the following research, different MNB generation principles will be considered. At the same time, the method will be applied to actual agricultural planting, which is the best way to verify the importance of the research results.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/ 10.3390/app112411638/s1. Table S1: Orthogonal experiment design and results; Table S2: Results of statistical models; Table S3. Regression model statistical analysis results; Table S4, The result of particle distribution of CO 2 -MNBs under optimal conditions; Figure S1, The particle size distribution of the CO2 MNBs under the optimized conditions.