3.1. Effect of the gas–liquid ratio in the container
To confirm the effect of the LVF on NB generation in the container, the NB generation test was performed for 5 min using an NB generation device. Figure 1 shows an image of the generated NBs in DI water using the NTA apparatus; the black background represents liquid and the shiny dots are NBs. The generated NBs increased as the amount of liquid in the container decreased. The image analysis results obtained through the NTA software and the zeta potential values for the generated NBs are shown in Fig. 2. Furthermore, as seen in Fig. 2 (a), the concentration of NBs increased dramatically in inverse proportion to the amount of liquid in the container. The NB generation device used in this experiment was of the gas–liquid mixing type. The mechanism of NB generation is as follows. When the liquid in the container strongly collided with the bottom or top of the container, which had a vertical reciprocating motion, the bubbles were trapped in the liquid (inflow by collision between the droplets and stuck in the container’s edge) to form foam. At the same time, the trapped bubbles were refined by the shear force (generated by the turbulent flow of the liquid) caused by the impact of the liquid. This process occurred repeatedly in the container that continuously reciprocated vertically, resulting in the trapped bubbles being reduced to nano-size. Therefore, we assumed that the two most important factors affecting NB generation in this device were (1) the impulse of the liquid colliding with the container surface and (2) shear force (generated by turbulent flow). Based on these assumptions, a theoretical analysis was performed on the NB generation results (Fig. 2 (a)). Under each experimental condition (i.e, LVF), the impulse and shear force values upon the liquid colliding with the bottom or top of the container were derived, and the relationship between these values and the generated NB concentration was investigated. The value of impulse was greatest when the LVF was 50% and was calculated in the order of 75%, 25%, and 100% (LVF: 25%—4.05×10− 3 kgžm/s; 50%—5.80 × 10− 3 kgžm/s; 75%—5.40 × 10− 3 kgžm/s; and 100%—0 kgžm/s). The Reynolds number of the liquid in the container was the largest at 25%, and it was confirmed that the conditions were sufficient to generate turbulence in the liquid in all cases except 100%. Therefore, when the LVF was 25%, strong turbulence occurred in the liquid in the container (magnitude of shear force—25% > 50% > 75% > 100%), and a high shear force was applied to the trapped bubbles. Although the above theoretical approaches were insufficient for analyzing quantitative values, it was assumed that comparing the rough magnitude (impulse, shear force) between the experimental conditions was not a problem. Based on this result, the magnitude of shear force applied to the liquid was proportional to the concentration of the generated NBs. In other words, the shear force applied to the trapped bubbles in the liquid played an important role in NB generation using the gas–liquid mixing method. However, since the NBs were generated not only at 25% but also at 50% and 75%, it was assumed that sufficient shear force was transmitted to generate NBs in all cases except 100%, meaning the amount of trapped bubbles (which are a prerequisite for NB generation) must be large to increase the concentration of NB generation. Regardless of the strength of shear force transmitted to the liquid in the container, if there were no bubbles trapped in the liquid, NBs could not be generated. When the container attached to the NB generation device descended at high speed, from the top position to the bottom position, the liquid in the container elongated (e.g., the head part of the sea wave) and collided with the bottom surface of the container. At this point, the gas was trapped inside the liquid to generate foam, and the amount of foam produced was proportional to the specific surface area of the liquid exposed to the gas when the liquid was elongated (Fig. 3). This phenomenon is similar to the generation of foam when sea waves break 22. Specifically, when the thickness of the head part of the wave was thin, a lot of foam was formed. Therefore, in this experiment, the amount of trapped foam increased at 25% with the smallest amount of liquid because the specific surface area of the liquid exposed to the gas was the largest (Fig. 3). Due to these reasons, it was assumed that the results of NB generation (Fig. 2 (a)) were caused by the amount of trapped bubbles and the shear force applied to the trapped bubbles. On the other hand, the mean diameter of the generated NBs was 200 nm or less, and there was no difference according to the LVF (Fig. 2 (b)). The zeta potential also exhibited a value between − 15–-20 mV in all cases except for the 100% gas–liquid ratio wherein NBs were not generated. Therefore, it can be assumed that the NBs generated through this NB generation technique were uniformly dispersed in liquid and existed in a relatively stable state.
3.2. Effect of generation time
To confirm the effect of generation time, an NB generation test, based on the generation times of 5, 10, and 20 min, was performed. Figure 4 shows the NTA images captured in all cases. Figure 5 shows the analysis results of the captured images with the NTA software and the zeta potential value of the generated NBs. In addition, the quantitative values for Fig. 5 are outlined in Table 1. As confirmed in Section 3.1, NBs were abundant when the LVF was 25% for each generation time condition (5, 10, and 20 min), and the concentration of the generated NBs decreased as the liquid content in the container increased. In addition, the concentration of the generated NBs increased in proportion to the generation time in all cases except for 100%. Since this NB generation device performed a vertical reciprocating motion (117 strokes/min), it could be assumed that the number of bubbles trapped in the liquid and the number of shear forces applied to the initially trapped bubbles were 234 times per min. In other words, for each NB generation time, the shear force was applied 1,170 times for 5 min, 2,240 times for 10 min, and 4,680 times for 20 min. Therefore, the NB concentration significantly increased because more bubbles were trapped in the liquid and more shear forces were applied to the trapped bubbles as the generation time increased. The generation time also affected the size of the generated NBs. As can be seen in Fig. 5 (b), in all cases except for 100%, the mean diameter of the generated NBs decreased as the generation time increased. This phenomenon can be confirmed more clearly in the size distribution graph of the NBs (Fig. 6). As shown in Fig. 6, the peak (mode diameter) of the graph shifted to the left as the generation time increased. Considering that the difference in mode diameter of the generated NBs based on the LVF was negligible, the size of the generated NBs was thought to be more affected by the generation time (i.e., the number of shear forces applied to the trapped bubbles) than the LVF (i.e., the magnitude of shear force applied to the trapped bubbles). In addition, since the difference in zeta potential values of the generated NBs was negligible in all cases except 100%, it could be confirmed that the NBs generated by this method existed in a relatively stable state regardless of the generation time and the LVF.
Table 1
The results of characteristics evaluation after NB generation according to the LVF and generation time.
Generation time
(min)
|
LVF
(%)
|
Concentration
(108 NBs/ml)
|
Mean diameter
(nm)
|
Zeta-potential
(mV)
|
5
|
25
|
3.21 ± 0.13
|
187.33 ± 18.01
|
-18.80 ± 1.58
|
50
|
1.48 ± 0.11
|
185.02 ± 14.73
|
-18.58 ± 1.80
|
75
|
0.75 ± 0.08
|
190.66 ± 23.54
|
-16.80 ± 4.56
|
100
|
0.10 ± 0.04
|
202.66 ± 13.05
|
-0.43 ± 0.44
|
10
|
25
|
6.56 ± 0.33
|
159.12 ± 12.16
|
-20.75 ± 2.22
|
50
|
2.78 ± 0.18
|
160.33 ± 4.93
|
-18.72 ± 0.63
|
75
|
1.23 ± 0.12
|
158.09 ± 11.35
|
-18.39 ± 1.45
|
100
|
0.13 ± 0.06
|
207.33 ± 22.14
|
-1.19 ± 1.65
|
20
|
25
|
10.65 ± 1.07
|
149.03 ± 5.56
|
-20.80 ± 1.72
|
50
|
5.01 ± 0.45
|
138.07 ± 9.53
|
-19.02 ± 2.15
|
75
|
3.15 ± 0.32
|
144.66 ± 18.55
|
-19.98 ± 1.78
|
100
|
0.12 ± 0.08
|
206.33 ± 25.69
|
-1.01 ± 0.61
|
3.3. NB stability
This study was designed to confirm the effect of the LVF and the generation time on NB generation and yielded meaningful results. In this section, we will investigate the long-term stability of the NBs generated by this device and the possibility of external contamination during the experiment.
First, the stability of the generated NBs was evaluated over time, for which the sample that generated the most NBs (LVF: 25%, generation time: 20 min) was used. NTA analysis and zeta potential measurements were then performed for seven days. In the NB research field, zeta potential is commonly used when discussing the stability of NBs because NBs with the same polarity repel each other and inhibit coalescence between neighboring bubbles 8,23,24. As shown in Fig. 7 (a), the NB concentration decreased significantly for one day but showed a small decrease after seven days. Although the NB concentration decreased over time, there were still a large number of NBs (approximately 57.72% of the initial concentration) after seven days. On the other hand, the zeta potential of NBs was approximately − 20 mV, and no significant change during the seven days of stability evaluation was observed (Fig. 7 (c)). Therefore, the impact of reducing NB concentration due to coalescence between adjacent bubbles was regarded as insignificant because there was no change in the zeta potential values of NBs over time. This phenomenon was also observed in the mean diameter measurement results of the NBs (Fig. 7 (b)). Although the mean diameter of the NBs increased over time, the increase was not significant, approximately 20–30 nm only. If the coalescence between the adjacent bubbles had predominantly occurred, the mean diameter of the NBs measured after seven days should have been at least twice the initial diameter. Therefore, it was assumed that the decrease in NB concentration was due to the disappearance of some NBs, which were maintained in a relatively unstable state and diffused into the liquid, rather than the effect of coalescence.
Next, the inflow of contaminants (solid particles), which may have occurred during the experiment, was confirmed. Studies that have explored the possibility of contaminant inflow share a lot of similarities with those proving the existence of NBs. In fact, many researchers who insist that NBs cannot exist in a stable state in liquid regard NBs, recorded by an NTA apparatus, as solid particles flowing in from the outside. The existence of spherical NBs has recently been proved using scanning electron microscopy 1, transmission electron microscopy 2, and the mass spectrometry method 4, using a resonance cantilever and microfluidic channel, which can distinguish NBs from solid particles. In addition to these direct approaches, indirect approaches proving the existence of NBs by removing the NBs in the liquid have also been reported 21,25,26. These studies exposed the NB solution to harsh environments (decompression, freezing, and thawing), causing the NBs to become unstable and diffuse into the liquid. Subsequently, NTA analysis was performed on an NB solution exposed to harsh environments, proving that the generated NBs were not solid particles. The freezing and thawing method was also used in this study to confirm the possibility of contaminant inflow. The test samples used for this experiment were selected from those applied for long-term stability evaluation (DI water, NB solution after generation, and NB solution after seven days). Each test sample was frozen in the freezer compartment of a home refrigerator for six hours and then left in the atmosphere until the temperature of the test sample became the same as room temperature. NTA analysis was performed on the test sample. The results are shown in Fig. 7 (d), where we can confirm that most of the initially present NBs disappeared in all the test samples after freezing and thawing. Therefore, it can be said that there was no possibility of inflow of contaminants from the outside during NB generation and the character evaluation that was performed in this study.