Plasma generation and its optical characteristics.
The applied voltage started to increase from − 2.0 kV to + 9.3 kV in 15 µs (Fig. 2(a)). The voltage then decreased to -3.2 kV, oscillated, and gradually rose to -2 kV, and the next pulse was applied. The pulse width of the applied voltage was 30 µs. During the application of the voltage pulse, there were an oscillated displacement current of up to about 0.03 A and several current pulses corresponding to the plasma discharge. The discharge current pulses were seen in both positive and negative polarities, with a maximum of about 0.04 A (in absolute value).
Focusing on the plasma emission during discharge, we can see that streamers were generated from the needle, which was the applying electrode, towards the inner surface of the glass tube with the ground electrode (Fig. 2(b)). The points at which these streamers were formed varied intermittently (Supplementary Movie S1). The spectra of the plasma emission exhibited peaks (Fig. 2(c)), and they were attributed to the second positive system of nitrogen [N2-SPS; N2(C–B)] and the first negative system of nitrogen [N2+-FNS; N2+(B–X)] [17]. These emission peaks are commonly found in atmospheric pressure plasma discharge [18].
The discharge appeared to be of a corona-like type, based on the observation of the discharge current, the light emission, and optical characteristics.
Dynamics and discharge characteristics in nano-sized mist.
We then infused UPW into the corona-like discharging electrode unit, and succeeded in generating a nano-sized mist to be blown out from the tip of the applying electrode (Fig. 3 and Supplementary Movie S2). Based on the movie taken with the 10× microscope lens, the mist particle size was measured to be at least less than 50 µm and at most less than 2.15 µm (Fig. 3(b)), i.e., smaller than the resolution of the optical system used in this study. The efficiency of mist generation depended on the infusion rate of the solution. Although the mist was generated even under conditions of high infusion rate (≥ 500 µL/min for UPW and PBS; ≥ 100 µL/min for CO, respectively), droplet formation at the tip of the unit occurred preferentially (Supplementary Fig. S1). The discharge characteristics at the time of mist generation exhibit that a number of discharge current pulses were induced during applying the voltage pulse (Fig. 4). The discharge current pulse was larger (up to about 0.06 A) than in the case where no solution was infused.
The developed nano-size mist generator can be applied to atomization not only of UPW but also of PBS with high electrical conductivity and CO with high viscosity (Fig. 5 and Supplementary Movies S2, S3, and S4). There was no significant difference in the conditions suitable for mist generation with respect to the type of solution. We could see some differences in the generation efficiency of mist itself and its finer size depending on the solution type, but it was generally possible to produce a mist of less than 2.15 µm. The discharge characteristics of each solution in mist generation showed that discharge current pulses occurred as in the case of UPW (Fig. 6). The pulses in PBS were larger than those in the other solutions, exceeding 0.1 A. In PBS, the pulse width of the applied voltage was stretched by about 5 µs, and the oscillations in both voltage and current were smaller. The pulse widths of the discharge current for CO were, in contrast, smaller than those of the other solutions. These differences in the discharge characteristics might be due to the electrical conductivity [19]. This notion is supported by the fact described below that there is a marked difference in physical and chemical characteristics between UPW or CO, which has a low conductivity [20, 21], and PBS with a high conductivity. Moreover, the changes in electrical conductivity of UPW by generation of chemical species induced a slight difference in the discharge characteristics between UPW and CO. We conclude that the physicochemical characteristics and the plasma-generated chemical species caused differences in the discharge characteristics and mist generation efficiency of each solution.
Physical and chemical characteristics of solutions from which the nano-sized mist was produced.
After passing UPW through the corona-like discharge, its pH value decreased from 6.5 to below 3.0, with some differences depending on the infusion rate (Fig. 7(a)). In contrast, the pH of PBS slightly decreased from 7.4 to about 6.8 due to the interaction with the plasma (Fig. 7(b)). The conductivity of each solution tended to increase with changes in its pH value, and reached the maximum value of 7.9 mS/cm (UPW, Fig. 7(c)) or 18.6 mS/cm (PBS, Fig. 7(d)), respectively, at the infusion rate of 50 µL/min. The changes in these characteristics can be attributed to the influence of chemical species produced in each solution by interaction with the plasma.
Various chemical reactions occurred at the gas-liquid interface where plasma discharge occurs, leading to the formation of oxidative products in the solution [16]. Our plasma source also produced oxidative compounds such as hydrogen peroxide, nitrite ion, and nitrate ion in each solution (Fig. 8). The amount of hydrogen peroxide produced in UPW increases with decreasing its infusion rate, reaching a maximum value (446.4 ± 19.0 mg/L) in the rate of 100 µL/min, and the produced amount (314.4 ± 17.6 mg/L) decreases under the condition with the lowest rate (50 µL/min) (Fig. 8(a)). Hydrogen peroxide is produced by the coupling of hydroxyl radicals, which are generated from water molecules in solution by the plasma, as shown in the following chemical reaction [22].
OH ∙ + OH ∙ → H2O2
Hydrogen peroxide also acts as an oxidant under acidic conditions as followed [23, 24].
H2O2 + 2H+ + 2e− → 2H2O
This half-reaction of hydrogen peroxide as an oxidant might induce a decrease in its produced amount at the 50 µL/min condition. This notion is supported by the result that the concentration of hydrogen peroxide in PBS, which had a small change in pH value, monotonically increased with decreasing the infusion rate (Fig. 8(b)). On the other hand, nitrite and nitrate ions are produced due to the reaction of nitric oxide, nitrogen dioxide, and hydroxyl radicals generated by the plasma, as shown in the following reactions [25].
NO + OH ∙ + M → HNO2 + M
NO2 + OH ∙ + M → HNO3 + M
2NO2 + H2O → HNO3 + HNO2
Here, M indicates the third body, which is typically H2O. The generated nitric acid and nitrous acid dissociate in the solution to yield nitrite and nitrate ions [26].
HNO2 ⇆ H+ + NO2−
HNO3 ⇆ H+ + NO3−
The smaller the infusion rate of the solution, the more nitrogen compounds were produced, with some exceptions. At the rate of 1000 µL/min, the concentration of the produced nitrite ions was 6.80 ± 0.75 mg/L in UPW (Fig. 8(c)) or 41.0 ± 3.3 mg/L in PBS (Fig. 8(d)), while it was less than 0.15 mg/L under other rate conditions. Nitrate ions were most abundantly produced under the condition of 50 µL/min solution infusion rate, and their concentrations were 2759.2 ± 489.5 mg/L in UPW (Fig. 8(e)). In PBS, the amount of nitrite ions produced was large, and it was difficult to remove them due to the effect of hydrogen peroxide produced at the same time. We therefore measured the total amount of nitrite and nitrate ions generated in PBS. As a result, their concentration was 774.0 ± 122.1 mg/L (Fig. 8(f)). The reason why there are fewer oxidative compounds produced by the plasma in PBS than those in UPW is probably because phosphoric acid compounds such as the disodium hydrogen phosphate and potassium dihydrogen phosphate in PBS acted as scavengers for them. The difference in the amount of nitrite and nitrate ions produced in UPW and PBS might be related to the slope of the equilibrium state depending on changes in pH value of each solution.