UV–vis spectroscopy study of plasma-activated water: Dependence of the chemical composition on plasma exposure time and treatment distance

A dielectric barrier discharge (DBD) helium plasma jet was used (see Fig. 1).^35–38) The plasma jet apparatus consisted of a 150-mm-long glass tube with a 3.7 mm inner diameter and was slightly tapered at the nozzle. Helium gas was fed into the glass tube at a fixed gas flow rate of 3 slpm through a flowmeter controller (FCON 1000). A high-voltage bipolar square pulse of 7 kV (peak to peak) at 10 kHz was applied to a metallic external electrode. The high voltage and discharge current were measured as previously reported,^26,28,39,40) employing a high-voltage probe (Tektronix P6015A) and a conventional current monitor (Pearson 2877), respectively. Figure 2 shows the bipolar high voltage and discharge current waveforms. Positive and negative current pulses per period were detected. A peak discharge current of 1.68 mA was measured for the positive current pulse with a pulse width of ∼5 µs, while −0.88 mA was measured for the negative pulse current with a relatively broad pulse width of ∼9 µs. The peak and average input powers (P_in) were 4.18 and 0.41 W, respectively, calculated using

$$\begin{equation} P_{\text{in}} = \int_{0}^{T}V(t) \times I_{\text{D}}(t)\,dt, \end{equation} \tag{ 1 }$$

where T is the time for half a period (50 µs), V is the applied voltage, and I_D is the discharge current. Both V and I_D are time-dependent.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Deionized water (DI water) (18.2 MΩ·cm at 25 °C) was obtained from a Millipore Direct-Q 3UV system. 500 mL of DI water was stored in a container in ambient atmosphere for one day. This storage procedure enabled the concentrations of dissolved gases within the DI water to equilibrate with the air, which ensured that the concentrations of dissolved gases were constant between experiments. For the plasma treatment, 3 mL of DI water was transferred to a 60-mm-diameter petri dish. Exposure time and treatment distance were varied up to 30 min and 80 mm, respectively.

Double-beam UV–vis–NIR spectrophotometers (Hitachi U-4100 and U-3900) were used in the transmittance measurement mode. The U-4100 model was for the measurement of a broad wavelength range from 200 to 2500 nm; the U-3900 model was used for shorter wavelength measurements up to 190 nm over a narrower range of 800 nm. The spectrophotometers were operated with a fixed spectral resolution of 0.2 nm and a fixed scan speed of 120 nm/min. Measurements were performed in quartz cuvettes (Hellma Analytics QS-101). The quartz cuvette guaranteed a high transmittance of over 87% on average over the whole visible and near-infrared wavelength ranges. At the shortest wavelength of 190 nm, the transmittance was 82% through the quartz cuvette. The optical path of light through the cuvette (l) was 10 mm.

The measurements presented in this study are sensitive to fluctuations in the intensity of the light from the lamp of the spectrophotometer. Thus, we measured background transmittance T(λ), recorded using empty cuvettes in both the reference I₀(λ) and sample I(λ) analysis chambers, to account for fluctuations in the intensity of the light sources of the spectrophotometers:

$$\begin{equation} T(\lambda) = I(\lambda)/I_{0}(\lambda). \end{equation} \tag{ 2 }$$

Next, the transmittances of DI water and PAW were measured in the sample analysis chamber with the empty cuvette as reference. The transmittance spectra of untreated DI water (T) and PAW (T') were converted into absorbance using

$$\begin{equation} \text{Absorbance} = {-}{\log}(T'/T). \end{equation} \tag{ 3 }$$

All UV–vis measurements were taken 5 min after the preparation of PAW. Owing to their short half-life of less than 1 s, the highly reactive molecules generated by atmospheric plasma in water, such as the OH, peroxynitrite (ONOO⁻), and other radicals, were expected to be extinguished in PAW before measurement. The longer-lived molecules such as H₂O₂, NO₂⁻, and NO₃⁻ will remain in PAW at stable concentrations owing to their much longer half-life; previously, these RONS were detected in PAW even days after plasma treatment.⁷⁾

In Fig. 3, the transmittances for DI water and PAW (plasma exposure time = 30 min; treatment distance = 15 mm) were measured using the U-4100 spectrophotometer over a wavelength range of 200–2500 nm. The UV–vis–NIR spectra were almost identical in the range of 340–2500 nm. Below 340 nm, the spectra were different. This difference is due to the concentrations of RONS and O₂ in PAW being different from those in DI water. Therefore, our subsequent measurements using the U-3900 model spectrophotometer focused on shorter wavelengths.

Zoom In Zoom Out Reset image size

Figure 4(a) shows typical spectra of PAW prepared with plasma exposure times of 10, 20, and 30 min at a fixed treatment distance of 25 mm. As expected, the intensity of the absorption spectra increased with increasing plasma exposure time. The profiles were similar but not identical. On closer inspection of the spectra, the central peak position slightly shifted from 205.8 nm after 10 min of plasma exposure to shorter wavelengths after prolonged plasma exposure (203.6 nm at 20 min and 203.0 nm at 30 min). In addition, the width (FWHM) of the peaks slightly varied from 31 nm at 10 min to 30 nm at 20 min and 36 nm at 30 min. These small differences in peak profiles indicate that not only did the concentration of RONS and O₂ change with plasma exposure, but so did the composition of RONS in PAW. The concentrations of RONS and O₂ were determined from the spectra using the Beer–Lambert law,

$$\begin{equation} \text{Abs}_{\lambda} = \varepsilon lc, \end{equation} \tag{ 4 }$$

where ε is the molar absorptivity of the chemical species at a certain wavelength λ, l is the optical path length, and c is the concentration of RONS and O₂.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Assuming that the total absorbance in the UV–vis spectra corresponds to the total RONS and O₂ concentrations, we suggest that the total RONS and O₂ concentration is equal to the total absorbance range of 190–340 nm. This wavelength range takes into consideration the peak shifts and the increase in the FWHM in PAW after prolonged plasma exposure. Therefore, the total RONS and O₂ concentration in PAW can be calculated according to the following equation:³²⁾

$$\begin{align} \text{Total absorbance} &\approx \int_{=190\text{nm}}^{=340\text{nm}}\text{Abs}(\lambda)\,d\lambda\\ &\approx \text{RONS and oxygen concentration}. \end{align} \tag{ 5 }$$

Using this equation, we plotted the total RONS and O₂ concentration in PAW following plasma exposure times of 1 to 30 min at a fixed treatment distance of 25 mm. Under these parameters, it is seen in Fig. 4(b) that the total RONS and O₂ concentration increased linearly as a function of plasma exposure time. Although the total RONS and O₂ concentration increased linearly, the relative concentration ratio of $\text{RONS}:\text{O}_{2}$ changed with treatment time. This is because the neutral helium gas flow during plasma treatment purges O₂ out of the DI water at the initial stages of treatment (up to 5 min) before the O₂ concentration begins to increase, whilst the plasma always increases the RONS concentration. Despite a decrease in the O₂ concentration in DI water, a linear increase is still seen in the total RONS and O₂ concentration in Fig. 4(b) because O₂ does not affect the total absorbance as much as RONS.⁴¹⁾

The plasma-induced aqueous liquid chemistry has been shown to be sensitive to small (mm) changes in treatment distance, i.e., the distance between the plasma jet nozzle and the surface of the target solution.⁴⁰⁾ In this study, we further investigated how the treatment distance influences the plasma jet generation of RONS and O₂ in DI water. The treatment distance was varied from 15 to 80 mm. Figure 5(a) shows the typical absorption spectra of PAW prepared using treatment distances of 15 mm (plasma plume in contact with the surface of the DI water), 25 mm (plasma plume not in contact but in close proximity to the DI water), and at a much larger distance of 80 mm (no contact, remote) for a fixed exposure time of 5 min. The intensity of the absorbance peak was highest in PAW prepared at 25 mm with a lower absorbance peak at 15 mm and the lowest at 80 mm. This indicates a nonlinear correlation between the total RONS and O₂ concentration in PAW and the treatment distance. The profile of the UV–vis absorption spectra were similar, but not identical. The central peak position shifted: 206.6 nm at 15 mm, 206.2 nm at 25 mm, and 206.8 nm at 80 mm. This indicates that the concentration of RONS and O₂ in PAW is influenced by treatment distance. We have identified four possible contributing factors to explain this.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

• (1)   
  There is an increase in mixing of the plasma effluent with ambient air at longer treatment distances that changes the RONS chemistry in the gas phase — Ito et al. showed that gas mixing enhances the plasma generation of NO₂⁻ in DI water.⁴²⁾
• (2)   
  Closer treatment distances result in a greater level of evaporation creating a more humid atmosphere that may affect the RONS concentration — e.g., Winter et al. showed that the Ar plasma jet generation of H₂O₂ in solution significantly increases with humidity.⁴³⁾
• (3)   
  A higher gas velocity impinging on the surface of the DI water at shorter treatment distances will significantly deoxygenate the DI water. This is due to the inert gas from the plasma jet purging dissolved O₂ in a process referred to as "sparging".⁴¹⁾
• (4)   
  Evaporation decreases the temperature of the DI water, which changes the solubilities of RONS and O₂.⁴⁴⁾

Figure 5(b) shows the total absorbance (total RONS and O₂ concentration) plotted as a function of treatment distance. The trend follows a parabolic relationship: the absorbance increased between 15 and 30 mm and then decreased from 30 to 80 mm. This trend can be explained by the different processes occurring at different treatment distances (as discussed directly above), which affect the plasma generation of RONS and O₂ in the DI water. However, in addition to enhanced mixing of the plasma effluent with ambient air at longer treatment distances, we also consider that moving the plasma jet (relative to the DI water surface) changes the solubility of RONS. Norberg et al.⁴⁵⁾ showed in their computational study that the concentrations of dissolved charged species and certain neutral RONS increase when the plasma is in contact with the water surface.

In order to quantify the concentrations of RONS and O₂ in PAW, it is necessary to perform a curve fitting routine to derive the contributions of the different RONS and O₂ to the UV–vis spectra. We have shown that the reference spectra of RONS and O₂ can be used to accurately and reproducibly curve-fit the UV–vis spectra of plasma-treated DI water.³⁴⁾ Since the accuracy of the reference spectra significantly affects the accuracy of the curve fitting routine, we compared our reference spectra of H₂O₂, NO₂⁻, and NO₃⁻ to the reference spectra kindly provided by Professors Katsuhisa Kitano (Osaka University) and Petr Lukeš (Institute of Plasma Physics CAS).⁴⁶⁾ The results are shown Figs. 6(a)–6(c). The spectra of H₂O₂, NO₂⁻, and NO₃⁻ in our work are in very good agreement with the spectra provided by Kitano and Lukeš. The peak of H₂O₂ is not shown in the graph because it is outside the wavelength range of the spectrophotometers — it was measured to be at 160 nm.^47,48) In addition, our earlier reports have shown that plasma treatment also changes the O₂ concentration in DI water.^33,34,49) Therefore, in order to curve-fit the changes in O₂ concentration, we obtained the reference spectrum of O₂ in Fig. 6(d). The absorption spectrum and concentration of O₂ were obtained by purging with He, as previously described.³³⁾ We were unable to find other O₂ spectra with which to compare our reference spectrum.

Zoom In Zoom Out Reset image size

We further investigated how the concentrations of H₂O₂, NO₂⁻, and NO₃⁻ might affect the UV–vis spectra (see Figs. S1–S3 in the online supplementary data at http://stacks.iop.org/JJAP/57/0102B9/mmedia). The concentrations of these RONS were varied: H₂O₂, 9.65–9666 mg/L; NO₂⁻, 1.66–1000 mg/L; and NO₃⁻, 2.5–10000 mg/L. The central peak of H₂O₂ was below 190 nm [Fig. S1(a) in the online supplementary data at http://stacks.iop.org/JJAP/57/0102B9/mmedia]. Therefore, we used the intensity of the peak position at the shortest wavelength of 190 nm to investigate how changes in H₂O₂ concentration affect the UV–vis absorption. The absorbance increased linearly as a function of increasing H₂O₂ concentration [Fig. S1(b) in the online supplementary data at http://stacks.iop.org/JJAP/57/0102B9/mmedia]. For NO₂⁻, the major peaks were observed at 209.8 nm (for lower concentrations of up to 33.2 mg/L) and at 354 nm (for higher concentrations of over 66.5 mg/L) [Fig. S2(a) in the online supplementary data at http://stacks.iop.org/JJAP/57/0102B9/mmedia]. The intensities of both absorption peaks at 209.8 and 354 nm increased linearly as functions of increasing concentration [Fig. S2(b) in the online supplementary data at http://stacks.iop.org/JJAP/57/0102B9/mmedia]. For NO₃⁻, we observed an absorption peak at 200.6 nm corresponding to the π–π^* electron transition up to 10 mg/L. Higher concentrations resulted in absorbances above the detection limit of the spectrophotometer. A peak at 302 nm (corresponding to the n–π electronic transition) emerged at higher concentrations of over 66.6 mg/L [Fig. S3(a) in the online supplementary data at http://stacks.iop.org/JJAP/57/0102B9/mmedia]. The intensities of both absorption peaks at 200.6 and 302 nm increased linearly as functions of increasing concentration [Fig. S3(b) in the online supplementary data at http://stacks.iop.org/JJAP/57/0102B9/mmedia]. For O₂, we only compared the spectra between DI water equilibrated with ambient air (cf. 8 mg/L) and 100% deoxygenated DI water.³³⁾

On the basis of the above measurements with the three reference solutions of H₂O₂, NO₂⁻, and NO₃⁻, we determined a linear relationship between peak intensity and RONS concentration. This linear relationship was observed at any given wavelength. However, in PAW, it is difficult to accurately use the UV–vis spectra over the entire wavelength range. The reason for this is that, at low RONS concentrations, the intensity of the UV–vis spectra is too low (below the noise level) at long wavelengths. On the other hand, at high RONS concentrations, short wavelengths are not suitable because the intensity over this range is too high (above the detector saturation level). Therefore, we used the intensity measurements taken at both long and short wavelengths to determine the RONS concentration in PAW.

Using the reference spectra of H₂O₂, NO₂⁻, NO₃⁻, and O₂, we curve-fitted the experimentally obtained UV–vis spectra of PAW by the following steps.

• (1)   
  The spectrum tail above 245 nm is first fitted with the reference H₂O₂ spectrum [Fig. 7(a)].
• (2)   
  The absorption shoulder between 230 and 240 nm is fitted with the reference NO₂⁻ spectrum [Fig. 7(b)].
• (3)   
  The peak between 200 and 210 nm is fitted using the NO₃⁻ reference spectrum [Fig. 7(c)].
• (4)   
  The peak below 200 nm is fitted with the reference spectrum of dissolved O₂ [Fig. 7(d)].

Zoom In Zoom Out Reset image size

The curve-fitting procedure for PAW prepared with 20 min of plasma treatment is relatively simple to fit. However, the method is too time-consuming for analyzing large datasets. Therefore, we developed an automated curve fitting program at Kochi University of Technology (Jun-Seok Oh and Akimitsu Hatta). This automated curve fitting program simultaneously calculates the concentrations of H₂O₂, NO₂⁻, NO₃⁻, and O₂ at four different wavelengths using a 4 × 4 matrix data structure.

The absorbance at the wavelength λ for molecule X is given by

$$\begin{equation} \text{Abs}_{X}(\lambda) = \varepsilon(\lambda)\cdot l\times c_{X}. \end{equation} \tag{ 6 }$$

When four different molecules (A, B, C, and D) contribute to the absorption spectra, the total absorbance per unit optical path length is given as the total sum of absorptivity multiplied by the concentration:

$$\begin{align} \text{Abs}_{X}(\lambda)/l &= \varepsilon_{\text{A}}(\lambda)\cdot c_{\text{A}} + \varepsilon_{\text{B}}(\lambda)\cdot c_{\text{B}} \\&\quad+ \varepsilon_{\text{C}}(\lambda)\cdot c_{\text{C}} + \varepsilon_{\text{D}}(\lambda)\cdot c_{\text{D}}. \end{align} \tag{ 7 }$$

Here, we chose four different wavelengths, λ₁ to λ₄, where the ratios between the absorptivity differed markedly for each molecule. The total absorbance at a given wavelength was defined as the product between the matrix for absorptivity and the vector for concentration as

$$\begin{equation} \begin{pmatrix} \text{Abs}(\lambda_{1})\\ \text{Abs}(\lambda_{2})\\ \text{Abs}(\lambda_{3})\\ \text{Abs}(\lambda_{4}) \end{pmatrix} \biggm/l= \begin{pmatrix} \varepsilon_{\text{A}}(\lambda_{1}) & \varepsilon_{\text{B}}(\lambda_{1}) & \varepsilon_{\text{C}}(\lambda_{1}) & \varepsilon_{\text{D}}(\lambda_{1})\\ \varepsilon_{\text{A}}(\lambda_{2}) & \varepsilon_{\text{B}}(\lambda_{2}) & \varepsilon_{\text{C}}(\lambda_{2}) & \varepsilon_{\text{D}}(\lambda_{2})\\ \varepsilon_{\text{A}}(\lambda_{3}) & \varepsilon_{\text{B}}(\lambda_{3}) & \varepsilon_{\text{C}}(\lambda_{3}) & \varepsilon_{\text{D}}(\lambda_{3})\\ \varepsilon_{\text{A}}(\lambda_{4}) & \varepsilon_{\text{B}}(\lambda_{4}) & \varepsilon_{\text{C}}(\lambda_{4}) & \varepsilon_{\text{D}}(\lambda_{4}) \end{pmatrix} \begin{pmatrix} c_{\text{A}}\\ c_{\text{B}}\\ c_{\text{C}}\\ c_{\text{D}} \end{pmatrix} . \end{equation} \tag{ 8 }$$

From the measured Abs(λ), the concentrations were directly determined using the inverse matrix for absorptivity:

$$\begin{equation} \begin{pmatrix} c_{\text{A}}\\ c_{\text{B}}\\ c_{\text{C}}\\ c_{\text{D}} \end{pmatrix} = \begin{pmatrix} \varepsilon_{\text{A}}(\lambda_{1}) & \varepsilon_{\text{B}}(\lambda_{1}) & \varepsilon_{\text{C}}(\lambda_{1}) & \varepsilon_{\text{D}}(\lambda_{1})\\ \varepsilon_{\text{A}}(\lambda_{2}) & \varepsilon_{\text{B}}(\lambda_{2}) & \varepsilon_{\text{C}}(\lambda_{2}) & \varepsilon_{\text{D}}(\lambda_{2})\\ \varepsilon_{\text{A}}(\lambda_{3}) & \varepsilon_{\text{B}}(\lambda_{3}) & \varepsilon_{\text{C}}(\lambda_{3}) & \varepsilon_{\text{D}}(\lambda_{3})\\ \varepsilon_{\text{A}}(\lambda_{4}) & \varepsilon_{\text{B}}(\lambda_{4}) & \varepsilon_{\text{C}}(\lambda_{4}) & \varepsilon_{\text{D}}(\lambda_{4}) \end{pmatrix} ^{-1} \begin{pmatrix} \text{Abs}(\lambda_{1})\\ \text{Abs}(\lambda_{2})\\ \text{Abs}(\lambda_{3})\\ \text{Abs}(\lambda_{4}) \end{pmatrix} \biggm/l. \end{equation} \tag{ 9 }$$

To improve the signal-to-noise ratio, instead of using four points at specific wavelengths, four different wavelength ranges were used to determine the absorbance and absorptivity. After calculating the concentration of each type of molecule, the correctness of the fitted curve was monitored by comparing the difference between the experimental data and the theoretical data obtained using Eq. (7).

In Sects. 3.5.1 and 3.5.2, we utilized the automated curve fitting routine to analyze how the RONS and O₂ composition changes depending on the exposure time and treatment distance.

3.5.1. Influence of exposure time.

Figure 8 shows the curve fitting results of PAW for plasma exposure times of 1, 10, and 30 min. After 1 min, a prominent negative absorbance was observed below 195 nm. The reduced absorbance is attributed to a reduced concentration of dissolved O₂ caused by the inert He gas of the plasma jet purging the DI water. When we took into account the He purging effect into the absorption spectra of PAW,³³⁾ the simulated curve was fitted with very good agreement with the measured spectrum, as shown in Fig. 8(a). (See Fig. S4 in the online supplementary data at http://stacks.iop.org/JJAP/57/0102B9/mmedia without consideration of O₂ in supporting information. This results in a poor curve fit. Therefore, we always fit the shorter wavelength region with O₂.) From the curve fitting results, we determined that the major RONS (for the specific RONS measured in this study) produced by the plasma jet was H₂O₂ at 0.132 mg/L with NO₂⁻ at 0.097 mg/L and NO₃⁻ at 0.021 mg/L, which are much lower concentrations. The O₂ concentration was 5.28 mg/L. Figure 8(b) shows the measured absorption spectrum and curve fitting results after 10 min of plasma exposure. The intensity of the spectrum significantly increased, corresponding to an increase in the RONS concentrations (H₂O₂: 1.132 mg/L; NO₂⁻: 0.752 mg/L; NO₃⁻: 0.208 mg/L) and a further decrease in the O₂ concentration to 3.670 mg/L. In Fig. 8(c), the intensity of the peak tripled, which directly corresponded to an increase in plasma exposure time from 10 to 30 min. Apart from the difference in the intensity, the 10 and 30 min spectra were identical. The change in the peak intensity corresponded to marked increases in each of the RONS concentrations, as determined by the curve fitting routine (H₂O₂: 3.429 mg/L; NO₂⁻: 1.689 mg/L; NO₃⁻: 0.787 mg/L). At 30 min, the O₂ concentration was higher at 5.52 mg/L; this increase is presumably due to the conversion of some of the plasma-generated RONS to O₂.

Zoom In Zoom Out Reset image size

The trends in the changes in RONS and O₂ concentrations as functions of plasma exposure time are summarized in Fig. 9. The plasma exposure time was varied from 1 to 30 min at a fixed distance of 25 mm. The RONS concentration increased linearly (R² = 0.9992) as a function of plasma exposure time. The H₂O₂ concentration always remained higher than the NO₂⁻ and NO₃⁻ concentrations. Conversely, the O₂ concentration decreased as a function of plasma exposure time up to 5 min, but it increased after that time point, presumably owing to the formation of O₂ by-products from the chemical reactions of the plasma-generated RONS in the DI water. The results show that the RONS concentrations (i.e., concentrations of H₂O₂, NO₂⁻, and NO₃⁻) could be adjusted by simply varying the plasma exposure time without significantly affecting the RONS composition of PAW (i.e., the concentration ratio of $\text{H}_{2}\text{O}_{2}:\text{NO}_{2}{}^{ - }:\text{NO}_{3}{}^{ - }$).

Zoom In Zoom Out Reset image size

3.5.2. Dependence of treatment distance on RONS and O₂ concentrations.

Figure 10 shows a summary of the curve fitting results for treatment distances of 15, 25, and 80 mm with a fixed exposure time of 5 min. In Fig. 10(a) for the distance of 15 mm (plasma in contact with DI water), the concentrations of H₂O₂, NO₂⁻, and NO₃⁻ were 0.377, 0.220, and 0.055 mg/L, respectively. At 25 mm (plasma not in contact with the DI water), as shown in Fig. 10(b), the H₂O₂, NO₂⁻, and NO₃⁻ concentrations were significantly higher at 0.550, 0.421, and 0.094 mg/L, respectively. At the farthest distance of 80 mm shown in Fig. 10(c), the concentrations of all RONS were lower: H₂O₂, 0.126 mg/L; NO₂⁻, 0.130 mg/L; and NO₃⁻, 0.012 mg/L. At 15 mm, the O₂ concentration was 5.02 mg/L. The lowest O₂ concentration of 3.55 mg/L was measured at 25 mm. However, a higher O₂ concentration of 7.44 mg/L was measured at 80 mm. The O₂ concentration was higher in PAW prepared at 15 mm than in PAW prepared at 25 mm because He gas purging of the DI water was not as effective at the shorter treatment distance. This is because the contact area of the He gas with the surface of the DI water increases up to a treatment distance of 35 mm, as shown by the Schlieren images in Fig. S5 in the online supplementary data at http://stacks.iop.org/JJAP/57/0102B9/mmedia.

Zoom In Zoom Out Reset image size

The RONS concentrations are plotted as a function of treatment distance from 15 to 80 mm in Fig. 11. We see that the RONS chemistry is tunable by simply changing the treatment distance. At closer treatment distances of 15–30 mm, the concentration of H₂O₂ far exceeds those of NO₂⁻ and NO₃⁻. At 40–80 mm, the concentration of NO₂⁻ is higher than that of H₂O₂. The NO₃⁻ concentration was always lower than the H₂O₂ and NO₂⁻ concentrations.

Zoom In Zoom Out Reset image size

Overall, the results show that the treatment distance affects not only the RONS (H₂O₂, NO₃⁻, and NO₂⁻) concentration but also the concentration ratio of these RONS. Although we have no direct evidence of the effect of treatment distance on the RONS composition, we can make reasonable assumptions on the basis of what we know from the literature. Firstly, H₂O₂ is the major RONS at closer distances of up to 30 mm. At such closer treatment distances, the rate of water evaporation is higher, creating a more humid environment. Humidity is known to significantly enhance the plasma jet generation of H₂O₂ in aqueous solution.⁴³⁾ We also expect that treatment at closer distances with the plasma jet in contact with the liquid surface will enhance the generation of ROS, including aqueous OH and H₂O₂.⁵⁰⁾ In addition to diffusion and solvation processes, OH can also be formed in the DI water through UV photolysis.^43,51,52) Recombination of OH can lead to further generation of H₂O₂ in the water via the following reaction:¹³⁾

$$\begin{equation} \text{OH${\cdot}$} + \text{OH${\cdot}$} \rightarrow \text{H$_{2}$O$_{2}$}. \end{equation} \tag{ R1 }$$

We expect the contributions of humidity and UV photolysis to decrease as the treatment distance increases. At longer treatment distances, there will be more mixing and reactions of the plasma effluent with ambient air before the solvation of RONS into the water. Since air comprises mainly N₂, we speculate that the generation of RNS is favored. Therefore, we see higher proportions of NO₂⁻ and NO₃⁻ and a concomitant decrease in the H₂O₂ concentration.

However, the concentration of NO₃⁻ peaked at a treatment distance of 25 mm compared with that of NO₂⁻ at 30 mm. We once again attribute this to greater humidity at the closer treatment distance. Humidity may also be an important parameter for the plasma jet generation of NO₃⁻, as shown in the following reactions:¹³⁾

$$\begin{align} &\text{NO$_{2}$} + \text{NO$_{2}$} + \text{H$_{2}$O} \rightarrow \text{NO$_{2}$$^{-}$} + \text{NO$_{3}$$^{-}$} + \text{2H$^{+}$}, \end{align} \tag{ R2 }$$

$$\begin{align} &\text{4NO$_{2}$} + \text{O$_{2}$} + \text{2H$_{2}$O} \rightarrow \text{4NO$_{3}$$^{-}$} + \text{4H$^{+}$}. \end{align} \tag{ R3 }$$

In addition, a higher concentration of H₂O₂ generated at closer treatment distances might also facilitate the generation of NO₃⁻ and the consumption of NO₂⁻ via¹³⁾

$$\begin{align} &\text{NO$_{2}$$^{-}$} + \text{H$_{2}$O$_{2}$} + \text{H$^{+}$} \rightarrow \text{NO$_{3}$$^{-}$} + \text{H$_{2}$O} + \text{H$^{+}$}, \end{align} \tag{ R4 }$$

$$\begin{align} &\text{NO$_{2}$$^{-}$} + \text{H$_{2}$O$_{2}$} + \text{H$^{+}$} \rightarrow \text{O=NOOH} + \text{H$_{2}$O}. \end{align} \tag{ R5 }$$

At longer treatment distances of >30 mm, the concentrations of all RONS decrease.