3.1 Dynamics of chemical species during plasma discharge
The temporal generation tendencies of O3, NO2, and NO in the plasma chamber are illustrated in Fig. 2. As shown, the generation of various chemical species from air during plasma discharge was dominant in the early stage and then gradually decreased as NO2 formation increased. When O3 was completely decomposed, the concentration of NO2 started to decrease, and NO formation began. Furthermore, the concentration of NO increased without reaching the saturation level. This phenomenon is widely referred to as a mode transition from the O3-focused generation mode to the NOx-focused generation mode in air plasma generation [31]. This transition occurs because the generation speed of RONS increases as the power used for plasma generation increases; however, the increase in temperature is also known to be more significant in the mode transition, according to the following reaction [32]:
O3 + M→ O + O2 + M, k = 3.92 × 10− 16 exp(− 11400/T). (2)
As the input voltage increases, not all the additional power is utilized in plasma discharge; instead, some of it is released as thermal energy to the dielectric barrier in the DBD device. Therefore, the voltage directly correlated with the gas temperature inside the plasma chamber. As shown in Fig. S2, the internal gas temperature in the chamber gradually increased to 60 ℃ due to the transmission of thermal energy from the DBD. Similarly, as shown in Fig. 2(a), the O3 decomposition rate accelerated (4 kV, 350 s) and the concentration of NO gradually increased as the operating voltage increased. In general, thermal plasma or nonthermal plasma with additional heating is used to obtain O3-free NO and NO2. However, heat treatment gas is unsuitable for treating fresh fruits. In fact, it is not suitable for processing fruits even after being passed through Peltier element-based coolers. Therefore, a method to promote the NO-enhanced mode without causing gas heating is needed.
UV irradiation is an alternative method for promoting NO generation without heating the gas because UV light not only photolyzes O3 but also promotes NO production by decomposing HONO, a protonated form of HNO2. The UV-radiation spectrum and the UV-absorption coefficients of O3 and NO are shown in Fig. S1. Compared with O3, the absorption coefficient of HONO weakly overlaps with the UV spectrum; therefore, active photolysis of O3 cannot be expected. Thus, under a 4 kV operating voltage, UV-light irradiation accelerated the O3 decomposition rate by suppressing the O3 production pathway, and the maximum concentration of O3 decreased to ~ 100 ppm, as shown in Fig. 2(b). However, the rate of increase in NO was not significantly altered, which could be attributed not only to negligible spectral overlap but also to the limited availability of HONO. Under 4 kV operating conditions, the measured concentration of HONO was less than 20 ppm, which is significantly lower than that of NO at 600 ppm. However, in the plot corresponding to the 4 kV operating condition with UV irradiation (Fig. 2(b)), the slope of the change in the NO2 concentration with time did not vary. Therefore, the inference that NO enhancement is difficult to monitor because of the low HONO concentration was disregarded because of the unchanged NO2 concentration. Notably, the concentration change of NO2 shown in Fig. 2(a, b) includes not only the change in NO2 concentration but also that of HONO because of the NOx analyzer used in the analysis. However, because the NO2 production behavior under UV irradiation only exhibited an increase in the production time but the total NO2 amount remained unchanged, the assumption that HONO photolysis did not occur during UV irradiation is reasonable.
The early transition to the NO-generation mode via the UV photolysis of the gas appears to be predominantly influenced by the photolysis of O3 rather than that of HONO. The generation time of NO increased by more than 100 s, and the total amount of NO generated increased by more than 200 ppm, as observed after 10 min of UV treatment.
3.2 Detailed reactions related to NO enrichment
The numerical calculation method was first validated using measured temporal concentrations of O3, NO2, and NO. Figure 3(a) compares the measured and calculated concentrations during 600 s of plasma discharge. The experimental concentrations differed slightly from the theoretical concentrations; however, the maximum concentration and generation/decomposition tendencies were in good agreement for O3 and NO. For NO2, the calculated concentration steadily increased, whereas the measured concentration decreased rapidly because of the measurement limitations of NOx analyzers, as mentioned
in the experimental section. It is reasonable to assume that the actual concentration of NO2 did not decrease and followed the behavior predicted by calculations. In addition, as the measured values were recorded by sampling air from NOx and O3 analyzers, measurement error due to the sample flow rate is possible, which can result in deviations from the calculated values.
The numerical calculations included 53 species and 623 reactions in the gas phase [30]. First, among these reactions, we considered the reactions that are closely related to O3 and NO generation, as shown below:
$${\text{N}}_{2}{\text{O}}_{3}+\text{M}\to \text{N}\text{O}+\text{N}{\text{O}}_{2}+\text{M}$$
R1
$$\text{N}{\text{O}}_{2}+\text{O}\to \text{N}\text{O}+{\text{O}}_{2}$$
R2
$$\text{H}\text{O}\text{N}\text{O} \underrightarrow{\text{h}\nu } \text{N}\text{O}+\text{O}\text{H}$$
R3
$${\text{O}}_{3}+\text{N}\text{O}\to {\text{O}}_{2}+\text{N}{\text{O}}_{2}$$
R4
$${\text{O}}_{3}+\text{N}{\text{O}}_{2}\to {\text{O}}_{2}+\text{N}{\text{O}}_{3}$$
R5
$${\text{O}}_{3}\underrightarrow{\text{h}\nu }{\text{O}}_{2}+\text{O}$$
R6
where M denotes neutral molecules, which are N2 and O2 in this study. Reactions R1 to R3 are involved in NO production, while reactions R4 to R6 are related to O3 decomposition. The other species involved in the O3 and NO reactions are atomic oxygen (O), nitrogen trioxide (NO3), and dinitrogen trioxide (N2O3). The concentrations of these species were measured experimentally (Fig. 2 and Fig. 3(a)), while those of other species that were generated in trace quantities and could not be quantified experimentally were calculated numerically (Fig. 3(b)). The other species involved in O3 and NO generation are presented in Fig. S3. Furthermore, the evolution of the rate of each reaction is depicted in Fig. 3(c,d). Under UV irradiation, the notable change in Fig. 3(b) is the increase in the concentration of atomic oxygen (O) from 10− 5 to 10− 3 due to active O3 photolysis (R6). After 200 s, the time at which O3 was completely decomposed, no source for O generation remained except for the plasma discharge itself. Therefore, after 200 s, the concentration of O under UV irradiation reached almost the same value as that of plasma generation without UV irradiation. The enhanced quantity of highly reactive O leads to various reactions, including R2, in which O reacts with NO2, generating NO and O2, thereby promoting NO generation. As shown in Fig. 3(c), the NO production rate from reaction R2 was the highest until 200 s; during this period, it increased by ~ 1000 times from 10− 8 to 10− 5 mol/m3∙s. However, because most of the NO generated by reaction R2 reacts with O3 (reaction R4) and is consumed immediately after its generation, NO accumulates and is not distributed in the chamber. The other change is that the inflection points of the NO3 and N2O3 concentrations advanced by ~ 100 s, which is comparable to the reduction in the O3 decomposition time. NO3 is mainly generated by the combination of NO2 and O (reaction R5), and the O3 used as a reactant is decomposed by UV photolysis; thus, the generation of NO3 is reduced according to the reduction in the O3 concentration. The reduced rate of reaction R5 due to UV irradiation (Fig. 3(d)) supports the effect of reduced NO3 concentration, considering that the time at which the NO3 concentration starts to decrease increased from ~ 350 to 250 s. Unlike in the case of NO3, the production of N2O3 was promoted by UV photolysis. Its concentration was saturated at ~ 10− 2 ppm at ~ 350 s in the absence of UV exposure, while it started to saturate from 250 s onward under UV exposure. This tendency is opposite to that of NO3 because of the reverse reaction of R1. R1 maintains the equilibrium state, in which the reactants and products are maintained at appropriate proportions, but the reaction proceeds in the reverse direction as more products are present. In this case, NO is the preferred product. As mentioned above, reaction R2 promotes the generation of NO, which can react with NO2 to produce N2O3. At this time, another product of reaction R1, NO2, is present in sufficient quantity to accelerate the reverse reaction of R1, which is recorded at several tens of ppm from the time O3 starts to decompose. Therefore, the concentration of N2O3 is also highly dependent on the concentration of O3; thus, the generation time of N2O3 also increases with decreasing O3 elimination time due to UV exposure. Similarly, the forward reaction of R1 is also advanced to the same extent as the N2O3 production is because of the promotion of the reverse reaction of R1, as shown in Fig. 3(c).
In addition to accelerated O3 decomposition, the decomposition of HONO also contributes to the rapid production of NO (reaction R3). However, as shown in Fig. 3(c), the NO production rate from R3 was much lower than that from R1 and R2 during the O3-dominant period. The HONO concentration was also extremely low during the O3-dominant period, and the HONO concentration exhibited an increasing trend with increasing NO2 because HNO2 lacked a decomposition source when O3 was present. When the O3 in the gas mixture was completely decomposed, the NO production rate increased to 10− 6 mol/m3∙s, similar to that of R2. Nevertheless, reaction R1 occurred to a greater extent (~ 105 times higher) than R2 and R3, and the photolysis of HONO did not seem to have a significant effect on the promotion of NO enrichment.
3.3 Changes in the appearance of tomatoes according to plasma treatment
The plasma-treated tomatoes were imaged every 3 days (Fig. 4). As observed with the naked eye, the appearance of plasma-treated tomatoes was distinctively different from that of nonplasma-treated tomatoes. The color and firmness of the samples were evaluated to analyze the change in appearance due to plasma treatment (Fig. 5). The color value, “a*,” which indicates how red or green the tomatoes were, did not change significantly until the operating voltage increased to 3.5 kV during the entire treatment period. When the operating voltage was increased to 4 kV, the color started to change after the 6th day, and the disparities markedly widened until the 12th day. On Day 12, the “a*” value of the plasma-treated tomatoes decreased from 40 to 20, indicating that plasma treatment can indeed prevent the reddening of tomatoes during the ripening period. Even when treated with 4 kV plasma, the reddening of the tomato plants was largely prevented, but UV irradiation under these conditions prevented the reddening
more significantly. The tomatoes belonging to the UV-irradiated plasma treatment group started to
exhibit differences from the other cases from Day 6 onward. On Days 9 and 12, the difference between the a* values of the UV-irradiated plasma group and the plasma-only group was ~ 10. In addition to the change in color, the change in the firmness of the tomatoes was noteworthy. According to Fig. 5(c), the
firmness of nonplasma-treated tomatoes (control) and tomatoes treated with plasma at 3.5 kV exhibited a similar tendency toward a decrease in firmness. The firmness decreased from 14 to 4 N over 12 days, while the samples treated with plasma at 4 kV exhibited a decrease in firmness from 13.5 to 5.5 N.
The effect of plasma treatment on the decrease in firmness could be observed more clearly upon UV irradiation. As shown in Fig. 5(d), the tomatoes subjected to plasma treatment with UV irradiation exhibited a much lower decrease in firmness (from 13.5 to 6 N). The decrease differed by 1 and 2 N compared with those of the control (nontreated) and plasma-only treatment groups, respectively, with the difference being significant. This difference falls within the same range as the values reported for the delayed aging of fruits using NO and ozone and other species. Therefore, the ripening of tomatoes was also delayed in our case [22–24].
3.4 Delayed ripening due to respiratory depression
Various studies have reported the prolongation of tomato ripening due to NO treatment. Therefore, the findings of the present study, which focused on the generation of highly concentrated NO from plasma discharge, can be explained by earlier studies. Based on the results obtained in this study, we attributed the effect of NO treatment to respiratory depression. Figure 6 shows the accumulation of CO2 emitted from tomatoes during the 10 min following plasma treatment. Nontreated tomatoes (control group) emitted more than 2000 ppm of CO2, while plasma-treated tomatoes only emitted 1700 ppm, which was outside the error range. Moreover, the UV-irradiated plasma groups emitted a much lower amount of CO2 (ca. 1600 ppm). The error range slightly overlapped between the plasma only and the plasma with UV-irradiation groups, but the tendency of the decrease in emitted CO2 agreed well with the changes in appearance. Steffens et al. reported that NO suppresses the respiration of fruits due to its toxicity, which causes low CO2 emissions [33]. Thus, the decrease in accumulated CO2 emission depending on regular plasma treatment and the NO-enhanced mode of plasma treatment suggested that the NO-enhanced plasma device can extend the ripening period of fruits by inhibiting their respiration.