Dissociation Kinetics of Gaseous Ozone in Onion (Allium Cepa L.) Bulbs

ABSTRACT Ozone decomposition kinetics varies significantly in the presence of biological commodity. This variation depends on many factors but primarily on the active sites in agricultural commodities and storage environments. Ozone decomposition kinetics in onion bulbs at temperatures (2, 10, and 18°C) and relative humidity (RH) (35 & 85%) for different ozone concentrations (100, 200, 300 ppm) and number of exposures (1, 2, 3) were investigated. Zero-order, first-order, and second-order kinetic models were fitted to the ozone decomposition data with respect to time. It was found that the decomposition kinetics follows a first-order reaction and the decomposition rate constant ($${{\rm{k}}_{\rm{d}}}$$kd) varies from 1.712 × 10−3 s−1 − 5.181 × 10−3 s−1 in the presence of onion. The half-life period ($${{\rm{t}}_{1/2}}$$t1/2) was found to increase with concentration and number of exposures and decrease with temperature and relative humidity. The activation energy was found to vary between 1.840 and 10.216 kJ mole−1 and 6.271–12.461 kJ mole−1 for the relative humidity of 85% and 35%, respectively.


Introduction
Onion, a semi-perishable horticultural crop, has vast economic significance in the world; it is required to be stored for 3-6 months, during which disinfection measures are to be taken up periodically.In a storage atmosphere, applying ozone at a proper concentration has proven to protect fruits and vegetables from several diseases with minimum physiological damage (Nadas, Olmo and Garcia 2003).Thus, an ozone fumigation system is required to be developed for commercial onion storage structures.
The ozone decomposition rate changes with surfaceactive sites present (Li, Ma and He 2020).The effect of ozone on commodities may be affected (directly and indirectly) by the decomposition rate, concentration, and frequency of ozone application (O'donnell et al. 2012).Nevertheless, to design an efficient ozonation system, it is imperative to determine the kinetics of ozone decomposition (Pandiselvam, Thirupathi and Anandakumar 2015).Sirohi et al. (2021) have also urged the current scientific community to study the (temporal) decomposition kinetics of ozone for different agricultural commodities that are of critical relevance (as ozone is a highly unstable gas) to resolve the issue of deciding suitable treatment conditions while considering moisture content of the commodities and half-life of ozone.Practically, no information is available on the temporal ozone decomposition kinetics in the storage of onion bulbs, a semi-perishable crop with a moisture level of 85-90% (wb), with the intended use as a fumigant.
During the fumigation, in some cases, there is a need to ensure the air-tightness of the storage chamber to increase the effectiveness of the treatment.In addition, most fumigants are dangerous to human beings when exposed directly and/or continuously, thus requiring automation for efficient management of the storage systems for agricultural commodities.Though automation is prevalent in many of the high-capacity and commercial grain storage systems for pest management, there is a lack of automated disinfection systems for fruit and vegetable storage systems.Thus, with detailed knowledge of the dissociation rate/temporal dissociation of the fumigant, fumigant flow behavior/dynamics, half-life period, and activation energy, an automated noninvasive fumigating system can be developed.
To thoroughly disinfect the commodities, it is essential to distribute ozone gas evenly throughout the structure before decomposing.Otherwise, the decomposition of ozone will occur before reaching its target.Therefore, the information on the temporal dissociation of ozone plays a significant role, which helps in finding the time required for complete dissociation and the decomposition constant.
Considering that ozone can be one of the potential disinfection measures, a detailed study needs to be taken to understand the decomposition behavior of gaseous ozone in the presence of onion bulbs.Therefore, a study was envisaged to determine the decomposition reaction kinetics of gaseous ozone in onion under different environmental conditions.

Materials and methods
Yellow onion bulbs harvested during the Rabi season in March 2019 were procured from the local market, Bhopal, India.The onions were sorted manually based on size and external injury.

Experimental setup and procedure
The experimental setup consists of an oxygen concentrator (Model: OZ-AIR HG 5, oxygen purity: >93%, creative Oz-Air (I) Pvt., Ltd., Noida, India) to concentrate oxygen and an ozone generator (Model: A-10, Southern Electronics Pvt. Ltd., Bengaluru, India) to generate ozone having capacity of 10 g h −1 .A voltage stabilizer unit (Servo Stabilizer, Remi Instruments Ltd., Mumbai) was used to stabilize the power supply.Oxygen (O 2 ) molecules split apart when they pass through a high-energy electric field and form very active atomic oxygen radicals.These radicals combine with intact O 2 molecules, forming O 3 (Equations 1 and 2).
For the study of ozone decomposition reaction kinetics, 0.50 kg of onion samples kept in glass jars (3 L volume) were stored in a walk-in cold room (Blue Star, 1-tonne capacity) at different temperatures (viz. 2 ± 1, 10 ± 1, 18 ± 1°C) and RH of 85 ± 5% (Melo et al. 2012) and 35 ± 5%, considering the fact that throughout India, annual RH varies from 85 to 35%, season and regionwise (Mukhopadhyay, Jaswal and Deshpande 2017).Glass jar along with onion bulbs were kept in a cold room for 2 h before treatment.During this period, the glass jar was open to the surrounding environment.It helped the onion bulbs come to equilibrium with the environmental condition of the cold room.
The sample was then subjected to gaseous ozone treatment until the targeted ozone concentration (100, 200, and 300 ppm) was achieved.Ozone was applied to the onion bulbs in pulse mode, i.e. number of exposures of 1, 2, and 3 for the concentrations mentioned earlier.The number of exposures indicate giving the same ozone concentration to the commodity once the previous ozone molecules have been completely dissociated.Evaluation of ozone decomposition started after reaching targeted ozone concentration.The ozone concentration was measured at an interval of 30 s using an ozone monitor.The concentration in the treatment glass jar was measured by an ozone monitor (Applied Techno System, India, Model: ATS-101 M, Range: 0-1000 PPM, operating condition: −15 to 55°C).Details of the experimental setup are given in Figure 1.The ozone decomposition rate measured in an empty glass jar of the same headspace was considered a control.

Decomposition rate constant (k d ) half-life (t 1=2 ), and activation energy (E a )
The general reaction mechanism of ozone decomposition on noble metal and transition metal oxide catalysts was described by Li, Ma and He (2020).During the ozonation process, at the initial stage, some O 3 molecules get adsorbed on the surface of the sample and dissociate into O 2 molecules and O− species through oxidation of the active sites.The left-out ozone molecules then react with the atomic oxygen species to form adsorbed peroxide (O 2 2− )/superoxide (O 2 −) species and O 2 molecules.Finally, the O 2 2− or O 2 − decomposes into oxygen and gets desorbed from the active site by a reduction reaction, leaving no harmful ions in the commodity.This chain reaction continues until the reactive species are present (Li, Ma and He 2020).Active sites present on biological commodity work as catalysts for ozone decomposition.
The kinetics of ozone decomposition was studied under static conditions.Measurement of the ozone concentration with time allowed the determination of the order (n) and the k d in accordance with the rate law equation (Equation 3) (Gurol and Singer 1982.Where r ½O 3 � -Rate of ozone decomposition, ppm s −1 ½O 3 � -Ozone concentration, ppm k d -Decomposition rate constant (unit depends on the order of reaction n) n -Order of reaction t -Time, s Zero-order, first-order, and second-order kinetics models (Table 1) were fitted to the ozone concentration data with respect to time (De Oliveira et al. 2020;Wright 2004).
The t 1=2 of ozone gives an insight into how long the reaction will proceed efficiently.Equation 4gives the generalized equation for the t 1=2 of n th order chemical reaction.
Where t 1=2 -Half-life period, s Temperature and the presence of any catalyst play a vital role in determining the reaction rate.This can be best explained by the E a concept.The reaction rate can be increased by increasing the temperature or adding a catalyst.Chemical theory predicting the relationship between the reaction rate and the temperature is as follows (Equation 5), Where T -Temperature, K E a -Activation energy, J mol −1 R -Universal gas constant (8.314J mol-K −1 ) m -Power term constant Since the exponential term is more sensitive than the power term, Equation 5 can be simplified to Equation 6.Thus, after finding the concentration dependency of the reaction rate, the temperature dependency can be expressed by an Arrhenius-type relationship (Jung et al. 2017).
Where, k 0 -Initial k d , s −1 If the k d value at different temperatures for any particular chemical reaction is known, then the E a can be calculated by plotting k d versus 1=T.

Prediction of K d
The regression equation approximates the true relationship between the dependent and independent variables and does not indicate the cause-effect relation effect of these variables.To assess the effect of the independent parameters on the dependent one, the multiple linear regression model (MLR) was developed using MATLAB statistical toolbox (MATLAB, version R2018a, MathWorks, Inc., USA).

Statistical analysis
The decomposition rate was calculated by plotting the concentration (ppm) with respect to time; the regression coefficients (k d ) were obtained from the least square method of analysis in Microsoft excel.The experiments were replicated three times, and the average data was used for representation.The coefficient of determination (R 2 ) was the main criterion, whereas the goodness of fit was also determined using statistical parameters chi-square (χ 2 ) (Equations 7) and root mean square error (RMSE) (Equations 8).
The mean relative percentage deviation modulus (E) (Equation 9) gives an idea of the mean divergence of the predicted data from the experimental data.

Decomposition rate constant (k d )
Ozone decomposition can occur indirectly, i.e., production of free radicals or direct, i.e. selective reaction with substance.The general equation is given as follows: Where, m, n, o, p, q, r are the number of molecules involved in the chemical reactions.CHO is the carbon, hydrogen, and oxygen component of any food materials, whereas X is any other non-hydrocarbon element participating in the chemical reaction, a1, a2, a3, a4 are the atoms involved in the chemical reaction.
Environmental humidity is observed to play a pivotal role in the ozone decomposition rate.The decrease in ozone concentration as a function of time for 100 ppm at 85% RH and 35% RH to onion is shown in Figures 2  and 3, respectively.It was observed that with an increase in humidity, k d value increases (p < 0.05).The k d increased from 2.017-3.233× 10 −3 s −1 to 3.271-5.181× 10 −3 s −1 for onion bulbs with an increase in humidity from 35% to 85%.(p < 0.05) (Table 2).The above phenomenon can be explained by the fact that the water molecules provide a reactive base for ozone decomposition, thus increases its reactivity (Strait 1998), which explicates a higher decomposition rate of ozone at higher RH (85%) in onion bulbs.Ozone molecules are highly soluble in moisture, resulting in their rapid disintegration into OH− radicals by an indirect reaction pathway.It is a very unstable molecule with a stronger oxidation potential than ozone, which immediately reacts with another molecule/active site (O'donnell et al. 2012;Jung, Kim and Choi 2004).Similar trends were also observed for all temperatures, number of exposures, and concentrations of 200 and 300 ppm (Figures 2 and Figure 3; Table 2).
The temporal decomposition of ozone gas (100 ppm) at different temperatures and exposures to onion and control is depicted in Figure 2 (85% RH) and Figure 3 (35% RH).With an increase in the temperature from 2°C to 18°C, the k d increased from 4.055-3.271× 10 −3 s −1 to 5.181-3.653× 10 −3 s −1 and 2.397-2.017× 10 −3 s −1 to 3.233-2.463× 10 −3 s −1 for onion at RH of 85% and 35%, respectively.At lower temperatures, the ozone molecules move slowly, which results in a lower number of collisions.Also, there will be less chance of a reaction occurring, as the momentum energy in the ozone molecules is less.The kinetic energy of molecules increases with temperature, which causes them to move faster, and thus the likelihood of collisions per unit time increases.This indicates that when a collision occurs, there is more probability that the collision will lead to a reaction (Hardin et al. 2010;Csefalvay, Nothe and Mizsey 2007).Similar trends were also observed for all the number of exposures and concentrations of 200 and 300 ppm (Figures 2 and 3; Table 2).
A decrease in the decomposition rate with several ozone exposures of the same concentration in the commodity was observed; the trend follows a linear relation.The results show that k d value decreased with an increase in exposures (p < 0.05) (Table 2).The k d for 200 ppm concentration at the number of exposures 1 to 3 at 85% of RH decreased from 2.880 × 10 −3 s −1 to 2.421 × 10 −3 s −1 at 2°C temperature.For ozone decomposition, the surface area, and active sites in the commodity play significant roles.As the active sites of the commodity inside the containers remain constant, there comes stagnation in the decomposition rate because the reactive components present on the surface of onion bulbs get used progressively with time.The initial ozone decomposition rate becomes higher in the case of first exposure, but with multiple numbers of exposures, the ozone consumption gets stabilized over time,   and the decomposition rate decreases.Besides, during first exposure, ozone decomposition might also have been affected by environmental factors inside the container; however, during subsequent exposures, these effects are diminished (Shelake et al., 2022b), resulting in a decrease in the decomposition rate.This result corroborates the findings of Kells et al. (2001) for maize, where during the first phase, inherent sites on maize reacted with ozone and were eliminated, resulting in a decrease in the degradation rate in the second phase.Similar trends were also observed for all temperatures and concentrations (Figures 2 and 3; Table 2).The decomposition rate was lower at higher concentrations of ozone (p < 0.05) (Figures 2 and 3).For the same concentration and temperature, the k d is found to be higher in the container with a commodity than that for the control (Table 2).The k d of ozone concentration for 100, 200, and 300 ppm at temperature 2 ºC for first exposure ranges between 4.054 × 10 −3 and 2.876 × 10 −3 s −1 for onion (Table 2) at 85% RH.The decomposition rate in the case of a commodity affected by the reactive components present on the surface of the onion may include chemical compounds of onion peel and flesh, surface characteristics of the onion, and the microorganisms present on the commodity.These reactive components (i.e.active sites) are readily oxidized by ozone, which has been observed through the change in the bioactive compounds, such as pyruvic acid, ascorbic acid, phenols, flavonoids, anthocyanins, etc., as well as structural modifications in onion and the microbes present on its surface, as observed from our earlier work Shelake et al. (2022bShelake et al. ( , 2022a)).Due to ozonation, the outer portion of the bulb gets exposed to oxidative stress, which results in membrane cracks in the epicuticular layer of the onion bulbs (Shelake et al. 2022a).The reaction of ozone with the commodity results in a shift in the primary pathway of ozone decomposition to surface-catalyzed decomposition, thus increasing the k d value (Hardin et al. 2010;Kells et al. 2001).In contrast, auto decomposition of ozone gas takes place in the control resulting in a lower rate of ozone dissociation.The effect of humidity, wall effect, and other catalytic effects may be the contributing factors to ozone degradation.Similar trends were also observed for all temperatures and concentrations (Figures 2 and 3; Table 2).

3.2Half-life (t 1=2 ), and activation energy (E a )
The calculated t 1=2 is shown in Table 2, and that for activation energy is presented in Table 3.With an increase in temperature and humidity, the decomposition rate increased, which resulted in a decreased t 1=2 of ozone (Table 2).At a concentration of 100 ppm (first exposure) and 85% RH, for an increase in the temperature from 2 to 18 ºC, the t 1=2 decreased from 171 s to 134 s for onion.With an increase in relative humidity from 35 to 85% for the concentration of 100 ppm (first exposure), the half-life period in the presence of onion bulbs decreased from 289 s to 171 s for onion at 2 ºC temperature.Similar trends were also observed for all temperatures and concentrations.It was also found that the t 1=2 increased with higher ozone concentration, and the number of exposures as the decomposition rate decreased.With an increase in concentration from 100 ppm to 300 ppm at 2 ºC temperature (first exposure), the half-life of ozone gas increased from 171 s to 241 s in the presence of onion bulbs.For 100 ppm concentration, with an increase in the number of exposures from 1 to 3, the half-life of ozone increased from 171 s to 212  s (at 85% RH and 2°C temperature).The t 1=2 of ozone is not only dependent on its chemically unstable nature but also affected by other catalytic factors like temperature, available active sites, and RH.The interaction of ozone with active sites available on the commodity surface, combined with environmental factors, resulted in a decrease in the half-life of ozone (Dos Santos et al. 2007).The reduction in the half-life period of ozone in the presence of a commodity can be related to the ozone decomposition rate, which has been discussed earlier.
It was observed that at 35% RH has a higher E a as compared to 85% (Table 3).The E a value was found to be 10.216-1.840kJ mole −1 and 12.461-6.271kJ mole −1 for RH of 85% and 35%, respectively.This may be due to lower metabolic activities at lower humidity evolving lower amount of energy, thus requiring higher E a for the reaction to take place and vice versa.It was also observed that at the same level of relative humidity, the E a decreased with an increase in the concentration and number of exposures.Since the k d observed to be directly proportional to the exponent of E a =RT, with an increase in concentration or by increasing the number of exposures, the number of particles will increase, resulting in frequent collisions, thereby decreasing the E a .

Kinetics of ozone decomposition
The R 2 , χ 2 , RMSE, and E values are given in Table 2 for the first-order reaction.Since R 2 values are close to 1 (0.951-0.998) and χ 2 (0.004-0.096),RMSE (0.066-0.297) values are close to 0 and E << 10% (0.037-6.369%); it can be inferred that the first-order rate reaction fits well with the experimental data.It was observed that the decomposition of ozone gas occurred in two distinct phases.During the first phase, the ozone decomposition rate was higher.A similar phenomenon has been witnessed by Dos Santos et al. (2007), who reported that the decomposition rate slows down after the saturation of active commodity sites.This explains the concentration dependency of reaction kinetics, making it a suitable fit for a first-order reaction.Further, the observations show that the order of reaction does not change with either concentration or number of exposures.The k d for R 2 , χ 2 , RMSE, and E values for zeroorder and second-order are provided in supplementary files (Table S1; Table S2).

Prediction of K d and temporal decomposition concentration
An empirical model was developed using multiple linear regression (MLR) for k d with respect to RH (R), temperature (T), concentration (C), and the number of exposures (N) (Equation 8).For the development of an empirical model, in addition to the aforementioned experiments, another set of experiments were conducted at a temperature of 34 ± 1 ºC and RH of 44 ± 5% at ozone concentrations of 100, 200, and 300 ppm in pulse mode (i.e.number of exposures of 1, 2, and 3).
By putting the value of k d in the first-order equation (equation 8), temporal decomposition concentration can be predicted.
Predicting the temporal concentration of ozone gas inside the storage chamber with time aids the user in accessing the decomposition time and duration for which the fumigating chamber needs to be undisturbed to achieve maximum fumigation efficiency.Furthermore, knowing the concentration below a threshold level helps restart the fumigation facility when applying ozone in pulsed mode.These results minimize man-fumigant contact, thus preventing health/environmental hazards.Additionally, this automatic prediction removed the need for installing multiple sensors at every storage structure, thereby reducing the cost of the storage facility.Using this model (equation ( 14)), we can work out the temporal concentration profile with changes in environmental conditions (i.e.RH and temperature), and treatment conditions (i.e.ozone concentration and number of exposures).

Validation of the model
The graph of experimental versus predicted temporal ozone decomposition is shown in Figure 4.The R 2 and RMSE values are observed to be 0. 981 & 0.196; 0.984 & 0.316 and 0.982 & 0.258 for the first condition (RH -45%, Temperature -6ºC, ozone concentration -100 ppm and Exposure 1); second condition (RH --60%, Temperature -15ºC, ozone concentration -150 ppm and Exposure 2) and third condition (RH -80%, Temperature -28ºC, ozone concentration -200 ppm and Exposure 3), respectively.This shows that for both k d and temporal concentration profile, predictability is good.

Conclusions
The ozone kinetics study was conducted under controlled environmental conditions in the presence of onion bulbs.The ozone decomposition kinetics in the presence of semi-perishable crops like onion bulbs followed the firstorder reaction.The decomposition rate is directly proportional to the temperature and relative humidity and inversely proportional to the concentration and number of exposures.With an increase in the number of exposures, the effect of other environmental factors would decrease, which may increase the effect of ozone on the microorganism.So, applying ozone gas in pulsed form would be a better practice.The reaction mechanism of ozone with particular agricultural commodities can be used to find the effect of ozone on the target commodity, to help optimize the design parameters of the ozone applicator, and to develop an ozone distribution system for its uniform and effective application.

Figure 2 .
Figure 2. Ozone concentration as a function of time for different treatment conditions at 85% RH (a) Exposure 1, (b) Exposure 2, (c) Exposure 3, and (d) Control.

Figure 3 .
Figure 3. Ozone concentration as a function of time for different treatment conditions at 35% RH (a) Exposure 1, (b) Exposure 2, (c) Exposure 3, and (d) Control.

Table 1 .
Kinetic models of decomposition.

Table 2 .
Decomposition rate constant (k d ) and half-life period (t 1=2 ) for first-order reaction.

Table 2 .
(Continued).The experiments were replicated three times and the average data was used for calculation of k d from the least square method of analysis.

Table 3 .
Activation energy (kJ mole −1 ) at different ozone concentrations and exposures.The experiments were replicated three times and the average data was used for calculation of k d .This k d was used for calculation of activation energy.