Kinetics of Heterogeneous Reaction of Ozone with Oleic Acid and Its Dependence on Droplet Size, Relative Humidity, and Ozone Concentration

Abstract

In this study, the heterogeneous reaction of oleic acid droplets with gas-phase ozone was studied by an ATR-FTIR flow reactor. The effects of droplet size, relative humidity, and ozone concentration on the reaction kinetics were carefully investigated. Specifically, the pseudo-first-order rate constant k_app and the uptake coefficient γ displayed a size dependence, with k_app decreasing from ~4.5 × 10⁻³ to ~3.2 × 10⁻³ and γ linearly increasing from ~4.4 × 10⁻⁵ to ~3.2 × 10⁻⁴ as the suspended droplet diameter increased from 0.1 to 1.0 μm. It is believed that the reaction kinetics were the major contributor to the reactive uptake in the reaction between the oleic acid droplets and gas-phase ozone observed in this study. In addition, RH showed no obvious influence on the heterogeneous reaction kinetics, in agreement with findings from previous studies. Furthermore, the k_app was found to display a Langmuir–Hinshelwood dependence on the gas-phase ozone concentration with K_O3 = (3.29 ± 0.46) × 10⁻¹⁵ molecules cm⁻³ and k[S] = 0.153 ± 0.007 s⁻¹, which is consistent with observations of the ozonolysis of unsaturated organic materials in the literature. Kinetics data related to the heterogeneous reaction of ozone and oleic acid under different conditions could be used in chemistry transport models and air quality models to better understand air pollutants’ adverse health impacts.

1. Introduction

Atmospheric aerosols are known to be an important component of the climate system. They are linked to various environmental issues ranging from air pollution, visibility, and human health to the Earth’s radiation balance, hydrological cycle, and climate change [1]. Aerosol-associated organic matter has been confirmed to be ubiquitous in the atmosphere, and contributes 20–60% of the total aerosol mass at the continental mid-latitudes and as high as 90% in tropical forest areas [2,3,4]. Due to its abundance as well as its diverse impacts on aerosol particles, such as modifying the hygroscopicity and cloud condensation nuclei (CCN) activity of inorganic aerosols, and changing the chemical compositions of aerosol particles continuously via heterogeneous reactions [4,5,6], aerosol-associated organic matter has received growing attention recently.

There are all kinds of organic compounds (e.g., carboxylic acids, alcohols, esters, and hydrocarbons) in atmospheric aerosols, which enter the atmosphere through various processes, including the combustion of fossil fuels, burning of biomasses, meat cooking, and the decay of plants, animals, and microbes [4,7]. Among all organic species, carboxylic acids are believed to comprise a significant fraction of the organic mass [8,9]. For example, oxalic acid, the smallest dicarboxylic acid and usually found to be the most abundant carboxylic acid in the ambient atmosphere, is present with concentrations within a low μg/m [3] range [10]. Sources of carboxylic acids comprise biogenic and anthropogenic emissions, which are also known to be the primary sources, and the chemical transformation of their precursors, that is, secondary sources [11]. Since carboxylic acids have a relatively high hydrophilicity, they may significantly modify the hygroscopic properties and cloud condensation nuclei (CCN) activity of aerosol particles, as well as their chemical composition [9]. Furthermore, reactions between unsaturated carboxylic acids and atmospheric oxidants, such as O₃, OH and NO₃ radicals, could lead to much more complicated effects on the physiochemical and radiative properties of aerosol particles [4]. These effects, however, are not yet well understood.

Oleic acid is an 18-carbon monounsaturated fatty acid that largely exists in edible oils [12]. It has often been utilized as a model aerosol system to better understand the heterogeneous oxidation process of unsaturated carboxylic acids and their environmental consequences [4]. Although the heterogeneous oxidation of oleic acid by O₃ has been extensively studied with respect to its kinetics, mechanism, and product identification [4,5,6,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28], kinetic data such as the uptake coefficient γ obtained from different experimental approaches sometimes differ by over one order of magnitude. In addition, although some reports on effects of temperature, physical state, and component mixing [6,13,27], how droplet size and relative humidity (RH) affect the ozone-initiated reaction kinetics have not been well explored. Furthermore, many heterogeneous reactions in the atmosphere are in accordance with Langmuir–Hinshelwood behavior [28,29,30,31]; to date, the dependence of the oxidation kinetics of oleic acid on O₃ concentration has not been clearly explained.

Lately, there has been growing interest in indoor air pollution sources such as household cooking and cleaning. Household cooking and cleaning generate a variety of organic compounds including volatile organic compounds (VOCs) and less volatile organic compounds (or semi-volatile organic compounds, SVOCs) which can react with indoor oxidants, including OH and O₃, to produce harmful secondary pollutants, consequently negatively impacting indoor air quality [32,33,34,35,36]. There are three major pathways for human exposure to pollutants: inhalation; dermal contact; and hand-to-mouth transfer. Recent studies have indicated that products from the ozonolysis of unsaturated organic compounds such as squalene, a major component of skin lipids, in an indoor environment can induce respiratory or skin irritation responses upon exposure [36,37,38,39,40]. Reaction kinetics data on oleic acid related to droplet size, ambient RH, and oxidant concentration could be utilized in chemistry transport models and air quality models for evaluating health risks for human beings in an indoor environment.

2. Experimental Section

2.1. Preparation of Oleic Acid Droplets with Uniform Sizes

Oleic acid droplets with uniform sizes were prepared using a micro-orifice uniform deposit impactor (Model 110, TSI Corporation, Shoreview, MN, USA), which has 10 different stages and can obtain uniform droplets with different sizes. Oleic acid (≥99% purity, Sigma, St. Louis, MO, USA) was first dissolved into ethanol (99% purity, the Science Company, Lakewood, CO, USA) to make a solution with a concentration of 0.05 M. A small amount of this solution (~20 mL) was then introduced into a nebulizer. By applying a flow of dry air (~5 L/min) to the nebulizer, droplets of the solution were generated. These non-uniform droplets then passed through a molecular sieve pipe, in which the ethanol solvent was eliminated and pure oleic acid droplets were formed. Finally, these oleic acid droplets were carried by the airflow to the micro-orifice uniform deposit impactor. By placing the ATR crystal onto the impactor in advance of one of the stages with a certain deposit time, uniform-sized oleic acid droplets were obtained. In this study, five different stages of the impactor were selected, in which uniform-sized oleic acid droplets with suspended diameters of 0.1, 0.18, 0.32, 0.56, and 1.0 μm were obtained.

2.2. Measurements for Droplet Morphologies

In this work, a Non-Contact Surface Profiler (Wyko NT2000, Veeco Metrology Group, Plainview, NY, USA) was used to measure the morphologies of the oleic acid droplets, which were deposited on the ATR crystal with different suspended diameters (0.1, 0.18, 0.32, 0.56 and 1.0 μm) with a micro-orifice uniform deposit impactor.

The optical surface profiler system, which is based on white light interferometry technology, enables precise 3-dimensional analyses of surfaces. It can measure heights from 0.1 nm to several millimeters, with a vertical resolution as low as 0.1 nm. It is equipped with both phase-shifting interferometry (PSI) and vertical scanning interferometry (VSI) modes. The phase-shifting interferometry (PSI) mode allows fairly smooth and continuous surfaces to be measured (0.1 nm < heights < 160 nm), while the vertical scanning interferometry (VSI) mode can measure rough surfaces and heights ranging between 160 nm and 2 mm. By switching between the modes in the software and by choosing the appropriate measurement optics, the instrument can accommodate a wide range of surface roughness measurements. In the present work, as the droplets studied here were all of micron size, the VSI mode was used. Detailed dimensional information including the radii and the heights of the oleic acid droplets was obtained.

2.3. Apparatus and Conditions for the IR Measurements

Figure 1 shows a schematic diagram of the experimental setup. Details about the experimental setup have been discussed elsewhere [41]. The heterogeneous reactions of oleic acid droplets with gas phase O₃ under different conditions were monitored in real-time by an ATR-FTIR spectrometer (Nicolet 6700, Thermo Scientific, Waltham, MA, USA) equipped with a liquid-nitrogen-cooled MCT detector. A ZnSe ATR crystal (refractive index: 2.4, 5 × 1 cm²), with a penetration depth of approximately 1.1 µm at 1700 cm⁻¹, was placed into a flow-through stainless steel chamber and employed as a substrate. A gas mixture, which was composed of O₃, H₂O, and air, was used for the reaction. The air was used to provide oxygen for ozone generation through a UV light source (Pen-Ray^®, Wheeling, IL, USA), to serve as a carrier gas for moisture generation through a dew point RH generator (V-Gen, INSTRUQUEST, Boca Raton, FL, USA), and to dilute the ozone and water vapor concentration via flow rates control by mass controllers. By adjusting the flow rates of the two air flows, the O₃ concentration (~350ppb) could be controlled and measured downstream by an O₃ monitor (InDevR 2B Technologies, Model 106-L, Boulder, CO, USA). The RH effects were studied over a wide range of 0–80%. For the humidification experiments, the dry air flow was switched to a different path through a dew point generator (V-gen, Instruquest, Boca Raton, FL, USA) and the RH in the flow reactor was controlled by adjusting dew point temperature. The setup allowed the O₃ and H₂O concentrations to be varied independently.

Before each reaction, the gas flow with O₃ was switched to the bypass instead of the ATR chamber. After taking a background spectrum of the clean ZnSe crystal, the ATR crystal was taken off the reaction chamber and placed onto the micro-orifice uniform deposit impactor during one of the stages to obtain oleic acid droplets with a certain uniform size. After that, the crystal was replaced in the reaction chamber, which was then sealed. The spectral measurements were started simultaneously by converting the O₃ flow to the ATR chamber. For the thin film experiments, a 3 mM solution was pre-prepared by dissolving oleic acid in carbon tetrachloride (99% purity, Aldrich, St. Louis, MO, USA). The thin film was made by dispensing 100 μL oleic acid solution directly onto the ZnSe ATR crystal using a micropipette. The amount of oleic acid and carbon tetrachloride mixture was sufficient to cover the whole ATR crystal surface completely and the thickness of the thin film was about 0.2 µm, much smaller than the penetration depth of the evanescent wave so that bulk of the film could be fully probed by the IR beam. An additional airflow was used to speed up the complete removal of carbon tetrachloride by evaporation, which can be confirmed by the disappearance of the C-Cl vibration mode in the IR spectra. In all experiments, IR spectra were recorded in the range of 650–4000 cm⁻¹ with a resolution of 2 cm⁻¹ and an average of 4–16 scans. For each reaction condition, experiments were repeated at least 3 times, during which standard deviations were determined.

3. Results and Discussion

3.1. Rate Constant and Uptake Coefficient

The method for processing a reaction as a pseudo-first-order reaction has been discussed in depth elsewhere [13,40]. In brief, for a given reaction, such as A + B → P (P = product), when A is present in great excess over B for pseudo-first-order conditions, the rate equation can be written as

$${\displaystyle \frac{\mathrm{d}[\mathrm{B}]}{\mathrm{d}t}}=k_{app}[\mathrm{B}], \mathrm{where} k_{app}=k\left[A\right]$$

(1)

k_app is the pseudo-first-order rate constant, and k is the second-order rate constant. In this case, the concentrations of reactant B and/or products are expected to change exponentially with time. Under our experimental conditions, the molar ratio of ozone to oleic acid was kept to at least 10.

According to the Beer–Lambert law, IR absorbance is linearly proportional to concentration. In previous studies on the oxidation kinetics of oleic acid, changes in the integrated absorbance intensity of the IR band located at ~1743 cm⁻¹, which is assigned to the C=O stretching of esters, was often used to describe changes in the concentrations of the reaction products, while the integrated intensity variations in another band nearby at ~1710 cm⁻¹, which is mainly attributed to the C=O stretching of carboxylic acids, were usually used to describe the concentration variation in the reactant, that is, the variation in the oleic acid concentration [5]. Although some reaction products such as ketones and aldehydes also contribute to the 1710 cm⁻¹ band, the behavior of oleic acid dominates the changes in this C=O stretching band, for the reason that the oscillation strength of carboxylic acid C=O groups is 100-fold greater than that of ketones and aldehydes [41].

In this work, we also utilized the 1743 and 1710 cm⁻¹ bands to investigate the kinetics of the reaction between oleic acid droplets and O₃. Figure 2 shows the spectral evolutions of the two bands with respect to reaction time. The 1743 cm⁻¹ band increased, whereas the 1710 cm⁻¹ band decreased with the increase in reaction time. Changes in the absorbance difference ΔA (ΔA = A_t − A_∞, where A_t and A_∞ are the integrated absorbance intensities at time t and time infinite; for simplification, the integrated absorbance intensity of the spectrum recorded at the end of each experiment was utilized as the A_∞ in this work) of these two bands with reaction time are also illustrated in Figure 2. Based on the curve fitting of the exponential changes in ΔA using a function of $\mathrm{\Delta }A=A_{\infty }{e}^{-k_{\mathrm{app}}t}$, k_app could be obtained [5]. As shown in Figure 2, the curve fittings yield an R² greater than 0.99 for both the 1743 and 1710 cm⁻¹ bands, indicating that the reactions took place under pseudo-first-order conditions. In addition, the k_app values derived from the 1743 cm⁻¹ and 1710 cm⁻¹ bands agree with each other well. For convenience of comparison, the results for k_app presented in this paper were all derived from the 1743 cm⁻¹ band unless it is indicated otherwise.

As for the ozone uptake coefficient γ, we estimated it using the following equation:

$${\displaystyle \frac{\mathrm{d}[\mathrm{OL}]}{\mathrm{d}t}}=-\gamma ({\displaystyle \frac{P_{{\mathrm{O}}_3}\overline{c}}{4RT}}){\displaystyle \frac{S_{\mathrm{A}}}{V}}$$

(2)

where d[OL]/dt = −k_app[OL], and [OL] [molecules cm⁻³] is the initial concentration of oleic acid; $\overline{c}$ [cm s⁻¹] is the mean speed of the ozone molecules in the gas phase; and S_A/V [cm⁻¹] is the surface area-to-volume ratio of the oleic acid droplets. It is worth mentioning that, as the oleic acid droplets studied in this work were deposited on the ATR crystal, the diffusion flux of O₃ was limited over a spherical cap rather than the entire droplet. Therefore, a simple correction was made by means of multiplying the γ calculated from Equation (2) by a factor of 2 to account for this effect, similar to the method used in the study of the kinetics of the heterogeneous reaction of deliquesced NaCl particles with gaseous HNO₃ [42].

3.2. Size Effect

3.2.1. Size of Oleic Acid Droplets

Figure 3 shows the sizes of the oleic acid droplets deposited on the ATR crystal with a suspended diameter of 0.56 μm. As seen, the size distribution has a good homogeneity. In addition, these droplets were found to be in the shape of spherical caps. Taking advantage of the information about the dimensions of the droplets, which was acquired from the morphology measurements, the surface area S_A and volume V of the droplets could be precisely estimated using the following equations:

S_A = 2πRh

(3)

V = πh²(3R − h)/3,

(4)

where R = (r² + h²)/2h and r and h are the observed radius and height of the droplet, respectively [13]. The observed r and h as well as the estimated S_A/V ratio for oleic acid droplets with different suspended diameters are listed in

Table 1. Summary of pseudo-first-order rate constants, k_app, and corresponding uptake coefficients, γ.

Suspended					[O3]	kapp (s−1) ×10−3		γ × 10−4
Diameter (μm)	r (μm)	h (μm)	SA/V (cm−1) × 105	RH (%)	Molecules/cm3	1743 cm−1	1710 cm−1	1743 cm−1	1710 cm−1
0.1	0.08 ± 0.02	0.05 ± 0.01	4.9 ± 1.5	0	8.8 × 1012	4.5 ± 0.1	4.6 ± 0.6	4.5 ± 0.1	4.6 ± 0.5
0.18	0.14 ± 0.03	0.09 ± 0.02	2.8 ± 0.8	0	8.8 × 1012	3.9 ± 0.1	4.0 ± 0.4	6.9 ± 0.3	7.1 ± 0.6
0.32	0.27 ± 0.05	0.14 ± 0.03	1.7 ± 0.5	0	8.8 × 1012	3.7 ± 0.1	3.8 ± 0.2	10.9 ± 0.3	1.2 ± 0.7
0.56	0.5 ± 0.1	0.22 ± 0.04	1.0 ± 0.3	0	1.3 × 1012	0.6 ± 0.1	0.7 ± 0.1	21.8 ± 2.1	22.5 ± 1.7
				0	6.3 × 1012	2.5 ± 0.3	2.6 ± 0.3	17.1 ± 1.8	17.5 ± 2.3
				40	6.3 × 1012	2.2 ± 0.2	2.4 ± 0.3	15.2 ± 1.5	16.3 ± 2.1
				80	6.3 × 1012	2.9 ± 0.4	2.4 ± 0.2	20.0 ± 2.5	16.0 ± 1.3
				0	8.8 × 1012	3.3 ± 0.1	3.5 ± 0.2	16.1 ± 0.5	17.1 ± 1.2
				0	2.5 × 1013	8.5 ± 0.3	9.1 ± 0.2	14.3 ± 0.5	15.4 ± 0.4
				0	1.3 × 1014	43.8 ± 1.6	41.9 ± 1.0	14.3 ± 0.5	13.7 ± 0.3
				0	6.3 × 1014	109 ± 25	99 ± 21	7.4 ± 1.7	6.7 ± 1.4
				0	1.4 × 1015	122 ± 10	109 ± 9	3.8 ± 0.3	3.4 ± 0.3
1.0	1.0 ± 0.2	0.31 ± 0.06	0.7 ± 0.2	0	8.8 × 1012	3.4 ± 0.2	3.5 ± 0.2	24.7 ± 1.3	25.2 ± 1.5

. Table 1 also summarizes the k_app and γ derived from the two C=O stretching bands (1743 and 1710 cm⁻¹) under different conditions.

3.2.2. Dependences of k_app and γ on Droplet Size

Oleic acid droplets with suspended diameters of 0.1, 0.18, 0.32, 0.56, and 1.0 μm (corresponding to observed radii of 0.08, 0.14, 0.27, 0.51 and 1.03 μm) were selected to investigate the effect of the kinetics on size under the conditions of [O₃] = ~350 ppb and T = 293 K. The plots of k_app and γ versus the observed droplet radius presented in Figure 4 demonstrate a distinct size dependence in the kinetics for oleic acid ozonolysis. In fact, within the size range we studied here, k_app decreased from ~4.5 × 10⁻³ to ~3.4 × 10⁻³, whereas γ increased by more than 5 fold from ~4.5 × 10⁻⁴ to ~2.5 × 10⁻³ as the droplet size increased. In addition, both k_app and γ seemed to flatten out gradually when the observed droplet radius increased to 0.51 μm. Data from the literature have shown that the variation in γ with particle size shows a characteristic rise-then-fall behavior [42]. In the size region in which γ increases, the reaction kinetics are the major contributor to the reactive uptake, while in the size region in which γ decreases, diffusion in both the gas phase and liquid phase becomes a limiting factor for the reactive uptake. It is interesting to note that the variation in γ with oleic acid droplet size in the work of Smith et al. showed a linear decrease with droplet diameters ranging from 1.36 to 4.90 μm [43]. Their explanation for the size dependence of γ was that the oleic acid diffusion within the particle limited the reaction. Thus, the results for the size dependence of the kinetics obtained in the present study are in good accordance with those of Smith et al., in that our data fall in the kinetic-controlled region.

In

Table 2. Summary of kinetics data for heterogeneous reaction between ozone and oleic acid.

		Oleic Acid Dimension	[O3]
Refs.	Sample Type	(Diameter or Thickness/μm)	(Molecules/cm−3)	γ × 10−4
[21]	suspended droplet	0.2–0.6	2.5 × 1012	16 ± 2
[28]	thin film	103	~1010	8.3 ± 0.2
[23]	suspended droplet	0.6–1.0	2.5 × 1015	7.5 ± 1.2
[19]	thin film	~50	1011–1012	8.0 ± 1.0
[25]	suspended droplet	polydisperse, submicron	7.0 × 1013	13 ± 2
[20]	thin film	103	3.0 × 1012	7.9 ± 0.3
[24]	suspended droplet	0.6–1.2	2.5 × 1014–2.5 × 1015	13.8 ± 0.6
[27]	suspended droplet	2	5.9 × 1014	3.4 ± 0.3
[13]	deposited droplet	10	1.1 × 1013	34.3 ± 11.4
this work	deposited droplet	0.16-2.01	8.8 × 1012	(4.5 ± 0.1)–(24.7 ± 1.3)
	thin film	0.2	8.8 × 1012	8.9 ± 0.6

, we also compare γ values obtained under the condition of T = 293 K, [O₃] ≈ 350 ppb in this work with those from previous studies under room temperature conditions and utilizing different experimental approaches. The oleic acid thin film data were obtained using the same method as our previous work [41]. It shows that the values of γ in previous studies range from (3.4 ± 0.3) × 10⁻⁴ to (34.3 ± 11.4) × 10⁻⁴, whereas the values of γ obtained in this work vary from (4.5 ± 0.1) × 10⁻⁴ to (24.7 ± 1.3) × 10⁻⁴ with increasing droplet size. Therefore, besides the obvious size dependence of γ we observed in this work, the γ values in our study are also consistent with previous studies regardless of the experimental methods.

3.3. RH Effect

The effect of RH on the kinetics of the reaction was investigated over a wide RH range from 0 to 80% with ozone concentration ~250 ppb and temperature 293 K. Uniform-sized oleic acid droplets with a suspended diameter of 0.56 μm (observed radius: 0.51 μm) were chosen for this study. Figure 5 displays the observed pseudo-first-order reaction rate constants k_app and the O₃ uptake coefficients γ for oleic acid droplets at three different RH (0, 40 and 80%). No significant variations in both k_app and γ were observed with increasing RH, indicating that water vapor has little influence on the kinetics of oleic acid ozonolysis. Similar results have been reported in heterogeneous ozonolysis reactions of cypermethrin and squalene thin films [29,31].

3.4. O₃ Concentration Effect

The observed k_app and corresponding γ values at various O₃ concentrations (10¹²–10¹⁵ molecules/cm³) are plotted in Figure 6a (note the log-log scale) and b. The k_app values were determined for oleic acid droplets with a suspended diameter of 0.56 μm (observed radius: 0.51 μm) at 293 K and 0% RH. The non-linear plot in Figure 6a shows a shape in accordance with the Langmiur–Hinshelwood reaction mechanism. Figure 6a shows the curve fitting plot using the following equation:

$$k_{\mathrm{app}}={\displaystyle \frac{k[\mathit{S}]K_{{\mathrm{O}}_3}{[{\mathrm{O}}_3]}_{\mathrm{gas}}}{1+K_{{\mathrm{O}}_3}{[{\mathrm{O}}_3]}_{\mathrm{gas}}}}$$

(5)

where k_app [s⁻¹] is the pseudo-first-order reaction rate constant, k [cm² s⁻¹ molecule⁻¹] is the second order reaction rate constant, [S] [molecules cm⁻²] is the total surface density of O₃ adsorption sites, K_O3 [cm³ molecule⁻¹] is the ratio between the O₃ adsorption and desorption rate coefficients, and [O₃]_gas is the gas-phase O₃ concentration. As illustrated in the Figure 6, the fitting curve (solid black line) is in agreement with the observed data for k_app with an R² of 0.996. This indicates that the reaction of ozone and oleic acid droplets involves two processes: firstly, ozone molecules are adsorbed on the droplet surface and quickly reach an equilibrium with the gas phase; then, the adsorbed ozone molecules react with oleic acid on the surface at a relatively slower rate.

It is reported that surface kinetics show a high dependence on the chemical and physical properties of the substrate [44,45,46], and the K_O3 values for different systems may range over several orders of magnitude [30]. In this work, the k[S] and K_O3 derived from the curve fitting in Figure 6a are 0.153 ± 0.007 s⁻¹ and (3.29 ± 0.46) × 10⁻¹⁵ molecules cm⁻³, respectively. The K_O3 obtained in the present work is of the same order of magnitude as those obtained previously in heterogeneous ozonolysis of anthracene film at the air–water interface [47], of benzo[a]pyrene on solid organic aerosols [48], and of squalene thin films on ZnSe crystals [31].

Furthermore, a non-linear curve fitting with a two-parameter exponential function was applied to the calculated γ using the following equation:

$$\begin{array}{l}\gamma ={\displaystyle \frac{4[\mathrm{OL}]}{\overline{c}{\mathit{S}}_{\mathrm{A}}/V}}\times {\displaystyle \frac{k[\mathit{S}]K_{{\mathrm{O}}_3}}{1+K_{{\mathrm{O}}_3}{[{\mathrm{O}}_3]}_{\mathrm{gas}}}}\end{array}$$

(6)

which was derived from Equations (2) and (6). As shown in Figure 6b, γ values decrease from 2.6 × 10⁻⁴ to 4.6 × 10⁻⁵ with an increasing ozone concentration. It is known that the uptake coefficient γ is the ratio of the number of collisions that lead to a reaction to the total number of collisions between a gas-phase molecule and a surface. As the ozone concentration in the gas phase increases, the ozone surface coverage would gradually become saturated, and the number of collisions leading to a reaction would remain unchanged, whereas the total number of collisions would increase. Therefore, the observed trend of γ decreasing with the increasing ozone concentration was consistent with the Langmuir–Hinshelwood surface reaction mechanism for saturation.

4. Conclusions

Oleic is an excellent proxy for understanding the ozone-initiated heterogeneous oxidation of unsaturated organic compounds in both outdoor and indoor environments. Kinetics data related to the heterogeneous reaction of ozone and oleic acid under different conditions could be used in chemistry transport models and air quality models to better understand air pollutants’ adverse health impact.

In this study, we used a flow reactor coupled with ATR-FTIR spectroscopy to carefully investigate the ozone-initiated heterogeneous oxidation of oleic acid droplets, aiming to elaborate on the effects of droplet size, relative humidity, and ozone concentration on reaction kinetics. Specific findings include the following: (1) the pseudo-first-order rate constant k_app and uptake coefficient γ displayed a size dependence, with k_app decreasing from ~4.5 × 10⁻³ to ~3.2 × 10⁻³ and γ linearly increasing from ~4.4 × 10⁻⁵ to ~3.2 × 10⁻⁴ as the suspended droplet diameter increased from 0.1 to 1.0 μm; (2) the reaction kinetics were the major contributor to the reactive uptake of the oleic acid droplets and ozone reaction; (3) the RH played an insignificant role in the kinetics; and (4) the k_app exhibited a Langmuir–Hinshelwood dependence on the gas-phase ozone concentration with K_O3 = (3.29 ± 0.46) × 10⁻¹⁵ molecules cm⁻³ and k[S] = 0.153 ± 0.007 s⁻¹, in agreement with previous observations on the ozonolysis of unsaturated organic materials.