Skip to main content
NCBI home page
ACS AuthorChoice logoLink to ACS AuthorChoice
. 2022 Nov 18;6(12):2900–2909. doi: 10.1021/acsearthspacechem.2c00206

Oxidation of Phenolic Aldehydes by Ozone and Hydroxyl Radicals at the Air–Solid Interface

Md Sohel Rana 1, Marcelo I Guzman 1,*
PMCID: PMC9762487  PMID: 36561198

Abstract

graphic file with name sp2c00206_0009.jpg

Biomass burning emissions contain abundant phenolic aldehydes (e.g., syringaldehyde, vanillin, and 4-hydroxybenaldehyde) that are oxidized during atmospheric transport, altering the physicochemical properties of particulates. Herein, the oxidative processing of thin films made of syringaldehyde, vanillin, and 4-hydroxybenaldehyde is studied at the air–solid interface under a variable O3(g) molar ratio (410 ppbv–800 ppmv) and relative humidity (0–90%). Experiments monitored the absorption changes of C=C, C=O, and —COOH vibration changes during the oxidation of thin films by transmission Fourier transform infrared spectroscopy (FTIR). Selected spectroscopic features of aromatic ring cleavage by O3(g) revealed the production of carboxylic acids. Instead, monitoring O—H stretching provided a comparison of a hydroxylation channel from in situ produced hydroxyl radical. The overall oxidation reactivity trend syringaldehyde > vanillin > 4-hydroxybenzladehyde can be explained based on the additional electron density from methoxide substituents to the ring. The reactive uptake coefficient of O3(g) increases for higher relative humidity, e.g., for syringaldehyde by 18 and 215 times at 74% and 90% relative humidity (RH), respectively, as compared to dry conditions. A Langmuir–Hinshelwood mechanism fits well the kinetics of oxidation under a variable O3(g) molar ratio at 74% RH, providing useful information that should be included in atmospheric chemistry models.

Keywords: biomass burning, methoxyphenols, ozone, hydroxyl radical, secondary organic aerosol

Introduction

Biomass burning contributes to the formation of secondary organic aerosol (SOA) affecting (1) air quality and (2) climate, directly through scattering and absorption of sunlight and indirectly by acting as cloud-condensation nuclei.1,2 Phenolic compounds with aldehyde substituents constitute a large fraction of lignin in biomass and thus of the products emitted to the atmosphere during biomass burning.3−6 Abundant phenolic aldehydes such as 4-hydroxybenzaldehdye, vanillin, and syringaldehyde are ubiquitous in aerosol and fog samples attributed to biomass burning.7−13 Furthermore, vanillin and syringaldehyde were quantified also over urban regions (e.g., in Belgrade, Serbia), and in remote Arctic aerosols during summer due to the transport of smoke from biomass burning.14,15 In the atmosphere, phenolic aldehydes can be transformed by common oxidants such as ozone (O3), hydroxyl radical (HO•), nitrate radical (NO3), providing SOA precursors with multifunctional groups, including carboxylic acids.16−18

The ozonolysis of syringaldehyde and vanillin on silica particles yielded syringic acid and vanillic acid, respectively,16 and potentially other products with increased mutagenicity based on oxidation studies of humic substances in water.19 The ozonolysis of 4-hydroxybenzaldehyde in water at pH 4 resulted in the formation of 1,4-dihydroxybenzene (hydroquinone) together with oxaloacetic acid, glyoxylic acid, and oxalic acid.20 The surface oxidation of syringaldehyde by gaseous hydroxyl radical proceeded with a lifetime of 15 h, based on kinetics measurements using ca. 10- to 50-times larger [HO•(g)] than the average in the atmosphere.21 Syringaldehyde and 4-hydroxybenzaldehyde were oxidized to their corresponding aromatic acid products by Fenton and UV-Fenton processes in bulk water under acidic conditions (pH 3.0).22 The direct photodegradation and photooxidation of phenolic aldehydes in water results in light absorbing (oligomeric yellowish compounds) SOA precursors of low volatility.23−25

Recently, we reported that the oxidation of phenolic aldehydes at the air–water interface under environmentally relevant molar ratios of 48–66 ppbv O3(g) could be controlled by in situ generated HO• and revealed the production of polyhydroxymethoxyphenols.26 Furthermore, higher molar ratios of O3(g) favored the generation of polyhydroxymethoxyphenols, which was suggested to precede the simultaneously observed ring cleavage ozonolysis products.26 Such reactivity was noted to increase with increasing pH from 5 to 10, which was supported by thermodynamic measurements of redox potentials at variable pH.26 However, a problem remaining is to distinguish whether the reactivity of phenolic aldehydes and their hydroxylated products at the air–solid interface under variable relative humidity (RH) is common or different to that observed at the air–water interface.27 Therefore, in this work, alternative reactors28 are used to study the heterogeneous reactions and advance the understanding of the oxidation of phenolic aldehydes. The results below provide new mechanistic information that explains the transformation of this class of compounds during the transport of biomass burning plumes.

Experimental Section

Thin-Film Preparation

Thin films containing 150.0 μg of either syringaldehyde (Alfa Aesar, 99.2%), vanillin (Alfa Aesar, 99.9%), or 4-hydroxybenzaldehyde (Alfa Aesar, 98.7%) were prepared following a previously developed method.28 Briefly, 50 μL of 3.0 mg mL–1 solution in isopropanol (Fisher Optima, 99.9%) of the selected molecule was deposited with a syringe (Hamilton, 705) drop by drop on a 2 mm thick polished ZnSe (PIKE) optical window with a diameter of 13.0 mm. After 3 h of solvent evaporation a thin film formed, which was verified to be stable by its constant infrared spectrum absorbance. Reported data and error bars in all figures correspond to the average and standard deviation of duplicate experiments, with each experiment including between 2 and 6 thin films of the same molecule, depending on the number of analyses needed.

Oxidation Experiments

The oxidation of thin films was conducted using an adapted protocol from an earlier study.28Figure 1 illustrates that a gas flow meter (Aalborg) was used to provide a constant 0.50 L min–1 O2(g) (Scott-Gross, UHP) dry flow, which was fed into a spark discharge O3 generator (Ozone solutions). The produced O3 was diluted with dry N2(g) afterward with the help of a gas flow meter (Aalborg) providing a 0.10–0.95 L min–1 flow of dry N2(g) (Scott-Gross, UHP) and later balanced with a variable 0–0.95 L min–1 flow of wet N2(g) (Aalborg flow meter) to adjust the relative humidity (RH) of the oxidizer. A portion of the diluted gas used to conduct experiments was diverted into a photodiode array UV–visible spectrophotometer (Evolution, Thermo Scientific) to quantify the concentration of O3, where the standard deviation remained stable (≤5.0% variation) during each experiment.28Figure 1 also illustrates how films were oxidized by a total 1.00 L min–1 O3(g) flow of known concentration through a 3.78 L borosilicate glass reactor with a Teflon coated lid at a fixed RH, which was monitored with a calibrated remote hygrometer (Traceable). Prior to starting an experiment by placing the films in the reactor, the flow-through system was allowed to equilibrate for at least 1 h. Control experiments in the absence of O3(g) were carried out following the same procedure. No unexpected or unusually high safety hazards were encountered during experiments.

Figure 1.

Figure 1

Open in a new tab

Flow-through oxidation reactor with diagram of mixed dry and wet gases. Reference flows also described in the text are provided here in units of liter per minute (LPM).

The low O3(g) molar ratios of 410 and 500 ppbv used herein are aimed at stimulating the level at which this pollutant is found within pollution plumes (e.g., in urban areas affected by combustion or biomass burning).29 Instead, the upper limit of 800 ppmv O3(g) studied here is simply needed to approach the reactivity limit for large surface coverage to model the plateau in the quadratic hyperbola of the Langmuir–Hinshelwood-type mechanism.

Fourier Transform Infrared (FTIR) Monitoring

Thin films were mounted on stainless steel plates, placed inside the equilibrated reactor (Figure 1), and withdrawn every 20 min intervals during 3 h of exposure to O3, unless stated otherwise, to record their FTIR spectra. A Nicolet iN10 infrared microscope (Thermo Scientific) with an iZ10 FTIR module was used to collect 64 scans (500–4000 cm–1, 4 cm–1 resolution), which were background subtracted using the optical material. OMNIC software (Thermo Scientific) was used for data collection and processing. The CD line of characteristic peaks was used to monitor reaction kinetics,28 which for product monitoring involved the subtraction of the initial spectrum to factor out the remaining precursor. The times series were analyzed using SigmaPlot 12.3 (Systat). Finally, the chamber to perform ozonolysis reactions with sampling at defined intervals in the present analytical approach has been carefully validated in previous work with catechol28 to provide valuable information, of comparable significance to other in-line laboratory studies using ATR-FTIR,30 Raman, and SFG31 monitoring.

Results and Discussion

Oxidation of Phenolic Aldehydes

Figure 2 shows the FTIR spectra of a thin film of syringaldehyde before and after 96 h of exposure to 500 ppbv O3(g) at 74% RH, a humidity that is comparable to the global-mean RH of 77%.32 The vibrational modes of syringaldehyde, vanillin, and 4-hydroxybenzaldehyde in Figure 2 are provided in Tables S1–S3 (Supporting Information), respectively. After exposure of the films to O3(g), the intensity drops for the characteristic stretching of O—H, C=O, and C=C, respectively, labeled as peaks 1, 2, and 3 in each panel of Figure 2, which indicate that the oxidative loss of syringaldehyde, vanillin, and 4-hydroxybenalyde occurs. The mentioned peaks 1, 2, and 3 (Tables S1–S3, Supporting Information) can be observed in Figure 2 at 3288, 1670, and 1608 cm–1 for syringaldehyde, at 3176, 1664, and 1587 cm–1 for vanillin, and at 3185, 1675, and 1601 cm–1 for 4-hydroxybenaldehyde, respectively.

Figure 2.

Figure 2

Open in a new tab

Fourier transform infrared (FTIR) spectra of thin films (red) before and (blue) after oxidation of (A) syringaldehyde, (B) vanillin, and (C) 4-hydroxynenzaldehyde. The green traces correspond to spectra for each compound after subtraction from the spectra of the remaining precursor. Experimental conditions: oxidation by 96 h exposure to 500 ppbv O3(g) at 74% relative humidity (RH).

Upon oxidation of the films, new stretching vibrations for O—H(peak 4), C=O (peak 5), C=C (peak 6), and =C—O (peak 7) are registered in Figure 2. The assignment of selected vibrational modes from the oxidized films of syringaldehyde, vanillin, and 4-hydroxybenzaldehyde (Figure 2) is provided in Tables S4–S6 (Supporting Information), respectively. Peaks 4, 5, 6, and 7 are noted in Figure 2 at 3397, 1740, 1642, and 1317 cm–1 for syringaldehyde, at 3393, 1705, 1644, and 1313 cm–1 for vanillin, and at 3384, 1698, 1646, and 1317 cm–1 for 4-hydroxybenaldehyde, respectively. The development of peaks in the infrared region for C=O in the oxidized films (Figure 2) agrees well with observed absorptions at ∼1700 cm–1 for aerosol particles collected from regions influenced by biomass burning and combustion emissions.33

Figure 3 and Tables S7–S16 (Supporting Information) display the FTIR spectra and vibrational modes assignment of thin films prepared with candidate products, based on our oxidation study on the surface of aqueous microdroplets,26 which are commercially available: cis,cis-muconic acid, fumaric acid monomethyl ester, maleic acid, glyoxylic acid, oxalic acid, syringic acid, vanillic acid, 5-hydroxyvanillin, 3,4-dihydroxybenzaldehye, and 4-hydroxybenzoic acid. The previous compounds possess stretching vibrations for O—H at ∼3400 cm–1, C=O in —COOH groups at ∼1740 cm–1, and aliphatic C=C at ∼1640 cm–1, corresponding to infrared features present (peaks 4–6) in the oxidized films of Figure 2. A broad peak 4 in the oxidized films at higher wavenumbers than peak 3 in the reactants suggests the generation of recalcitrant carboxylic acid compounds such as glyoxylic and oxalic acids. A feasible alternative for explaining peak 4 in the oxidized film of Figure 2C indicates that 4-hydroxybenzoic acid contributes the peak at ∼3391 cm–1, as it overlaps well with the subtracted spectrum for oxidized 4-hydroxybenzaldehyde. However, this acid product alone is not enough to fully explain the changes registered in the spectrum. In other words, the oxidized films (Figure 2) are spectroscopically more complex than any single spectrum of the candidate products presented in Figure 3.

Figure 3.

Figure 3

Open in a new tab

FTIR spectra of films of standards, as labeled, with common features present in the oxidized films.

The experiment in Figure 2 for 96 h of exposure to 500 ppbv O3(g) corresponds to an oxidizer molar ratio representative of pollution plumes (e.g., in urban areas affected by combustion or biomass burning).29 Thus, the data in Figure 2 provides valuable information for interpreting the oxidation of phenolic aldehydes at the air–solid interface in a polluted environment. However, the long experiments needed at 500 ppbv O3(g) cause a bias in the kinetics registered from sublimation loss of the phenolic aldehydes during the 96 h period. For example, a control experiment with pure N2(g) flow at 74% RH display a loss of syringaldehyde, vanillin, and 4-hydroxybenzaldehyde by sublimation after 96 h of 2%, 85%, and 11%, respectively. Instead for 3 h of controls with N2(g) flow, the loss of syringaldehyde and 4-hydroxybenzaldehyde was <2%, while for vanillin it remained <6%, a limit assessed arbitrarily as acceptable. Therefore, experiments lasting 3 h, but generally under a higher oxidant level, are used below to provide valuable initial reaction rate constants. Gathering such information can enable the extrapolation back to realistic low [O3(g)] by studying the reaction as a function of RH and [O3(g)].

Effect of RH on Oxidation Kinetics

Figure 4 shows the loss of phenolic aldehydes with time monitored during 3 h of oxidations with 207 ppmv O3(g) by FTIR. The experiments in Figure 4 used a fixed [O3(g)] = 5.21 × 1015 molecules cm–3 but variable RH (0–90%). Inspection of Figure 4 points to the accelerating effect, caused by adsorbed water molecules, on the surface oxidation of the phenolic aldehyde films. The normalized concentrations ([reactant]/[reactant]0) in the y-axes of Figure 4 for syringaldehyde, vanillin, and 4-hydroxybenzaldehyde were obtained from measurements of the CD line (or corrected peak height at the local baseline) of peak 3 relative to their initial value, CD0. The relative decay of the C=C aromatic stretching at ∼1606 cm–1 for the phenolic aldehydes (Ph(C=O)OH) displayed in Figure 4 generally follows the fitted linear function of zero-order kinetics represented with dashed lines. However, for the cases of syringaldehyde at 74% and 90% RH and vanillin at 90% RH in Figure 4, a first-order regime is followed as fitted with solid lines represented by a two-parameter exponential decay eq 2:

graphic file with name sp2c00206_m001.jpg 2

where a is the pre-exponential constant, t represents time, and kPh(C=O)OH+O3 is the reaction rate constant of the phenolic aldehyde available in Table S17 (Supporting Information). A unified linear treatment of all data sets in Figure 4 is provided with the dashed lines using initial reaction rate constants (Inline graphic) for the three nonlinear cases. Thus, during the early stage of the reaction with k0, Ph(C=O)OH+O3, the first order kinetic regime is described by eq 3:34

graphic file with name sp2c00206_m003.jpg 3

The information provided by the simplified treatment above yields k0(Ph(C=O)OH+O3) for the initial linear region34 of syringaldehyde at 90% and 74% RH and vanillin at 90% RH in Figure 4.

Figure 4.

Figure 4

Open in a new tab

Time series for the normalized loss of (A) syringaldehyde (SA), (B) vanillin (VL), and (C) 4-hydroxynezaldehyde (4-HBA), monitored from peak 3 by FTIR, for experiments under 207 ppmv O3(g) at RH of (blue circle) 0%, (red triangle) 25%, (pink square) 50%, (green star) 74%, and (purple diamond) 90%. Solid lines represent a nonlinear fitting (eq 2) of the data points, and dashed lines correspond to a linear fitting of the data or for initial times of nonlinear cases (eq 3).

Good linear correlations for initial times are observed for the dashed lines in Figure 4. Because the prepared films for each molecules at variable RH have a reproducible initial corrected absorbance value, the determination of k0, Ph(C=O)OH+O3 is achieved with small experimental errors. It must be noted that the determined initial kinetics have been arbitrarily constrained to reflect linear subsets of the data defined to keep coefficients of determination r2 ≥ 0.96. In other words, a cumulative experimental error of <4% is considered acceptable for keeping linearity. Although the intercept of eq 3 should be ideally 1 (after dividing each concentration by its initial value), the assumed initial surface reactions may be affected to some extent by minor losses of phenolic aldehyde from deeper layers, which together with any experimental errors can contribute to cause small deviations to the linearity over initial times. The previous uncertainties explain any small variation in the value of the intercept observed when fitting eq 3. In other words, the dashed lines (linear fittings for first order kinetics) may minimally deviate from 1 due to experimental errors and/or to any contribution of bulk chemistry to the initial kinetics.

Consideration should be given to the possibility that, while the most external surface layers of the phenolic aldehyde are processed, the attack of O3(g) to the bulk (deeper layers) may not be fully prevented, especially for the highest RH cases of syringaldehyde (90% and 74%) and vanillin (90%) in Figure 4. In such cases, it is not all clear that only the surface of the solid phenolic aldehyde is active, and the kinetics of the reactions become more complex and difficult to interpret. Thus, for simplicity in the treatment of such heterogeneous systems, the active sites where oxidation takes place are assumed to be confined to the surface in the beginning of the reaction.34 The zero-order kinetics and linear range of first-order kinetics data observed in Figure 4 should correspond to heterogeneous oxidations that are surface initiated.34 Moreover, if the oxidations were entirely occurring at the surface, the reaction rate constants should be proportional to the extent of surface (or as we expressed it, they should be inversely proportional to the thickness of the film δ–1 = surface/volume ratio) as demonstrated in experiments that will be discussed below.

Based on the discussion above, the oxidation reactions should proceed by several of five steps: (1) O3(g) molecules diffuse from the gas phase to the solid surface. (2) O3(g) molecules are adsorbed on the surface of the film with phenolic aldehyde. (3) Chemical reactions take place between adsorbed ozone and the phenolic aldehyde. (4) Product molecules resulting from oxidation can remain on the surface or become desorbed. (5) Desorbed product molecules can diffuse away from the solid surface. Because the diffusion of molecules to and from the solid surface is rapid for gases and the oxidizer has no limitation to access the surface, the overall reaction rates observed are not diffusion controlled (i.e., steps 1 and 5 are not the slow steps). Thus, the control of these heterogeneous reactions must be restricted to steps 2, 3, and 4, which can be described by two models for the nature of the adsorbed layer of O3 at the surface of the phenolic aldehyde. In one of them, the adsorbed layer of O3(g) is assumed to be loosely bound to the surface and can migrate relatively freely from one surface site to another. The mobile layer can be pictured in the extreme case as representing a two-dimensional gas sorbed on the surface of phenolic aldehyde. Instead, the second model corresponds to a strongly bound surface layer of adsorbed O3 molecules that forms chemical bonds with the surface species. In such a localized layer, the migration of O3 molecules may take place slowly either by surface diffusion or by desorption and readsorption. The relative slowness of these latter processes can make them rate-controlling for these experiments, e.g., in Figure 4.

The uptake coefficient of O3(g) molecules (γO3) impinging on the phenolic aldehyde’s surface provides the fraction of reactive collisions relative to the total rate of collisions based on eq 4:35

graphic file with name sp2c00206_m004.jpg 4

where the gas constant is R = 8.314 J K–1 mol–1; the mean thermal velocity of O3(g) at T = 298 K is vO3 = 394 m s–1; the partial pressure of O3(g) is PO3 = 20.97 Pa for the experiments at 207 ppm of O3(g); Table S17 (Supporting Information) includes the initial reaction rate constant of reactant decay k0, Ph(C=O)OH+O3, determined for the linear fittings (dashed lines) in Figure 4; and the effective film thickness, δeff, for syringaldehyde, vanillin, and 4-hydroxybenzaldehyde are estimated to be δeff = 1.125 × 10–7, 1.072 × 10–7, and 9.239 × 10–8 m, respectively. The film thickness values were estimated for uniform surface coverage of 150 × 10–9 kg deposited for each phenolic aldehyde (4.96 × 1017 molecules of syringaldehyde, 5.94 × 1017 molecules of vanillin, and 7.40 × 1017 molecules of 4-hydroxybenzaldehyde) over the IR-transparent crystal of area SA = 1.32 × 10–4 m2; [syringaldehyde] = 5.543 × 103 molecules m–3, [vanillin] = 6.969 × 103 molecules m–3, and [4-hydroxybenzaldehyde] = 1.007 × 104 molecules m–3 based on their density, dsyringaldehyde = 1010, dvanillin = 1060, and d4-hydroxybenzaldehyde = 1230 kg m–3, and respective formula mass of 182.2 × 10–3, = 152.1 × 10–3, and 122.1 × 10–3 kg mol–1. The determined γO3 at variable RH are displayed in Figure 5.

Figure 5.

Figure 5

Open in a new tab

Reactive uptake coefficient of O3(g) (γO3) by thin films of (A) syringaldehyde, (B) vanillin, and (C) 4-hydroxybenzaldehyde exposed to 207 ppmv O3(g) under variable relative humidity. Dashed lines are provided as a guide to the eye only.

The γO3 values in Figure 5 remain relative flat from 0% to 50% RH, start to slightly increase at 74% RH, and monotonically rise to maxima for 90% RH, in agreement with the exponential behavior previously observed for thin films of catechol.28 The large increase of γO3 for 90% RH indicates that, as more water molecules are adsorbed on the organic film surface, the interfacial reaction of ozone is facilitated on a hydrated environment that should resemble aqueous reactivity pathways. This type of dependence of γO3 on increasing RH has been attributed in related studies with shikimic acid to the enhanced bulk-surface partitioning of O3(g), and the packing effect and higher mobility that facilitates for the favorable orientation between reactant molecules and the incoming oxidant at higher RH.36−38

Mechanistically, the ozonolysis of the phenolic aldehydes starts with the formation of a primary ozonide (a trioxyl diradical), which creates a Criegee intermediate (CI) that reacts with water molecules39 in a process accelerated at a higher RH. Indeed, as the surface coverage by water molecules grows with RH, the degradation of the phenolic aldehydes is simultaneously enhanced for several reasons. Among them, hydrogen bond formation between the phenolic aldehydes and water can lower the activation energy needed to form the primary ozonide,40 while solvation also enables proton and electrons transfer reactions that provide additional degradation pathways through the generation of hydroxyl radical.41 In addition, water also works as a critical reactant in multiple stages of the oxidative processing of secondary and tertiary reaction products, i.e., maleic acid, generating multiple highly oxygenated compounds.42 Finally, the rate of in situ production of the secondary oxidant H2O2 increases at a high RH due to the more favorable generation of an organic hydroperoxide precursor in the reaction CI + H2O.

From the viewpoint of the products generated, the kinetic derived values of γO3 are plotted versus RH in Figure S2 (Supporting Information). For this purpose, a similar treatment of the data was applied to that described above for the reactant. Specifically, peak 5 (1695–1745 cm–1) in the infrared spectra of Figure 2 can be assigned to the total ester and carboxylic acid products from ring cleavage by direct ozonolysis. Instead, peak 7 (1310–1320 cm–1) represents all functionalized products from hydroxylation reactions before ring fragmentation. Thus, for the case of 90% RH and as a first approximation, we can estimate the branching ratio of direct ozonolysis and ring functionalization channels based on the uptake values by ignoring any minor overlapping contributed by other molecules. The uptake coefficients for peaks 5 and 7 at 90% RH of syringaldehyde, vanillin, and 4-hydroxybenzaldehyde are 1.2 × 10–7 and 3.5 × 10–8, 3.3 × 10–8 and 7.9 × 10–9, and 1.6 × 10–9 and 1.9 × 10–9, respectively. Therefore, the estimated branching ratios for direct ozonolysis to ring functionalization are 0.77 to 0.23 for syringaldehyde, 0.81 and 0.19 for vanillin, and 0.45 to 0.55 for 4-hydroxybenzaldehyde. The presence of electron donating —OCH3 groups imprints a similar branching ratio for syringaldehyde and vanillin for ring cleavage and hydroxylation of the aromatic ring. Instead, the absence of a —OCH3 group results in a considerable increase for hydroxylation of the aliphatic aldehyde group to form 4-hydroxybenzoic acid.

Kinetic Regime of Surface Reaction

A set of syringaldehyde films prepared with different thicknesses (δ) was exposed to 100.0 ppmv O3(g) (2.39 × 10–15 O3 molecules cm–3) at 74% RH to evaluate the time interval controlled exclusively by surface reactivity. The initial reaction rate constants, k0, with units of s–1 from these films were obtained from the early linear region in Figure 6A and plotted as a function of 1/δ in Figure 6B. The excellent linear fitting for k0 = 2.41 × 10–6 s–1 + 3.93 × 10–5 μm s–1 × 1/δ with coefficient of determination r2 = 0.996 in Figure 6B implies that the heterogeneous reaction is limited (1) to proceed on the film’s surface and/or (2) by the diffusion of gaseous O3 through the material.43 The change in the kinetic regime (accelerated reaction rates) starting from 40 to 90 min for the thinnest to thickest films, respectively, corresponds to the degradation of deeper layers of syringaldehyde that become accessible to the oxidizer. In consequence and due to the reduced steric hindrance of the new surface molecules, the deeper unreacted syringaldehyde layers experience an enhanced uptake of the oxidizer than possible otherwise.

Figure 6.

Figure 6

Open in a new tab

(A) Normalized decay of SA reacting with 100.0 ppmv O3 at 74% RH for the variable film thickness (δ) provided in the legend versus time. The initial reaction times used to measure the initial reaction rate constant (k0) fitted with a linear regression are represented by the solid lines, while the dashed lines for the rest of the data are simple eye guides. (B) Values of k0 vs (bottom axis) 1/δ and (top axis) δ. The solid green line represents the linear fitting with parameters provided in the text. (C) Example of the linkage among two SA molecules in the crystalline structure through intermolecular and intramolecular hydrogen bonds marked with bold blue dashes for the bond lengths in red font, and geometrical distances in green font as indicated by the green hollow arrowheads. Two limit cases for the orientation of O3 are provided to the top as referred in the text for the horizontal and vertical geometries to the left- and right-hand sides, respectively.

Figure S1 confirms that the lattice vibrations of syringaldehyde in the mid-infrared region are in close agreement for the FTIR transmission spectra of prepared films and the reagent crystals registered by attenuated total reflectance (ATR). The minor differences between the relative absorbance of peaks in the infrared of syringaldehyde (Figure S1) are due to the nature of comparing transmission and ATR spectra. Overall, the large agreement on individual position and width of peaks confirms the similarity between both samples. Thus, Figure 6C serves as a representation of the layers in the crystalline structure with a dimer of SA orderly connected by intermolecular O—H···O hydrogen bonds. Specifically, the hydrogen bonds occur between the hydroxy group in carbon 4 of the ring and the aldehyde function in carbon 1, forming infinite linear chains separated by only 2.05 Å (H···O). This arrangement involves a 5-center ring with interaction between the aldehyde hydrogen and the phenolic oxygen, which keeps the two oxygen atoms 2.71 Å apart. In addition, the H atom on the hydroxy group of carbon 4 also participate in an intramolecular O—H···O interaction with a neighboring methoxy O atom with a length of 2.22 Å (H···O).44 Such an arrangement makes the delocalized electron density of the SA rings above the dimer more available for ozonolysis, favoring the interaction of impinging O3 with the aromatic region, while likely decreasing its diffusion into deeper layers in the bulk. Instead, this assembling disfavors the ozonolysis of SA rings buried below the dimer. Figure 6C also depicts two orientations for the molecule of O3, which has a C2v symmetry with O—O bond distances of 1.27 Å, an O—O—O angle of 116.8°, and a separation between the two terminal O atoms of 2.17 Å.45 The latter molecular dimension is similar to the length of the H-bond of two SA molecules in the infinite chain of the crystalline structure. Therefore, the direct diffusion of gaseous O3 through deeper layers appears to be practically hindered for the horizontal O3 arrangement displayed in Figure 6C. Thus, the diffusion process into the film would be constrained to the population fraction with ca. vertical geometries (Figure 6C) that reach the surface.

Considering longer oxidation times than used to obtain k0 in Figure 6A, the degradation of the surface layer can become significant enough, as it noted earlier for the thinnest film. In more detail, after 40 min, there is a clear change of kinetic regime with an accelerated degradation of syringaldehyde for the film with δ = 0.375 μm. This acceleration is likely due to in situ creation of reactive species and defects in the external layers for O3 to diffuse under reduced hindrance and facilitate the deeper processing of the film. The obtained k0 (with units of s–1) from Figure 6A can be used to calculate the half-life, τ1/2 = (Ln 2)/k0, for the heterogeneous oxidation of SA exposed to 100 ppmv O3 at 74% RH, which range between 1.78 and 11.8 h from the thinnest to the thickest film.

The depth of penetration of the oxidizer under surface dominated reactivity can be estimated from the initial kinetics. For this purpose, a diffusion coefficient at ∼ 74% RH for O3(g) DO3 = 7.12 × 10–9 cm2 s–1 can be considered, which as a first approximation is interpolated from the s-shape data set for shikimic acid.37 We must note that the depths of penetration of O3(g) molecules into the solid film, Inline graphic, predicted under O3 diffusion limitation for the initial regime of Figure 6A are physically unreasonable as they would range between 96 and 246 μm. The previous observation is not surprising because the diffusion coefficients of syringaldehyde must differ from the reliable value determined for shikimic acid. Furthermore, even for the first 40 min, the calculated L40 min = 58.5 μm cannot be correct as it is larger than the dimensions of the films prepared.

Instead, the slope of the best straight line for k0 vs 1/δ, m = 3.93 × 10–5 μm s–1 in Figure 6A can be used to estimate the effective film length of SA loss during the first 40 min of reaction exposed to 100.0 ppm of O3 at 74% RH to be only Leff,40 min = 94.8 nm. The only way to reconcile the values of Leff,40 min and L40 min would be for the actual species to limit the diffusion into the film to be syringaldehyde, which in the semisolid state46 should have an estimated diffusion coefficient DSA = 1.87 × 10–14 cm2 s–1 at ∼ 74% RH. For comparison, the proposed DSA is ∼ 26-times smaller than the fitted value for shikimic acid (Dshikimic = 4.85 × 10–13 cm2 s–1) predicted at 74% RH by a model run including product effects consistent with a Vignes-type fit.37 For the case with DSA = 1.87 × 10–14 cm2 s–1 at ∼74% RH, the corrected depths of penetration during the half-live are 155, 311, 355, and 399 nm for the films with thickness of 0.375, 1.125, 2.250, and 3.375 μm, respectively. Thus, it appears reasonable to assume that only for the initial kinetic regime the bulk of the films may remain unreactive, since the data fitted with a straight line for k0 vs δ–1 in Figure 6B suggests a surface dominated reaction. However, due to the difficulty in resolving the exact transition from surface to bulk reactions even for limit cases,37 it can be safely concluded that pure surface chemistry only controls the initial kinetics.

Langmuir–Hinshelwood Dependence of k0 on [O3(g)]

Figure 7 displays the time series for the loss of phenolic aldehydes exposed to variable [O3(g)]. While vanillin and 4-hydroxybenzaldehye decay with zero order kinetics during the 3 h reaction, syringaldehyde only obeys the same behavior for molar ratios of O3 ≤ 55 ppmv. Instead for higher [O3(g)], syringaldehyde follows a first-order kinetics, which only maintains linearity for the first ∼30 min. Thus, the data in Figure 7 is fitted for zero-order kinetics with dashed lines for either the course of the entire reaction or the initial linear regime of the exemption cases. As explained in the previous section, using initial reaction rates not only assures a description of the surface dominated kinetic regime based on eq 3 but also facilitates a unified treatment of all data in Figure 7.

Figure 7.

Figure 7

Open in a new tab

Normalized decay of (A) syringaldehyde (SA), (B) vanillin (VL), and (C) 4-hydroxybenzaldehyde (4-HBA) at 74% RH under exposure to the variable molar ratios of O3(g) reported in the legend to the right of each panel in units of ppmv. Dashed lines correspond to a linear fitting of the data or for initial times, as described by eq 3, of nonlinear cases displayed with solid lines fitted with eq 2.

Figure 8 shows the dependence of the k0 values from Figure 7 on increasing [O3(g)], which is described by a Langmuir–Hinshelwood (LH) surface mechanism with eq 5:47

graphic file with name sp2c00206_m006.jpg 5

with nonlinear fitting parameters for the maximum decay rate constant, k0,max, predicted for high [O3(g)] to saturate the surface of syringaldehyde, vanillin and 4-hydroxybenzaldeyde of 7.784 × 10–4, 6.218 × 10–5, and 2.163 × 10–5 s–1, respectively. In this mechanism first O3(g) molecules are adsorbed on the surface to then react with the phenolic aldehydes,

graphic file with name sp2c00206_m007.jpg 6
graphic file with name sp2c00206_m008.jpg 7

reaching eventually a LH equilibrium in a process where the uptake of ozone and the loss of organic material are limited by the surface reaction.37 The constant KL in eq 5 represents the [O3(g)] needed to cover a half of the phenolic aldehyde surface with adsorbed molecules, which was fitted in Figure 8 panels A through C to 3.681 × 10,15 5.151 × 10,15 and 1.136 × 1016 molecules cm–3, respectively. A previous second frequency generation (SFG) study has showed that the heterogeneous ozonolysis of the C=C bond of cyclohexene displayed two reactivity modes, at high- and low-ozone levels, following also a Langmuir–Hinshelwood mechanism.31

Figure 8.

Figure 8

Open in a new tab

Initial loss rate constant (k0) of (A) syringaldehyde, (B) vanillin, and (C) 4-hydroxybenzaldehyde for variable molar ratios of O3(g) at 74% RH. Data obtained fitting the equation ([Ph(C=O)OH]/[Ph(C=O)OH]0) = a – k0(Ph(C=O)OH+O3) × t, with a = 1, to the linear regimes in Figure 7. The dashed lines correspond to a nonlinear regression described by eq 5 for the Langmuir–Hinshelwood mechanism with parameters provided in the text.

The half-life under 500 ppbv O3(g) at 74% RH in the LH mechanism of Figure 8 for syringaldehyde, vanillin, and 4-hydroxybenzaldehyde can be calculated from eq 8,

graphic file with name sp2c00206_m009.jpg 8

to be 4.46, 78, and 494 days, revealing the important oxidative acceleration caused by —CH3 substituents.48 However, based on the experiments at variable RH (Figures 4 and 5), the previous reaction rates of syringaldehyde, vanillin, and 4-hydroxybenzaldehyde at 90% RH can be enhanced by 6.24-, 2.73-, and 3.45-times, reducing τ1/2 at 90% RH to 0.72, 28.6, and 143.3 days, respectively.

Conclusions and Atmospheric Implications

Phenolic aldehydes are efficiently oxidized under atmospheric relevant conditions on a variety of environmental surfaces (e.g., aerosol particles, cloudwater, grime, etc.). Heterogenous oxidation experiments at the air–solid interface were performed under variable relative humidity, and the formation of hydroxylated species (polyhydroxyphenols) and polyfunctional carboxylic acids was monitored. The presence of O3(g) not only contributes an oxidizer for the processing of aromatic pollutants but also facilitates an electron transfer process that creates HO• in situ on the reacting surface.

The reactive uptake of O3(g) by the phenolic aldehydes showed a strong nonlinear enhancement as RH tends to 90% RH, as observed before for catechol.28 Therefore, the role of adsorbed water molecules may be key on the surface reactivity of O3, because they lower the viscosity on the surface of the film by creating an air–water interface49 and can participate in the reactivity of intermediates.50,51 As demonstrated in this and our previous work, this aromatic family reacts on confined liquid26 and solid surfaces to undergo electrophilic attack by O3(g), resulting in a variety of products that should involve the participation of multiple intermediates. The sequential decomposition of hydroperoxide intermediates that release H2O2 in situ on the surface provides an additional Baeyer–Villiger oxidation channel. The resulting carboxylic acids from this type of reactions should increase the acidity of the surface.52

The reactive uptake coefficient at 90% RH for syringaldehyde γO3 = 5.00 × 10–6, vanillin γO3 = 6.60 × 10–7, and 4-hydroxybenzaldehyde γO3 = 1.25 × 10–7 are within a close range to that we determined at this RH for catechol (γO3 = 7.49 × 10–6).28 The range of γO3 values registered for the phenolic aldehydes at 90% RH indicates that, while the aldehyde group deactivates the aromatic ring, the presence of two methoxide substituents accelerates the oxidation. The consecutive addition of one and two —OCH3 groups used to compare reactivity26 increased γO3 of syringaldehyde by 7.6-fold that of vanillin and 40-fold that of 4-hydroxybenzaldehyde.

Based on the ratio of γO3,ring cleavage/γO3,hydroxylation determined from Figure S2 for all molecules, both processes are competitive, with ozonolysis dominating on average for the whole RH range by ∼3.3 (±2.2) times over hydroxylation. We arrive to a similar conclusion by comparing the ratio of predicted reaction rates for O3 versus HO• on the surface (A) of particles with a diameter of 100 nm covered by syringaldehyde. For 72 ppbv O3 ([O3(g)] = 1.77 × 1012 molecules cm–3) competing with a mean [OH(g)] = 1.6 × 106 molecules cm–3, the ratio of reaction rates is (γO3[O3(g)] νO3A/4)/(γHO•[HO•(g)] νHO•A/4) =3.3. For the previous calculation, room temperature was used, γO3 = 5.00 × 10–6 for syringaldehyde at 90% RH, γHO• = 1 for HO•, and the mean thermal velocities of O3(g) and HO•(g) are νO3 = 394 m s–1 and νHO• = 661 m s–1, respectively. Therefore, ozonolysis may be ∼3-times faster than hydroxylation for the loss of syringaldehyde by HO• during these interfacial reactions.

The surface oxidations studied based on initial reactions rates are supported by the fact that they follow a Langmuir–Hinshelwood type mechanism for increasing ozone levels. The fitting of eq 5 for the Langmuir–Hinshelwood mechanism enabled the reliable extrapolation of data sets for the three phenolic aldehydes to low molar ratio of O3 to estimate the lifetime against the reaction with O3 on atmospheric surfaces. For example, the lifetime of syringaldehyde in a pollution plume experiencing a 90% RH is only 17.3 h. Future work should explore the role of nitrate radicals on atmospheric interfaces, adapt the approach to characterize the generated brown carbon by alternative spectroscopies, and integrate present mechanistic observations for phenolic aldehydes and methoxyphenols into atmospheric chemistry models.

Acknowledgments

Support from the U.S.A. National Science Foundation under award 1903744 to M.I.G. is gratefully acknowledged.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsearthspacechem.2c00206.

  • Figures of FTIR spectra of crystalline syringaldehyde and reactive uptake from product and tables of vibrational assignment and kinetic parameters at variable RH (PDF)

The authors declare no competing financial interest.

Supplementary Material

sp2c00206_si_001.pdf (357.4KB, pdf)

References

  1. IPCC. Climate change 2013: The physical science basis. Contribution of working group I to the fifth assessment report of the intergovernmental panel on climate change; Cambridge University Press, 2013. 10.1017/CBO9781107415324 [DOI] [Google Scholar]
  2. De Gouw J.; Jimenez J. L. Organic Aerosols in the Earth’s Atmosphere. Environ. Sci. Technol. 2009, 43 (20), 7614–7618. 10.1021/es9006004. [DOI] [PubMed] [Google Scholar]
  3. Simoneit B. R. T. Biomass burning — a review of organic tracers for smoke from incomplete combustion. Appl. Geochem. 2002, 17 (3), 129–162. 10.1016/S0883-2927(01)00061-0. [DOI] [Google Scholar]
  4. Simoneit B. R. T.; Rogge W. F.; Mazurek M. A.; Standley L. J.; Hildemann L. M.; Cass G. R. Lignin pyrolysis products, lignans, and resin acids as specific tracers of plant classes in emissions from biomass combustion. Environ. Sci. Technol. 1993, 27 (12), 2533–2541. 10.1021/es00048a034. [DOI] [Google Scholar]
  5. Kjällstrand J.; Ramnäs O.; Petersson G. Methoxyphenols from burning of Scandinavian forest plant materials. Chemosphere 2000, 41 (5), 735–741. 10.1016/S0045-6535(99)00427-0. [DOI] [PubMed] [Google Scholar]
  6. Schauer J. J.; Kleeman M. J.; Cass G. R.; Simoneit B. R. T. Measurement of emissions from air pollution sources. 3. C1–C29 organic compounds from fireplace combustion of wood. Environ. Sci. Technol. 2001, 35 (9), 1716–1728. 10.1021/es001331e. [DOI] [PubMed] [Google Scholar]
  7. Dron J.; El Haddad I.; Temime-Roussel B.; Jaffrezo J. L.; Wortham H.; Marchand N. Functional group composition of ambient and source organic aerosols determined by tandem mass spectrometry. Atmos. Chem. Phys. 2010, 10 (15), 7041–7055. 10.5194/acp-10-7041-2010. [DOI] [Google Scholar]
  8. Decesari S.; Fuzzi S.; Facchini M. C.; Mircea M.; Emblico L.; Cavalli F.; Maenhaut W.; Chi X.; Schkolnik G.; Falkovich A.; et al. Characterization of the organic composition of aerosols from Rondônia, Brazil, during the LBA-SMOCC 2002 experiment and its representation through model compounds. Atmos. Chem. Phys. 2006, 6 (2), 375–402. 10.5194/acp-6-375-2006. [DOI] [Google Scholar]
  9. Gilardoni S.; Massoli P.; Paglione M.; Giulianelli L.; Carbone C.; Rinaldi M.; Decesari S.; Sandrini S.; Costabile F.; Gobbi G. P.; et al. Direct observation of aqueous secondary organic aerosol from biomass-burning emissions. Proc. Natl. Acad. Sci. U. S. A. 2016, 113 (36), 10013–10018. 10.1073/pnas.1602212113. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Czech H.; Miersch T.; Orasche J.; Abbaszade G.; Sippula O.; Tissari J.; Michalke B.; Schnelle-Kreis J.; Streibel T.; Jokiniemi J.; et al. Chemical composition and speciation of particulate organic matter from modern residential small-scale wood combustion appliances. Sci. Total Environ. 2018, 612, 636–648. 10.1016/j.scitotenv.2017.08.263. [DOI] [PubMed] [Google Scholar]
  11. Brege M.; Paglione M.; Gilardoni S.; Decesari S.; Facchini M. C.; Mazzoleni L. R. Molecular insights on aging and aqueous-phase processing from ambient biomass burning emissions-influenced Po Valley fog and aerosol. Atmos. Chem. Phys. 2018, 18 (17), 13197–13214. 10.5194/acp-18-13197-2018. [DOI] [Google Scholar]
  12. Miyazaki Y.; Kondo Y.; Shiraiwa M.; Takegawa N.; Miyakawa T.; Han S.; Kita K.; Hu M.; Deng Z. Q.; Zhao Y.; et al. Chemical characterization of water-soluble organic carbon aerosols at a rural site in the Pearl River Delta, China, in the summer of 2006. J. Geophys. Res. 2009, 114 (D14), D14208. 10.1029/2009JD011736. [DOI] [Google Scholar]
  13. Srivastava D.; Tomaz S.; Favez O.; Lanzafame G. M.; Golly B.; Besombes J.-L.; Alleman L. Y.; Jaffrezo J.-L.; Jacob V.; Perraudin E.; et al. Speciation of organic fraction does matter for source apportionment. Part 1: A one-year campaign in Grenoble (France). Sci. Total Environ. 2018, 624, 1598–1611. 10.1016/j.scitotenv.2017.12.135. [DOI] [PubMed] [Google Scholar]
  14. Zangrando R.; Barbaro E.; Zennaro P.; Rossi S.; Kehrwald N. M.; Gabrieli J.; Barbante C.; Gambaro A. Molecular Markers of Biomass Burning in Arctic Aerosols. Environ. Sci. Technol. 2013, 47 (15), 8565–8574. 10.1021/es400125r. [DOI] [PubMed] [Google Scholar]
  15. Zangrando R.; Barbaro E.; Kirchgeorg T.; Vecchiato M.; Scalabrin E.; Radaelli M.; Đorđević D.; Barbante C.; Gambaro A. Five primary sources of organic aerosols in the urban atmosphere of Belgrade (Serbia). Sci. Total Environ. 2016, 571, 1441–1453. 10.1016/j.scitotenv.2016.06.188. [DOI] [PubMed] [Google Scholar]
  16. Net S.; Alvarez E. G.; Gligorovski S.; Wortham H. Heterogeneous reactions of ozone with methoxyphenols, in presence and absence of light. Atmos. Environ. 2011, 45 (18), 3007–3014. 10.1016/j.atmosenv.2011.03.026. [DOI] [Google Scholar]
  17. Seinfeld J. H.; Pandis S. N.. Atmospheric Chemistry and Physics: From Air Pollution to Climate Change; John Wiley & Sons, 1998. [Google Scholar]
  18. Tomaz S.; Cui T.; Chen Y.; Sexton K. G.; Roberts J. M.; Warneke C.; Yokelson R. J.; Surratt J. D.; Turpin B. J. Photochemical cloud processing of primary wildfire emissions as a potential source of secondary organic aerosol. Environ. Sci. Technol. 2018, 52 (19), 11027–11037. 10.1021/acs.est.8b03293. [DOI] [PubMed] [Google Scholar]
  19. Hibino M.; Matsuda H.; Sato T.; Ose Y.; Nagase H.; Kito H. Generation of mutagenicity by ozonation of humic substances’ components. Sci. Total Environ. 1992, 116 (1), 1–13. 10.1016/0048-9697(92)90361-U. [DOI] [PubMed] [Google Scholar]
  20. Andreozzi R.; Caprio V.; D’Amore M. G.; Insola A. p-Coumaric acid abatement by ozone in aqueous solution. Water Res. 1995, 29 (1), 1–6. 10.1016/0043-1354(94)00142-T. [DOI] [Google Scholar]
  21. Liu C.; Zeng C. Heterogeneous kinetics of methoxyphenols in the OH-initiated reactions under different experimental conditions. Chemosphere 2018, 209, 560–567. 10.1016/j.chemosphere.2018.06.131. [DOI] [PubMed] [Google Scholar]
  22. Kang J.; Irmak S.; Wilkins M. Conversion of lignin into renewable carboxylic acid compounds by advanced oxidation processes. Renew. Energy 2019, 135, 951–962. 10.1016/j.renene.2018.12.076. [DOI] [Google Scholar]
  23. Smith J. D.; Kinney H.; Anastasio C. Phenolic carbonyls undergo rapid aqueous photodegradation to form low-volatility, light-absorbing products. Atmos. Environ. 2016, 126, 36–44. 10.1016/j.atmosenv.2015.11.035. [DOI] [Google Scholar]
  24. Li Y. J.; Huang D. D.; Cheung H. Y.; Lee A. K. Y.; Chan C. K. Aqueous-phase photochemical oxidation and direct photolysis of vanillin – a model compound of methoxy phenols from biomass burning. Atmos. Chem. Phys. 2014, 14 (6), 2871–2885. 10.5194/acp-14-2871-2014. [DOI] [Google Scholar]
  25. Yee L. D.; Kautzman K. E.; Loza C. L.; Schilling K. A.; Coggon M. M.; Chhabra P. S.; Chan M. N.; Chan A. W. H.; Hersey S. P.; Crounse J. D.; et al. Secondary organic aerosol formation from biomass burning intermediates: phenol and methoxyphenols. Atmos Chem. Phys. 2013, 13 (16), 8019–8043. 10.5194/acp-13-8019-2013. [DOI] [Google Scholar]
  26. Rana M. S.; Guzman M. I. Oxidation of phenolic aldehydes by ozone and hydroxyl radicals at the air–water interface. J. Phys. Chem. A 2020, 124 (42), 8822–8833. 10.1021/acs.jpca.0c05944. [DOI] [PubMed] [Google Scholar]
  27. Ram M. S.; Stanbury D. M. Reactions of the tris(3,4,7,8-tetramethylphenanthroline)iron(II,III) redox couple in nitrous acid. J. Am. Chem. Soc. 1984, 106 (26), 8136–8142. 10.1021/ja00338a022. [DOI] [Google Scholar]
  28. Pillar-Little E. A.; Zhou R.; Guzman M. I. Heterogeneous oxidation of catechol. J. Phys. Chem. A 2015, 119 (41), 10349–10359. 10.1021/acs.jpca.5b07914. [DOI] [PubMed] [Google Scholar]
  29. Seinfeld J. H.; Pandis S. N.. Atmospheric Chemistry and Physics: From Air Pollution to Climate Change, 2nd ed.; Wiley, 2006; p 1232. [Google Scholar]
  30. Dubowski Y.; Vieceli J.; Tobias D. J.; Gomez A.; Lin A.; Nizkorodov S. A.; McIntire T. M.; Finlayson-Pitts B. J. Interaction of gas-phase ozone at 296 k with unsaturated self-assembled monolayers: A new look at an old system. J. Phys. Chem. A 2004, 108 (47), 10473–10485. 10.1021/jp046604x. [DOI] [Google Scholar]
  31. Stokes G. Y.; Chen E. H.; Walter S. R.; Geiger F. M. Two reactivity modes in the heterogeneous cyclohexene ozonolysis under tropospherically relevant ozone-rich and ozone-limited conditions. J. Phys. Chem. A 2009, 113 (31), 8985–8993. 10.1021/jp904104s. [DOI] [PubMed] [Google Scholar]
  32. Hartmann D. L.Global Physical Climatology; Academic Press, 1994. [Google Scholar]
  33. Zhang Q.; Shen Z.; Zhang L.; Zeng Y.; Ning Z.; Zhang T.; Lei Y.; Wang Q.; Li G.; Sun J.; et al. Investigation of primary and secondary particulate brown carbon in two chinese cities of Xi’an and Hong Kong in wintertime. Environ. Sci. Technol. 2020, 54 (7), 3803–3813. 10.1021/acs.est.9b05332. [DOI] [PubMed] [Google Scholar]
  34. Benson S. W.The Foundations of Chemical Kinetics; McGraw-Hill, 1960. [Google Scholar]
  35. Hung H.-M.; Ariya P. Oxidation of oleic acid and oleic acid/sodium chloride(aq) mixture droplets with ozone: Changes of hygroscopicity and role of secondary reactions. J. Phys. Chem. A 2007, 111 (4), 620–632. 10.1021/jp0654563. [DOI] [PubMed] [Google Scholar]
  36. Steimer S. S.; Berkemeier T.; Gilgen A.; Krieger U. K.; Peter T.; Shiraiwa M.; Ammann M. Shikimic acid ozonolysis kinetics of the transition from liquid aqueous solution to highly viscous glass. Phys. Chem. Chem. Phys. 2015, 17 (46), 31101–31109. 10.1039/C5CP04544D. [DOI] [PubMed] [Google Scholar]
  37. Berkemeier T.; Steimer S. S.; Krieger U. K.; Peter T.; Pöschl U.; Ammann M.; Shiraiwa M. Ozone uptake on glassy, semi-solid and liquid organic matter and the role of reactive oxygen intermediates in atmospheric aerosol chemistry. Phys. Chem. Chem. Phys. 2016, 18 (18), 12662–12674. 10.1039/C6CP00634E. [DOI] [PubMed] [Google Scholar]
  38. Vieceli J.; Ma O. L.; Tobias D. J. Uptake and collision dynamics of gas phase ozone at unsaturated organic interfaces. J. Phys. Chem. A 2004, 108 (27), 5806–5814. 10.1021/jp0494584. [DOI] [Google Scholar]
  39. Tillmann R.; Hallquist M.; Jonsson Å. M.; Kiendler-Scharr A.; Saathoff H.; Iinuma Y.; Mentel T. F. Influence of relative humidity and temperature on the production of pinonaldehyde and OH radicals from the ozonolysis of alpha-pinene. Atmos. Chem. Phys. 2010, 10 (15), 7057–7072. 10.5194/acp-10-7057-2010. [DOI] [Google Scholar]
  40. Barnum T. J.; Medeiros N.; Hinrichs R. Z. Condensed-phase versus gas-phase ozonolysis of catechol: A combined experimental and theoretical study. Atmos. Environ. 2012, 55, 98–106. 10.1016/j.atmosenv.2012.02.019. [DOI] [Google Scholar]
  41. Pillar-Little E. A.; Camm R. C.; Guzman M. I. Catechol oxidation by ozone and hydroxyl radicals at the air–water interface. Environ. Sci. Technol. 2014, 48 (24), 14352–14360. 10.1021/es504094x. [DOI] [PubMed] [Google Scholar]
  42. Gallimore P. J.; Achakulwisut P.; Pope F. D.; Davies J. F.; Spring D. R.; Kalberer M. Importance of relative humidity in the oxidative ageing of organic aerosols: case study of the ozonolysis of maleic acid aerosol. Atmos. Chem. Phys. 2011, 11 (23), 12181–12195. 10.5194/acp-11-12181-2011. [DOI] [Google Scholar]
  43. Milsom A.; Squires A. M.; Woden B.; Terrill N. J.; Ward A. D.; Pfrang C. The persistence of a proxy for cooking emissions in megacities: a kinetic study of the ozonolysis of self-assembled films by simultaneous small and wide angle X-ray scattering (SAXS/WAXS) and Raman microscopy. Faraday Discuss. 2021, 226 (0), 364–381. 10.1039/D0FD00088D. [DOI] [PubMed] [Google Scholar]; DOI: 10.1039/D0FD00088D.
  44. Kolev T.; Wortmann R.; Spiteller M.; Sheldrick W. S.; Heller M. 4-Hydroxy-3,5-dimethoxybenzaldehyde (syringaldehyde). Acta Cryst. E 2004, 60 (8), o1387–o1388. 10.1107/S1600536804017015. [DOI] [Google Scholar]
  45. Herzberg G.Electronic spectra and electronic structure of polyatomic molecules; Van Nostrand, 1966. [Google Scholar]
  46. Shiraiwa M.; Pöschl U. Mass accommodation and gas–particle partitioning in secondary organic aerosols: dependence on diffusivity, volatility, particle-phase reactions, and penetration depth. Atmos. Chem. Phys. 2021, 21 (3), 1565–1580. 10.5194/acp-21-1565-2021. [DOI] [Google Scholar]
  47. Zhang T.; Yang W.; Han C.; Yang H.; Xue X. Heterogeneous reaction of ozone with syringic acid: Uptake of O3 and changes in the composition and optical property of syringic acid. Environ. Pollut. 2020, 257, 113632 10.1016/j.envpol.2019.113632. [DOI] [PubMed] [Google Scholar]
  48. Pillar-Little E. A.; Guzman M. I. Oxidation of substituted catechols at the air–water interface: Production of carboxylic acids, quinones, and polyphenols. Environ. Sci. Technol. 2017, 51 (9), 4951–4959. 10.1021/acs.est.7b00232. [DOI] [PubMed] [Google Scholar]
  49. Wu D. G.; Malec A. D.; Head-Gordon M.; Majda M. Viscosity of the Aqueous Liquid/Vapor Interfacial Region: 2D Electrochemical Measurements with a Piperidine Nitroxy Radical Probe. J. Am. Chem. Soc. 2005, 127 (12), 4490–4496. 10.1021/ja042969j. [DOI] [PubMed] [Google Scholar]
  50. Chao W.; Hsieh J.-T.; Chang C.-H.; Lin J. J.-M. Direct kinetic measurement of the reaction of the simplest Criegee intermediate with water vapor. Science 2015, 347 (6223), 751–754. 10.1126/science.1261549. [DOI] [PubMed] [Google Scholar]
  51. Kumar M.; Zhong J.; Zeng X. C.; Francisco J. S. Reaction of Criegee Intermediate with Nitric Acid at the Air–Water Interface. J. Am. Chem. Soc. 2018, 140 (14), 4913–4921. 10.1021/jacs.8b01191. [DOI] [PubMed] [Google Scholar]
  52. Eugene A. J.; Pillar-Little E. A.; Colussi A. J.; Guzman M. I. Enhanced acidity of acetic and pyruvic acids on the surface of water. Langmuir 2018, 34 (31), 9307–9313. 10.1021/acs.langmuir.8b01606. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

sp2c00206_si_001.pdf (357.4KB, pdf)

Articles from ACS Earth & Space Chemistry are provided here courtesy of American Chemical Society

RESOURCES