An Overview of Dynamic Heterogeneous Oxidations in the Troposphere

Abstract

Due to the adverse effect of atmospheric aerosols on public health and their ability to affect climate, extensive research has been undertaken in recent decades to understand their sources and sinks, as well as to study their physical and chemical properties. Atmospheric aerosols are important players in the Earth’s radiative budget, affecting incoming and outgoing solar radiation through absorption and scattering by direct and indirect means. While the cooling properties of pure inorganic aerosols are relatively well understood, the impact of organic aerosols on the radiative budget is unclear. Additionally, organic aerosols are transformed through chemical reactions during atmospheric transport. The resulting complex mixture of organic aerosol has variable physical and chemical properties that contribute further to the uncertainty of these species modifying the radiative budget. Correlations between oxidative processing and increased absorptivity, hygroscopicity, and cloud condensation nuclei activity have been observed, but the mechanisms behind these phenomena have remained unexplored. Herein, we review environmentally relevant heterogeneous mechanisms occurring on interfaces that contribute to the processing of aerosols. Recent laboratory studies exploring processes at the aerosol–air interface are highlighted as capable of generating the complexity observed in the environment. Furthermore, a variety of laboratory methods developed specifically to study these processes under environmentally relevant conditions are introduced. Remarkably, the heterogeneous mechanisms presented might neither be feasible in the gas phase nor in the bulk particle phase of aerosols at the fast rates enabled on interfaces. In conclusion, these surface mechanisms are important to better understand how organic aerosols are transformed in the atmosphere affecting the environment.

1. Atmospheric Aerosols: Classification and Importance to Societies

Atmospheric aerosols, or the liquid and solid particulates suspended in air, can be emitted to the atmosphere or be generated in situ. Particles emitted directly to the atmosphere are called primary aerosols, and can originate from either natural or anthropogenic processes [1]. Wind-blown mineral, dust, sea spray ejected from breaking waves, volcanic ash, pollen, and transported soil particles, among others, are all examples of natural particulates found in the atmosphere [2]. Smoke and soot from industry, controlled agricultural burns, and fires from burning wood, dominate anthropogenic primary aerosol emissions. Secondary aerosols are mainly generated through the condensation or coagulation of gaseous species or ultrafine particulate into larger particles or onto preexisting aerosols, such as black carbon or sea spray aerosols [3]. Secondary aerosols can also be formed through chemical reactions between volatile organics emitted from plants and a wide host of anthropogenic sources, and trace gas species such as nitrogen oxides (NO_x) emissions from fossil fuel combustion, dimethyl sulfide emissions over the oceans, and sulfur dioxide emissions from fossil fuel combustion and volcanoes [3,4,5,6]. These physical and chemical processes can alter the residence time of species in the atmosphere as well as change their optical and cloud condensation nuclei (CCN) properties.

It has been proposed that the natural aerosol loading of the planet can be modulated directly through anthropogenic emissions, or indirectly by influencing biogeochemical cycles [1,7,8]. Human-driven changes to the climate system, such as increased temperatures, and higher carbon dioxide (CO₂) or ozone (O₃) levels [9,10], can cause fluctuations in the aerosol emissions budget, as well. For instance, rising global temperature and CO₂ concentrations cause an increase in biogenic volatile organic emissions (BVOCs) [11,12]. However, expected emissions of BVOCs could be diminished because of deforestation, land usage changes, and losses of viable crop land due to aridification and salinification of soils [13,14]. Human activities, such as fossil fuel burning, industrial development, and controlled agricultural burns, can increase the amounts of sulfur, black carbon, and organic aerosols in the atmosphere [15]. Surface O₃ concentrations should reasonably increase with growing BVOC emissions [16]. However, increasing atmospheric moisture, driven by rising global temperatures, can partially counteract this trend by converting O₃ into hydroxyl radical (HO^•) [16,17]. Reductions in greenhouse gas emissions have, likewise, cut the amounts of anthropogenic aerosols released to the atmosphere [1]. In fact, some of the warming predicted by the end of this century might be attributable to smaller magnitudes of negative radiative forcing from aerosol–cloud and aerosol–radiation effects, due to a reduction in the total atmospheric aerosol loading [18,19,20]. Therefore, accounting for the impact of feedback loops on total aerosol loadings and trace gas species is important for predicting the effect of aerosols in the future climate.

Particle size highly affects the fate of atmospheric particles because it dictates the residence time of aerosols against natural removal from the atmosphere through conversion or deposition. In the atmosphere, existing particles typically have diameters from 1 nm to 100 μm. These particles can generally be grouped into four classes. In the first class, called nucleation mode (1–10 nm), aerosol physical processes—such as gas to particle conversion and the condensation of particles with water vapor—occur, resulting in the generation of larger particles. In the second class, named Aitken mode, particles are defined in the range of 10 to 100 nm in diameter, and can undergo further coagulation and condensation to form accumulation mode particles (the third class), which range from 0.1 to 1 μm in diameter. In the fourth class are windblown mineral dust, sea spray, pollen, and volcanic ash, as well as cloud-scavenged accumulation mode particles, that can range from 1 to 100 μm in size, and are termed coarse mode aerosols [3,21].

The size regime of atmospheric aerosols plays a key role in determining for how long the particulates reside in the atmosphere. Nucleation and Aitken-sized aerosols have short lifespans, on the order of a few hours to a few days, because they tend to coagulate into larger particles. Studies have demonstrated that these nucleation events have strong diurnal and seasonal patterns that are driven by BVOCs from forests and other vegetation. For example, the concentration of nanoparticles is the highest in the morning and early afternoon, when sunlight-driven gas-phase chemistry between trace gases and BVOCs promotes new particle formation [22,23]. By mid-afternoon, the mass concentration of the smallest particles decreases as they coagulate into larger particles [24]. This behavior was confirmed by observing slight increases in the mass concentration of accumulation and coarse-sized aerosols during the evening and overnight hours [22].

Coarse mode particulates have been demonstrated to have lifetimes shorter than a few days, which is two orders of magnitude smaller than the lifespan of accumulation mode particulates [25]. This is primarily due to gravity-driven deposition being the main route of atmospheric removal for coarse particles, regardless of the type of deposition surface [26]. Other studies have demonstrated that the time scales of formation and deposition of accumulation mode aerosols over vegetated areas are on the same order of magnitude [27]. The previous finding agrees with measurements collected over a botanical garden that demonstrated the amplitude of diurnal changes in the particle number concentration of accumulation mode particulates was quite small comparing to nucleation and Aitken-sized aerosols [22]. This is why the 0.1–1 μm particles are given the name “accumulation mode”, because they tend to collect and linger in the atmosphere for 1–3 weeks before they are removed via deposition [3,25,28]. Therefore, when reviewing below important heterogeneous atmospheric mechanisms, processes relevant to these accumulation mode aerosols are central for understanding the changes in composition and associated properties during aerosol aging.

Regardless of their source or fate, atmospheric aerosols have many direct and indirect impacts on the health of the Earth’s inhabitants and atmosphere [29]. The most obvious of these explicit effects are the deleterious health impacts observed from the inhalation of particulate matter (PM) with diameters less than 10 μm (PM₁₀). As the main component of urban smog, these small and ubiquitous particles pass into the lungs after being inhaled, leading to respiratory complications. Fine particulate matter with diameters smaller than 2.5 μm (PM_2.5) often enters the bloodstream, and contributes to systemic inflammation that can lead to hypertension, even after short exposure times [30]. Long term exposure to PM_2.5 has been linked to cardiovascular disease, respiratory diseases, asthma, low birth weight, lung cancer, stroke, and even diabetes [31,32]. These complications arising from PM_2.5 exposure have been estimated to cause 3.2 million premature deaths per year [33]. Consequently, the World Health Organization has set exposure guidelines used to shape regulatory policies based on total mass concentrations of PM_2.5. However, recent studies have suggested that the toxicity of PM can vary greatly based on composition and that using non-differentiated PM mass may result in an underestimation of its impact on mortality [34]. For instance, carbonaceous aerosols were determined to be up to 5 times more harmful than their inorganic counterparts and, thus, responsible for up to a quarter of all air pollution-related deaths annually [35,36]. Furthermore, it was determined that emissions from residential and commercial energy sources, such as cooking and heating fires, generators, and waste incineration, could represent 60–70% of PM-related mortalities in the populous areas of South Asia [36].

2. The Effect of Anthropogenic Aerosols on the Earth’s Radiative Balance

In addition to their implications on human health, aerosols can directly alter the radiation balance of the Earthsecondary by interacting with sunlight [37]. The Earth’s radiative balance describes the energetic equilibrium between the incoming solar radiation, and what is absorbed or reflected to space by the atmosphere and the surface. In this regard, radiative forcing is a measure of how much a particular atmospheric constituent traps (warms) or reflects (cools) incoming solar and outgoing terrestrial radiation [28]. The warming effect of methane, carbon dioxide, chlorofluorocarbons, and other greenhouse gases, is generally well-understood and documented at all but very low altitudes [15]. Thus, current efforts are focused in the utilization of small unmanned aerial systems (sUAS) capable of providing high spatiotemporal resolution of the level of greenhouse gases at low altitudes, to characterize emissions and provide information of the boundary layer needed to calibrate instruments in satellites [38]. Additionally, there is still a large degree of uncertainty in the net magnitude and direction (positive or negative) of the contribution of aerosols toward the radiative forcing budget, mainly attributable to their variable composition, size, optical properties, and hygroscopicity (Figure 1) [15,39,40,41].

In addition to directly modulating solar radiation, aerosols can indirectly impact the radiative balance by altering cloud properties and formation dynamics. Studies have demonstrated that increases in aerosol concentrations often translates to increased cloud droplet number concentrations, resulting in higher cloud albedo due to enhanced cloud formation [42,43]. Elevated aerosol concentrations also decrease the size of cloud droplets. These clouds do not produce precipitation efficiently and, thus, have longer lifetimes and increased horizontal and vertical extension that enhance the albedo as well [44]. However, it is still poorly understood how aerosol composition impacts aerosol-driven cloud enhancement.

Inorganic aerosols, such as sulfates, nitrates, and sea salt aerosols, are known to be excellent drivers of a net cooling radiative forcing, and act as excellent CCN by increasing the amount of water associated with the particles. This promotes enhanced cloud formation and thus heightens atmospheric reflectivity [39,45,46]. However, water is a potent infrared absorber and, thus, the magnitude of cooling effects exerted on the atmosphere by inorganic aerosols may be diminished [15]. Mineral dust, carried by winds from deserts, such as the Sahara, has been demonstrated to scatter incoming radiation to assist in cooling the planet [47]. Nevertheless, particles, including photoactive components, such as iron and titanium oxides, can increase the absorptive power of sand and dust particles, resulting in a net neutral or slightly positive radiative forcing [48]. Another important atmospheric aerosol constituent is black carbon (BC), which is produced from the incomplete combustion of fossil fuels, biofuels, and vegetation. BC is an efficient absorber of radiation and, thus, contributes a net positive effect on aerosol radiative forcing (Figure 1) [49]. This warming effect also continues once BC is deposited onto the Earth’s surface, reducing the amount of radiation reflected back to outer space via absorption. This phenomenon is thought to be contributing to the acceleration of ice and snow melting in the polar regions through heat trapping in BC deposits [50,51].

Unlike the extensive knowledge about the sources and properties of inorganic and BC aerosols, far less information is available regarding the fates of organic aerosols (OA) in the atmosphere [4,52,53,54]. Models presently underestimate the OA loading in the troposphere, particularly from biomass-burning sources, by 24% relative to field observations [55]. The OA loading and other key aerosol properties, such as optical absorptivity and hygroscopicity, seem to be driven by seasonal and regional variabilities [55,56]. Most previous attempts to constrain uncertainties from OA have focused on the formation of secondary organic aerosol (SOA) from biogenic sources, such as isoprene and several terpenes [57]. To increase the certainty in future global temperature and climate forecasts made by climate models, knowledge of the fate of anthropogenically driven organic compounds in the atmosphere is needed [58]. In addition, better inventories of sources and sinks of atmospheric oxidizers are necessary to predict how oxidation contributes to aerosol aging [1,59]. The study of the transformations of overlooked sources of SOA, such as biomass-burning aerosols and polycyclic aromatic hydrocarbons under atmospherically relevant conditions, provided information needed to constrain the uncertainties that OA contribute to radiative forcing [60,61,62]. Correlations between the oxidative processing and increased absorptivity, hygroscopicity, and CCN activity of aerosols have been proposed [63]. Oxidation also amplifies the ability of OA to absorb radiation in the near-ultraviolet range. These competing factors make the impact of OA on the net radiative forcing budget complicated [40,41]. Thus, the continuous transformation of OA into aged secondary organic aerosol during atmospheric transport [64] changes their physical and chemical properties. As a result of these changes, there is a high level of uncertainty associated to the radiative forcing due to aerosol–radiation interactions (RF_ari) [15]; see Figure 1.

Recent studies have also demonstrated that as SOA undergoes subsequent stages of aging, it becomes increasingly water soluble and absorbs more light [65,66]. This aged absorptive organic aerosol fraction is often called brown carbon, after the light brown color of collected samples [67]. Field observations suggest that reactions taking place in clouds are critical for the secondary formation of brown carbon, which is thought to be an underrepresented source of positive radiative forcing [41,68]. To sum up, the brown carbon contribution from SOA has been identified as one of the key uncertainties of the anthropogenic effect of aerosols on climate [15]. Moreover, heterogeneous reactions have been proposed as important for the aging of aerosols, increasing the complexity of the problem [69]. Therefore, in order to better constrain the impacts of SOA properties on climate, a better understanding needs to be gained on the heterogeneous processing of aerosols, the products of these reactions, and how these products transform the optical and hygroscopic properties of SOA.

3. Processing at Atmospheric Interfaces

Much of the chemical processing of aerosols occurs at the air–particle interface, which can be the region between the gas and either a solid, a semisolid, or a liquid. Different products may result from reactions on each kind of interface [70]. The interaction and uptake of a gas molecule by an aerosol particle occurs in a similar manner regardless of the phase, but additional parameters are needed to describe the processes on solid or semisolid interfaces, which are summarized next, and illustrated in Figure 2. In a general model, the aerosol surface can be seen as interacting with a number of gas molecules colliding on it. This collisional flux can be defined as

$$J_{coll}= \frac{\upsilon {\left[X\right]}_g}{4},$$

(1)

where [X]_g is the concentration of the species X in the gas phase, and $\upsilon$ is the mean thermal velocity of X [71]. As gas molecules collide with the aerosol surface, a portion of these equilibrate with the surface. This accommodation process is dictated by the temperature, composition, and pressure of the gas, as well as the phase and state of the surface [72].

Experimental observations are often corrected with a thermal accommodation coefficient, ${\alpha }_T$, which can be described by the relationship

$${\alpha }_T= \frac{E_{in}-E_{re}}{E_{in}-E_w},$$

(2)

where E_in is the incident energy flux, E_re is the reflected energy flux, and E_w is the flux of emitted molecules if they were in thermal equilibrium with the surface [72,73,74]. Molecules that do not immediately bounce off the surface can be held by van der Waals interactions or hydrogen bonds. This process, known as adsorption, is often termed physisorption because it represents the contribution of physical processes to the sticking of gaseous species to the interface. The adsorption flux can be estimated by the following equation:

$$J_{ads}= \frac{{\alpha }_s \upsilon }{4\left(1-\theta \right){\left[X\right]}_g},$$

(3)

with the surface accommodation coefficient α_s, which is the ratio of accommodated molecules to the number of molecular collisions, and θ is the fractional surface coverage by the species of interest [71,75].

The fate of the adsorbed molecule can be one of the next three processes: (1) absorption, (2) chemisorption, and (3) desorption. A portion of the adsorbed molecules are taken up and accommodated into the aerosol in a process called absorption. The absorption flux can be defined as

$$J_{abs}= {\alpha }_b J_{coll},$$

(4)

where ${\alpha }_b$ is the bulk accommodation coefficient [71] that is determined by the relationship

$${\alpha }_b= {\alpha }_s\frac{k_{sb}}{k_{sb}+ k_{des}}.$$

(5)

In Equation (5), $k_{sb} \mathrm{and}{k}_{des}$are the surface to bulk transfer rate constant and the desorption rate constants, respectively [75]. Alternatively, adsorbed molecules can become chemisorbed to the aerosol if they undergo chemical reactions with other adsorbed species or with aerosol constituents. Chemisorption processes, such as these, are apparent through a hyperbolic relationship between the concentration of the gas and the reaction rate. This is often described by the Langmuir–Hinshelwood adsorption mechanism, which provides information regarding the partitioning of the reactant gas between the gas phase and the particle phase, followed by the reaction between the adsorbed gas and species in and on the aerosol [75,76].

Finally, a portion of the adsorbed molecules can detach or desorb from the surface back into the gas phase. The desorption flux can be expressed as

$$J_{des}= {\left[X\right]}_g {\tau }_d{}^{-1},$$

(6)

where ${\tau }_d$ is the desorption lifetime [71]. The net uptake of the molecules into the aerosols, J_net, describes the difference between the adsorption (J_ads) and desorption (J_des) fluxes. However, in practice, it is extremely challenging to isolate and measure any of the individual processes needed to define the adsorption and desorption fluxes. Therefore, many studies have focused on determining the reactive uptake of gases by aerosol surfaces (γ). This quantity describes the overall uptake for all surface processes combined, regardless if they are physical or chemical in nature [77,78]. While γ cannot be measured directly, it can be described using the experimentally observed reactive uptake of a gas into the interfacial region, ${\gamma }_{eff}$ Equation (7):

$$\frac{1}{{\gamma }_{eff}}= \frac{1}{{\Gamma }_g}+ \frac{1}{\gamma },$$

(7)

where ${\Gamma }_g$is a correction factor defined as

$${\Gamma }_g=8D_g{\nu }^{-1}{d}_p{}^{-1},$$

(8)

which includes the gas-phase diffusion coefficient for the compound of interest (D_g) and the particle diameter (d_p) [71]. Therefore, the net flux between a gas and the surface of an aerosol, J_net [78], can also be defined as

$$J_{net}= \gamma J_{coll}.$$

(9)

In order to better comprehend the oxidative aging processes that aerosols undergo, it is imperative to understand how oxidants interact with aerosol surfaces [78,79]. Competition between chemical reactions during adsorption and absorption can lead to radical chemistry in both the bulk and the interface [70]. This complex interaction between aerosols and the atmosphere gives rise to multiple reaction pathways “in” and “on” aerosols, leading to the complex compositions observed in field observations [70]. With a deeper understanding of how the processes occur at the surface of aerosols, the resulting physical and chemical outcomes affecting optical and chemical transformations can be revealed.

4. Aging of Aerosols in the Atmosphere

Understanding the atmospheric aging of aerosol altering its properties is a topic of current interest to atmospheric chemists and climate scientists [69]. Gas-phase reactions between atmospheric oxidizers and volatile organic species have been studied extensively. In general, common atmospheric oxidants, like ozone, nitric oxide, and peroxides, react with gas-phase organic molecules to generate less volatile and more polar products that partition to the particle phase, forming SOA [4,63,79]. Studying how SOA behaves as it is transported through the atmosphere is an important matter to describe the way urban pollution becomes more oxygenated (higher O/C ratio) during the aging process [63]. Despite the varied sources of organic aerosols, oxalic acid has been identified as a pervasive component of aged SOA, representing about 50% of the total OA mass in aerosols collected in a rural location downwind of agricultural and urban areas [80]. Thus, there must be a common mechanism of oxidative chemical processing that generates oxalic acid as an end product in the atmosphere [60,61,62,80,81,82,83]. In order to elucidate such a mechanism of oxidative aging, it is imperative to understand the major routes atmospheric aerosol processing can take, and how they can contribute to aerosol complexity. Even though physical processing (Section 3) drives the formation of new aerosols from the emission of natural and anthropogenic particles, the fate of these newborn particles during transport can be extremely complex, due to the many forces acting on them.

The chemical processing of SOA generating aged SOA can proceed through oxidative pathways, non-oxidative pathways, or photochemical pathways [64]. During oxidative processing, trace gases react with organic species on the surface of the aerosol to produce alkyl radicals R^• (Figure 3). Molecular oxygen (O₂) then adds to the radical to form the highly reactive alkylperoxy radical RO₂^•, which then can undergo reactions with a host of atmospheric radicals (NO, NO₂, RO₂^•, HO₂^•, NO₃^•) [64]. These reactions result in the addition of polar functional groups, such as hydroperoxyl (HOO–), nitrate (–NO₃), nitro (–NO₂), hydroxyl (–OH), carboxylic acid (–COOH), carbonyls (>C=O), among others [84]. These added polar groups, in addition to increasing hygroscopicity, decrease molecule volatility [85]. These oxidation processes enhance further SOA formation via condensation and coagulation.

Semi-volatile species produced from oxidation reactions can undergo accretion reactions in aerosols to form heavier species after partitioning into the particle phase [64,86]. This non-oxidative pathway can take many forms, but most, if not all, accretion reactions involve polymerization or radical recombination of organic compounds. The resulting products are heavier and nonvolatile complex oligomeric species unlikely to be formed by oxidation in the gas phase. Based on observations from field studies, it has been inferred that accretion reactions are enhanced in acidic aerosols by acid-catalyzed nucleophilic additions to aldehydes, carbonyls, and alcohols [86]. Additionally, alkenes, such as limonene, have also been shown to more readily undergo accretion reactions such as dimerization in the presence of strong acids [87,88]. Two pathways dominated the oxidation of β-pinene via nitrate radicals (NO₃^•), the reactions RO₂^• + NO₃^• and RO₂^• + HO₂^•, with similar SOA yields from both channels [89]. As many atmospheric aerosols are comprised of acidic components, these reactions represent a common pathway for the formation of non-volatile species in aerosols [90].

In addition to chemical reactions, sunlight can also facilitate the processing, accretion, and aging chemistry in and on atmospheric particles [82,83,91,92,93,94,95,96,97]. Noteworthy, when gaseous carboxylic acid partition from air into the particle phase, they exert enhanced acidities to the water interface that are at least larger than 1.7 pH-units than predicted by bulk pK_a values [95]. Organic and inorganic molecules can act as chromophores, or light absorbing moieties, that absorb visible and ultraviolet radiation. These compounds, such as conjugated double bonds or carbonyls, lead to the formation of excited state molecules that can undergo homolytic bond cleavage, radical recombination, electron transfer, decomposition, hydrogen abstraction, photosensitization, and proton coupled electron transfer [82,83,93,94,96,97,98,99,100,101]. The resulting products, which are similar to what is observed with particle-phase accretion chemistry, dehydrate or cyclize during the night to form oligomers or supramacromolecular structures. When the sun rises the next day, photons are absorbed by these complexes, leading to their dissociation. This pattern of diurnal cycling is commonly seen with α-dicarbonyls, which are prevalent in aged OA, and is a possible source of positive radiative forcing from OA [82,83,92,97].

Sunlight can also accomplish aerosol processing indirectly via the photochemical production of hydroxyl radical (HO^•), which is one of the most abundant tropospheric oxidizers [102]. Volatile organics often first undergo gas-phase oxidation by HO^•, decreasing volatility and encouraging coagulation to form SOA. Photochemistry can also alter the balance of available trace oxidizers, such as ozone and nitrogen dioxide, through the cycling of halogen radicals [103]. Therefore, photochemical processes are an important factor in the direct and indirect production of SOA through oxidative and non-oxidative means.

All the aerosol processes discussed above can change the physiochemical properties of SOA. It has been demonstrated that higher O/C ratios in aerosols enhances hygroscopicity due to the greater hydrophilic character of the newly formed polar functional groups [4]. Highly oxygenated SOA can serve as CCN at lower relative humidity, whereas more hydrophobic OA can only do this in very moist environments. Diminished compound volatility has also been observed relative to an elevated O/C ratio. As aging proceeds, volatility can be altered further via oligomerization and fragmentation [104]. All of these processes can amplify the number of light absorbing functional groups, which increase the absorbing power of aged SOA [91]. Oligomerization is also thought to be a contributing factor to the creation of light absorbing brown carbon in humic-like substances (HULIS) [105,106]. Through understanding the interplay between chemical processing and aerosol properties, more complete mechanistic models can be developed to describe the evolution of aerosols, including their composition and absorptivity [64,91].

5. Environmental Information from Laboratory Studies of Heterogeneous Reactions

While the physical and chemical processes that lead to aerosol formation in the environment are a current focus of study, it remains very unclear how the compositional complexity observed in field measurements is generated. Gas phase and aqueous phase (bulk) studies can guide the initial prediction of the products expected during interfacial reactions, but are physically limited and often underestimate the variety of possible products and pathways of atmospheric significance [60,61,107]. Cloud chambers [108,109,110], Knudsen cells [111,112], droplet train flow reactors [113,114,115], flow tube reactors [116,117,118,119], and molecular dynamics simulations [120] have been used to investigate different properties of aerosols and the results of atmospheric processing. Multiple studies utilizing the methods outlined above indicated that a Langmuir–Hinshelwood adsorption mechanism generally controls the reactive uptake of oxidants by the surface of aerosols [115,116,118]. Additionally, it was determined that ozonolysis on the particle surfaces could be a key source of gas-phase aldehyde production [117]. However, the previous methods do not always provide atmospherically relevant results reflecting aerosol surface to volume ratios, interfacial diffusion rates and mass accommodation values, the presence of organic coatings and water associated with particles, or the hydrophobicity of the components [70].

X-ray Photoelectron Spectroscopy (XPS) has been used to probe the composition of ambient aerosols at the surface of collected particles, demonstrating its strong correlation with size [121]. For instance, carbon content is elevated in accumulation mode aerosols [121], which suggests that these particulates are likely dominated by OA born in nucleation events. Furthermore, chemical species associated with mineral dusts and soil (e.g., silicates, carbonates, and iron oxides) were only observed in coarse mode particulates [121]. XPS was also applied to analyze BC aerosol generated in a laboratory chamber showing that sulfur associated with diesel particles was present only as sulfate [122]. Further efforts to couple XPS with mass spectrometry (MS) enabled a better understanding of the differences in chemical composition and speciation of aerosol surfaces with respect to the bulk particle [123]. Moreover, the XPS MS technique previously demonstrated how deposition of inorganic species on soot and BC aerosols could decrease their overall hydrophobicity [124]. Overall, XPS analysis of aerosols provided relevant information to describe how changes in composition at the interface can improve CCN activity or increased absorptivity of aerosols.

Collected aerosols from field studies have been characterized using aerosol time-of-flight mass spectrometers [125,126,127], electron microscopy [128,129], X-ray diffraction [130,131], and many other techniques to determine the outcome of atmospheric aging. These techniques provide insight into the composition of the aerosol, but generally do not permit tracking of how the particles arrived to that state. Recently, the combination of multiple MS techniques was used to examine several generations of organic aerosols derived from the gas-phase oxidation of α-pinene, showing a drop in the volatility of OA constituents during aging [132]. However, no mechanism was proposed as to how these reactions occur. Therefore, new instrumental methods should be also capable of providing mechanistic insights of the aging process (Section 4), e.g., by probing the heterogeneous oxidation of aerosol proxies such as iodide and substituted catechols [60,62,107]. The main outcome of this new surface sensitive MS method, developed with proxies, was to have identified reaction pathways that predominate in the interface of air with water [60,62,107] and, in addition, to have compared the surface oxidation of solid catechol in air under variable relative humidity [61]. The last findings showed the mechanistic pathways [60,61,62] by which organic precursors are created to change the optical properties of aerosols [82,83,91,92].

The oxidation of aqueous particles has been investigated using droplet generation systems paired with inline detection systems, to allow for the observation of intermediates that might be lost in traditional collection and offline processing methods. Nebulizers coupled to flow tubes interfaced with mass spectrometers, particle sizers, and infrared spectrometers, have been employed in the laboratory to examine how transformations might occur in clouds [116,133]. However, interfacial reactions often proceed too quickly to capture reactive intermediates in these systems. To bridge this gap, online electrospray ionization mass spectrometry (OESI-MS) has been developed (Figure 4) [60,62,95,107]. In OESI-MS, a nebulizing system interfaced with an electrically grounded ESI system provided aqueous microdroplets that can probe reactions at the air–water interface with shortened analysis times, as the contact time of the colliding gas with the aqueous surface only lasts for a few microseconds [60,107].

The major characteristic of these pneumatically assisted aerosolization systems has been described to be the high ratio of nebulizing gas to liquid flow (1.2 × 10³) [134]. Ions formed after the microsecond-lasting reaction are ejected off the droplet surfaces in a series of cascade desolvation steps [135], to be detected by a mass spectrometer within 1 millisecond. As a result, this analytical method allows for selectively sampling the reactants and products in the microdroplets interface. The customized instrument has allowed for the delivery of gaseous ozone to examine the in situ oxidation of inorganic and organic species on the surface of aqueous microdroplets under atmospheric relevant conditions [60,62,107].

The interfacial conversion of iodide to hypoiodous acid and molecular iodine on the surface of aqueous microdroplets exposed to O₃ was explored in detail by OESI-MS [107]. This study proposed the important role of reactions occurring on the surface of sea spray aerosols and their implications in tropospheric ozone depletion [107]. A related study demonstrated that iodide and bromide can be respectively enriched by 7.3 × 10³ and 7.5 × 10² times at the air–water interface relative to bulk seawater (Figure 5), as it happens when sea spray aerosol is formed during bubble bursting and wave breaking processes [134]. In turn, tropospheric ozone can react with the surface available iodide to produce reactive halogen species (RHS), molecular iodine, and hypoiodous acid. This mechanism of RHS production was proposed as important to provide the missing light sensitive precursors needed to deplete ozone over the open ocean [107].

The fast oxidation by ozone and in situ-produced hydroxyl radicals of important hydroxylated aromatics, such as catechol, was studied in detail at the air–water interface by OESI-MS [60]. Catechol, 1,2-diihydroxybenzene, is widely found in biomass-burning emissions. While catechol was previously thought to primarily undergo ring-opening reactions when exposed to ozone [136], the application of OESI-MS resolved the simultaneous contribution to oxidation by an indirect electron transfer-initiated oxidation, providing HO^• radicals [60]. In addition to observing common catechol ozonolysis products [137], it was proposed, for the first time, that oxidation by HO^• caused the production of polyhydroxylated aromatic compounds and quinones in atmospheric particles [60]. This mechanism provided an explanation for the production of light absorbing brown carbon, along with how complex mixtures of organic compounds are generated [60,61]. Carefully performed analysis spiking standards indicated that when microdroplets containing 100 µM catechol at pH 8.0 were exposed to 3.18 ppmv O₃ (g) during a contact time of 1 µs, the concentration of the main low molecular weight carboxylic acids produced at the interface was [glyoxylic acid] = 200 nM > [oxalic acid] = 174 nM > [cis,cis-muconic acid] = 129 nM > [3-oxopropanoic acid] = 109 nM [60].

Furthermore, the previous concepts were expanded to a series of substituted catechols that served as proxies to explore how anthropogenically emitted phenols (or oxy-aromatics) derived from benzene, anisole, and toluene, are oxidized at the air–water interface [62]. A variety of some important polyphenols and quinones produced during the oxidation of substituted catechols at the air–water interface are included in Figure 6 [62], as an example of the many different species accounted by these mechanisms [60,61,62]. The work predicted, thermodynamically, the trend of favorable electron transfer reactions of the substituted catechols with ozone, and concluded that stronger electron donors to the aromatic system lead to an enhancement in the rate of product formation during oxidation [62]. The interfacial oxidation of catechol [60,61,62] has also been proposed to be affected by Baeyer–Villiger (BV) oxidation of the unsaturated dicarboxylic acids, generating unsaturated, oxo-, and dicarboxylic acids [60,61]. A pioneer constrain for the interfacial production of HO^• of (1.1 ± 0.3) × 10¹³ radicals cm⁻³ was determined when microdroplets with 100 μM of substituted catechol, at pH 7.8, reacted with 2.7 ppmv O₃(g) [62].

For the study of reactions on solid surfaces in the laboratory, the preparation of thin films with proxy species found in solid atmospheric particles has been useful to investigate the uptake of reactive gases and characterize the transformations that alter the initial composition. Studies in aerosol chambers and flow tubes [138,139,140,141,142] have demonstrated how to use gas-phase reactions to generate SOA in controlled environments. The resulting aerosols are analyzed by different techniques, e.g., optical spectroscopy, before they are collected for further characterization by chromatography and mass spectrometry [139]. Variations in the physical state of the aerosol, reaction time, size, and the procedure for measuring rate of loss, can lead to the dissimilar reaction rates and uptake coefficients reported by different groups studying the same molecules [70]. These approaches typically capture the variations in composition in the early stages of aging, but are not necessarily representative of the extended aging periods experienced during transport [61].

More recently, thin films have been used to simulate relevant reactions on the aerosol surface modeling the air–particle interface [61,116,136,143]. This approach has proven quite useful to monitor the degradation of the proxy materials and evolution of condensed and gas phase products in gas cells or environmental chambers [144]. While fast processes in the microsecond timescale at the air–water interface were studied in the customized OESI-MS set-up described above [60], most likely, some products that form over longer times could have remained hidden. Therefore, an additional system was developed to study oxidative processing on surfaces that could pair the real-time monitoring from OESI-MS [60,62] with offline compositional analysis by multiple techniques [61]. A clear picture of reactions on solid surfaces was provided by this new flow-through reaction chamber (Figure 7) [61]. Oxidation resistant infrared holders, with optical windows supporting the thin films of proxy species, were placed into a reaction chamber. Ozone under variable relative humidity (0–90%) could be circulated through the reaction chamber at a constant flow rate. In such a system, samples could be regularly scanned, spectroscopically, in the infrared, to monitor the oxidation process [61]. Of note is that protocols were also developed to extract the reacted films and analyze their products by chromatography and mass spectrometry [61].

This flow-through reaction chamber was used to examine the heterogeneous oxidation of catechol at the air–solid interface. The work employed infrared spectroscopy, UV–visible spectroscopy, chromatography, and mass spectrometry to analyze the reaction products. Not only did this experimental design validate findings using the air–water interface technique, but also provided new details of intermediates and products that escaped detection in the microsecond contact time of OESI-MS [60,61]. The longer exposure times in the air–solid interface study enabled the observation of organic acids produced from direct ozonolysis, and polyhydroxylated aromatics, bi- and terphenyls products via the crosslinking of reactive semiquinone intermediates [61]. Relative humidity was found to be a key driver in the reaction rate at the air–solid interface [61], suggesting that adsorbed water facilitates ring cleavage and electron transfer as observed in the air–water interface study [60].

6. Outcomes Learned from Recent Interfacial Studies

This review has discussed recent important heterogeneous atmospheric mechanisms that contribute to the understanding of the composition, processing, and properties of aerosols. The first important conclusion from these recent heterogeneous oxidation studies is that ozone is an important player in tropospheric chemistry, which enables the cycling of radicals and other trace atmospheric species by previously unconsidered mechanisms. This secondary pollutant formed through a series of gas phase reactions [29,145,146] is widely available to oxidize many atmospheric species as described above. When ozone reacts either with iodide or a substituted catechol, a wide range of reactive intermediates are produced alongside the oxidized products [60,61,62,107]. The generation of radicals can further propagate chain reactions through previously unexpected channels, as evidenced by the hydroxylation pathway observed during the oxidation catechol by ozone under humid conditions [60,61,62]. Both inorganic and organic systems revealed that tropospheric pollution can impact atmospheric composition. Elevated emissions of volatile organic compounds (VOCs) and NO_x in urban areas favor high O₃ levels and, thus, impact the oxidation mechanism on aerosols surfaces [146].

A second important takeaway from examining heterogeneous reactions at the atmosphere–aerosol interface is that the oxidation of organic aerosols by trace species can dramatically alter their composition, while increasing their hygroscopicity and UV–visible absorptivity. It was determined that SOA and some primary OA are prone to undergo multistep chemical reactions with trace atmospheric oxidizers (e.g., HO^•, O₃, NO_x) and sunlight, that alter their composition [60,61,91,92,93,107]. Each successive step of oxidation has the ability to add polar functional groups, decreasing the volatility of the constituents [4]. This process leads to the partitioning of organics into the aqueous phase of particles, where further processing can occur [147,148]. Additionally, with the incorporation of polar functional groups, new chromophores can be generated in these secondary organic aerosols along with organic aerosols generated from combustion or biomass-burning events [60,61,62]. As a consequence, the processed organics can absorb ultraviolet and visible radiation [149].

The third remark is that investment in designing, developing, and testing advanced instrumentation and reactors in the laboratory is a key aspect to make progress in this area of research. Without such efforts, it would be impossible to reveal the mechanistic information needed to predict the effect of aerosols. The use of the advanced and creative methods, summarized above, has pushed the boundaries of current knowledge about how aerosols undergo oxidation reactions at atmospheric interfaces. OESI-MS has been a fundamental tool for studying surface chemistry by actually examining the structure of the interfacial region and reactions occurring in this boundary region [60,62,95,107,134,150,151]. The experiments studying the oxidation of iodide [107], catechol [60], and substituted catechols [62] constitute the forefront measurements to observe the products of chemical aging at the air–water interface. The combined use of FTIR and UV–visible spectroscopies, mass spectrometry, and chromatography to characterize aerosol proxies reacting at the air–solid interface under variable humidity has been demonstrated [61]. The setup utilized a flow-through reactor to provide the reactive uptake coefficient of ozone on catechol films [61] under variable humidity conditions, and elucidated the formation of quinones as well as biphenyl and terphenyl rings [61]. Present and future laboratory, chamber, and field studies to examine the product mixtures resulting from these mechanisms are being inspired [152,153,154,155,156,157,158,159,160,161,162,163] by this work, to gather information for inclusion in the next generation of computational models to predict the effects of aerosols on climate.

Finally, this review has provided examples describing how to design laboratory studies to collect and interpret the data of oxidative aging of aerosols. Aerosol and cloud water samples collected at different geographical locations often display similarities in composition, despite significant differences in parent species [80,164,165,166]. The mechanisms discussed above explain the processing of different organic atmospheric precursors, which are found in both biomass burning [167] and anthropogenic emissions [168], and likely proceed through similar pathways in the atmosphere [60,61,62]. Oxy-aromatics can generate a rich product mixture that absorbs more light, forms better nuclei to form clouds, and partitions into clouds easily [60,61,62]. The absorptivity of the oxidation products grows in the visible region both in films and aqueous solutions, contributing to the “browning” of aerosols during the aging process [149]. The highly oxidized products contain carboxylic acid and hydroxyl moieties that justify the extremely hygroscopic nature of this class of aerosols from aromatic precursors [45,46]. Therefore, these kind of products are expected to easily partition into cloud waters and serve as precursors to the formation of humic-like materials [169,170]. Future efforts should explore some of these ideas further in the laboratory, by developing new experimental methods and instrumentation capable of investigating reactions at atmospheric interfaces.