Notice: Wiley Online Library will be unavailable on Saturday 30th July 2016 from 08:00-11:00 BST / 03:00-06:00 EST / 15:00-18:00 SGT for essential maintenance. Apologies for the inconvenience.
Department of Atmospheric Sciences, Center for Atmospheric Chemistry and Environment, Texas A&M University, College Station, Texas, USA
Corresponding authors: S. D. Brooks and R. Zhang, Department of Atmospheric Sciences, Center for Atmospheric Chemistry and Environment, Texas A&M University, College Station, TX 77843, USA. (firstname.lastname@example.org), (email@example.com)
 Although secondary organic aerosol formation is well studied, the extent to which oxidation products of biogenic volatile organic compounds condense onto primary aerosols and modify their cloud-nucleating properties remains highly uncertain. Here we show that water-soluble organic acids produced from the reaction between α-pinene and ozone rapidly accumulate onto preexisting particles forming coatings of organic materials that reach a mass fraction of 80–90% within a time period of 30 to 60 min for the reactant conditions of 7 to 37 ppbv α-pinene and 20 ppbv ozone. Cloud condensation nuclei (CCN) measurements reveal that the initially hydrophobic aerosols are rapidly converted to efficient CCN at a supersaturation of 0.22%. Our results imply that changes in the activation potential of a significant fraction of the atmospheric aerosol population are controlled by the formation and composition of coatings formed during the aging process, rather than by the original particle size or composition.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
 Globally, emissions of volatile organic compounds (VOCs) from the biosphere are estimated to be about 1000 Tg yr−1, much larger than those from anthropogenic sources [Guenther et al., 1995]. Biogenic nonmethane VOCs mainly include isoprene, sesquiterpenes, and monoterpenes, of which α-pinene is the most abundant monoterpene. Products of the photochemical oxidation of biogenic VOCs represent a major secondary organic aerosol (SOA) precursor in the atmosphere (up to 90% or more by mass) [Tunved et al., 2006; Johnson and Marston, 2008; Hallquist et al., 2009; Jimenez et al., 2009]. Further, in the presence of primary aerosols, such as biomass burning, fossil fuel combustion, soil, or plant wax particles, much of the condensable material formed through VOC reactions may not form new particles [Zhang et al., 2009, 2012] but instead may condense onto existing aerosols and alter their microphysical properties [Hudson et al., 2004; Sareen et al., 2013]. An understanding of aerosol aging is pivotal to assess the roles of emissions from the biosphere in the direct and indirect forcing of aerosols on climate [Pöschl et al., 2010; Intergovernmental Panel on Climate Change(IPCC), 2013].
 Conventionally, Köhler theory is employed to predict the critical supersaturation of aerosols, a threshold above which particle growth by water vapor uptake and formation of a stable cloud droplet becomes thermodynamically favorable [Pruppacher and Klett, 2000]. However, recent field measurements reveal conflicting evidence regarding the relative importance of particle size and composition in predicting CCN activation. For example, measurements over a continental field site suggest that CCN concentrations are mainly determined by the aerosol size distribution [Dusek et al., 2006]. In contrast, measurements in a wide variety of locations indicate that the chemical composition is important in determining the cloud nucleating ability of aerosols [Hudson, 2007; Furutani et al., 2008; Quinn et al., 2008; Twohy and Anderson, 2008; Topping and McFiggans, 2012].
 Freshly emitted soot particles are hydrophobic and must undergo aging prior to becoming cloud condensation nuclei [Zhang et al., 2008; Khalizov et al., 2009; Pagels et al., 2009]. Many global climate models address the aging of soot empirically by assuming that fresh particles are converted into hydrophilic CCN on a moderate time scale of 1 day or more [Jacobson, 2001; Cooke et al., 2002; Koch et al., 2009a; Bauer et al., 2010]. Despite much recent focus on the role of soot in climate models, aging processes are represented by oversimplified parameterizations and the interactions between soot particles and biogenic emissions are excluded [Bond et al., 2013]. Even in the most recent climate studies, soot may be divided into hydrophilic soot and hydrophobic soot categories, but no explicit calculations of the aging processes are included [Oshima et al., 2009]. In a recent study, the heterogeneous oxidation of soot was treated in a more detailed approach using a particle-resolved aerosol model [Kaiser et al., 2011]. Clearly, it is necessary to incorporate biogenic aging processes into atmospheric chemical transport and climate models.
 In addition to evidence that SOA can independently act as CCN [Engelhart et al., 2008, 2011; King et al., 2009; Asa-Awuku et al., 2010; Ervens et al., 2011; Frosch et al., 2011; Lambe et al., 2011; Pierce et al., 2012], the forming of condensed phase products may modify a large number of preexisting primary aerosols. For example, biomass burning particles alone comprise 33% of particles over North America [Hudson et al., 2004]. Since primary aerosols are often hydrophobic and require a critical supersaturation higher than saturation levels obtainable in the atmosphere, the formation of coatings that reduce the critical supersaturation may greatly improve the propensity of the aerosol to become CCN [Saathoff et al., 2003; Schnaiter et al., 2005; Xue et al., 2009; Tritscher et al., 2011; Vaden et al., 2011].
 In this study, we investigated the mechanism and time scale of primary aerosol conversion to SOA. In an environmental chamber, monodisperse soot and polystyrene latex (PSL) particles were used as model primary aerosols to evaluate their transformation to SOA by exposure to the oxidation products from α-pinene ozonolysis for the reactant concentrations 7 to 37 ppbv α-pinene and 20 ppbv ozone.
 Experiments were conducted in a fluoropolymer environmental chamber equipped with a soot generation system, a differential mobility analyzer for producing size classified soot aerosols, a reactant injection and monitoring system, and measurement systems to determine particle size, and mass, as used in our recent studies (supporting information, Figure S1) [Wang et al., 2010; Qiu et al., 2012; Khalizov et al., 2013]. In addition, cloud activation by aerosol was quantified using combined measurements of a second postchamber differential mobility analyzer, a CCN counter (Droplet Measurement Technologies), and a condensation nuclei counter. In a separate series of experiments, coated soot and PSL aerosols from the chamber were analyzed using thermal desorption–ion drift chemical ionization mass spectrometry (TD-ID-CIMS) [Zhang et al., 2009; Wang et al., 2010]. Experiment details are provided in the supporting information.
3 Results and Discussion
 Changes in the particle coating thickness and other physical properties occurring during the aging of soot aerosol were quantified throughout the experiments. Figures 1a and 1b show the size and mass growth factors (D/D0 and m/m0), defined as the ratio of the mean particle diameter or mass at a given time to that of the particle initial core diameter or mass. The size and mass growth factors rapidly increase shortly after the ozonolysis reaction begins for all initial core particle sizes. The particle organic mass fraction, (m − m0)/m, also increases significantly and reaches ~80–90% within 30 min (Figure 1c). Measurements of the fresh and coated particle mass and diameter are used to estimate the coating thickness, Δrve (Figure 1d). For all initial core sizes, the values of Δrve increase similarly to ~14, 27, and 41 nm, at 15, 30, and 60 min, respectively. Hence, the oxidation products of the ozonolysis reaction of α-pinene rapidly accumulate onto preexisting particles to develop organic coatings that convert the particles almost entirely to SOA. This process is much more efficient than photochemical oxidation of other anthropogenic (i.e., toluene) and other biogenic compounds (i.e., isoprene) observed in similar chamber studies [Qiu et al., 2012; Khalizov et al., 2013].
 To identify the species responsible for the rapid conversion of primary particles to SOA in these experiments, we analyzed the chemical composition of organic materials coated on the soot cores using TD-ID-CIMS [Zhang et al., 2009; Wang et al., 2010]. The formation of pinalic acid, norpinic acid, pinonic acid, pinic acid, and hydroxyl-pinonic acid was confirmed by the presence of their complexes at m/z = 202, 204, 216, 218, and 232 in the spectra, respectively (see supporting information, Figure S2). The observation of organic acids in the particle coating is consistent with their low volatilities, and their relative abundances are likely attributed to a combination of their reaction yields, saturation vapor pressures, equilibrium partition coefficients, and water solubility [Pankow, 2007; Pankow et al., 2001].
 The conversion of hydrophobic soot particles into CCN during the aerosol aging process is illustrated by the temporal variation in the number of soot particles with a core diameter of 103 nm that activate as CCN at 0.18% (blue) and 0.22% supersaturation (red) (Figure 2a). Fresh soot particles are ineffective CCN, as evidenced by the near-zero concentration of CCN detected prior to injection of ozone and α-pinene. For the 0.22% and 0.18% supersaturation levels, particles start to activate in as short as ~10 and 15 min, respectively, and complete CCN activation is achieved at ~30 and 60 min, respectively. Coating thicknesses of about 10 and 15 nm are sufficient to noticeably increase the percentage of aerosols acting as CCN at 0.22% and 0.18% supersaturation, respectively (Figure 2b).
 The critical supersaturations required for cloud activation of coated soot particles are summarized in Figure 3. For comparison, the critical supersaturations of pure particles composed of fresh soot, pure ammonium sulfate, and single component organic acids are also shown. For particles of identical composition, an increase in the particle size causes a reduction in the critical saturation, because of a reduced curvature effect and an increased soluble mass [Pruppacher and Klett, 2000]. The effect of the chemical composition of single-component aerosols on critical saturation is illustrated by the difference between water-soluble ammonium sulfate particles, which require modest supersaturations to activate, and fresh hydrophobic soot particles, which require much higher supersaturations. The critical supersaturations of pure acids, i.e., pinic acid and norpinic acid, observed here are in good agreement with critical supersaturations of purely SOA particles in the literature [Prenni, 2007 and references within]. In general, CCN activation by SOA can be modeled to within 10–15% using Köhler theory [Engelhart et al., 2008].
 Figure 3 shows that the critical supersaturation of soot particles is considerably reduced as coatings develop on the particles, since coating contributes to both increased soluble mass and overall particle size. The critical supersaturation of a 100 nm coated particle with a 83 nm soot core is higher than that of pure pinic or norpinic acid particles of 100 nm diameter (Figure 3), implying that there is insufficient soluble material to achieve activation. For the same core size, a coated particle size of 120 nm diameter exhibits a similar activation to the pure organic acid particle. Hence, for sufficiently coated primary particles, there is a nearly complete loss of the original microphysical properties (i.e., the initial core size and chemical composition) of the fresh soot and the CCN properties are indistinguishable from those of the pure pinic and norpinic acids of the similar particle size. Although the present work is focused on soot particles, our results are applicable to other types of primary aerosols. For instance, in experiments using PSL spheres, we also measure rapid development of organic coatings and significant increase in the particle CCN activation.
 The rate of conversion of hydrophobic soot particles into active CCN depends on the ambient aerosol precursor concentrations and the supersaturation. While the ozone concentration (20 ppb) employed in our experiments is comparable to that in the atmosphere, the experimental α-pinene concentration of 37 ppbv is higher than that in the atmosphere (a few ppbv) [Hallquist et al., 2009]. Also, the typical supersaturation in cumulus clouds is in the range of 0.1 to 0.7 [Pruppacher and Klett, 2000]. Given an ozone concentration of 20 ppbv and 0.31% supersaturation, conversion of 103 soot cores into active CCN activation at 37 ppbv α-pinene, occurs in ~0.6 h (Figure 2). In separate experiments with 22 and 7 ppbv α-pinene and all other conditions the same, activation requires substantially longer times, i.e., about 1.2 and 3.1 h, respectively. The corresponding coating thicknesses in our experiments are in the range of 20 to 30 nm for the α-pinene concentrations of 7, 22, and 37 ppbv. Atmospheric measurements have shown a particle growth rate of 3–13 nm h−1 following photochemical oxidation of α- and β-pinenes [Russell et al., 2007], which is lower than those of 22 and 37 ppbv α-pinene, but comparable to that of 7 ppbv α-pinene in our experiments.
 Our results imply that under ambient conditions the transformation rate of primary particles may be greatly underestimated and their indirect climate forcing effect is likely much larger than that predicted by current atmospheric models [Jacobson, 2000; Riemer et al., 2009]. For example, in global models the time scale for converting hydrophobic soot particles into hydrophilic CCN is often empirically assumed to be much longer, typically more than 1 day [Cooke et al., 2002; Jacobson, 2001; Koch et al., 2009b, 2009c].
 Our measurements indicate that a fast transformation of primary aerosols occurs because of coatings of α-pinene ozonolysis products, leading to a significant enhancement in the aerosol cloud-nucleating ability. Due to biogenic aging processes, the CCN activation abilities of a large fraction of total atmospheric aerosol may become distinctly different from those of the primary aerosols. In addition, our results suggest that the microphysical properties of aerosols present in continental air masses are mainly controlled by the composition and thickness of coatings formed during aerosol aging processes, rather than by the size or composition of the original primary particles. Hence, depending on the extent of aging, atmospheric aerosols may exhibit CCN activation abilities either similar to or distinctly different than those of their corresponding primary aerosols. This suggests that revisiting analysis of previous field results with the considerations of the proximity of aerosol sampling to source and atmospheric aging would likely reconcile the observed variable dependence of CCN activation on particle size versus chemical composition under diverse atmospheric conditions [Dusek et al., 2006; Furutani et al., 2008; Hudson, 2007; Quinn et al., 2008].
 Recent modifications of Köhler theory provide a single parameter (κ) representation of the cloud nucleating ability of aerosols containing multiple water-soluble components [Petters and Kreidenweis, 2007, 2008]. However, the dynamic effect of aerosol aging on CCN activation has yet to be accounted for in atmospheric models. Cloud-nucleating abilities may be driven mainly by the thickness and properties of coatings formed from condensable oxidation products of inorganic and organic compounds of anthropogenic and biogenic origins, rather than by the composition or size of the original primary particles. In particular, rapid transformation of primary particles in the atmosphere due to biogenic emissions must be included in global assessment of aerosol-cloud interactions and their impacts on climate forcings.
 Y. M. acknowledges support from the National Natural Science Foundation of China (21377059, 41275142, and 41030962) and Jiangsu Natural Science Foundation (BK2012861). R. Z. acknowledges additional support by the Ministry of Science and Technology of China (2013CB955800). This work was supported by the National Science Foundation (CAREER-0548075, AGS-0938352 and CBET-0932705) and the Robert A. Welch Foundation (A-1417).
 The Editor thanks two anonymous reviewers for their assistance in evaluating this paper.