Primary marine organic aerosol: A dichotomy of low hygroscopicity and high CCN activity

Authors


Abstract

[1] High-time resolution measurements of primary marine organic sea-spray physico-chemical properties reveal an apparent dichotomous behavior in terms of water uptake: specifically sea-spray aerosol enriched in organic matter possesses a low hydroscopic Growth Factor (GF∼1.25) while simultaneously having a cloud condensation nucleus/condensation nuclei (CCN/CN) activation efficiency of between 83% at 0.25% supersaturation and 100% at 0.75%. In contrast, the activation efficiency of particles dominated by non-sea-salt (nss)-sulfate ranged between 48–100% over supersaturation range of 0.25%–1%. Simultaneous retrieval of Cloud Droplet Number Concentration (CDNC) during primary organic aerosol plumes reveals CDNC concentrations of 350 cm−3 for organic mass concentrations 3–4 μg m−3. It is demonstrated that the retrieved high CDNCs under clean marine conditions can only be explained by organic sea-spray and corroborates the high CCN activation efficiency associated with primary organics. It is postulated that marine hydrogels are responsible for this dichotomous behavior.

1. Introduction

[2] The global importance of marine aerosol particles to the CCN formation was postulated several decades ago by Charlson et al. [1987]; however, this study overlooked the existence of biogenic organic matter and centered on the marine sulfur cycle. Later studies by O'Dowd et al. [2004] demonstrated a significant, if not dominant organic mass fraction in submicron marine aerosol, while the most recent study by Ovadnevaite et al. [2011] reported primary spray plumes enriched in organic matter with mass loadings up to 3–4 μg m−3 (i.e., far greater than sulfate mass concentrations in clean marine air). Organic mass loadings of this magnitude, occurring in CCN sizes, suggest that the organic matter may contribute notably to the CCN population. In addition, satellite observations [Meskhidze and Nenes, 2006; Sorooshian et al., 2009] indicated concurrent increases in CCN activity and/or cloud droplet concentrations in marine stratocumulus overlying chlorophyll-rich regions, pointing a possible link between organic marine aerosol and cloud droplet activation; however, it has been more difficult to determine whether this link is through secondary organic aerosol or primary (sea-spray) organic aerosol. The picture is somewhat complicated in that spray produced from biologically rich water is expected to be water insoluble [Facchini et al., 2008; Vesna et al., 2008], while that produced from secondary processes is expected to be more oxygenated and hence water soluble. Further, several recent laboratory studies [Fuentes et al., 2011; Irwin et al., 2010; Prisle et al., 2010] show a reduction of CCN activity due to hydrophobic properties of biogenic organics. Here we present comprehensive real time ambient measurements of aerosol CCN activation, hygroscopicity, chemical composition and size distributions which reveal the important role of primary marine organics in CCN activation.

2. Experiment

[3] Continuous marine aerosol physico-chemical measurements were undertaken at the Mace Head atmospheric research station (54°19′N, 9°54′W), located on the west coast of Ireland [O'Connor et al., 2008]. Aerosol measurements were performed by sampling ambient particles through a community air-sampling duct 10 m above ground level. The size resolved non-refractory chemical composition of submicron aerosol particles is measured with an Aerodyne High Resolution Time of Flight Aerosol Mass Spectrometer deployed in standard mode [DeCarlo et al., 2006]. Aerosol size distributions were measured by a TSI scanning mobility particle sizer (SMPS) and a relative humidity of 40%. CCN concentrations were determined by a Droplet Measurements Technology CCN counter [Lance et al., 2006] operated at supersaturations of 0.25%, 0.5%, 0.75% and 1%. Aerosol GFs were measured using a Hygroscopic Tandem Differential Mobility Analyzer (HTDMA) and follows the EUSAAR standard installation and accuracy. GFs at 90% RH were determined for dry size particles of 35, 50, 75, 110 and 165 nm. Clean marine air is only considered in this analysis and previously described in many studies [e.g., O'Dowd et al., 2004].

3. Results

[4] Two distinct seasons were selected from the long-term aerosol composition, size and hygroscopicity measurements to cover both primary sea-spray and secondary (sulphate dominated) aerosol production episodes. The selected periods include spray events with significant primary organic matter enrichment, typically occurring in the late summer, and secondary (predominantly sulphate) aerosol formation periods, prevalent in late spring, early summer. The dataset spanned from 11th – 28th August 2009, 14th July – 12th August 2010, and 2nd – 27th May 2009, within which the aerosol was categorized according to the AMS analysis.

[5] CCN activation efficiency as a function of particle hygroscopicity, composition and size is presented in Figure 1 for all periods. The color scale in Figure 1a represents the fraction of organic mass to total particle mass, while in Figure 1b, it represents the fraction of sulphate mass to total mass. The size of the circle represents weighted-average particle size of the whole aerosol population for the averaging period (that is a 1-hour integration period). For example, for a bimodal size distribution with a mode at 65 nm and 200 nm, both with equal number concentrations, would have a weighted average particle size of 132 nm. Examples of actual organic and sulphate dominated size distributions, and their resultant weighted diameter are also shown in Figures 1c and 1d, respectively.

Figure 1.

(a and b) CCN0.75% activity (CCN/CN) as a function of GF (at 90% RH), chemical composition (colour scale) and weighted average particle size (size of the circle). CN is the total particle number above 20 nm in diameter; the colour scale represents the dominance of a given chemical species. Measurement periods: 02nd – 27th May 2009, 11th – 28th August 2009 and 14th July – 12th August 2010. Note, the measurement periods cover periods much longer than individual plume events. In Figure 1a, the boxed region highlights particles dominated by primary organic matter while Figure 1b highlights the particles dominated by sulphate. Particles to the extreme right of both figures are dominated by sea-salt mass. (c) Two organic-dominated size distributions (on 00:00 UTC - 22:00 UTC 16th August 2009 and 13:30 UTC - 16:30 UTC 05th August 2010) and their resultant weighted diameters and (d) the same for sulphate-dominated distributions (on 21:00 UTC - 22:00 UTC 02nd August 2010 and 06:00 UTC - 10:00 UTC 09th August 2010).

[6] Aerosol dominated by organics possessed a low GF of 1.25 or less, as emphasized by the boxed area in Figure 1a, while sulfate dominance led to a GF∼1.65, as emphasized by the boxed area in Figure 1b. The points to the extreme right of each boxed area represent sea-salt dominated particles with the highest GF of >1.8. CCN activation was dependant on aerosol chemical composition, but did not follow typical hygroscopicity dependence pattern – that is, particles with the lowest GF were more readily activated than particles with the highest GF. As a result, particles clustered in the upper left corner in the graph (Figure 1), represent high organic enrichment, low GF, and high CCN activation. These particles also posses the largest mean weighted size. By contrast, particles with high sulfate or sea salt mass fractions posses a higher GF but greater ranges of mean weighted size and CCN activation efficiency. The periods during which primary marine organic matter dominated particles were associated with maritime air masses arriving from the regions of high biological activity and elevated wind speeds over the N.E. Atlantic and extended up to 30 hours in duration [Ovadnevaite et al., 2011] with organic mass concentrations approaching ∼4 μg m−3.

[7] The data reveal a pattern of high average weighted diameter, low solubility and high CCN activation for particles with high organic mass fraction and a broad range of weighted diameters, high solubility and diverse CCN activity for high inorganic mass fraction. The Probability Distribution Function (PDF) of growth factors (Figure 2) revealed an external mixture of different particle populations determined by three distinct GF modes during typical organic plume event. A less-hygroscopic mode (GF peak at 1.25) was very prominent and evident across all sizes and always contributed to at least 51% of the total particle number. However, the largest contribution to the 1.25 GF peak was from 75 nm particles (up to 75%) indicating the maximum hydrophobic organic-matter enrichment factor for that size range. A more-hygroscopic mode (GF between 1.36 and 1.8) was less prominent and subject to greater variability between particle sizes, although it contributed more to the smallest and largest measured particles (35 and 165 nm accordingly) rather than mid-sized particles. Finally, a relatively-pure sea salt mode (GF between 1.8 and 2.3) was evident (although only contributing to 10% concentration) in the smallest measured dry size (35 nm) particles. The GF-PDF for a sulphate dominated aerosol is also shown for comparison. The CCN activation efficiency for the organic dominated and sulfate dominated populations (as classified by the AMS) was examined as a function of supersaturation in Figure 3. For organic dominated particles, the activation efficiency ranged from 83% at 0.25% supersaturation to 100% at 1% supersaturation. Given that the less hygroscopic mode always contained at least 51% of the total particle number, a large fraction of particles with GF =1.25, the majority of which were ≤75 nm, must have been activated at 0.25% supersaturation. In contrast, particles dominated by sulfate mass exhibited a lower activation efficiency at 0.25% (48% activation for sulfate).

Figure 2.

(a) Hygroscopic growth dependence on the particle size represented as probability density functions and average growth factor for one typical primary organic plume event on 00:00 UTC-22:00 UTC 16th August 2009 and (b) GF-PDF for typical sulphate dominated aerosol on 21:00 UTC-22:00 UTC 02nd August 2010.

Figure 3.

CCN/CN activity versus supersaturation for particles of different percentage mass composition (dominant mass percentage highlighted in brackets in figure key). Bars represent standard deviations.

[8] Stratocumulus microphysical properties are presented for one of the primary organic plume events (4th–5th August 2010). CDNC derived from ground-based remote sensing measurements [Martucci and O'Dowd, 2011] ranged from 100 to 350 cm−3 (Figure 4). Taking the weighted average diameter from the SMPS size distribution and the AMS-derived nss-sulfate and sea-spray mass concentrations, assuming both sulphate and sea-spray have the same weighted diameter but are externally mixed (where sea-spray is an internal mix of sea salt and organics), the number of nss-sulfate and sea-spray potential cloud nuclei was calculated. Assumed densities were 1.4 g cm−3 for organics, 2.17 g cm−3 for sea salt, 1.83 g cm−3 for sulfuric acid and 1.77 g cm−3 for ammonium sulfate. The validity of these assumptions is reflected in the derived nuclei concentration being within 10% of the total aerosol concentration for D > 20 nm. The calculated total nuclei concentration corresponded, almost exactly, to the CCN concentration at 0.75% supersaturation. Sulfate nuclei concentration initially was of the order of 200 cm−3 but then fell off to ∼100 cm−3 where it remained relatively constant. Sea-spray nuclei peaked at 450 cm−3, although there was no overlapping CDNC data for these peak concentrations. For overlapping CDNC, both CDNC and sea-spray peaked at ∼350 cm−3, and both the absolute spray concentration and its trend matched that of CDNC with a correlation co-efficient of r = 0.76. Nss-sulfate concentrations were insufficient to account for the CDNC while the latter can be almost completely accounted for by the sea-spray concentration.

Figure 4.

(a) CDNC, measured CCN0.75%, calculated sea-spray, sulphate and total nuclei concentration. (b) CDNC as a function of sea-spray particle concentration.

4. Discussion

[9] These results show a dichotomous behavior for primary marine organic aerosol: that is a low hygroscopic GF but high CCN activation efficiency. Such behavior is determined not only by particle size effects [Dusek et al., 2006], but also by chemical composition influences. Low hygroscopicity in sea-spray aerosol is determined by the specific chemical composition of organic matter transferred from the ocean surface during sea-spray production and comprising a broad spectrum of biopolymers including polysaccharides, proteins, nucleic acids, lipids and so called exopolymer substances [Facchini et al., 2008; Verdugo and Santschi, 2010]. The dichotomous behavior may have one of a number of explanations, e.g., limited water solubility in all but the most dilute solutions and surface active properties of sea-spray organics is one possibility [Charlson et al., 2001; Facchini et al., 2008; Mircea et al., 2005; Moore et al., 2008].

[10] Surface activity plays an important role in the Köhler equation which consists of two terms, the Raoult term and the Kelvin term - the former being more important for sub-saturated conditions, and the latter being more important for supersaturated conditions [Irwin et al., 2010]. The balance between the two terms determines the resulting critical supersaturation needed for the particle activation. When surface tension counteracts water activity, the net effect is an enhancement of particle CCN activity. Fuentes et al. [2011] and Irwin et al. [2010] argued that a moderate reduction in surface tension at activation produced by marine organics is too low to compensate the impact on the Raoult term, but this concept also led to a disagreement between measured and modeled CCN activity when applied to the ambient environment [Irwin et al., 2010]. Moreover, surface tension measurements of aerosols containing organic matter have been limited to mainly water soluble organics. Such surface tension lowering, however, inadequately represents insoluble organics and/or colloidal material. A recent study by Ekstrom et al. [2010] indicated a much higher degree of surface tension suppression by marine organics compared to what was previously considered.

[11] Alternatively, adsorption of water on wettable surfaces can dominate over hygroscopic uptake, leading to adsorption theory replacing Kholer Theory [Kumar et al., 2011]; however, the lack of experimental data on water vapor adsorption on wettable organic surfaces prevents a quantitative evaluation in this study. Nonetheless, the GF and activation efficiency for the organic plumes here are consistent with the activation of ∼75 nm low-soluble particles at supersaturations of the order of 0.2–0.3% [Kumar et al., 2011].

[12] The organic compounds transferred in sea-spray are potentially the same as those involved in the formation of marine gels in seawater [Chin et al., 1998; Engel et al., 2004]. The gel-forming properties of marine biopolymers may also play a role in the observed dichotomous behavior of organics-rich sea-spray particles, assuming that a three dimensional structure, similar to that of marine hydrogels, is present in these particles. At sub-saturated conditions the three-dimensional structure of the gel in sea-spray particles is expected to be more condensed and to behave like a water insoluble core, while during the particle activation the gel-like structure may be partially solubilized as an equilibrium between the gel and the truly dissolved phase is established at the higher dilution typical of this regime, as observed in sea water [Chin et al., 1998; Orellana et al., 2007]. This, together with the observed surface active properties of sea-spray organics, can contribute to reduce the particle critical supersaturation [Charlson et al., 2001; Mircea et al., 2005]. Moreover, the strong affinity of amphiphilic residues for the air-water interface, associated with rising bubbles leading to spray production, could result in surface partitioning of organic moieties which can cause a specific molecular arrangement within the gel network, with hydrophilic groups preferentially in the aerosol core and the hydrophobic groups oriented toward the air medium. A physical model describing the forces causing this specific arrangement, although simplistic with respect to the complex three-dimensional configuration of marine hydrogels, was introduced by [Ellison et al., 1999] in terms of an inverted micelle model and single particle mass spectrometry measurements lend credence to such a model [Tervahattu et al., 2002]. This specific arrangement of organic molecules as hydrogels and pseudo-inverted micelles would explain the observed CCN behavior of primary marine aerosol and its low hygroscopicity. Additionally, such a model would allow water processing, even though the particle would have a hydrophobic surface concomitant with significant lowering of surface tension [Chakraborty and Zachariah, 2007; Ekstrom et al., 2010; Facchini et al., 2008].

[13] If, as assumed, marine hydrogels are transferred into the aerosol particles through bubble bursting, their peculiar physico-chemical properties may not only influence the properties of the sea-spray aerosol, but it would also affect the residual spray particle dry size. For instance, the gel network could reduce the water evaporation of spray drops, and along with the displacement of sea salt solution by larger organic molecules, could lead to a net increase in “dry” sea-spray particle size. Nonetheless, size cannot be the main factor determining activation efficiency since for some of these events, the Aitken mode dominated particle concentration over and above the accumulation mode. Given that CCN/CN activation ratio ranged from 83–100% over the 0.25–1% supersaturation range, the majority of the dominant, and organically-enriched Aitken mode (∼70 nm) would have had to be activated, emphasizing the fact that chemical composition, and particularly, organic matter enrichment, particularly in the smallest potential nuclei sizes are important in the droplet activation process.

5. Conclusions

[14] This study highlights the importance of primary marine organic aerosol to the marine CCN population. Despite being predominately hydrophobic in nature (GF∼1.25), primary organic marine particles have effectively a higher activation efficiency than more soluble inorganic aerosol. Sea-spray aerosol enriched in primary organic matter possessed CCN activation efficiency of between 83% (at 0.25% supersaturation) and 100% (at 0.75% and 1% supersaturation). In contrast, particles dominated by non-sea-salt (nss)-sulfate ranged between 48–100% over the same supersaturation range. The CCN activity of sea-spray enriched in organic matter is corroborated by simultaneous measurements of very high CDNCs (N∼350 cm−3) and where a correlation coefficient of 0.76 was calculated between CDNC and inferred sea-spray CCN concentrations. The vast majority of the CDNC could only be explained by activation of sea-spray aerosol enriched in organic matter. We suggest that this phenomena relates to the enrichment of marine hydrogels in sea-spray aerosol.

Acknowledgments

[15] This work was supported by the Science Foundation Ireland (grant 08/RFP/GEO1233), HEA-PRTLI4 Environment and Climate Change: Impact and Responses programme, European Commission FP7 EUCAARI, and EPA Ireland.

[16] The Editor thanks two anonymous reviewers for their assistance in evaluating this paper.