J. Phys. Chem. A 2010, 114, 2837–2844
2837
Surface Morphology and Phase Transitions in Mixed NaCl/MgSO4 Aerosol Particles Ephraim Woods,* Daniel Chung, Howard M. Lanney, and Benjamin A. Ashwell Department of Chemistry, Colgate UniVersity, 13 Oak DriVe, Hamilton, New York 13346 ReceiVed: NoVember 23, 2009; ReVised Manuscript ReceiVed: January 15, 2010
Probe molecule spectroscopy characterizes the surface environment of mixed NaCl/MgSO4 (0.01-50 wt % MgSO4) aerosol particles as a model for marine aerosol. Two complementary measurements, the probe’s excited state spectroscopy and photoionization efficiency, measure the electronic properties of the particle surface and monitor phase changes that are driven by changes in relative humidity (RH). The results illustrate that over a wide range of composition, these particles have a layered structure with NaCl in the core and primarily hydrated MgSO4 at the surface. Modeling the spectroscopic data reveals that the surface layer is not a uniform shell and that the coumarin 314 probe molecules partition selectively to the MgSO4 domains. The surface layer has a π* value of 1.7, indicative of a very high interfacial polarity. In cases where MgSO4 is a minor component (e10 wt %), the NaCl component crystallizes at 44% RH, consistent with the single salt NaCl result. Deliquescence-mode experiments with these particles show that the MgSO4 component forms a solution at 42% RH, prior to the full deliquescence of the particle. For mixed particles with 50 wt % MgSO4, the crystallization of NaCl occurs at 35% RH, and the predeliquescence of MgSO4 occurs at 38% RH owing to the contribution of MgCl2 in the surface layer. A model surfactant, SDS, slightly lowers the RH of the NaCl formation to approximately 42% and leads to the formation of a thin soap film that persists to low values of RH. I. Introduction Marine aerosol particles represent an important contributor to the particle mass of the troposphere. As much as 1012 or 1013 kg of sea salt enters the atmosphere annually in the form of seawater aerosols.1 These particles influence climate through light scattering and cloud condensation, and they affect the composition of the atmosphere through heterogeneous chemistry. Each aspect depends strongly on the physical and chemical characteristics of the particle surfaces, which are generally complex in both their composition and morphology. Sea salt aerosols contain a variety of inorganic components, including Na+, Cl-, Mg2+, SO42-, Ca2+, K+, HCO3-, Br-, and Sr2+, and on order 10% by mass organic matter.2 The organic component includes primarily surface active molecules, including some long-chain fatty acids.3 While pure NaCl has been a successful laboratory proxy for sea salt in many respects (for example, the heterogeneous chemistry involving Cl-), the complex composition of these particles does have some important consequences. In particular, composition can determine a particle’s hygroscopic properties, which, in turn, affects both the light scattering cross section and ability to be a cloud condensation nucleus. Further, water controls many important heterogeneous reactions in the atmosphere. For example, the uptake of HNO3 onto NaCl particles is strongly RH-dependent,4,5 and even surface adsorbed water can enhance the reaction probability of HNO3 with NaCl particles by as much as an order of magnitude over that of very dry particles.6 Characterizing the surface composition and morphology of sea salt aerosols as a function of water content, then, yields insight into both cloud condensation processes and the heterogeneous chemistry of the atmosphere. Of the many inorganic constituents of seawater besides Na+ and Cl-, Mg2+ is among the most intriguing and potentially * Corresponding author. E-mail:
[email protected].
has the greatest effect on the surface morphology of marine particles. Recent characterizations of Mg-containing aerosol particles reveal several interesting features. First, though Mg salts are readily soluble in water, they do not always show typical deliquescence and efflorescence behavior. Rather, they can take up or lose water continuously with changing relative humidity.7 Second, Mg-containing aerosols retain significant amounts of water at very low values of RH.8,9 The increased water content can enhance heterogeneous reaction rates, as with the aforementioned HNO3 uptake.4,5 Lastly, the Mg-containing fraction of artificial and natural seawater particles segregate to the surface under low RH conditions, making it an important constituent of seawater to consider despite its relatively low concentration.4,10 Wise et al.10 used an environmental transmission electron microscope (ETEM) to examine the morphological changes that accompany changes in relative humidity for natural and artificial seawater. They found that both kinds of particles have a lower onset of morphological changes with increasing RH than does pure NaCl. Energy-dispersive X-ray spectrometry (EDS) shows that the “liquid layer” produced prior to full deliquescence is enriched in Mg, S, and O relative to the solid particle core. One implication of this work is that many heterogeneous reactions involving marine aerosol may occur in or on MgSO4enriched solutions. The literature concerning aqueous MgSO4 aerosol particles provide some insight into the possible structure of the Mgcontaining portion of these particles. Several researchers have studied the hygroscopic properties of MgSO4 aerosol particles using electrodynamic balance,11,12 IR,9 and Raman13,14 techniques. A commonality among the conclusions is that supersaturated aqueous MgSO4 forms gels that contribute to mass transfer limitations during the dehydration of the particles. These gels can explain the higher-than-expected water content of lowRH seawater aerosols. Supersaturated MgSO4 has contributions
10.1021/jp911133j 2010 American Chemical Society Published on Web 02/04/2010
2838
J. Phys. Chem. A, Vol. 114, No. 8, 2010
from several structures including simple contact ion pairs (CIPs), polymeric chains of CIPs, solvent-shared ion pairs (SIPs), and double solvent-shared ion pairs (2SIPs). Polymeric chains of CIPs are the primary component of the gel phase. There is some evidence also of CIPs at subsaturated concentrations, though the lifetime of these structures is short.15 Zhang and co-workers9 track changes in IR spectroscopic features of water and SO42- as a function of RH to characterize changes in MgSO4 aerosol. They divide the full range of RH into four regions with distinct hygroscopic properties. In region 1 (80%, which is adequate to fully deliquesce the particles in the flow. A 3.0 slpm, RH-controlled flow then dilutes the flow and decreases the RH to its ultimate value. In efflorescence mode, the size-selected then deliquesced particles never experience a
Surface Morphology of NaCl/MgSO4 Particles lower RH than the nominal measurement RH. For both schemes, the particles pass through an empty cylinder, allowing them to equilibrate at their ultimate RH for approximately 30 s prior to analysis. For the photoionization spectroscopy experiment, the particle flow interacts with two copropagating laser beams in an ionization cell. Tunable visible light from a Nd:YAG-pumped optical parametric oscillator excites the S1 state of C314, and 355 nm light produced by frequency tripling a separate Nd: YAG laser ionizes the electronically excited molecules. A small electric field (∼10 V/cm) removes the nascent photoelectrons from the flow without significantly altering the trajectory of the aerosol particles, and an aerosol electrometer (TSI model 3065A) measures the remaining charge in the flow following the ionization. Monitoring the amount of charge in the flow while scanning the visible laser’s wavelength produces the aerosol photoionization spectrum. Because only the electronically excited molecules contribute to the two-laser signal, the photoionization spectrum reflects the absorption spectrum of the excited state. The photoelectric charging efficiency experiment is similar. In this case, we fix the wavelength of the excitation laser to the maximum of the photoionization spectrum and measure the signal intensity as a function of RH. The signal is linear with particle concentration over the small range of concentrations sampled in our experiment. For each value of RH, we normalize the signal to the particle concentration, which is measured by the electrometer current in the absence of lasers. This ratio, which we refer to as the photoelectric charging efficiency (φ), reflects the average number of charges per particle following the laser ionization. This measurement, in general, depends on the experimental conditions, especially the laser alignment and power; however, we only compare φ values within a particular data set where the conditions are consistent. III. Results and Discussion Morphology. One of the useful aspects of our probe molecule spectroscopy is that it is possible to deduce the location of the probe in a heterogeneous system when the different phases produce sufficiently different spectroscopic signatures. For example, C314 adsorbed to solid NaCl has a λmax at 446 nm,26 while C314 adsorbed to MgSO4 has a λmax of 458 nm. Further, because the probe molecule is surface-active in the supersaturated droplets that form during the drying stage, the spectroscopy of the probe molecule primarily reveals the composition of the surfaces layers of the particle. Figure 1 shows the photoionization spectra measured at two values of RH for two different mixtures of NaCl and MgSO4. The top panel shows the results using 10 wt % MgSO4 particles, and the bottom panel shows data for 1 wt % MgSO4. (The reported compositions refer to dry particles.) These are efflorescence mode experiments; therefore, the particles approach their ultimate RH from the high-RH limit. For both mixtures, the figure shows spectra collected under dry conditions (
