Ellis S. Robinson

Organic aerosol mixing observed by single-particle mass spectrometry.

We present direct measurements of mixing between separately prepared organic aerosol populations in a smog chamber using single-particle mass spectra from the high-resolution time-of-flight aerosol mass spectrometer (HR-ToF-AMS). Docosane and docosane-d46 (22 carbon linear solid alkane) did not show any signs of mixing, but squalane and squalane-d62 (30 carbon branched liquid alkane) mixed on the time scale expected from a condensational-mixing model. Docosane and docosane-d46 were driven to mix when the chamber temperature was elevated above the melting point for docosane. Docosane vapors were shown to mix into squalane-d62, but not the other way around. These results are consistent with low diffusivity in the solid phase of docosane particles. We performed mixing experiments on secondary organic aerosol (SOA) surrogate systems finding that SOA derived from toluene-d8 (a surrogate for anthropogenic SOA (aSOA)) does not mix into squalane (a surrogate for hydrophobic primary organic aerosol (POA)) but does mix into SOA derived from α-pinene (biogenic SOA (bSOA) surrogate). For the aSOA/POA, the volatility of either aerosol does not limit gas-phase diffusion, indicating that the two particle populations do not mix simply because they are immiscible. In the aSOA/bSOA system, the presence of toluene-d8-derived SOA molecules in the α-pinene-derived SOA provides evidence that the diffusion coefficient in α-pinene-derived SOA is high enough for mixing on the time scale of 1 min. The observations from all of these mixing experiments are generally invisible to bulk aerosol composition measurements but are made possible with single-particle composition data.

Mixing of secondary organic aerosols versus relative humidity

Significance Recent studies called into question whether diffusion in “glassy” atmospheric secondary organic aerosols (SOA) is fast enough for phase partitioning to equilibrate, as commonly assumed in models; however, most of those studies relied on rheological measurements or uptake of small molecules such as water and ammonia. Here, we conduct unique SOA mixing experiments using single-particle mass spectrometry to probe the diffusion of semivolatile organics into SOA derived from toluene and α-pinene at different relative humidities. Results show that, if the SOA systems are representative, equilibrium partitioning likely does have time to occur in the boundary layer. This work directly probes uptake of semivolatile organics by SOA particles over a wide humidity range. Atmospheric aerosols exert a substantial influence on climate, ecosystems, visibility, and human health. Although secondary organic aerosols (SOA) dominate fine-particle mass, they comprise myriad compounds with uncertain sources, chemistry, and interactions. SOA formation involves absorption of vapors into particles, either because gas-phase chemistry produces low-volatility or semivolatile products that partition into particles or because more-volatile organics enter particles and react to form lower-volatility products. Thus, SOA formation involves both production of low-volatility compounds and their diffusion into particles. Most chemical transport models assume a single well-mixed phase of condensing organics and an instantaneous equilibrium between bulk gas and particle phases; however, direct observations constraining diffusion of semivolatile organics into particles containing SOA are scarce. Here we perform unique mixing experiments between SOA populations including semivolatile constituents using quantitative, single-particle mass spectrometry to probe any mass-transfer limitations in particles containing SOA. We show that, for several hours, particles containing SOA from toluene oxidation resist exchange of semivolatile constituents at low relative humidity (RH) but start to lose that resistance above 20% RH. Above 40% RH, the exchange of material remains constant up to 90% RH. We also show that dry particles containing SOA from α-pinene ozonolysis do not appear to resist exchange of semivolatile compounds. Our interpretation is that in-particle diffusion is not rate-limiting to mass transfer in these systems above 40% RH. To the extent that these systems are representative of ambient SOA, we conclude that diffusion limitations are likely not common under typical ambient boundary layer conditions.

Vapor wall loss of semi-volatile organic compounds in a Teflon chamber

ABSTRACT We have investigated the vapor wall loss of semi-volatile organic compounds (SVOCs) in a Teflon smog chamber. We studied the vapor wall loss of seven SVOCs with known saturation concentrations, including alkanes (hexacosane, pentacosane, docosane, eicosane, and d62-squalane), an organic acid (oleic acid), and a polyol (levoglucosan) in single-component and binary-component (organic) systems, using ammonium sulfate (AS) seeds to constrain the particle wall loss. We coated inorganic particles with SVOCs and measured the loss of organics from those particles to constrain the wall losses, observing loss rates proportional to the saturation concentrations of the SVOCs. The loss rate of oleic acid mixed with d62-squalane was proportional to its mole fraction in the mixture. Our results show that the vapor wall-loss rates of SVOCs are significant, quasi-irreversible, and proportional to the SVOC vapor concentrations. The vapor wall-loss rate constant of the SVOCs that we studied in the CMU chamber is 3.8 ± 0.3 h−1; this is comparable to values in other chambers with similar surface area to volume ratios. Our results are also consistent with a relatively high mass accommodation coefficient for SVOCs, αorg > 0.1.

Photochemical Aging of Secondary Organic Aerosols Generated from the Photooxidation of Polycyclic Aromatic Hydrocarbons in the Gas-Phase

Aging processes of secondary organic aerosol (SOA) may be a source of oxygenated organic aerosols; however, the chemical processes involved remain unclear. In this study, we investigate photochemical aging of SOA produced by the gas-phase oxidation of naphthalene by hydroxyl radicals and acenaphthylene by ozone. We monitored the SOA composition using a high-resolution time-of-flight aerosol mass spectrometer. We initiated SOA aging with UV photolysis alone and with OH radicals in the presence or absence of light and at different NOx levels. For naphthalene, the organic composition of the particulate phase seems to be dominated by highly oxidized compounds such as carboxylic acids, and aging data may be consistent with diffusion limitations. For acenaphthylene, the fate of oxidized products and the moderately oxidized aerosol seem to indicate that functionalization reactions might be the main aging process were initiated by the cumulative effect of light and OH radicals.

Probing the Evaporation Dynamics of Mixed SOA/Squalane Particles Using Size-Resolved Composition and Single-Particle Measurements.

An analysis of the formation and evaporation of mixed-particles containing squalane (a surrogate for hydrophobic primary organic aerosol, POA) and secondary organic aerosol (SOA) is presented. In these experiments, one material (D62-squalane or SOA from α-pinene + O3) was prepared first to serve as surface area for condensation of the other, forming the mixed-particles. The mixed-particles were then subjected to a heating-ramp from 22 to 44 °C. We were able to determine that (1) almost all of the SOA mass is comprised of material less volatile than D62-squalane; (2) AMS collection efficiency in these mixed-particle systems can be parametrized as a function of the relative mass fraction of the components; and (3) the vast majority of D62-squalane is able to evaporate from the mixed particles, and does so on the same time scale regardless of the order of preparation. We also performed two-population mixing experiments to directly test whether D62-squalane and SOA from α-pinene + O3 form a single solution or two separate phases. We find that these two OA types are immiscible, which informs our inference of the morphology of the mixed-particles. If the morphology is core-shell and dictated by the order of preparation, these data indicate that squalane is able to diffuse relatively quickly through the SOA shell, implying that there are no major diffusion limitations.

Uptake of Semivolatile Secondary Organic Aerosol Formed from α-Pinene into Nonvolatile Polyethylene Glycol Probe Particles.

Semivolatile organic compounds (SVOCs) play an essential role in secondary organic aerosol (SOA) formation, chemical aging, and mixing of organic aerosol (OA) from different sources. Polyethylene glycol (PEG400) particles are liquid, polar, and nearly nonvolatile; they provide a new vehicle to study the interaction between SVOCs with OA. With a unique fragment ion C4H9O2(+) (m/z 89), PEG400 can be easily separated from α-pinene SOA in aerosol mass spectra. By injecting separately prepared PEG probe particles into a chamber containing SOA coated on ammonium sulfate seeds, we show that a substantial pool of SVOCs exists in equilibrium with the original SOA particles. Quantitative findings are based on bulk mass spectra, size-dependent composition, and the evolution of individual particle mass spectra, which we use to separate the two particle populations. We observed a larger fraction of SVOC vapors with increased amounts of reacted α-pinene. For the same amount of reacted α-pinene, the SOA formed from α-pinene oxidized by OH radicals had a higher fraction of SOA vapors than SOA formed by α-pinene ozonolysis. Compared to the PEG400 probe particles, we observed a lower mass fraction of SVOCs in poly(ethylene glycol) dimethyl ether (MePEG500) probe particles under otherwise identical conditions; this may be due to the lower polarity of the MePEG500 or caused by esterification reactions between the PEG400 and organic acids in the SOA.

A dual‐chamber method for quantifying the effects of atmospheric perturbations on secondary organic aerosol formation from biomass burning emissions

Biomass burning (BB) is a major source of atmospheric pollutants. Field and laboratory studies indicate that secondary organic aerosol (SOA) formation from BB emissions is highly variable. We investigated sources of this variability using a novel dual-smog-chamber method that directly compares the SOA formation from the same BB emissions under two different atmospheric conditions. During each experiment, we filled two identical Teflon smog chambers simultaneously with BB emissions from the same fire. We then perturbed the smoke with UV-lights, UV-lights plus HONO, or dark ozone in one or both chambers. These perturbations caused SOA formation in nearly every experiment with an average organic aerosol (OA) mass enhancement ratio of 1.78 ± 0.91 (mean ± 1σ). However, the effects of the perturbations were highly variable ranging with OA mass enhancement ratios ranging from 0.7 (30% loss of OA mass) to 4.4 across the set of perturbation experiments. There was no apparent relationship between OA enhancement and perturbation type, fuel type, and modified combustion efficiency. To better isolate the effects of different perturbations, we report dual-chamber enhancements (DUCE), which quantity the effects of a perturbation relative to a reference condition. DUCE values were also highly variable, even for the same perturbation and fuel type. Gas measurements indicate substantial burn-to-burn variability in the magnitude and composition of SOA precursor emissions, even in repeated burns of the same fuel under nominally identical conditions. Therefore, the effects of different atmospheric perturbations on SOA formation from BB emissions appear to be less important than burn-to-burn variability.

Spatial Variability of Sources and Mixing State of Atmospheric Particles in a Metropolitan Area

Characterizing intracity variations of atmospheric particulate matter has mostly relied on fixed-site monitoring and quantifying variability in terms of different bulk aerosol species. In this study, we performed ground-based mobile measurements using a single-particle mass spectrometer to study spatial patterns of source-specific particles and the evolution of particle mixing state in 21 areas in the metropolitan area of Pittsburgh, PA. We selected sampling areas based on traffic density and restaurant density with each area ranging from 0.2 to 2 km2. Organics dominate particle composition in all of the areas we sampled while the sources of organics differ. The contribution of particles from traffic and restaurant cooking varies greatly on the neighborhood scale. We also investigate how primary and aged components in particles mix across the urban scale. Lastly we quantify and map the particle mixing state for all areas we sampled and discuss the overall pattern of mixing state evolution and its implications. We find that in the upwind and downwind of the urban areas, particles are more internally mixed while in the city center, particle mixing state shows large spatial heterogeneity that is mostly driven by emissions. This study is to our knowledge, the first study to perform fine spatial scale mapping of particle mixing state using ground-based mobile measurement and single-particle mass spectrometry.

Mass accommodation coefficients of fresh and aged biomass-burning emissions

ABSTRACT Most chemical transport models treat the partitioning of semi-volatile organic compounds (SVOCs) with the assumption of instantaneous thermodynamic equilibrium. However, the mass accommodation coefficients, α, of biomass-burning organic aerosol (BBOA) are largely unconstrained. During the FLAME-IV campaign, we thermally perturbed aged and fresh BBOA with a variable residence time thermodenuder and measured the resulting change in particle mass concentration to restore equilibrium. We used this equilibration profile to retrieve an effective α for components of BBOA that dictated this profile and found that the mass accommodation coefficients lie within the range 0.1 ≪ α ⩽ 1. A simple plume dilution model shows a maximum of only a 7% difference between a dynamical and an instantaneous equilibrium partitioning model using our best-estimate value for α. This supports continued use of the equilibrium assumption to treat partitioning of biomass-burning emissions in chemical-transport models. Copyright

Evaporation rate of particles in the vaporizer of the Aerodyne aerosol mass spectrometer

ABSTRACT We present calculations for evaporation rates of particles collected on the vaporizer of the Aerodyne aerosol mass spectrometer (AMS). These calculations provide insight on certain observed phenomena associated with the size-resolved mass spectrum (MS), because the time width of the MS signal from a particle can be limited by its evaporation rate upon contact with the vaporizer. We show that the counterintuitive weak dependence of observed MS signal widths (evaporation rates) on particle volatility is due to suppression of evaporation rates induced by latent heat release, which is more prominent at high volatilities. The same physics is responsible for the observed diminishing returns associated with increasing the vaporizer temperature to achieve narrower single particle pulses. We also show that the vaporizer typical operating temperature of 600°C is sufficient to evaporate extremely low volatility organic compounds (ELVOCs) rapidly enough to obtain reliable measurements for particles smaller than approximately 600 nm. However, the sizing resolution is compromised for large (near-micron) sizes regardless of particle volatility. Finally, our calculations indicate that the observed delayed particle signals, which lead to an artificial tail in AMS mass distributions, are not due to slow evaporation of particles deposited on a surface with lower temperature than the vaporizer, but particles bouncing in the ionizer cage and finally depositing on the vaporizer. Copyright